/* A program for displaying comments from the GPv3 drafting process on gplv3.fsf.org.
Copyright (C) 2021 FSF
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see . */
const COMMENT_JSON_XML = {"2875:2961:3118": "\n\n\n\n<p style="display: inline">Following is a summary of the GPLv3d3 in 2 pages. The numbers in brackets indicate the related sections in the GPLv3d3. Numbers in inner brackets are the corresponding GPLv2 sections. The Preamble, definitions, license revisions, and applying the license sections are not included.</p><p>\nPlease consider this layout for the GPLv3.</p><p>\n1. Receiving Works and Licences</p><p>\n1.1. Grant of License: Upon your receipt of the Program, each contributor grants you a license for the copyrights and patents they hold covering their contributions. [2, 10[6], 11]</p><p>\na. License Acceptance: Use of covered works beyond that allowed in Section 2.1 indicates your acceptance of this License. [9[5]]</p><p>\nb. Warranty: There is no warranty for the Program, unless otherwise agreed to in writing in return for a fee. [15[11,12]]</p><p>\nc. Warranty Additional Restrictions: You must accept contributors' additional liability disclaimers. [7a]</p><p>\nd. Names and Marks Additional Restrictions: You must accept contributors' limitations on the use of trademarks, etc. as permitted by law. [7d]</p><p>\ne. Local Law Additional Restrictions: You must accept contributors' additional terms drafted for compatibility with local law, in so far as they uphold the intent of this License. [7]</p><p>\nf. License Adherence: Other obligations do not excuse you from the conditions of this License. [12[7]]</p><p>\ng. License Termination: Contributors may terminate your right to use their contributions if you violate this License. [8[4]]</p><p>\n1.2. Additional Terms: Additional permissions may have been granted for the Program or parts of it, and you are not bound by purported additional restrictions beyond those allowed by this License. [7]</p><p>\n1.3. Receiving Source Code: You could be entitled to obtain the Program's source code from the party that conveyed you the object code.</p><p>\n2. Basic Permissions.</p><p>\n2.1 Using the Program: You may receive or run the Program without accepting this License, and you are entitled to fair use. [2, 9[5]]</p><p>\n2.2 Making and Using Covered Works: You may make and use covered works. [2]</p><p>\n3. Conveying Source.</p><p>\n3.1. Conveying Verbatim Copies: You may convey verbatim copies of the Program's source code. [4[1]]</p><p>\na. Licensing: You may not sub-license the Program or impose further restrictions on the rights granted by this License, and you are not responsible for enforcing compliance with this License. [2, 10[6]]</p><p>\nb. Patents: You may not initiate patent litigation covering the Program, and if you arrange or knowingly rely on patent licenses, you must protect all recipients of covered works. [10[6], 11]</p><p>\nc. DRM: You waive any legal right to forbid circumvention of technical measures that limit the use of covered works under the License. [3]</p><p>\nd. Notices: You must keep intact license notices and warranty disclaimers, publish a copyright notice on each copy, and give recipients a copy of this License. [4[1]]</p><p>\ne. Transfer of Control: Transfer of control of a covered work is considered to be conveying. [10[6]]</p><p>\n3.2. Conveying Modified Source Versions: You may convey a work based on the Program in source code form. [5[2], 7]</p><p>\na. Above Terms: You must adhere to the terms of Section 3.1. [5[2]]</p><p>\nb. Licensing: You must license the derived work, excluding linked AGPL works, under this License, although you may remove additional permissions, or place additional terms on your contribution as permitted by this License. [5c[2], 7, 13]</p><p>\nc. Notices: The derived work must carry copyright notices, licensing notices, including applicable additional terms, warranty disclaimers, and notices of your modification. [5abd[2], 7]</p><p>\nd. Notices Additional Restrictions: You must adhere to contributors' additional requirements to preserve notices and mark modified versions. [7bc]</p><p>\n4. Conveying Non-Source Forms: You may convey covered works in object code form. [6[3]]</p><p>\na. Above Terms: You must adhere to the terms of Sections 3.1 and 3.2. [6[3]]</p><p>\nb. Source Code: You must provide Corresponding Source. [6[3]]</p><p>\nc. Installation Information: If you convey object code for use in a User Product, you must provide Installation Information for modified versions. [6[3]]</p>\ntitle.0.0\njamesgnz\nlogin to agree\n0\n2875\n\n\n<p style="display: inline">I understand that it was actually a stated goal of the GPLv3 to internationalize the license. Surely then, it should be in international English. So, for starters, License -> Licence (when the word is used as a noun).</p><p>\nI suggested this in Comment 2159: S -> C (international English), but didn't receive any feedback. Are there any reasons not to do this?</p>\ntitle.0.0\njamesgnz\nlogin to agree\n1\n2961\n\n\n<p style="display: inline">Write the SFDL as a set of additional permissions to the GPL (like the LGPL). Maybe call it the Printing GPL or Document GPL.</p><p>\nThe SFDL states that it is intended for "any work of authorship meant for human appreciation, rather than machine execution". Whereas the GPL caters for executable works that are practical, rather than for human appreciation. However the lines are poorly drawn.</p><p>\nIt is not difficult to imagine that someone might want to use fiction as the basis for a game (in fact, it happens often), nor that someone might want to use non-fiction as the basis for interactive teaching material. However, in each case, the former is likely to be licensed under the SFDL, whereas the later would require the GPL.</p><p>\nGames are intended for both machine execution and human appreciation. Like fiction, they often have (to a greater or lesser extent) characters, dialogue, and a plot which will inevitably convey a moral. Interactive teaching material is also intended for both machine execution and human appreciation. And like non-fiction, it conveys concepts and ideas. So if the SFDL addresses important issues regarding works intended for human appreciation, then the GPL is flawed in not addressing these issues.</p><p>\nOther than that, the SFDL also allows extra permissions for printing documents. However, this function could be performed by a set of additional permissions to the GPL, the way the LGPL works.</p><p>\n(I originally posted this comment on the SFDL as a reply to Comment 2701: GPL compatibility, but since it applies to both licenses, I'm posting it here too.)</p>\ntitle.0.0\njamesgnz\nlogin to agree\n1\n3118\n\n\n", "2875:2961:3118:3134": "\n\n\n\n<p style="display: inline">Following is a summary of the GPLv3d3 in 2 pages. The numbers in brackets indicate the related sections in the GPLv3d3. Numbers in inner brackets are the corresponding GPLv2 sections. The Preamble, definitions, license revisions, and applying the license sections are not included.</p><p>\nPlease consider this layout for the GPLv3.</p><p>\n1. Receiving Works and Licences</p><p>\n1.1. Grant of License: Upon your receipt of the Program, each contributor grants you a license for the copyrights and patents they hold covering their contributions. [2, 10[6], 11]</p><p>\na. License Acceptance: Use of covered works beyond that allowed in Section 2.1 indicates your acceptance of this License. [9[5]]</p><p>\nb. Warranty: There is no warranty for the Program, unless otherwise agreed to in writing in return for a fee. [15[11,12]]</p><p>\nc. Warranty Additional Restrictions: You must accept contributors' additional liability disclaimers. [7a]</p><p>\nd. Names and Marks Additional Restrictions: You must accept contributors' limitations on the use of trademarks, etc. as permitted by law. [7d]</p><p>\ne. Local Law Additional Restrictions: You must accept contributors' additional terms drafted for compatibility with local law, in so far as they uphold the intent of this License. [7]</p><p>\nf. License Adherence: Other obligations do not excuse you from the conditions of this License. [12[7]]</p><p>\ng. License Termination: Contributors may terminate your right to use their contributions if you violate this License. [8[4]]</p><p>\n1.2. Additional Terms: Additional permissions may have been granted for the Program or parts of it, and you are not bound by purported additional restrictions beyond those allowed by this License. [7]</p><p>\n1.3. Receiving Source Code: You could be entitled to obtain the Program's source code from the party that conveyed you the object code.</p><p>\n2. Basic Permissions.</p><p>\n2.1 Using the Program: You may receive or run the Program without accepting this License, and you are entitled to fair use. [2, 9[5]]</p><p>\n2.2 Making and Using Covered Works: You may make and use covered works. [2]</p><p>\n3. Conveying Source.</p><p>\n3.1. Conveying Verbatim Copies: You may convey verbatim copies of the Program's source code. [4[1]]</p><p>\na. Licensing: You may not sub-license the Program or impose further restrictions on the rights granted by this License, and you are not responsible for enforcing compliance with this License. [2, 10[6]]</p><p>\nb. Patents: You may not initiate patent litigation covering the Program, and if you arrange or knowingly rely on patent licenses, you must protect all recipients of covered works. [10[6], 11]</p><p>\nc. DRM: You waive any legal right to forbid circumvention of technical measures that limit the use of covered works under the License. [3]</p><p>\nd. Notices: You must keep intact license notices and warranty disclaimers, publish a copyright notice on each copy, and give recipients a copy of this License. [4[1]]</p><p>\ne. Transfer of Control: Transfer of control of a covered work is considered to be conveying. [10[6]]</p><p>\n3.2. Conveying Modified Source Versions: You may convey a work based on the Program in source code form. [5[2], 7]</p><p>\na. Above Terms: You must adhere to the terms of Section 3.1. [5[2]]</p><p>\nb. Licensing: You must license the derived work, excluding linked AGPL works, under this License, although you may remove additional permissions, or place additional terms on your contribution as permitted by this License. [5c[2], 7, 13]</p><p>\nc. Notices: The derived work must carry copyright notices, licensing notices, including applicable additional terms, warranty disclaimers, and notices of your modification. [5abd[2], 7]</p><p>\nd. Notices Additional Restrictions: You must adhere to contributors' additional requirements to preserve notices and mark modified versions. [7bc]</p><p>\n4. Conveying Non-Source Forms: You may convey covered works in object code form. [6[3]]</p><p>\na. Above Terms: You must adhere to the terms of Sections 3.1 and 3.2. [6[3]]</p><p>\nb. Source Code: You must provide Corresponding Source. [6[3]]</p><p>\nc. Installation Information: If you convey object code for use in a User Product, you must provide Installation Information for modified versions. [6[3]]</p>\ntitle.0.0\njamesgnz\nlogin to agree\n0\n2875\n\n\n<p style="display: inline">I understand that it was actually a stated goal of the GPLv3 to internationalize the license. Surely then, it should be in international English. So, for starters, License -> Licence (when the word is used as a noun).</p><p>\nI suggested this in Comment 2159: S -> C (international English), but didn't receive any feedback. Are there any reasons not to do this?</p>\ntitle.0.0\njamesgnz\nlogin to agree\n1\n2961\n\n\n<p style="display: inline">Write the SFDL as a set of additional permissions to the GPL (like the LGPL). Maybe call it the Printing GPL or Document GPL.</p><p>\nThe SFDL states that it is intended for "any work of authorship meant for human appreciation, rather than machine execution". Whereas the GPL caters for executable works that are practical, rather than for human appreciation. However the lines are poorly drawn.</p><p>\nIt is not difficult to imagine that someone might want to use fiction as the basis for a game (in fact, it happens often), nor that someone might want to use non-fiction as the basis for interactive teaching material. However, in each case, the former is likely to be licensed under the SFDL, whereas the later would require the GPL.</p><p>\nGames are intended for both machine execution and human appreciation. Like fiction, they often have (to a greater or lesser extent) characters, dialogue, and a plot which will inevitably convey a moral. Interactive teaching material is also intended for both machine execution and human appreciation. And like non-fiction, it conveys concepts and ideas. So if the SFDL addresses important issues regarding works intended for human appreciation, then the GPL is flawed in not addressing these issues.</p><p>\nOther than that, the SFDL also allows extra permissions for printing documents. However, this function could be performed by a set of additional permissions to the GPL, the way the LGPL works.</p><p>\n(I originally posted this comment on the SFDL as a reply to Comment 2701: GPL compatibility, but since it applies to both licenses, I'm posting it here too.)</p>\ntitle.0.0\njamesgnz\nlogin to agree\n1\n3118\n\n\nI appreciate that GPL3 (like other GPLs before it) is generally regarded as a bare license. In other words, breach me, and you no longer have permission to do anything with the code which would require permission under applicable copyright law. Under English law, at least, this does have some disadvantages. The remedies available to the copyright holder are (1) an account of profits (difficult to assess, as the work is likely to be combined with others); (2) diminution of overall value of the software (if it's available gratis, then the court is unlikely to be willing to assess monetary damages under this head) and (3) injunction (which is at the discretion of the court anyway). If GPL3 is stated to be a licence agreement (which it can easily be made to be, and is, arguably, already) (as a unilateral contract) then there are additional remedies available to the copyright holder and any third party. In brief, it would be possible to structure the license chain so that any third party would have the right, as a third party beneficiary, to apply to the court for an order requiring that any conveyor of modified covered code provide the source code in accordance with the terms of the license. As I understand it, currently, it is only the copyright holder which has any effective remedy against the party in breach in these circumstances. This relatively small change would strengthen the copyleft provision significantly (albeit at some cost, particularly in relation to the liability disclaimers)\ntitle.0.0\nandrewk\nlogin to agree\n0\n3134\n\n\n", "2566": "\n\n\n\nWhy is the license text copyrighted? I know license proliferation is not helping the Free Software Community, but if someone wants to create his own license (for software or for something else), and he wants to use clauses similar to those of the GPL, why would we want to prevent him using the same (copied) wording? Using copyright on legal texts and licenses seems to me to be even worse a misuse of copyright law as software patents are a misuse of the patents system. If the license must be copyrighted, the meta-license (currently only verbatim copies allowed) should be more loose/liberal/free. [On the other hand, I might understand it if the FSF wanted to avoid other people calling their licenses something akin to the term General Public License or GPL. Maybe we need a trade mark or something similar on that name.]\ncopyright.0.0\nadhemar\nlogin to agree\n5\n2566\n\n\n", "2566:2917:3060": "\n\n\n\nWhy is the license text copyrighted? I know license proliferation is not helping the Free Software Community, but if someone wants to create his own license (for software or for something else), and he wants to use clauses similar to those of the GPL, why would we want to prevent him using the same (copied) wording? Using copyright on legal texts and licenses seems to me to be even worse a misuse of copyright law as software patents are a misuse of the patents system. If the license must be copyrighted, the meta-license (currently only verbatim copies allowed) should be more loose/liberal/free. [On the other hand, I might understand it if the FSF wanted to avoid other people calling their licenses something akin to the term General Public License or GPL. Maybe we need a trade mark or something similar on that name.]\ncopyright.0.0\nadhemar\nlogin to agree\n5\n2566\n\n\nUse the correct copyright symbol.\ncopyright.0.0\nnslater\nlogin to agree\n0\n2917\n\n\nThe issue here isn't that the licence is copyrighted. After all, software under the GPL is copyrighted, but it's still free (as in freedom), as it respects people's freedom. However, I think (as do others), that the licence should have more liberal terms of use, specifically that distributing modified versions should be allowed. There's more comments about this which refer to the "but changing it is not allowed" part.\ncopyright.0.0\njohnston\nlogin to agree\n0\n3060\n\n\n", "2435:2604:2741": "\n\n\n\nThis URL is missing a trailing '/'. It should read "http://fsf.org/" or "http://www.fsf.org/".\ncopyright.0.0\nfraggle\nlogin to agree\n16\n2435\n\n\n-> "http://www.fsf.org/"\ncopyright.0.0\nskquinn\nlogin to agree\n2\n2604\n\n\n<p style="display: inline">Angled brackets are generally better for URL delimiters. Some URLs contain rounded brackets, so rounded brackets can not be used as a general method of delimiting URLs without potentially causing confusion in some instances. No URL contains angled brackets.</p><p>\nIt's also recommended by the W3C\nhttp://www.w3.org/Addressing/URL/5.1_Wrappers.html\n(I would put angled brackets on this URL, but I don't know whether the comment system would think it was an HTML tag.)</p><p>\nAnd personally, I think they look nicer.</p>\ncopyright.0.0\njamesgnz\nlogin to agree\n6\n2741\n\n\n", "2641": "\n\n\n\nThe address has changed once before and will likely change again during the life of the license. Suggest instead something like 'Free Software Foundation, Inc., an association incorporated in the USA'. The purpose is to specifically define the legal entity, not provide a mailing address (and in this context would an email address not be more appropriate in the vast majority of cases?).\ncopyright.0.0\nsamj\nlogin to agree\n5\n2641\n\n\n", "2580:2641": "\n\n\n\nNeeds a full stop between USA and Everyone: "...USA. Everyone..."\ncopyright.0.0\nsanjoy\nlogin to agree\n9\n2580\n\n\nThe address has changed once before and will likely change again during the life of the license. Suggest instead something like 'Free Software Foundation, Inc., an association incorporated in the USA'. The purpose is to specifically define the legal entity, not provide a mailing address (and in this context would an email address not be more appropriate in the vast majority of cases?).\ncopyright.0.0\nsamj\nlogin to agree\n5\n2641\n\n\n", "2580:3018": "\n\n\n\nNeeds a full stop between USA and Everyone: "...USA. Everyone..."\ncopyright.0.0\nsanjoy\nlogin to agree\n9\n2580\n\n\n<p style="display: inline">The very first sentence is a condition that says:</p><p>\n"Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed."</p><p>\n1. If the first sentence goes anywhere, it should go under Section 14, "Revised Versions of this License". \n\n2. The first sentence conflicts with part or all of Section 7, "Additional Terms". \n\n3. The first sentence should be shortened as follows: "Anyone may copy and redistribute this license without modification."</p>\ncopyright.0.0\njkoenig\nlogin to agree\n1\n3018\n\n\n", "3018": "\n\n\n\n<p style="display: inline">The very first sentence is a condition that says:</p><p>\n"Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed."</p><p>\n1. If the first sentence goes anywhere, it should go under Section 14, "Revised Versions of this License". \n\n2. The first sentence conflicts with part or all of Section 7, "Additional Terms". \n\n3. The first sentence should be shortened as follows: "Anyone may copy and redistribute this license without modification."</p>\ncopyright.0.0\njkoenig\nlogin to agree\n1\n3018\n\n\n", "2501:2567:2642:2952:2957:2988:3018:3033": "\n\n\n\n\tFree software needs a free license! The GNU GPL alone will never be sufficient as an all-encompassing document able to cover every program. It must allow reasonable modifications. The open source revolution will fail if the software industry giants continue to attack developers in court for harming their monopolies.\n\tOpponents to allowing modifications to the GNU GPL say that one could release a non-free license as the GNU GPL, or that it is immoral to edit the GNU GPL. Obviously, the GNU GPL must require modified versions to say \u00c3\u00a2\u00c2\u0080\u00c2\u009cThis is not the GNU GPL\u00c3\u00a2\u00c2\u0080\u00c2\u009d. Also, modified versions of the GPL could not allow modifications, because this would cause so many strange versions of the GPL to be created that some would look nothing like the GPL. Because the GNU GPL doesn\u00c3\u00a2\u00c2\u0080\u00c2\u0099t allow modifications, it is non-free, thus, Richard Stallman contradicts himself by releasing the GNU GPL as non-free intellectual property and violates his own community's definition of morality. Others say that an exception file can change the GNU GPL without editing it. However, if anyone does not notice the exception file, and offends Microsoft\u00c3\u0082\u00c2\u00ae, the developer of the program will be required by a lawsuit filed by Microsoft\u00c3\u0082\u00c2\u00ae to stop distributing the program, or worse, the Free Software Foundation could be charged for supporting a program intended to break laws.\n\tThe GPL does not fit every program. Most open source programmers are not lawyers alongside being programmers, and cannot write their own licenses. Some programs, particularly encryption programs, need a license specifically stating that they are the intellectual property of someone who is a citizen of a country outside of the United States, to get around our export laws. This is not needed for most programs. For a program which copies CDs and DVDs, the license ought to say \u00c3\u00a2\u00c2\u0080\u00c2\u009cThis program may not be used to illegally copy anything.\u00c3\u00a2\u00c2\u0080\u00c2\u009d Otherwise, a user would make an illegal copy of five movies, two music CDs, and a video game, resulting in the developer being sued because he wrote a tool for software piracy. For example, in the DeCSS lawsuit, a programmer was sued for freely distributing a legal program. The problem was that the program could be used to illegally copy DVDs over the Internet. This time, the programmer won, but if he had lost, the Business Software Alliance\u00c3\u0082\u00c2\u00ae would have sued whoever had the most money, most likely the Free Software Foundation, because very few open source programmers can pay extravagant damages in lawsuits.(Landley) The GPL cannot prevent all of these problems and still allow programs which cannot possibly be used in an illegal way to be entirely free.\n\t\tVery strong licenses are needed to protect programmers from lawsuits. Clearly, Microsoft\u00c3\u0082\u00c2\u00ae doesn\u00c3\u00a2\u00c2\u0080\u00c2\u0099t like open source, and will do anything they can to stop it. Look at all of the antitrust lawsuits. How can Microsoft\u00c3\u0082\u00c2\u00ae survive despite its obvious guilt? They must be bribing judges to lessen their penalties for monopolizing the American operating system market. Think about it, if you have fifty two billion dollars and someone tried to take all of it away, what would you give to keep most of it? If you were a judge, how high a bribe would you accept? Would five million dollars be enough? If you were fighting your most powerful enemy, the Free Software Foundation, what would you give? If Microsoft\u00c3\u0082\u00c2\u00ae bribes judges in a fight against the Free Software Foundation, the innocence of the developers involved and the Free Software Foundation must be obvious or the judge could accept the bribe without raising suspicion and destroy free software. If all programmers could make it impossible for themselves to be successfully sued, open source\u00c3\u00a2\u00c2\u0080\u00c2\u0099s replacement of commercial software would be inevitable, and soon no one would pay for software, or pirate it, again.\n\tThe open source revolution is based on freedom. Developers will feel more comfortable with their own licenses applied to their software, giving them more freedom as to what one can do with their programs. However, they will want the license to prevent themselves from being sued. So, they will each need a lawyer, which few programmers can afford. The solution is to add an exceptions file. But then, twenty people will ignore the exceptions file and do something illegal with the program, and Microsoft\u00c3\u0082\u00c2\u00ae will sue the programmer. The only remaining solution is to let programmers modify the GPL. It will cause far more open source programs to be released. Also, it will keep the people who really fuel the free software revolution satisfied.\n\tWhat is to be lost in allowing reasonable modifications to the GPL? More has been lost already in banning modifications to the GPL. When modifications are allowed, there will be a surge of programs released which were too powerful to be safely released under the old GPL. Over time, more free programs will be written because developers will feel more secure in releasing them. The Free Software Foundation will be safer from lawsuits arising from illegal activities, mostly software piracy and bad hacking, using free software. Morally, the GPL must be as free as the programs it is applied to. Many new free software users or developers have left free software and reverted to using Windows\u00c3\u0082\u00c2\u00ae or writing shareware because of this irony and moral error. They will not come back until it is corrected, or possibly not at all. The longer the mistake is left uncorrected, the more people will leave free software forever. There is no adequate substitute for a truly free license. There is nothing wrong with allowing reasonable modifications to the GPL. What are we waiting for? It is time to allow modifications to the GPL now!\ncopyright.0.0\ndakdevel\nlogin to agree\n1\n2501\n\n\nIs Lawrence Lessig not constantly preaching that Law is Code (or something very similar) (with its own syntax, and logic, and side effects, etc)? By that logic, should the GPL not be under a free license? Maybe even the GPL itself, or something similar.\ncopyright.0.0\nadhemar\nlogin to agree\n6\n2567\n\n\nShouldn't others be permitted to change the content but be restricted from doing so under the name 'GNU General Public License'? Is the GFDL appropriate here?\ncopyright.0.0\nsamj\nlogin to agree\n2\n2642\n\n\nI think that modifying the GPL (and all other copyright licences) should definitely be allowed, as long as it's made completely clear that it's a different licence. A copyright licence is a functional work, and I think all functional works should be free in the same sense as free software. I've heard RMS say something similar to that in his speeches before. I'd possibly even go as far as saying the FSF is doing something unethical by not allowing modified versions. As the preamble isn't functional, as it's not part of the section which actually acts as a copyright licence, I think prohibiting people from modifying it is acceptable. I think that notice should be changed to something along the lines of "Everyone is permitted to copy and distribute verbatim copies of this license document. Modified versions may also be made and distributed as long as the name of the licence is changed, the terms 'GNU' and 'General Public License' are removed, and the preamble is removed." In fact, the FSF gives permission to do this (with an extra condition about the instructions-for-use) at http://www.gnu.org/licenses/gpl-faq.html#ModifyGPL. However, people aren't very likely to come across that, so that permission should be given in the licence text. The FSF is contradicting itself and being misleading by claiming that "changing it is not allowed".\ncopyright.0.0\njohnston\nlogin to agree\n5\n2952\n\n\n<p style="display: inline">Free software needs a free license.</p><p>\nHow about something like this:</p><p>\n"This license document is licensed to everyone under the GNU General Public License, version 3 or later. (Please note that any derivative documents are not the GNU General Public License version 3 and must not pretend to be: only the unmodified document is the GNU General Public License.)"\n</p>\ncopyright.0.0\nneroden\nlogin to agree\n2\n2957\n\n\n<p style="display: inline">The other people who remarked on this point are quite right. On the one hand, you wish people to have control over the software on their machines, but on the other, apparently you would rather they didn't create modified versions of your legal work?!</p><p>\nSince this is all about software developers abdicating the right to control derivative works, how about your lawyers do so too?</p>\ncopyright.0.0\nalastair\nlogin to agree\n4\n2988\n\n\n<p style="display: inline">The very first sentence is a condition that says:</p><p>\n"Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed."</p><p>\n1. If the first sentence goes anywhere, it should go under Section 14, "Revised Versions of this License". \n\n2. The first sentence conflicts with part or all of Section 7, "Additional Terms". \n\n3. The first sentence should be shortened as follows: "Anyone may copy and redistribute this license without modification."</p>\ncopyright.0.0\njkoenig\nlogin to agree\n1\n3018\n\n\n<p style="display: inline">This statement needs to stay. If people start making their own versions of the GPL (even if they give it a new name, and make it clearly different), we will quite likely have people adding more restrictions that will make their resulting programs not GPL-compatible.</p><p>\nThat defeats the entire purpose of this document. If someone wants to make a license that is based upon the ideas of the GPL, that's one thing. But letting them change a few lines and re-release is a bad plan...</p>\ncopyright.0.0\nLovasco\nlogin to agree\n2\n3033\n\n\n", "2435:2501:2566:2567:2580:2604:2641:2642:2741:2917:2952:2957:2988:3018:3033:3060": "\n\n\n\nThis URL is missing a trailing '/'. It should read "http://fsf.org/" or "http://www.fsf.org/".\ncopyright.0.0\nfraggle\nlogin to agree\n16\n2435\n\n\n\tFree software needs a free license! The GNU GPL alone will never be sufficient as an all-encompassing document able to cover every program. It must allow reasonable modifications. The open source revolution will fail if the software industry giants continue to attack developers in court for harming their monopolies.\n\tOpponents to allowing modifications to the GNU GPL say that one could release a non-free license as the GNU GPL, or that it is immoral to edit the GNU GPL. Obviously, the GNU GPL must require modified versions to say \u00c3\u00a2\u00c2\u0080\u00c2\u009cThis is not the GNU GPL\u00c3\u00a2\u00c2\u0080\u00c2\u009d. Also, modified versions of the GPL could not allow modifications, because this would cause so many strange versions of the GPL to be created that some would look nothing like the GPL. Because the GNU GPL doesn\u00c3\u00a2\u00c2\u0080\u00c2\u0099t allow modifications, it is non-free, thus, Richard Stallman contradicts himself by releasing the GNU GPL as non-free intellectual property and violates his own community's definition of morality. Others say that an exception file can change the GNU GPL without editing it. However, if anyone does not notice the exception file, and offends Microsoft\u00c3\u0082\u00c2\u00ae, the developer of the program will be required by a lawsuit filed by Microsoft\u00c3\u0082\u00c2\u00ae to stop distributing the program, or worse, the Free Software Foundation could be charged for supporting a program intended to break laws.\n\tThe GPL does not fit every program. Most open source programmers are not lawyers alongside being programmers, and cannot write their own licenses. Some programs, particularly encryption programs, need a license specifically stating that they are the intellectual property of someone who is a citizen of a country outside of the United States, to get around our export laws. This is not needed for most programs. For a program which copies CDs and DVDs, the license ought to say \u00c3\u00a2\u00c2\u0080\u00c2\u009cThis program may not be used to illegally copy anything.\u00c3\u00a2\u00c2\u0080\u00c2\u009d Otherwise, a user would make an illegal copy of five movies, two music CDs, and a video game, resulting in the developer being sued because he wrote a tool for software piracy. For example, in the DeCSS lawsuit, a programmer was sued for freely distributing a legal program. The problem was that the program could be used to illegally copy DVDs over the Internet. This time, the programmer won, but if he had lost, the Business Software Alliance\u00c3\u0082\u00c2\u00ae would have sued whoever had the most money, most likely the Free Software Foundation, because very few open source programmers can pay extravagant damages in lawsuits.(Landley) The GPL cannot prevent all of these problems and still allow programs which cannot possibly be used in an illegal way to be entirely free.\n\t\tVery strong licenses are needed to protect programmers from lawsuits. Clearly, Microsoft\u00c3\u0082\u00c2\u00ae doesn\u00c3\u00a2\u00c2\u0080\u00c2\u0099t like open source, and will do anything they can to stop it. Look at all of the antitrust lawsuits. How can Microsoft\u00c3\u0082\u00c2\u00ae survive despite its obvious guilt? They must be bribing judges to lessen their penalties for monopolizing the American operating system market. Think about it, if you have fifty two billion dollars and someone tried to take all of it away, what would you give to keep most of it? If you were a judge, how high a bribe would you accept? Would five million dollars be enough? If you were fighting your most powerful enemy, the Free Software Foundation, what would you give? If Microsoft\u00c3\u0082\u00c2\u00ae bribes judges in a fight against the Free Software Foundation, the innocence of the developers involved and the Free Software Foundation must be obvious or the judge could accept the bribe without raising suspicion and destroy free software. If all programmers could make it impossible for themselves to be successfully sued, open source\u00c3\u00a2\u00c2\u0080\u00c2\u0099s replacement of commercial software would be inevitable, and soon no one would pay for software, or pirate it, again.\n\tThe open source revolution is based on freedom. Developers will feel more comfortable with their own licenses applied to their software, giving them more freedom as to what one can do with their programs. However, they will want the license to prevent themselves from being sued. So, they will each need a lawyer, which few programmers can afford. The solution is to add an exceptions file. But then, twenty people will ignore the exceptions file and do something illegal with the program, and Microsoft\u00c3\u0082\u00c2\u00ae will sue the programmer. The only remaining solution is to let programmers modify the GPL. It will cause far more open source programs to be released. Also, it will keep the people who really fuel the free software revolution satisfied.\n\tWhat is to be lost in allowing reasonable modifications to the GPL? More has been lost already in banning modifications to the GPL. When modifications are allowed, there will be a surge of programs released which were too powerful to be safely released under the old GPL. Over time, more free programs will be written because developers will feel more secure in releasing them. The Free Software Foundation will be safer from lawsuits arising from illegal activities, mostly software piracy and bad hacking, using free software. Morally, the GPL must be as free as the programs it is applied to. Many new free software users or developers have left free software and reverted to using Windows\u00c3\u0082\u00c2\u00ae or writing shareware because of this irony and moral error. They will not come back until it is corrected, or possibly not at all. The longer the mistake is left uncorrected, the more people will leave free software forever. There is no adequate substitute for a truly free license. There is nothing wrong with allowing reasonable modifications to the GPL. What are we waiting for? It is time to allow modifications to the GPL now!\ncopyright.0.0\ndakdevel\nlogin to agree\n1\n2501\n\n\nWhy is the license text copyrighted? I know license proliferation is not helping the Free Software Community, but if someone wants to create his own license (for software or for something else), and he wants to use clauses similar to those of the GPL, why would we want to prevent him using the same (copied) wording? Using copyright on legal texts and licenses seems to me to be even worse a misuse of copyright law as software patents are a misuse of the patents system. If the license must be copyrighted, the meta-license (currently only verbatim copies allowed) should be more loose/liberal/free. [On the other hand, I might understand it if the FSF wanted to avoid other people calling their licenses something akin to the term General Public License or GPL. Maybe we need a trade mark or something similar on that name.]\ncopyright.0.0\nadhemar\nlogin to agree\n5\n2566\n\n\nIs Lawrence Lessig not constantly preaching that Law is Code (or something very similar) (with its own syntax, and logic, and side effects, etc)? By that logic, should the GPL not be under a free license? Maybe even the GPL itself, or something similar.\ncopyright.0.0\nadhemar\nlogin to agree\n6\n2567\n\n\nNeeds a full stop between USA and Everyone: "...USA. Everyone..."\ncopyright.0.0\nsanjoy\nlogin to agree\n9\n2580\n\n\n-> "http://www.fsf.org/"\ncopyright.0.0\nskquinn\nlogin to agree\n2\n2604\n\n\nThe address has changed once before and will likely change again during the life of the license. Suggest instead something like 'Free Software Foundation, Inc., an association incorporated in the USA'. The purpose is to specifically define the legal entity, not provide a mailing address (and in this context would an email address not be more appropriate in the vast majority of cases?).\ncopyright.0.0\nsamj\nlogin to agree\n5\n2641\n\n\nShouldn't others be permitted to change the content but be restricted from doing so under the name 'GNU General Public License'? Is the GFDL appropriate here?\ncopyright.0.0\nsamj\nlogin to agree\n2\n2642\n\n\n<p style="display: inline">Angled brackets are generally better for URL delimiters. Some URLs contain rounded brackets, so rounded brackets can not be used as a general method of delimiting URLs without potentially causing confusion in some instances. No URL contains angled brackets.</p><p>\nIt's also recommended by the W3C\nhttp://www.w3.org/Addressing/URL/5.1_Wrappers.html\n(I would put angled brackets on this URL, but I don't know whether the comment system would think it was an HTML tag.)</p><p>\nAnd personally, I think they look nicer.</p>\ncopyright.0.0\njamesgnz\nlogin to agree\n6\n2741\n\n\nUse the correct copyright symbol.\ncopyright.0.0\nnslater\nlogin to agree\n0\n2917\n\n\nI think that modifying the GPL (and all other copyright licences) should definitely be allowed, as long as it's made completely clear that it's a different licence. A copyright licence is a functional work, and I think all functional works should be free in the same sense as free software. I've heard RMS say something similar to that in his speeches before. I'd possibly even go as far as saying the FSF is doing something unethical by not allowing modified versions. As the preamble isn't functional, as it's not part of the section which actually acts as a copyright licence, I think prohibiting people from modifying it is acceptable. I think that notice should be changed to something along the lines of "Everyone is permitted to copy and distribute verbatim copies of this license document. Modified versions may also be made and distributed as long as the name of the licence is changed, the terms 'GNU' and 'General Public License' are removed, and the preamble is removed." In fact, the FSF gives permission to do this (with an extra condition about the instructions-for-use) at http://www.gnu.org/licenses/gpl-faq.html#ModifyGPL. However, people aren't very likely to come across that, so that permission should be given in the licence text. The FSF is contradicting itself and being misleading by claiming that "changing it is not allowed".\ncopyright.0.0\njohnston\nlogin to agree\n5\n2952\n\n\n<p style="display: inline">Free software needs a free license.</p><p>\nHow about something like this:</p><p>\n"This license document is licensed to everyone under the GNU General Public License, version 3 or later. (Please note that any derivative documents are not the GNU General Public License version 3 and must not pretend to be: only the unmodified document is the GNU General Public License.)"\n</p>\ncopyright.0.0\nneroden\nlogin to agree\n2\n2957\n\n\n<p style="display: inline">The other people who remarked on this point are quite right. On the one hand, you wish people to have control over the software on their machines, but on the other, apparently you would rather they didn't create modified versions of your legal work?!</p><p>\nSince this is all about software developers abdicating the right to control derivative works, how about your lawyers do so too?</p>\ncopyright.0.0\nalastair\nlogin to agree\n4\n2988\n\n\n<p style="display: inline">The very first sentence is a condition that says:</p><p>\n"Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed."</p><p>\n1. If the first sentence goes anywhere, it should go under Section 14, "Revised Versions of this License". \n\n2. The first sentence conflicts with part or all of Section 7, "Additional Terms". \n\n3. The first sentence should be shortened as follows: "Anyone may copy and redistribute this license without modification."</p>\ncopyright.0.0\njkoenig\nlogin to agree\n1\n3018\n\n\n<p style="display: inline">This statement needs to stay. If people start making their own versions of the GPL (even if they give it a new name, and make it clearly different), we will quite likely have people adding more restrictions that will make their resulting programs not GPL-compatible.</p><p>\nThat defeats the entire purpose of this document. If someone wants to make a license that is based upon the ideas of the GPL, that's one thing. But letting them change a few lines and re-release is a bad plan...</p>\ncopyright.0.0\nLovasco\nlogin to agree\n2\n3033\n\n\nThe issue here isn't that the licence is copyrighted. After all, software under the GPL is copyrighted, but it's still free (as in freedom), as it respects people's freedom. However, I think (as do others), that the licence should have more liberal terms of use, specifically that distributing modified versions should be allowed. There's more comments about this which refer to the "but changing it is not allowed" part.\ncopyright.0.0\njohnston\nlogin to agree\n0\n3060\n\n\n", "2680:3017:3019:3110": "\n\n\n\n<p style="display: inline">This preamble needs a complete overhaul.</p><p>\n1) Introduce the license</p><p>\n2) Introduce the license authors (FSF et al)</p><p>\n3) Introduce the reader</p><p>\nIt is talking to the LICENSEE and all members of the public who are potential licensees - not the user of the software. 'You' and 'Your' refer to the licensee.</p><p>\nDon't use terms that confuse, or distract attention from, the licensee: developer, author, individual, user, manufacturer, recipient, anyone, etc.</p><p>\nFor example, instead of 'developer' say "If you develop". Instead of 'manufacturer' say "If you manufacture".</p><p>\nNo-one else is important. We are only talking to 'you' and what 'licensees' such as 'you' need to do in order that 'your' liberty is restored. So if 'you' use, write, develop, manufacture, copy, modify, receive, give, sell, etc.</p><p>\nRights: Life, privacy, truth, liberty.\nPrivileges: copyright, patent.\nPermissions: what this license lets you do.\nConditions: what this license requires you to do if you wish to enjoy its permissions.\nFreedom: What you had prior to its suspension by copyright and patent law, and what is restored back to you by this license.\n\n\n4) Objectives</p><p>\nIt needs to state the ACHIEVABLE objectives of the GPL - not aspirations that the GPL only partially achieves.</p><p>\n5) Comparison with other licenses - and no license</p><p>\n6) What it is doing and how</p><p>\nIt is RESTORING liberty to all members of the public who become licensees, because it is only by becoming a licensee that the GPL can restore liberty (otherwise suspended by copyright and patent law).</p><p>\nThe vector by which the restorating of liberty is transmitted is via conveyance of copies - either sale/gift of authorised copies (transfer of ownership), or creation and sale/gift of authorised copies by licensees.</p><p>\nTransfer of a copy's ownership also occurs when it is intentionally combined with a computing device and the device is sold or given.</p><p>\n7) Caveats</p><p>\nThe FSF must make it clear that it is only concerned with restoring liberty.</p><p>\nIt does not stand in the way of commercial exploitation, or the ability of software developers to sell their software, or copies thereof.</p><p>\nMoreover, it does not compel the publication of software modifications by way of obligatory reciprocation.</p><p>\nFree software may be copied, modified, and exploited privately or publicly, and copies may be resold to any extent without royalty or cap on profit.</p><p>\n8) Precisely why the GPL is unique</p><p>\nWe have the usual description of how this copyleft license works to restore liberty to all downstream licensees, how it liberates and only constrains each licensee to preserve the same liberty for licensees to their derivatives and so on.</p><p>\n9) Brief discussion of nefarious circumventions of the GPL and the exotic counter-measures required</p>\ngpl3.preamble.0.0\ncrosbie\nlogin to agree\n1\n2680\n\n\nThe present preamble to the GPLv3 is 10 times longer than the Preamble to the United States Constitution! It is 4 times longer than the preamble to the Communist Manifesto. It has the same number of paragraphs as the entire U.S. Bill of Rights. Explanations in the current GPLv3 preamble look like comments to source code. Since the GPLv2 has a 27 page FAQ, there should be no need to explain what GPLv3 intends --- just move the GPLv3 preamble to the GPLv3 FAQ. Moreover, anyone who needs to determine the intent of the drafters should simply read the excellent GPLv3 drafting summary from the FSF.\ngpl3.preamble.0.0\njkoenig\nlogin to agree\n3\n3017\n\n\n<p style="display: inline">The GPLv3 is verbose. The Preamble alone is over 500 words. That's a waste of time and energy. Below is a replacement one paragraph preamble. It should be enough for people to get the point. The freedoms enumerated below are derived from the FSF description itself.</p><p>\n"The GPL Version 3 is a "free" license for use with copyright protected works such as software. "Free" here means the recipients of a work are given numerous freedoms allowing them for instance, to study the work, to change the work, to use the work for any purpose (including commercial purposes), and to publicly distribute originals or changes to the work. To ensure that all the freedoms carry forward with a work for others to likewise enjoy, the GPL Version 3 contains a condition called "copyleft" which requires source code availability of the modifications in any covered work that is redistributed. Only through this and the other license conditions below are the freedoms preserved for everyone."</p>\ngpl3.preamble.0.0\njkoenig\nlogin to agree\n3\n3019\n\n\n<p style="display: inline">We at the FSF believe that the right to use and convey information and ideas, like freedom of speech, is an important personal liberty. The GPL is intended, for the works it covers, to provide the maximum freedom to use and convey information and ideas compatible with a like freedom for all.</p><p>\nWhen a work is licensed under the GPL, everyone who obtains a copy of the work receives all the rights to use and convey the work that would normally be denied by copyright, patent, and TPM law, so long as they do not deny others these rights, and, to a certain extent, take an active role in ensuring these rights are upheld.</p><p>\nThe precise terms and conditions follow.</p><p>\n(Note: I've drawn from the other preamble rewrites that have been posted.)</p>\ngpl3.preamble.0.0\njamesgnz\nlogin to agree\n0\n3110\n\n\n", "2687:2843:3148:3186": "\n\n\n\nEven though it's in the preamble, you shouldn't use an uncommon word like copyleft without a definition or explanation.\ngpl3.preamble.p0.s1\nsepreece\nlogin to agree\n9\n2687\n\n\nThis copyleft is very flexible and the most useful for \t\nwhichever type of software. I think that the software with this free licence has many possibilities to be diffused and improved. Very good work!\ngpl3.preamble.p0.s1\nnavback\nlogin to agree\n0\n2843\n\n\nwhat bothers me is that copyleft and copyright seems to be thesis and antithesis in the classical Kantian sense, and I am just a businessperson who wants to use some open source things to stay competitive in a global market economy, not adhere to either side which seem to run, ideologically speaking, blinder and blinder in our society. I honestly believe both licenses are just as harmful to businesses, especially when your lawyer is saying Google doesn't have to reveal any code modifications in GPL software that are considered "critical" to their business, while we small businesses are being told if we commission some changes to an application under the gpl, in order to more readily compete, we must return the sourcecode to "the community" --in this case, all my competitors. Quite frankly, I think both groups -- proprietary and GPL are two faced.\ngpl3.preamble.p0.s1\nsourcev\nlogin to agree\n0\n3148\n\n\n<p style="display: inline">I'm not sure what 'copyleft' is supposed to mean, but it seems to often be taken as referring to 'left' in a political sense, conjuring up notions of dictorial central control of 'IP assets', aligning the GPL with communists and gift-economy nutters, thereby legitimizing IP trolls as in: Comment 3148: uneasiness with a communistic vs market ideological battle</p><p>\nThe GPL is not about central control at all, it is the exact opposite, it is about liberty. Rather, the Microsoft line on control of the desktop resembles Communist ideals in terms of control: we must control you because we know what is best for you, if you are given a choice you will stuff it up. (And Sun said much the same thing regarding middleware for a while.) The GPL, if anything, has more in common with the political right in this regard--i.e. a free market.</p><p>\nThe term 'copyleft' seems simply to cause confusion, I suggest it be dropped.</p>\ngpl3.preamble.p0.s1\njamesgnz\nlogin to agree\n0\n3186\n\n\n", "3148": "\n\n\n\nwhat bothers me is that copyleft and copyright seems to be thesis and antithesis in the classical Kantian sense, and I am just a businessperson who wants to use some open source things to stay competitive in a global market economy, not adhere to either side which seem to run, ideologically speaking, blinder and blinder in our society. I honestly believe both licenses are just as harmful to businesses, especially when your lawyer is saying Google doesn't have to reveal any code modifications in GPL software that are considered "critical" to their business, while we small businesses are being told if we commission some changes to an application under the gpl, in order to more readily compete, we must return the sourcecode to "the community" --in this case, all my competitors. Quite frankly, I think both groups -- proprietary and GPL are two faced.\ngpl3.preamble.p0.s1\nsourcev\nlogin to agree\n0\n3148\n\n\n", "2733:2946:2969": "\n\n\n\nIt mentions "other kinds of works" here, and yet later on only software is mentioned. I guess it's not a serious problem, since the software-like terms are redefined in a more general context. Still, why even say this? This license is clearly geared toward software.\ngpl3.preamble.p0.s1\nmkorman\nlogin to agree\n0\n2733\n\n\nFor concision, I'd think that "other works" is sufficient here. \ngpl3.preamble.p0.s1\nabhay\nlogin to agree\n0\n2946\n\n\n\nIn a software category such as entertainment software, it would be useful to have artwork, digitized sounds, and music that are licensed under the GPL or in another manner that is compatible with the GPL. There might be other categories of software that incorporate even more different kinds of works. It might be possible for the GPL to be compatible with some of the Creative Commons licenses (http://creativecommons.org) but that might not happen. It would seem that some software does not consist entirely of code.\n\n\ngpl3.preamble.p0.s1\nmezzanine1\nlogin to agree\n0\n2969\n\n\n", "2733:2839:2946:2969": "\n\n\n\nIt mentions "other kinds of works" here, and yet later on only software is mentioned. I guess it's not a serious problem, since the software-like terms are redefined in a more general context. Still, why even say this? This license is clearly geared toward software.\ngpl3.preamble.p0.s1\nmkorman\nlogin to agree\n0\n2733\n\n\nPerhaps change "other kinds of works," to "other practical works." It answers the question, "what kinds of other works"? And would be parallel with the next paragraph.\ngpl3.preamble.p0.s1\njosh\nlogin to agree\n0\n2839\n\n\nFor concision, I'd think that "other works" is sufficient here. \ngpl3.preamble.p0.s1\nabhay\nlogin to agree\n0\n2946\n\n\n\nIn a software category such as entertainment software, it would be useful to have artwork, digitized sounds, and music that are licensed under the GPL or in another manner that is compatible with the GPL. There might be other categories of software that incorporate even more different kinds of works. It might be possible for the GPL to be compatible with some of the Creative Commons licenses (http://creativecommons.org) but that might not happen. It would seem that some software does not consist entirely of code.\n\n\ngpl3.preamble.p0.s1\nmezzanine1\nlogin to agree\n0\n2969\n\n\n", "2687:2733:2839:2843:2946:2969:3148:3186": "\n\n\n\nEven though it's in the preamble, you shouldn't use an uncommon word like copyleft without a definition or explanation.\ngpl3.preamble.p0.s1\nsepreece\nlogin to agree\n9\n2687\n\n\nIt mentions "other kinds of works" here, and yet later on only software is mentioned. I guess it's not a serious problem, since the software-like terms are redefined in a more general context. Still, why even say this? This license is clearly geared toward software.\ngpl3.preamble.p0.s1\nmkorman\nlogin to agree\n0\n2733\n\n\nPerhaps change "other kinds of works," to "other practical works." It answers the question, "what kinds of other works"? And would be parallel with the next paragraph.\ngpl3.preamble.p0.s1\njosh\nlogin to agree\n0\n2839\n\n\nThis copyleft is very flexible and the most useful for \t\nwhichever type of software. I think that the software with this free licence has many possibilities to be diffused and improved. Very good work!\ngpl3.preamble.p0.s1\nnavback\nlogin to agree\n0\n2843\n\n\nFor concision, I'd think that "other works" is sufficient here. \ngpl3.preamble.p0.s1\nabhay\nlogin to agree\n0\n2946\n\n\n\nIn a software category such as entertainment software, it would be useful to have artwork, digitized sounds, and music that are licensed under the GPL or in another manner that is compatible with the GPL. There might be other categories of software that incorporate even more different kinds of works. It might be possible for the GPL to be compatible with some of the Creative Commons licenses (http://creativecommons.org) but that might not happen. It would seem that some software does not consist entirely of code.\n\n\ngpl3.preamble.p0.s1\nmezzanine1\nlogin to agree\n0\n2969\n\n\nwhat bothers me is that copyleft and copyright seems to be thesis and antithesis in the classical Kantian sense, and I am just a businessperson who wants to use some open source things to stay competitive in a global market economy, not adhere to either side which seem to run, ideologically speaking, blinder and blinder in our society. I honestly believe both licenses are just as harmful to businesses, especially when your lawyer is saying Google doesn't have to reveal any code modifications in GPL software that are considered "critical" to their business, while we small businesses are being told if we commission some changes to an application under the gpl, in order to more readily compete, we must return the sourcecode to "the community" --in this case, all my competitors. Quite frankly, I think both groups -- proprietary and GPL are two faced.\ngpl3.preamble.p0.s1\nsourcev\nlogin to agree\n0\n3148\n\n\n<p style="display: inline">I'm not sure what 'copyleft' is supposed to mean, but it seems to often be taken as referring to 'left' in a political sense, conjuring up notions of dictorial central control of 'IP assets', aligning the GPL with communists and gift-economy nutters, thereby legitimizing IP trolls as in: Comment 3148: uneasiness with a communistic vs market ideological battle</p><p>\nThe GPL is not about central control at all, it is the exact opposite, it is about liberty. Rather, the Microsoft line on control of the desktop resembles Communist ideals in terms of control: we must control you because we know what is best for you, if you are given a choice you will stuff it up. (And Sun said much the same thing regarding middleware for a while.) The GPL, if anything, has more in common with the political right in this regard--i.e. a free market.</p><p>\nThe term 'copyleft' seems simply to cause confusion, I suggest it be dropped.</p>\ngpl3.preamble.p0.s1\njamesgnz\nlogin to agree\n0\n3186\n\n\n", "2976": "\n\n\n\nIt is great that changes have been made to make sure the GPL is applicable to non-software works. However, further work is required to generalise the language, particularly in the preamble. Editing such famous and evocative text feels like handling the crown jewels, but here is my attempt at para 2:\n\n\n Some licenses for software and other works are designed to take away your freedom to share and change them. By contrast, the GNU General Public License is intended to guarantee that you and all other users have and keep those freedoms. We, the Free Software Foundation, use the GNU General Public License for most of our software; it applies also to any other work whose authors commit to using it. You can apply it to your works too.\ngpl3.preamble.p1.s1\ngerv\nlogin to agree\n1\n2976\n\n\n", "2426:2516:2732:2936:2945:2976:3062": "\n\n\n\n<p style="display: inline">I don't understand the word "most" on the preamble. Is it really true that most software has non-free licenses? When the preamble was originally written this made sense, for sure, but now it sounds unnecessarily pessimist.</p><p>\nMaybe something along the lines of "The licenses for proprietary software are designed to take away your freedom to share..." would be clearer.</p>\ngpl3.preamble.p1.s1\neml\nlogin to agree\n14\n2426\n\n\nChange this to "The licenses for many software applications..." I'm not even sure "most" is accurate given how much software is released under various free software licenses. \ngpl3.preamble.p1.s1\njring\nlogin to agree\n3\n2516\n\n\nMost people do not understand that free software exists. When they hear the term "software license", they assume that it is a document that takes away their rights. I believe the average user would think that "most" software licenses are proprietary.\ngpl3.preamble.p1.s1\nmkorman\nlogin to agree\n0\n2732\n\n\nWhile I have no statistics to back up my belief, I do think that most software in use today is proprietary. While lots of fields in computing are well provided, there are also gigantic amounts of embedded systems that are proprietary, such as all sorts of appliances as well as more niche equipment, like space monitoring robots.\ngpl3.preamble.p1.s1\nmsikma\nlogin to agree\n1\n2936\n\n\nStrive to make the wording in a way to keep it valid in 220 years, as if carved in stone (back in 1787 no one was considering frivolous usage of the word "most")! Either make it declarative, not comparative, or at least differentiate between copyright and copyleft licenses.\ngpl3.preamble.p1.s1\ngoldie\nlogin to agree\n0\n2945\n\n\nIt is great that changes have been made to make sure the GPL is applicable to non-software works. However, further work is required to generalise the language, particularly in the preamble. Editing such famous and evocative text feels like handling the crown jewels, but here is my attempt at para 2:\n\n\n Some licenses for software and other works are designed to take away your freedom to share and change them. By contrast, the GNU General Public License is intended to guarantee that you and all other users have and keep those freedoms. We, the Free Software Foundation, use the GNU General Public License for most of our software; it applies also to any other work whose authors commit to using it. You can apply it to your works too.\ngpl3.preamble.p1.s1\ngerv\nlogin to agree\n1\n2976\n\n\n<p style="display: inline">Most software running on my computers at home and at work is Free. SourceForge alone hosts about 200,000 projects with Free and Open Source licenses.</p><p>\nThis could be re-worded as "Most licenses for software and other practical works are designed to take away..." This is certainly true, since most proprietary software comes with its own license.</p>\ngpl3.preamble.p1.s1\nedscho\nlogin to agree\n0\n3062\n\n\n", "2945:2976": "\n\n\n\nStrive to make the wording in a way to keep it valid in 220 years, as if carved in stone (back in 1787 no one was considering frivolous usage of the word "most")! Either make it declarative, not comparative, or at least differentiate between copyright and copyleft licenses.\ngpl3.preamble.p1.s1\ngoldie\nlogin to agree\n0\n2945\n\n\nIt is great that changes have been made to make sure the GPL is applicable to non-software works. However, further work is required to generalise the language, particularly in the preamble. Editing such famous and evocative text feels like handling the crown jewels, but here is my attempt at para 2:\n\n\n Some licenses for software and other works are designed to take away your freedom to share and change them. By contrast, the GNU General Public License is intended to guarantee that you and all other users have and keep those freedoms. We, the Free Software Foundation, use the GNU General Public License for most of our software; it applies also to any other work whose authors commit to using it. You can apply it to your works too.\ngpl3.preamble.p1.s1\ngerv\nlogin to agree\n1\n2976\n\n\n", "2500:2945:2976": "\n\n\n\nWho decides what practical works are? I don't think we should to put this term in the GPL. \ngpl3.preamble.p1.s1\njastiv\nlogin to agree\n8\n2500\n\n\nStrive to make the wording in a way to keep it valid in 220 years, as if carved in stone (back in 1787 no one was considering frivolous usage of the word "most")! Either make it declarative, not comparative, or at least differentiate between copyright and copyleft licenses.\ngpl3.preamble.p1.s1\ngoldie\nlogin to agree\n0\n2945\n\n\nIt is great that changes have been made to make sure the GPL is applicable to non-software works. However, further work is required to generalise the language, particularly in the preamble. Editing such famous and evocative text feels like handling the crown jewels, but here is my attempt at para 2:\n\n\n Some licenses for software and other works are designed to take away your freedom to share and change them. By contrast, the GNU General Public License is intended to guarantee that you and all other users have and keep those freedoms. We, the Free Software Foundation, use the GNU General Public License for most of our software; it applies also to any other work whose authors commit to using it. You can apply it to your works too.\ngpl3.preamble.p1.s1\ngerv\nlogin to agree\n1\n2976\n\n\n", "2513:2643:2976": "\n\n\n\nIt is the default mode of copyright law that takes away the freedoms, not licenses. Licenses are designed to grant the minimum amount of freedoms required to make money.\ngpl3.preamble.p1.s1\nraulir\nlogin to agree\n2\n2513\n\n\nDo you 'take away' freedom or 'restrict' it?\ngpl3.preamble.p1.s1\nsamj\nlogin to agree\n0\n2643\n\n\nIt is great that changes have been made to make sure the GPL is applicable to non-software works. However, further work is required to generalise the language, particularly in the preamble. Editing such famous and evocative text feels like handling the crown jewels, but here is my attempt at para 2:\n\n\n Some licenses for software and other works are designed to take away your freedom to share and change them. By contrast, the GNU General Public License is intended to guarantee that you and all other users have and keep those freedoms. We, the Free Software Foundation, use the GNU General Public License for most of our software; it applies also to any other work whose authors commit to using it. You can apply it to your works too.\ngpl3.preamble.p1.s1\ngerv\nlogin to agree\n1\n2976\n\n\n", "2583": "\n\n\n\n\ngpl3.preamble.p1.s2\nsanjoy\nlogin to agree\n0\n2583\n\n\n", "2428:2822:2918:2937": "\n\n\n\n<p style="display: inline">The initial part of the sentence is fine. Why spoil it?</p><p>\nIf you must, try this:</p><p>\n"By contrast, the GNU General Public License is intended to guarantee your freedom to share and change free software--to make sure the recipients of such software are free."</p><p>\nIt's not the software that must be free, but those members of the public who are in receipt of it.</p><p>\n'free software' is a descriptive term to identify the class of software. The thing that is free is the human being - NOT the software.</p>\ngpl3.preamble.p1.s2\ncrosbie\nlogin to agree\n3\n2428\n\n\n<p style="display: inline">People who are interacting with a web service are "users" of the software powering the service, so the freedoms you desire should be extended to them.</p><p>\nHowever, in your definitions section you state that, "mere interaction with a user through a computer network, with no transfer of a copy, is not conveying," and therefore the license protections don't apply.</p><p>\nThis is exactly what I need to be protected in your license if I'm to use it for my software. If you continue to allow the "software as a service loophole," I will be forced to draft my own license, which is not a good thing.</p><p>\nIf you think that the Affero GPL solves this problem, you are wrong. I have written a huge (100,000+ lines) C++ class library that I want to release as free software. It wouldn't qualify under section 2d of the AGPL since it is not a program, and therefore can't offer up its source code via HTTP.</p><p>\nMike\n</p>\ngpl3.preamble.p1.s2\nmiked\nlogin to agree\n0\n2822\n\n\nUse the correct em dash symbol.\ngpl3.preamble.p1.s2\nnslater\nlogin to agree\n0\n2918\n\n\n<p style="display: inline">I understand that you'd rather not use the proper em dash symbol, as this is a plain text version of the license. However, I do think that people are bound to change that when the license is transcluded in a work. For example, using the license in LaTeX would cause this plain text em dash to be rendered as an en dash instead.</p><p>\nPerhaps three hyphens should be used instead?</p>\ngpl3.preamble.p1.s2\nmsikma\nlogin to agree\n0\n2937\n\n\n", "2428:2822": "\n\n\n\n<p style="display: inline">The initial part of the sentence is fine. Why spoil it?</p><p>\nIf you must, try this:</p><p>\n"By contrast, the GNU General Public License is intended to guarantee your freedom to share and change free software--to make sure the recipients of such software are free."</p><p>\nIt's not the software that must be free, but those members of the public who are in receipt of it.</p><p>\n'free software' is a descriptive term to identify the class of software. The thing that is free is the human being - NOT the software.</p>\ngpl3.preamble.p1.s2\ncrosbie\nlogin to agree\n3\n2428\n\n\n<p style="display: inline">People who are interacting with a web service are "users" of the software powering the service, so the freedoms you desire should be extended to them.</p><p>\nHowever, in your definitions section you state that, "mere interaction with a user through a computer network, with no transfer of a copy, is not conveying," and therefore the license protections don't apply.</p><p>\nThis is exactly what I need to be protected in your license if I'm to use it for my software. If you continue to allow the "software as a service loophole," I will be forced to draft my own license, which is not a good thing.</p><p>\nIf you think that the Affero GPL solves this problem, you are wrong. I have written a huge (100,000+ lines) C++ class library that I want to release as free software. It wouldn't qualify under section 2d of the AGPL since it is not a program, and therefore can't offer up its source code via HTTP.</p><p>\nMike\n</p>\ngpl3.preamble.p1.s2\nmiked\nlogin to agree\n0\n2822\n\n\n", "2581": "\n\n\n\nThe fraction of software for which FSF uses it is not relevant or that interesting. Delete the whole sentence.\ngpl3.preamble.p1.s3\nsanjoy\nlogin to agree\n1\n2581\n\n\n", "2582": "\n\n\n\nIf you delete the preceding sentence, then delete this bit too.\ngpl3.preamble.p1.s4\nsanjoy\nlogin to agree\n0\n2582\n\n\n", "2582:2583:2643:2732:2822:2918:2936:2937:2945:2976:3062": "\n\n\n\nIf you delete the preceding sentence, then delete this bit too.\ngpl3.preamble.p1.s4\nsanjoy\nlogin to agree\n0\n2582\n\n\n\ngpl3.preamble.p1.s2\nsanjoy\nlogin to agree\n0\n2583\n\n\nDo you 'take away' freedom or 'restrict' it?\ngpl3.preamble.p1.s1\nsamj\nlogin to agree\n0\n2643\n\n\nMost people do not understand that free software exists. When they hear the term "software license", they assume that it is a document that takes away their rights. I believe the average user would think that "most" software licenses are proprietary.\ngpl3.preamble.p1.s1\nmkorman\nlogin to agree\n0\n2732\n\n\n<p style="display: inline">People who are interacting with a web service are "users" of the software powering the service, so the freedoms you desire should be extended to them.</p><p>\nHowever, in your definitions section you state that, "mere interaction with a user through a computer network, with no transfer of a copy, is not conveying," and therefore the license protections don't apply.</p><p>\nThis is exactly what I need to be protected in your license if I'm to use it for my software. If you continue to allow the "software as a service loophole," I will be forced to draft my own license, which is not a good thing.</p><p>\nIf you think that the Affero GPL solves this problem, you are wrong. I have written a huge (100,000+ lines) C++ class library that I want to release as free software. It wouldn't qualify under section 2d of the AGPL since it is not a program, and therefore can't offer up its source code via HTTP.</p><p>\nMike\n</p>\ngpl3.preamble.p1.s2\nmiked\nlogin to agree\n0\n2822\n\n\nUse the correct em dash symbol.\ngpl3.preamble.p1.s2\nnslater\nlogin to agree\n0\n2918\n\n\nWhile I have no statistics to back up my belief, I do think that most software in use today is proprietary. While lots of fields in computing are well provided, there are also gigantic amounts of embedded systems that are proprietary, such as all sorts of appliances as well as more niche equipment, like space monitoring robots.\ngpl3.preamble.p1.s1\nmsikma\nlogin to agree\n1\n2936\n\n\n<p style="display: inline">I understand that you'd rather not use the proper em dash symbol, as this is a plain text version of the license. However, I do think that people are bound to change that when the license is transcluded in a work. For example, using the license in LaTeX would cause this plain text em dash to be rendered as an en dash instead.</p><p>\nPerhaps three hyphens should be used instead?</p>\ngpl3.preamble.p1.s2\nmsikma\nlogin to agree\n0\n2937\n\n\nStrive to make the wording in a way to keep it valid in 220 years, as if carved in stone (back in 1787 no one was considering frivolous usage of the word "most")! Either make it declarative, not comparative, or at least differentiate between copyright and copyleft licenses.\ngpl3.preamble.p1.s1\ngoldie\nlogin to agree\n0\n2945\n\n\nIt is great that changes have been made to make sure the GPL is applicable to non-software works. However, further work is required to generalise the language, particularly in the preamble. Editing such famous and evocative text feels like handling the crown jewels, but here is my attempt at para 2:\n\n\n Some licenses for software and other works are designed to take away your freedom to share and change them. By contrast, the GNU General Public License is intended to guarantee that you and all other users have and keep those freedoms. We, the Free Software Foundation, use the GNU General Public License for most of our software; it applies also to any other work whose authors commit to using it. You can apply it to your works too.\ngpl3.preamble.p1.s1\ngerv\nlogin to agree\n1\n2976\n\n\n<p style="display: inline">Most software running on my computers at home and at work is Free. SourceForge alone hosts about 200,000 projects with Free and Open Source licenses.</p><p>\nThis could be re-worded as "Most licenses for software and other practical works are designed to take away..." This is certainly true, since most proprietary software comes with its own license.</p>\ngpl3.preamble.p1.s1\nedscho\nlogin to agree\n0\n3062\n\n\n", "2646": "\n\n\n\n<p>Would it be better to quote the four freedoms (from GNU's definition of free software) here?</p>\n<p> * The freedom to run the program, for any purpose (freedom 0).</p>\n<p> * The freedom to study how the program works, and adapt it to your needs (freedom 1). Access to the source code is a precondition for this.</p>\n<p> * The freedom to redistribute copies so you can help your neighbor (freedom 2).</p>\n<p> * The freedom to improve the program, and release your improvements to the public, so that the whole community benefits (freedom 3). Access to the source code is a precondition for this.</p>\n\ngpl3.preamble.p2.s2\nilly\nlogin to agree\n1\n2646\n\n\n", "2585:2646": "\n\n\n\n\ngpl3.preamble.p2.s2\nsanjoy\nlogin to agree\n0\n2585\n\n\n<p>Would it be better to quote the four freedoms (from GNU's definition of free software) here?</p>\n<p> * The freedom to run the program, for any purpose (freedom 0).</p>\n<p> * The freedom to study how the program works, and adapt it to your needs (freedom 1). Access to the source code is a precondition for this.</p>\n<p> * The freedom to redistribute copies so you can help your neighbor (freedom 2).</p>\n<p> * The freedom to improve the program, and release your improvements to the public, so that the whole community benefits (freedom 3). Access to the source code is a precondition for this.</p>\n\ngpl3.preamble.p2.s2\nilly\nlogin to agree\n1\n2646\n\n\n", "2584:2646": "\n\n\n\nCan shorten to "may"\ngpl3.preamble.p2.s2\nsanjoy\nlogin to agree\n0\n2584\n\n\n<p>Would it be better to quote the four freedoms (from GNU's definition of free software) here?</p>\n<p> * The freedom to run the program, for any purpose (freedom 0).</p>\n<p> * The freedom to study how the program works, and adapt it to your needs (freedom 1). Access to the source code is a precondition for this.</p>\n<p> * The freedom to redistribute copies so you can help your neighbor (freedom 2).</p>\n<p> * The freedom to improve the program, and release your improvements to the public, so that the whole community benefits (freedom 3). Access to the source code is a precondition for this.</p>\n\ngpl3.preamble.p2.s2\nilly\nlogin to agree\n1\n2646\n\n\n", "2646:2867": "\n\n\n\n<p>Would it be better to quote the four freedoms (from GNU's definition of free software) here?</p>\n<p> * The freedom to run the program, for any purpose (freedom 0).</p>\n<p> * The freedom to study how the program works, and adapt it to your needs (freedom 1). Access to the source code is a precondition for this.</p>\n<p> * The freedom to redistribute copies so you can help your neighbor (freedom 2).</p>\n<p> * The freedom to improve the program, and release your improvements to the public, so that the whole community benefits (freedom 3). Access to the source code is a precondition for this.</p>\n\ngpl3.preamble.p2.s2\nilly\nlogin to agree\n1\n2646\n\n\nIt is generally understood that "charging for distribution" means simply covering the cost of any physical distribution media (such as CD or DVD) plus any post and package. However, some people may assume that if they are providing downloads from a website then they can charge for this distribution service even though no physical media is involved. This loophole should be closed by explicitly stating that charging is only allowed when physical media is involved, and NOT when accessing the software from a website or other server.\ngpl3.preamble.p2.s2\ntony32\nlogin to agree\n0\n2867\n\n\n", "2868": "\n\n\n\n<p style="display: inline">If you change any GPL'd software and release this modified software to others then it must be for free. This implies that you are not allowed to charge any sort of fee from those who access your software. To my mind allowing a third party to run a copy of the software which resides on your own computer (i.e. a central server) is just the same as providing them with a copy of the software which they can run on their computer.</p><p>\nIf you allow other people to access your software then, in order to comply with this cause, it must be for free. Saying that it need not be free if the software is "accessed" but not "distributed" is a contradiction that needs to be cleared up.</p>\ngpl3.preamble.p2.s2\ntony32\nlogin to agree\n0\n2868\n\n\n", "2584:2585:2646:2867:2868": "\n\n\n\nCan shorten to "may"\ngpl3.preamble.p2.s2\nsanjoy\nlogin to agree\n0\n2584\n\n\n\ngpl3.preamble.p2.s2\nsanjoy\nlogin to agree\n0\n2585\n\n\n<p>Would it be better to quote the four freedoms (from GNU's definition of free software) here?</p>\n<p> * The freedom to run the program, for any purpose (freedom 0).</p>\n<p> * The freedom to study how the program works, and adapt it to your needs (freedom 1). Access to the source code is a precondition for this.</p>\n<p> * The freedom to redistribute copies so you can help your neighbor (freedom 2).</p>\n<p> * The freedom to improve the program, and release your improvements to the public, so that the whole community benefits (freedom 3). Access to the source code is a precondition for this.</p>\n\ngpl3.preamble.p2.s2\nilly\nlogin to agree\n1\n2646\n\n\nIt is generally understood that "charging for distribution" means simply covering the cost of any physical distribution media (such as CD or DVD) plus any post and package. However, some people may assume that if they are providing downloads from a website then they can charge for this distribution service even though no physical media is involved. This loophole should be closed by explicitly stating that charging is only allowed when physical media is involved, and NOT when accessing the software from a website or other server.\ngpl3.preamble.p2.s2\ntony32\nlogin to agree\n0\n2867\n\n\n<p style="display: inline">If you change any GPL'd software and release this modified software to others then it must be for free. This implies that you are not allowed to charge any sort of fee from those who access your software. To my mind allowing a third party to run a copy of the software which resides on your own computer (i.e. a central server) is just the same as providing them with a copy of the software which they can run on their computer.</p><p>\nIf you allow other people to access your software then, in order to comply with this cause, it must be for free. Saying that it need not be free if the software is "accessed" but not "distributed" is a contradiction that needs to be cleared up.</p>\ngpl3.preamble.p2.s2\ntony32\nlogin to agree\n0\n2868\n\n\n", "2467": "\n\n\n\nThis paragraph is overly complicated. Forbid anyone to deny you these rights has multiple negatives, and would be cleaner as Forbid anyone from denying you these rights. \n\nThe whole paragraph could be re-written as:\nTo protect these rights you in turn have certain responsibilities if you distribute or modify copies of the software. \n\nSo much cleaner. \ngpl3.preamble.p3.s1\naugustz\nlogin to agree\n2\n2467\n\n\n", "2431:2467:2572": "\n\n\n\n<p style="display: inline">The GPL protects the freedom that it contrives for the licensee.</p><p>\nThe law (rather than the GPL) is supposed to protect our rights.</p><p>\nUnfortunately, the law in the form of copyright suspended people's right to share and build upon public works.</p><p>\nSo, if anything the GPL restores the licensee's right and protects this restoration.</p><p>\nRights are what people start off with (supposedly inalienable).</p><p>\nCopyright and patents are commercial privileges that suspend rights (forgivably from a few corporations, but unethically from the public).</p><p>\nLicenses grant permission, i.e. selective/conditional restoration of rights suspended by privilege.</p><p>\nLet's get this terminology coherent and consistent.</p><p>\nThe GPL has already demonstrated against abusive redefinition of license from contract (EULA), and free 'as in freedom not as in free beer', let's see it assert the proper definition for 'right' too.</p>\ngpl3.preamble.p3.s1\ncrosbie\nlogin to agree\n3\n2431\n\n\nThis paragraph is overly complicated. Forbid anyone to deny you these rights has multiple negatives, and would be cleaner as Forbid anyone from denying you these rights. \n\nThe whole paragraph could be re-written as:\nTo protect these rights you in turn have certain responsibilities if you distribute or modify copies of the software. \n\nSo much cleaner. \ngpl3.preamble.p3.s1\naugustz\nlogin to agree\n2\n2467\n\n\nI would replace "rights" with "freedoms" and maybe even repeat the 4 basic freedoms with respect to software as explained by the FSF (even if this means the risk of making the introduction+license still longer).\ngpl3.preamble.p3.s1\nadhemar\nlogin to agree\n0\n2572\n\n\n", "2467:2535": "\n\n\n\nThis paragraph is overly complicated. Forbid anyone to deny you these rights has multiple negatives, and would be cleaner as Forbid anyone from denying you these rights. \n\nThe whole paragraph could be re-written as:\nTo protect these rights you in turn have certain responsibilities if you distribute or modify copies of the software. \n\nSo much cleaner. \ngpl3.preamble.p3.s1\naugustz\nlogin to agree\n2\n2467\n\n\nshould still be "these rights".\ngpl3.preamble.p3.s1\nhkbst\nlogin to agree\n1\n2535\n\n\n", "2586:2587": "\n\n\n\n"Therefore" sounds a bit funny because it implies too direct a connection. The previous sentence describes the need to make requirements to help "you". But this sentence then switches to talking about your obligations. Using "Likewise" instead of "Therefore" gives a smoother and more accurate connection.\ngpl3.preamble.p3.s2\nsanjoy\nlogin to agree\n0\n2586\n\n\nThe following "or if" makes the parallel with "if you distribute", so the comma is unneeded.\ngpl3.preamble.p3.s2\nsanjoy\nlogin to agree\n0\n2587\n\n\n", "2592": "\n\n\n\n\ngpl3.preamble.p4.s1\nsanjoy\nlogin to agree\n0\n2592\n\n\n", "2588": "\n\n\n\n\ngpl3.preamble.p4.s2\nsanjoy\nlogin to agree\n0\n2588\n\n\n", "2589": "\n\n\n\nthe "so that" indicating "with the intention that" (vs a mere "so" meaning "with the consequence that")\ngpl3.preamble.p4.s3\nsanjoy\nlogin to agree\n2\n2589\n\n\n", "2536": "\n\n\n\n"(1) asserting" and "(2) offering" would make this sentence better. \ngpl3.preamble.p5.s1\nhkbst\nlogin to agree\n2\n2536\n\n\n", "2536:3027": "\n\n\n\n"(1) asserting" and "(2) offering" would make this sentence better. \ngpl3.preamble.p5.s1\nhkbst\nlogin to agree\n2\n2536\n\n\nOne should define an acronym before using it; one of the previous mentions of "General Public License" should have something to the effect of "General Public License (GPL)".\ngpl3.preamble.p5.s1\nsalnet\nlogin to agree\n0\n3027\n\n\n", "2504:2536": "\n\n\n\nShould just say "and".\ngpl3.preamble.p5.s1\nraulir\nlogin to agree\n7\n2504\n\n\n"(1) asserting" and "(2) offering" would make this sentence better. \ngpl3.preamble.p5.s1\nhkbst\nlogin to agree\n2\n2536\n\n\n", "2504:2536:3027": "\n\n\n\nShould just say "and".\ngpl3.preamble.p5.s1\nraulir\nlogin to agree\n7\n2504\n\n\n"(1) asserting" and "(2) offering" would make this sentence better. \ngpl3.preamble.p5.s1\nhkbst\nlogin to agree\n2\n2536\n\n\nOne should define an acronym before using it; one of the previous mentions of "General Public License" should have something to the effect of "General Public License (GPL)".\ngpl3.preamble.p5.s1\nsalnet\nlogin to agree\n0\n3027\n\n\n", "2666:2919": "\n\n\n\nIsn't just "authors'" sufficient?\ngpl3.preamble.p6.s1\nsanjoy\nlogin to agree\n3\n2666\n\n\nUse the correct closing quotation symbol.\ngpl3.preamble.p6.s1\nnslater\nlogin to agree\n0\n2919\n\n\n", "2666": "\n\n\n\nIsn't just "authors'" sufficient?\ngpl3.preamble.p6.s1\nsanjoy\nlogin to agree\n3\n2666\n\n\n", "2590": "\n\n\n\nHow clear it is is a matter of opinion that the reader should decide without editorial comment from the license author. So delete this adverb.\ngpl3.preamble.p6.s1\nsanjoy\nlogin to agree\n2\n2590\n\n\n", "2468": "\n\n\n\nCould simply read: The GPL requires that modified versions be marked as changed, so that problems will not be associated erroneously with a previous version. \n\nPerhaps I don't use or modify the software, but do research on its source code. Why shouldn't I also care about the version history? In other words, users and authors is both unnecessary and too specific. Even other software programs may find it useful to grep though and identify changes. \ngpl3.preamble.p6.s2\naugustz\nlogin to agree\n1\n2468\n\n\n", "2660": "\n\n\n\nHow must these changes be marked? \ngpl3.preamble.p6.s2\nrobmyers\nlogin to agree\n0\n2660\n\n\n", "2590:2660:2666:2919": "\n\n\n\nHow clear it is is a matter of opinion that the reader should decide without editorial comment from the license author. So delete this adverb.\ngpl3.preamble.p6.s1\nsanjoy\nlogin to agree\n2\n2590\n\n\nHow must these changes be marked? \ngpl3.preamble.p6.s2\nrobmyers\nlogin to agree\n0\n2660\n\n\nIsn't just "authors'" sufficient?\ngpl3.preamble.p6.s1\nsanjoy\nlogin to agree\n3\n2666\n\n\nUse the correct closing quotation symbol.\ngpl3.preamble.p6.s1\nnslater\nlogin to agree\n0\n2919\n\n\n", "3071": "\n\n\n\nA dash or a semi-colon is required here to make it flow correctly.\ngpl3.preamble.p7.s1\njamesj\nlogin to agree\n0\n3071\n\n\n", "2591": "\n\n\n\n"This" needs a noun following it, which also helps the reader track the meaning. e.g. "This denial is..." or "This restriction is...".\ngpl3.preamble.p7.s2\nsanjoy\nlogin to agree\n0\n2591\n\n\n", "2461": "\n\n\n\n<p style="display: inline">Who is the license talking to?\nIt should be the licensee.\nThe GPL may strive to preserve liberty for all recipients of free software, but 'you' and 'your' are terms that identify the licensee. EVEN IN THE PRE-AMBLE.\nIf the license is to indicate 'members of the public who are in receipt of free software' then it should use an appropriate identifier because 'you' means 'the licensee'.</p><p>\nWhy suddenly introduce terms such as 'author' and 'developer'?</p><p>\nEven the term 'user' is highly suspect.\nIs it the user of the source code, the user of the license, or the user of the device currently executing the software?</p><p>\nThe license may, where it can, address use of the software, but it does not address the user.</p><p>\nThe GPL can only address the licensee, and the licensee is someone otherwise constrained by copyright, i.e. a person able to duplicate or create derivatives of the software - someone who we may deduce is in receipt of a copy of the software.</p><p>\nThe GPL cannot grant privileges to users, it can only grant privileges (even if conditionally constrained) to licensees. Licensees are those who legitimately receive a copy of the software.</p><p>\nUsers of someone else's device do not receive a copy of the software.</p><p>\nTherefore the GPL restores liberty (suspended by copyright and patent) to all immediate and descendant licensees of the GPL. Not necessarily all users of such software.</p><p>\nThere are many licensees who are users, but users of GPL software are not necessarily licensees.</p><p>\nUsers who would like to be licensees need only contrive to become a legitimate recipient of the software, i.e. buy or be given a copy (whether on a medium or stored within a consumer device).</p><p>\nIt is necessary to be rigorous concerning the distinction between user and licensee in order to protect the human right to privacy (which encompasses the right to private property).</p><p>\nThe public's liberties concerning free software are delimited by the boundary of everyone else's private domain.</p>\ngpl3.preamble.p7.s2\ncrosbie\nlogin to agree\n0\n2461\n\n\n", "2425:2517:2594": "\n\n\n\nMight suggest 'This abuse occurs systematically in products designed for individual use...'\ngpl3.preamble.p7.s3\ntieguy\nlogin to agree\n8\n2425\n\n\nRemove "The" because it doesn't only occur in this area. \ngpl3.preamble.p7.s3\njring\nlogin to agree\n0\n2517\n\n\n=> "Such abuse occurs systematically..."\ngpl3.preamble.p7.s3\nsanjoy\nlogin to agree\n1\n2594\n\n\n", "2425:2594": "\n\n\n\nMight suggest 'This abuse occurs systematically in products designed for individual use...'\ngpl3.preamble.p7.s3\ntieguy\nlogin to agree\n8\n2425\n\n\n=> "Such abuse occurs systematically..."\ngpl3.preamble.p7.s3\nsanjoy\nlogin to agree\n1\n2594\n\n\n", "2425": "\n\n\n\nMight suggest 'This abuse occurs systematically in products designed for individual use...'\ngpl3.preamble.p7.s3\ntieguy\nlogin to agree\n8\n2425\n\n\n", "2425:2593": "\n\n\n\nMight suggest 'This abuse occurs systematically in products designed for individual use...'\ngpl3.preamble.p7.s3\ntieguy\nlogin to agree\n8\n2425\n\n\n\ngpl3.preamble.p7.s3\nsanjoy\nlogin to agree\n0\n2593\n\n\n", "2425:2469": "\n\n\n\nMight suggest 'This abuse occurs systematically in products designed for individual use...'\ngpl3.preamble.p7.s3\ntieguy\nlogin to agree\n8\n2425\n\n\nchange for individuals to use to\nfor individual use\ngpl3.preamble.p7.s3\naugustz\nlogin to agree\n0\n2469\n\n\n", "2595": "\n\n\n\nNothing in life, esp. in legal matters, is ever pricise.\ngpl3.preamble.p7.s3\nsanjoy\nlogin to agree\n0\n2595\n\n\n", "2710": "\n\n\n\nSaying it's most unacceptable in one situation implies it's unacceptable everywhere. If that's the case, it should be banned everywhere. Having the ability to change software doesn't require it, so I don't agree with the rationale for limiting it to consumer products.\ngpl3.preamble.p7.s3\nflaschen\nlogin to agree\n0\n2710\n\n\n", "2832": "\n\n\n\nMultimedia can use a liberating license.\ngpl3.preamble.p7.s5\nBrylie\nlogin to agree\n0\n2832\n\n\n", "3080": "\n\n\n\n\nFor some, the statement about the threat of software patents might be seen as being an opinion of the Free Software Foundation as opposed to a factual statement. To be sure, there might be documented cases where patents have interfered with software development. It might be useful to consider the wording of the statement and to reword the statement, if necessary, so that it is clear as to whether the statement is an opinion or not.\n\n\ngpl3.preamble.p8.s1\nmezzanine1\nlogin to agree\n0\n3080\n\n\n", "2596:3121": "\n\n\n\n"States" often means a US state, so use "Jurisdictions" for clarity.\ngpl3.preamble.p8.s2\nsanjoy\nlogin to agree\n3\n2596\n\n\n\nConsider the statement about states and software patents. Presumably, this statement is an opinion of the FSF. In this case, it might be useful to change the statement so that it is obvious to the reader that the statement is an opinion.\n\n\ngpl3.preamble.p8.s2\nmezzanine1\nlogin to agree\n0\n3121\n\n\n", "2712:3121": "\n\n\n\nIf the FSF truly opposes software patents, why don't they use the license as a tool to weaken them as much as possible?\ngpl3.preamble.p8.s2\nflaschen\nlogin to agree\n0\n2712\n\n\n\nConsider the statement about states and software patents. Presumably, this statement is an opinion of the FSF. In this case, it might be useful to change the statement so that it is obvious to the reader that the statement is an opinion.\n\n\ngpl3.preamble.p8.s2\nmezzanine1\nlogin to agree\n0\n3121\n\n\n", "2712": "\n\n\n\nIf the FSF truly opposes software patents, why don't they use the license as a tool to weaken them as much as possible?\ngpl3.preamble.p8.s2\nflaschen\nlogin to agree\n0\n2712\n\n\n", "2598": "\n\n\n\nI meant "in places where" => "where"\ngpl3.preamble.p8.s2\nsanjoy\nlogin to agree\n4\n2598\n\n\n", "2597:2598": "\n\n\n\njust "where" is enough\ngpl3.preamble.p8.s2\nsanjoy\nlogin to agree\n0\n2597\n\n\nI meant "in places where" => "where"\ngpl3.preamble.p8.s2\nsanjoy\nlogin to agree\n4\n2598\n\n\n", "2599": "\n\n\n\n"this" needs a noun after it, e.g. "this outcome"\ngpl3.preamble.p8.s3\nsanjoy\nlogin to agree\n3\n2599\n\n\n", "3080:3121": "\n\n\n\n\nFor some, the statement about the threat of software patents might be seen as being an opinion of the Free Software Foundation as opposed to a factual statement. To be sure, there might be documented cases where patents have interfered with software development. It might be useful to consider the wording of the statement and to reword the statement, if necessary, so that it is clear as to whether the statement is an opinion or not.\n\n\ngpl3.preamble.p8.s1\nmezzanine1\nlogin to agree\n0\n3080\n\n\n\nConsider the statement about states and software patents. Presumably, this statement is an opinion of the FSF. In this case, it might be useful to change the statement so that it is obvious to the reader that the statement is an opinion.\n\n\ngpl3.preamble.p8.s2\nmezzanine1\nlogin to agree\n0\n3121\n\n\n", "2667": "\n\n\n\nNeeds a comma after "distribution"\ngpl3.preamble.p9.s1\nsanjoy\nlogin to agree\n0\n2667\n\n\n", "2427:3141": "\n\n\n\nA party's "essential patent claims" in a work are all patent claims owned or controlled by the party, including but not limited to the taking of any promise, pledge, undertaking or other obligation to protect from or refrain from enforcement of the potential or actual patent claim of another, whether already acquired or to be acquired, that would be infringed by making, using, or selling the work.\n\ngpl3.terms.0.0\naaronmf\nlogin to agree\n0\n2427\n\n\nA version of this clause is often added to companies Memoranda of Association under English law, it seems to fit in perfectly with the GPL.\r\n\r\n--------\r\n\r\n16. IT IS HEREBY DECLARED that in the interpretation of this licence the powers conferred on the user by any section paragraph or clause shall not be restricted by reference to any other section paragraph or clause or by the juxtaposition of any two or more sections paragraphs or clauses and in the event of any ambiguity this clause and every other shall be construed in such a way as to widen and not restrict and thus more perfectly guarantee the freedom of the end user to share and change the free software covered by said licence.\ngpl3.terms.0.0\nhandyman\nlogin to agree\n0\n3141\n\n\n", "3049": "\n\n\n\nStarting the document section numbering with "0" solely benefits a small group of condescending programmers. There are good reasons why this sophomoric joke is inappropriate in the GPLv3. Chiefly it invites the contempt of non-programmers including judges and juries in copyright infringement cases who will have little interest in or tolerance for geek humor. Do we really want to agitate people who may be rendering verdicts on the meaning of the GPLv3 and looking for any little reason to hammer the hacker subculture? \ngpl3.definitions.0.0\njkoenig\nlogin to agree\n1\n3049\n\n\n", "2688": "\n\n\n\nIn this draft it appears there was a conscious move towards doing definitions at the beginning of the individual sections. Perhaps these definitions should similarly be moved to where they are first used.\n\ngpl3.definitions.0.0\nsepreece\nlogin to agree\n3\n2688\n\n\n", "2688:3049": "\n\n\n\nIn this draft it appears there was a conscious move towards doing definitions at the beginning of the individual sections. Perhaps these definitions should similarly be moved to where they are first used.\n\ngpl3.definitions.0.0\nsepreece\nlogin to agree\n3\n2688\n\n\nStarting the document section numbering with "0" solely benefits a small group of condescending programmers. There are good reasons why this sophomoric joke is inappropriate in the GPLv3. Chiefly it invites the contempt of non-programmers including judges and juries in copyright infringement cases who will have little interest in or tolerance for geek humor. Do we really want to agitate people who may be rendering verdicts on the meaning of the GPLv3 and looking for any little reason to hammer the hacker subculture? \ngpl3.definitions.0.0\njkoenig\nlogin to agree\n1\n3049\n\n\n", "2920:3050": "\n\n\n\nUse the correct open/close quotation symbols.\ngpl3.definitions.p0.s1\nnslater\nlogin to agree\n0\n2920\n\n\n<p style="display: inline">There is no reason to include the adjective "This" in the definition of "License" or in any other definition in the GPLv3. "License" alone is fine because sometimes it is necessary to refer to the License as "This License" or "The License" at the beginning of a sentence, or as "this License" or "the License" in other parts of a sentence. Obviously there is no utility in restricting the definition "License" only to force incorrect usage later in the GPLv3. The adjective "This" is completely unnecessary and incorrect, since no less than two paragraphs later in the definition of a "contributor" an incorrect reference to "this License" occurs. So, instead of perpetuating a lot more drafting errors, just use the following definition:</p><p>\n"License" refers to version 3 of the GNU General Public License. </p>\ngpl3.definitions.p0.s1\njkoenig\nlogin to agree\n0\n3050\n\n\n", "2920": "\n\n\n\nUse the correct open/close quotation symbols.\ngpl3.definitions.p0.s1\nnslater\nlogin to agree\n0\n2920\n\n\n", "3051": "\n\n\n\n<p style="display: inline">There is fundamentally no need to define the term "Copyright". Even if there actually was, it is very bad drafting to say what something "also" is before you define what it is in the first place. Anyway, the definition of "Copyright" should be dropped entirely. The real goal is to define the terms "work" and "works". The present "Copyright" definition fails even to accomplish that small task. The next line says "The Program refers to any copyrightable work". That clearly doesn't faithfully apply the defined term "Copyright" to either the term Program or the term "work". So this whole definition of "Copyright" and its usage is flawed and needs be immediately re-directed to the null file. It should be replaced with the following:</p><p>\n"Work" and "works" means any copyrightable material and any comparable subject matter with sui generis protection such as a semiconductor mask or database."</p>\ngpl3.definitions.p1.s1\njkoenig\nlogin to agree\n0\n3051\n\n\n", "2600:2659:3051": "\n\n\n\nAs I started reading this phrase, I hoped it would include the UK's so-called "copyright in typographical designs" and maybe compilation copyright or (in the EU) database rights. But the "that apply to OTHER kinds of works" dashed my hopes, since those rights also apply to the usual works that copyright applies to. \n\nPerhaps instead "...copyright-like laws, such as those covering semiconductor masks". Or is that too broad (perhaps catching patents by mistake)?\ngpl3.definitions.p1.s1\nsanjoy\nlogin to agree\n2\n2600\n\n\nIf I get a GPL-ed work that has US boat hull or mask rights what happens when I export that work to the EU? What happens when I export EU Database right GPL work to the US?\ngpl3.definitions.p1.s1\nrobmyers\nlogin to agree\n0\n2659\n\n\n<p style="display: inline">There is fundamentally no need to define the term "Copyright". Even if there actually was, it is very bad drafting to say what something "also" is before you define what it is in the first place. Anyway, the definition of "Copyright" should be dropped entirely. The real goal is to define the terms "work" and "works". The present "Copyright" definition fails even to accomplish that small task. The next line says "The Program refers to any copyrightable work". That clearly doesn't faithfully apply the defined term "Copyright" to either the term Program or the term "work". So this whole definition of "Copyright" and its usage is flawed and needs be immediately re-directed to the null file. It should be replaced with the following:</p><p>\n"Work" and "works" means any copyrightable material and any comparable subject matter with sui generis protection such as a semiconductor mask or database."</p>\ngpl3.definitions.p1.s1\njkoenig\nlogin to agree\n0\n3051\n\n\n", "2600:2659:2970:3051": "\n\n\n\nAs I started reading this phrase, I hoped it would include the UK's so-called "copyright in typographical designs" and maybe compilation copyright or (in the EU) database rights. But the "that apply to OTHER kinds of works" dashed my hopes, since those rights also apply to the usual works that copyright applies to. \n\nPerhaps instead "...copyright-like laws, such as those covering semiconductor masks". Or is that too broad (perhaps catching patents by mistake)?\ngpl3.definitions.p1.s1\nsanjoy\nlogin to agree\n2\n2600\n\n\nIf I get a GPL-ed work that has US boat hull or mask rights what happens when I export that work to the EU? What happens when I export EU Database right GPL work to the US?\ngpl3.definitions.p1.s1\nrobmyers\nlogin to agree\n0\n2659\n\n\n\nIn a software category such as entertainment software, it would be useful to have artwork, digitized sounds, and music that are licensed under the GPL or in another manner that is compatible with the GPL. There might be other categories of software that incorporate even more different kinds of works. It might be possible for the GPL to be compatible with some of the Creative Commons licenses (http://creativecommons.org) but that might not happen. It would seem that some software does not consist entirely of code.\n\n\ngpl3.definitions.p1.s1\nmezzanine1\nlogin to agree\n0\n2970\n\n\n<p style="display: inline">There is fundamentally no need to define the term "Copyright". Even if there actually was, it is very bad drafting to say what something "also" is before you define what it is in the first place. Anyway, the definition of "Copyright" should be dropped entirely. The real goal is to define the terms "work" and "works". The present "Copyright" definition fails even to accomplish that small task. The next line says "The Program refers to any copyrightable work". That clearly doesn't faithfully apply the defined term "Copyright" to either the term Program or the term "work". So this whole definition of "Copyright" and its usage is flawed and needs be immediately re-directed to the null file. It should be replaced with the following:</p><p>\n"Work" and "works" means any copyrightable material and any comparable subject matter with sui generis protection such as a semiconductor mask or database."</p>\ngpl3.definitions.p1.s1\njkoenig\nlogin to agree\n0\n3051\n\n\n", "2659:3051": "\n\n\n\nIf I get a GPL-ed work that has US boat hull or mask rights what happens when I export that work to the EU? What happens when I export EU Database right GPL work to the US?\ngpl3.definitions.p1.s1\nrobmyers\nlogin to agree\n0\n2659\n\n\n<p style="display: inline">There is fundamentally no need to define the term "Copyright". Even if there actually was, it is very bad drafting to say what something "also" is before you define what it is in the first place. Anyway, the definition of "Copyright" should be dropped entirely. The real goal is to define the terms "work" and "works". The present "Copyright" definition fails even to accomplish that small task. The next line says "The Program refers to any copyrightable work". That clearly doesn't faithfully apply the defined term "Copyright" to either the term Program or the term "work". So this whole definition of "Copyright" and its usage is flawed and needs be immediately re-directed to the null file. It should be replaced with the following:</p><p>\n"Work" and "works" means any copyrightable material and any comparable subject matter with sui generis protection such as a semiconductor mask or database."</p>\ngpl3.definitions.p1.s1\njkoenig\nlogin to agree\n0\n3051\n\n\n", "2659:2703:3051": "\n\n\n\nIf I get a GPL-ed work that has US boat hull or mask rights what happens when I export that work to the EU? What happens when I export EU Database right GPL work to the US?\ngpl3.definitions.p1.s1\nrobmyers\nlogin to agree\n0\n2659\n\n\nI think the example 'semiconductor masks' is fairly confusing to most people (myself included), who don't know what a semiconductor mask is.\ngpl3.definitions.p1.s1\nn0dalus\nlogin to agree\n0\n2703\n\n\n<p style="display: inline">There is fundamentally no need to define the term "Copyright". Even if there actually was, it is very bad drafting to say what something "also" is before you define what it is in the first place. Anyway, the definition of "Copyright" should be dropped entirely. The real goal is to define the terms "work" and "works". The present "Copyright" definition fails even to accomplish that small task. The next line says "The Program refers to any copyrightable work". That clearly doesn't faithfully apply the defined term "Copyright" to either the term Program or the term "work". So this whole definition of "Copyright" and its usage is flawed and needs be immediately re-directed to the null file. It should be replaced with the following:</p><p>\n"Work" and "works" means any copyrightable material and any comparable subject matter with sui generis protection such as a semiconductor mask or database."</p>\ngpl3.definitions.p1.s1\njkoenig\nlogin to agree\n0\n3051\n\n\n", "2752:3052": "\n\n\n\nSince the GPL is designed to be applicable to any work of authorship (not only computer programs), I would suggest using a more neutral term than "the Program". Something like "the Work" would avoid misleading many many people into thinking that the GPL can only be applied to computer programs.\ngpl3.definitions.p1.s2\nfrx\nlogin to agree\n2\n2752\n\n\n<p style="display: inline">Yet Another Unnecessary Use Of An Adjective. There is no point in using "The" in the defined term "The Program". You can just say "Program". But to get it exactly right, the definition should be like this: \n\n"Program" means any "work" (as defined above) governed by this License. \n\nOf course this new definition of "Program" must be used in conjunction with the previously submitted a new definition of "Copyright" as follows:</p><p>\n"Work" and "works" means any copyrightable material and any comparable subject matter with sui generis protection such as a semiconductor mask or database.</p>\ngpl3.definitions.p1.s2\njkoenig\nlogin to agree\n0\n3052\n\n\n", "3052": "\n\n\n\n<p style="display: inline">Yet Another Unnecessary Use Of An Adjective. There is no point in using "The" in the defined term "The Program". You can just say "Program". But to get it exactly right, the definition should be like this: \n\n"Program" means any "work" (as defined above) governed by this License. \n\nOf course this new definition of "Program" must be used in conjunction with the previously submitted a new definition of "Copyright" as follows:</p><p>\n"Work" and "works" means any copyrightable material and any comparable subject matter with sui generis protection such as a semiconductor mask or database.</p>\ngpl3.definitions.p1.s2\njkoenig\nlogin to agree\n0\n3052\n\n\n", "2464:3052": "\n\n\n\nI would have thought that 'Any program' refers to 'any copyrightable work...', but that 'The program' refers to 'the copyrightable work...'\n\ngpl3.definitions.p1.s2\ncrosbie\nlogin to agree\n2\n2464\n\n\n<p style="display: inline">Yet Another Unnecessary Use Of An Adjective. There is no point in using "The" in the defined term "The Program". You can just say "Program". But to get it exactly right, the definition should be like this: \n\n"Program" means any "work" (as defined above) governed by this License. \n\nOf course this new definition of "Program" must be used in conjunction with the previously submitted a new definition of "Copyright" as follows:</p><p>\n"Work" and "works" means any copyrightable material and any comparable subject matter with sui generis protection such as a semiconductor mask or database.</p>\ngpl3.definitions.p1.s2\njkoenig\nlogin to agree\n0\n3052\n\n\n", "2454:2573": "\n\n\n\n<p style="display: inline">"Organizations" seems to apply to two or more individuals here but used in Section 10 to refer to parts of the work.</p><p>\nAlso, when describing two or more individuals where the concepts of "propagate" or "convey," DO NOT apply, then it seems that the boundary conditions of the organization need to be specified.</p><p>\nDoes the term organization refer to strictly legal organizations such as corporations, charities, etc. or does it refer to informal organizations such as clubs, schools, etc.</p><p>\nWhat makes an individual a member of the organization? How long must a member be in the organization before the GPL terms "propagate" and "convey" do not apply? To circumvent the GPL, can organizations (i.e., companies) make their customers members?</p><p>\nWhat about organizations such as the federal or state governments? Does the GPL terms "propagate" and "convey" apply to the Army sharing GPL'd based code classified as SECRET with the Navy? Since government employees are public employees and cannot copyright a work, does the GPL apply to any software created by government employees? Under what conditions is a private contractor a member of the government organization? Does the GPL apply to GPL'd based code classified as SECRET and written for the government?</p>\ngpl3.definitions.p1.s4\nmaginnis\nlogin to agree\n1\n2454\n\n\nIf organizations can be considered a *single* licensee or recipient party with respect to the GPL, thus this mean that intra-organization copying (even of modified code) is not subject to this license?\ngpl3.definitions.p1.s4\nadhemar\nlogin to agree\n1\n2573\n\n\n", "2600:2659:2703:2752:2970:3051:3052": "\n\n\n\nAs I started reading this phrase, I hoped it would include the UK's so-called "copyright in typographical designs" and maybe compilation copyright or (in the EU) database rights. But the "that apply to OTHER kinds of works" dashed my hopes, since those rights also apply to the usual works that copyright applies to. \n\nPerhaps instead "...copyright-like laws, such as those covering semiconductor masks". Or is that too broad (perhaps catching patents by mistake)?\ngpl3.definitions.p1.s1\nsanjoy\nlogin to agree\n2\n2600\n\n\nIf I get a GPL-ed work that has US boat hull or mask rights what happens when I export that work to the EU? What happens when I export EU Database right GPL work to the US?\ngpl3.definitions.p1.s1\nrobmyers\nlogin to agree\n0\n2659\n\n\nI think the example 'semiconductor masks' is fairly confusing to most people (myself included), who don't know what a semiconductor mask is.\ngpl3.definitions.p1.s1\nn0dalus\nlogin to agree\n0\n2703\n\n\nSince the GPL is designed to be applicable to any work of authorship (not only computer programs), I would suggest using a more neutral term than "the Program". Something like "the Work" would avoid misleading many many people into thinking that the GPL can only be applied to computer programs.\ngpl3.definitions.p1.s2\nfrx\nlogin to agree\n2\n2752\n\n\n\nIn a software category such as entertainment software, it would be useful to have artwork, digitized sounds, and music that are licensed under the GPL or in another manner that is compatible with the GPL. There might be other categories of software that incorporate even more different kinds of works. It might be possible for the GPL to be compatible with some of the Creative Commons licenses (http://creativecommons.org) but that might not happen. It would seem that some software does not consist entirely of code.\n\n\ngpl3.definitions.p1.s1\nmezzanine1\nlogin to agree\n0\n2970\n\n\n<p style="display: inline">There is fundamentally no need to define the term "Copyright". Even if there actually was, it is very bad drafting to say what something "also" is before you define what it is in the first place. Anyway, the definition of "Copyright" should be dropped entirely. The real goal is to define the terms "work" and "works". The present "Copyright" definition fails even to accomplish that small task. The next line says "The Program refers to any copyrightable work". That clearly doesn't faithfully apply the defined term "Copyright" to either the term Program or the term "work". So this whole definition of "Copyright" and its usage is flawed and needs be immediately re-directed to the null file. It should be replaced with the following:</p><p>\n"Work" and "works" means any copyrightable material and any comparable subject matter with sui generis protection such as a semiconductor mask or database."</p>\ngpl3.definitions.p1.s1\njkoenig\nlogin to agree\n0\n3051\n\n\n<p style="display: inline">Yet Another Unnecessary Use Of An Adjective. There is no point in using "The" in the defined term "The Program". You can just say "Program". But to get it exactly right, the definition should be like this: \n\n"Program" means any "work" (as defined above) governed by this License. \n\nOf course this new definition of "Program" must be used in conjunction with the previously submitted a new definition of "Copyright" as follows:</p><p>\n"Work" and "works" means any copyrightable material and any comparable subject matter with sui generis protection such as a semiconductor mask or database.</p>\ngpl3.definitions.p1.s2\njkoenig\nlogin to agree\n0\n3052\n\n\n", "2753:3053:3091": "\n\n\n\nThe definitions of "modify", "modified version", work "based on" another work, and "covered work" are slightly changed with respect to GPLv3draft2, but they are still clear. It's good that the definition of "modified version" exploits applicable copyright law without trying to rewrite it. This ensures that the license does not place restrictions on activities that do not require permission under applicable copyright law.\n\ngpl3.definitions.p2.s1\nfrx\nlogin to agree\n0\n2753\n\n\n<p style="display: inline">Maybe we can just replace this sentence and skip the discussion about the unfortunate choice of the word "fashion". Replace with this:</p><p>\nTo "modify" a work means to copy from or adapt all or part of the work in a manner subject to a copyright holder's permission.</p><p>\nOne additional point: Making a "verbatim copy" is not under any copyright interpretation a "modification", so don't use that phrase. Just eliminate the "verbatim copy" reference entirely. It adds nothing.</p>\ngpl3.definitions.p2.s1\njkoenig\nlogin to agree\n0\n3053\n\n\n<p style="display: inline">The GPLv2 explicitly permitted modifying while not distributing modified copies, for example a company maintaining a set of patches that they only use internally and do not distribute outside of themselves.</p><p>\nIf this is still intended to be allowed, it should be explicitly mentioned, as it probably isn't allowed in the copyright laws of all jurisdictions.</p>\ngpl3.definitions.p2.s1\ntfelker\nlogin to agree\n0\n3091\n\n\n", "2473:2753:3053:3091": "\n\n\n\nI would use "permissible under applicable copyright law" rather than "copyright permission". Otherwise it's not clear if you mean local copyright law or any country's.\ngpl3.definitions.p2.s1\nmerijn\nlogin to agree\n1\n2473\n\n\nThe definitions of "modify", "modified version", work "based on" another work, and "covered work" are slightly changed with respect to GPLv3draft2, but they are still clear. It's good that the definition of "modified version" exploits applicable copyright law without trying to rewrite it. This ensures that the license does not place restrictions on activities that do not require permission under applicable copyright law.\n\ngpl3.definitions.p2.s1\nfrx\nlogin to agree\n0\n2753\n\n\n<p style="display: inline">Maybe we can just replace this sentence and skip the discussion about the unfortunate choice of the word "fashion". Replace with this:</p><p>\nTo "modify" a work means to copy from or adapt all or part of the work in a manner subject to a copyright holder's permission.</p><p>\nOne additional point: Making a "verbatim copy" is not under any copyright interpretation a "modification", so don't use that phrase. Just eliminate the "verbatim copy" reference entirely. It adds nothing.</p>\ngpl3.definitions.p2.s1\njkoenig\nlogin to agree\n0\n3053\n\n\n<p style="display: inline">The GPLv2 explicitly permitted modifying while not distributing modified copies, for example a company maintaining a set of patches that they only use internally and do not distribute outside of themselves.</p><p>\nIf this is still intended to be allowed, it should be explicitly mentioned, as it probably isn't allowed in the copyright laws of all jurisdictions.</p>\ngpl3.definitions.p2.s1\ntfelker\nlogin to agree\n0\n3091\n\n\n", "2506:2753:3053:3091": "\n\n\n\nSomeone might interpret this as "other than the making of a verbatim copy [of part of the work]". Stating explicitly that it is "a verbatim copy [of the whole work]" would make it crystal clear to the SCOs of the world.\ngpl3.definitions.p2.s1\nraulir\nlogin to agree\n0\n2506\n\n\nThe definitions of "modify", "modified version", work "based on" another work, and "covered work" are slightly changed with respect to GPLv3draft2, but they are still clear. It's good that the definition of "modified version" exploits applicable copyright law without trying to rewrite it. This ensures that the license does not place restrictions on activities that do not require permission under applicable copyright law.\n\ngpl3.definitions.p2.s1\nfrx\nlogin to agree\n0\n2753\n\n\n<p style="display: inline">Maybe we can just replace this sentence and skip the discussion about the unfortunate choice of the word "fashion". Replace with this:</p><p>\nTo "modify" a work means to copy from or adapt all or part of the work in a manner subject to a copyright holder's permission.</p><p>\nOne additional point: Making a "verbatim copy" is not under any copyright interpretation a "modification", so don't use that phrase. Just eliminate the "verbatim copy" reference entirely. It adds nothing.</p>\ngpl3.definitions.p2.s1\njkoenig\nlogin to agree\n0\n3053\n\n\n<p style="display: inline">The GPLv2 explicitly permitted modifying while not distributing modified copies, for example a company maintaining a set of patches that they only use internally and do not distribute outside of themselves.</p><p>\nIf this is still intended to be allowed, it should be explicitly mentioned, as it probably isn't allowed in the copyright laws of all jurisdictions.</p>\ngpl3.definitions.p2.s1\ntfelker\nlogin to agree\n0\n3091\n\n\n", "2444:3055": "\n\n\n\n<p style="display: inline">This still does not contain a definition of "based on". What does "based on" mean?</p><p>\nWe all know the common "linking" stuff... but "based on" is a very subjective term. It needs to be clarified. Does this mean linking in the sense of copying original work into new work or "using" in layman terms. If I write some software which talks to a GPLed piece of software through a pipe, does that make my work "based on" the other? I would certainly think it would. I am using their code to accomplish a job. There is nothing here disimilar to linking except on a technical level. Yet, people commonly consider this 'ok'.</p><p>\nPlease clarify the term. Thanks. </p>\ngpl3.definitions.p2.s2\nwasabi\nlogin to agree\n0\n2444\n\n\n<p style="display: inline">The definitions of of "modified version", "covered work" and work "based on" are poorly constructed. The definition of "contributor" and "contribution" are easily collapsed into one sentence. The definition of "based on" is wrong, because the later usage is actually "work based on" and so it should be stated that way in the first place. The language below replaces the 6th, 7th and 8th sentences in the Definitions section:</p><p>\nA "covered work" means either the original Program under this License or a version of Program that has been modified. The modifications made to a covered work (thereby creating a "modified version") constitute a "contribution" and the person or organization making such contribution is the "contributor". A "work based on" a covered work means a modified version of such covered work.</p>\ngpl3.definitions.p2.s2\njkoenig\nlogin to agree\n0\n3055\n\n\n", "2444:2472:3055": "\n\n\n\n<p style="display: inline">This still does not contain a definition of "based on". What does "based on" mean?</p><p>\nWe all know the common "linking" stuff... but "based on" is a very subjective term. It needs to be clarified. Does this mean linking in the sense of copying original work into new work or "using" in layman terms. If I write some software which talks to a GPLed piece of software through a pipe, does that make my work "based on" the other? I would certainly think it would. I am using their code to accomplish a job. There is nothing here disimilar to linking except on a technical level. Yet, people commonly consider this 'ok'.</p><p>\nPlease clarify the term. Thanks. </p>\ngpl3.definitions.p2.s2\nwasabi\nlogin to agree\n0\n2444\n\n\nThere's now a perfectly good definition of "modify" and "modified work". By keeping the "based on" things just get muddled.\n\n\ngpl3.definitions.p2.s2\nmerijn\nlogin to agree\n0\n2472\n\n\n<p style="display: inline">The definitions of of "modified version", "covered work" and work "based on" are poorly constructed. The definition of "contributor" and "contribution" are easily collapsed into one sentence. The definition of "based on" is wrong, because the later usage is actually "work based on" and so it should be stated that way in the first place. The language below replaces the 6th, 7th and 8th sentences in the Definitions section:</p><p>\nA "covered work" means either the original Program under this License or a version of Program that has been modified. The modifications made to a covered work (thereby creating a "modified version") constitute a "contribution" and the person or organization making such contribution is the "contributor". A "work based on" a covered work means a modified version of such covered work.</p>\ngpl3.definitions.p2.s2\njkoenig\nlogin to agree\n0\n3055\n\n\n", "2473:2506:2753:3053:3055:3091": "\n\n\n\nI would use "permissible under applicable copyright law" rather than "copyright permission". Otherwise it's not clear if you mean local copyright law or any country's.\ngpl3.definitions.p2.s1\nmerijn\nlogin to agree\n1\n2473\n\n\nSomeone might interpret this as "other than the making of a verbatim copy [of part of the work]". Stating explicitly that it is "a verbatim copy [of the whole work]" would make it crystal clear to the SCOs of the world.\ngpl3.definitions.p2.s1\nraulir\nlogin to agree\n0\n2506\n\n\nThe definitions of "modify", "modified version", work "based on" another work, and "covered work" are slightly changed with respect to GPLv3draft2, but they are still clear. It's good that the definition of "modified version" exploits applicable copyright law without trying to rewrite it. This ensures that the license does not place restrictions on activities that do not require permission under applicable copyright law.\n\ngpl3.definitions.p2.s1\nfrx\nlogin to agree\n0\n2753\n\n\n<p style="display: inline">Maybe we can just replace this sentence and skip the discussion about the unfortunate choice of the word "fashion". Replace with this:</p><p>\nTo "modify" a work means to copy from or adapt all or part of the work in a manner subject to a copyright holder's permission.</p><p>\nOne additional point: Making a "verbatim copy" is not under any copyright interpretation a "modification", so don't use that phrase. Just eliminate the "verbatim copy" reference entirely. It adds nothing.</p>\ngpl3.definitions.p2.s1\njkoenig\nlogin to agree\n0\n3053\n\n\n<p style="display: inline">The definitions of of "modified version", "covered work" and work "based on" are poorly constructed. The definition of "contributor" and "contribution" are easily collapsed into one sentence. The definition of "based on" is wrong, because the later usage is actually "work based on" and so it should be stated that way in the first place. The language below replaces the 6th, 7th and 8th sentences in the Definitions section:</p><p>\nA "covered work" means either the original Program under this License or a version of Program that has been modified. The modifications made to a covered work (thereby creating a "modified version") constitute a "contribution" and the person or organization making such contribution is the "contributor". A "work based on" a covered work means a modified version of such covered work.</p>\ngpl3.definitions.p2.s2\njkoenig\nlogin to agree\n0\n3055\n\n\n<p style="display: inline">The GPLv2 explicitly permitted modifying while not distributing modified copies, for example a company maintaining a set of patches that they only use internally and do not distribute outside of themselves.</p><p>\nIf this is still intended to be allowed, it should be explicitly mentioned, as it probably isn't allowed in the copyright laws of all jurisdictions.</p>\ngpl3.definitions.p2.s1\ntfelker\nlogin to agree\n0\n3091\n\n\n", "2668:2743:2845:2947": "\n\n\n\n<p style="display: inline">I've read this sentence five times and still don't understand it. The idea is that you contribute by taking a program and modifying it (using the GPL as a license) and then licensing your modification.</p><p>\nOkay, but the "on which the Program is based" gets the sequence backwards. Shouldn't it be "based on the Program"?</p>\ngpl3.definitions.p3.s1\nsanjoy\nlogin to agree\n2\n2668\n\n\nLet P and Q be unrelated programs, J be a patent holder and I be a hypothetical self. I and J are active contributors to P, and P doesn't infringe upon any of J's patents.\nI'm working on Q, and it will infringe upon J's patents. I'd like to get some patent protection for myself and other future users of Q, but P is not willing to offer me a suitable license.\nSince P, as distributed by J, does not implement any of J's patents, there's no patent license as per section 11.\nHowever, the last paragraph of section 10 appears to provide for what I need, because I know P is critical to J's business.\nConsider that I take any code whatsoever from P, as distributed by J, and add it to Q, even though it plays no functional role at all. Then, if J were to sue anyone for patent infringement in Q, I'd put it on notice and threaten to terminate its license for P.\nI understand we'd love if software patents didn't affect Free Software in any way, but it appears to me that the consequences of the above, if I go it right, might be for patent holders to stay away from development and distribution of any GPLv3 programs whatsoever. This could be a strategic loss.\nSome questions:\nIs my reading correct?\nIs posting patches (perhaps under more permissive licenses) enough to avoid the risks above for the patent holder?\nShould we perhaps try to find some way to accommodate such concerns, such as by limiting the scope of the termination for litigation such that it applies only to patents implemented in A's version of P (distributed or not), rather than to any work whatsoever that contains code from P?\nIt could be a bit wider than the provisions for implied patent licensing, in that it could still cover versions not distributed by P, but I'm not sure it's fair, or even strategically useful, to enable such tricks to bring in patent protection to unrelated projects.\n\ngpl3.definitions.p3.s1\noliva\nlogin to agree\n0\n2743\n\n\nA better way of saying this is: A 'contributor' is a party who licenses under this License a work based on the Program.\ngpl3.definitions.p3.s1\njkoenig\nlogin to agree\n0\n2845\n\n\n<p style="display: inline">Rationale says "contributors include all the copyright holders for the Program, other than copyright holders of material originally licensed under non-GPL terms and later incorporated into a GPL-covered work."</p><p>\nIf so, current text is wrong, because the Program itself is dropped off. (The term "based on" is premised on modifying.)</p><p>\nI think, the above rationale's text is more appropriate.\n</p>\ngpl3.definitions.p3.s1\nyusuke\nlogin to agree\n1\n2947\n\n\n", "2432:2668:2743:2845:2947": "\n\n\n\nThink that distinction should be made between code contributer and patent contributer. Are two different roles. Suspect that licence would be simpler if treated as two roles.\ngpl3.definitions.p3.s1\ngustavb\nlogin to agree\n0\n2432\n\n\n<p style="display: inline">I've read this sentence five times and still don't understand it. The idea is that you contribute by taking a program and modifying it (using the GPL as a license) and then licensing your modification.</p><p>\nOkay, but the "on which the Program is based" gets the sequence backwards. Shouldn't it be "based on the Program"?</p>\ngpl3.definitions.p3.s1\nsanjoy\nlogin to agree\n2\n2668\n\n\nLet P and Q be unrelated programs, J be a patent holder and I be a hypothetical self. I and J are active contributors to P, and P doesn't infringe upon any of J's patents.\nI'm working on Q, and it will infringe upon J's patents. I'd like to get some patent protection for myself and other future users of Q, but P is not willing to offer me a suitable license.\nSince P, as distributed by J, does not implement any of J's patents, there's no patent license as per section 11.\nHowever, the last paragraph of section 10 appears to provide for what I need, because I know P is critical to J's business.\nConsider that I take any code whatsoever from P, as distributed by J, and add it to Q, even though it plays no functional role at all. Then, if J were to sue anyone for patent infringement in Q, I'd put it on notice and threaten to terminate its license for P.\nI understand we'd love if software patents didn't affect Free Software in any way, but it appears to me that the consequences of the above, if I go it right, might be for patent holders to stay away from development and distribution of any GPLv3 programs whatsoever. This could be a strategic loss.\nSome questions:\nIs my reading correct?\nIs posting patches (perhaps under more permissive licenses) enough to avoid the risks above for the patent holder?\nShould we perhaps try to find some way to accommodate such concerns, such as by limiting the scope of the termination for litigation such that it applies only to patents implemented in A's version of P (distributed or not), rather than to any work whatsoever that contains code from P?\nIt could be a bit wider than the provisions for implied patent licensing, in that it could still cover versions not distributed by P, but I'm not sure it's fair, or even strategically useful, to enable such tricks to bring in patent protection to unrelated projects.\n\ngpl3.definitions.p3.s1\noliva\nlogin to agree\n0\n2743\n\n\nA better way of saying this is: A 'contributor' is a party who licenses under this License a work based on the Program.\ngpl3.definitions.p3.s1\njkoenig\nlogin to agree\n0\n2845\n\n\n<p style="display: inline">Rationale says "contributors include all the copyright holders for the Program, other than copyright holders of material originally licensed under non-GPL terms and later incorporated into a GPL-covered work."</p><p>\nIf so, current text is wrong, because the Program itself is dropped off. (The term "based on" is premised on modifying.)</p><p>\nI think, the above rationale's text is more appropriate.\n</p>\ngpl3.definitions.p3.s1\nyusuke\nlogin to agree\n1\n2947\n\n\n", "2668:2682:2683:2743:2845:2947": "\n\n\n\n<p style="display: inline">I've read this sentence five times and still don't understand it. The idea is that you contribute by taking a program and modifying it (using the GPL as a license) and then licensing your modification.</p><p>\nOkay, but the "on which the Program is based" gets the sequence backwards. Shouldn't it be "based on the Program"?</p>\ngpl3.definitions.p3.s1\nsanjoy\nlogin to agree\n2\n2668\n\n\n\ngpl3.definitions.p3.s1\nmerijn\nlogin to agree\n0\n2682\n\n\nA contributor contributes by adding things to the work. The converse doesn't make sense.\ngpl3.definitions.p3.s1\nmerijn\nlogin to agree\n0\n2683\n\n\nLet P and Q be unrelated programs, J be a patent holder and I be a hypothetical self. I and J are active contributors to P, and P doesn't infringe upon any of J's patents.\nI'm working on Q, and it will infringe upon J's patents. I'd like to get some patent protection for myself and other future users of Q, but P is not willing to offer me a suitable license.\nSince P, as distributed by J, does not implement any of J's patents, there's no patent license as per section 11.\nHowever, the last paragraph of section 10 appears to provide for what I need, because I know P is critical to J's business.\nConsider that I take any code whatsoever from P, as distributed by J, and add it to Q, even though it plays no functional role at all. Then, if J were to sue anyone for patent infringement in Q, I'd put it on notice and threaten to terminate its license for P.\nI understand we'd love if software patents didn't affect Free Software in any way, but it appears to me that the consequences of the above, if I go it right, might be for patent holders to stay away from development and distribution of any GPLv3 programs whatsoever. This could be a strategic loss.\nSome questions:\nIs my reading correct?\nIs posting patches (perhaps under more permissive licenses) enough to avoid the risks above for the patent holder?\nShould we perhaps try to find some way to accommodate such concerns, such as by limiting the scope of the termination for litigation such that it applies only to patents implemented in A's version of P (distributed or not), rather than to any work whatsoever that contains code from P?\nIt could be a bit wider than the provisions for implied patent licensing, in that it could still cover versions not distributed by P, but I'm not sure it's fair, or even strategically useful, to enable such tricks to bring in patent protection to unrelated projects.\n\ngpl3.definitions.p3.s1\noliva\nlogin to agree\n0\n2743\n\n\nA better way of saying this is: A 'contributor' is a party who licenses under this License a work based on the Program.\ngpl3.definitions.p3.s1\njkoenig\nlogin to agree\n0\n2845\n\n\n<p style="display: inline">Rationale says "contributors include all the copyright holders for the Program, other than copyright holders of material originally licensed under non-GPL terms and later incorporated into a GPL-covered work."</p><p>\nIf so, current text is wrong, because the Program itself is dropped off. (The term "based on" is premised on modifying.)</p><p>\nI think, the above rationale's text is more appropriate.\n</p>\ngpl3.definitions.p3.s1\nyusuke\nlogin to agree\n1\n2947\n\n\n", "2505:2668:2682:2683:2743:2845:2947": "\n\n\n\nAccording to the "based" language here, the latest party to modify the Program is not a contributor until someone else further bases their work on the Program. "Contributor" normally means anyone who has work in the Program. Having the definitions match common usage makes for a clearer text.\ngpl3.definitions.p3.s1\nraulir\nlogin to agree\n0\n2505\n\n\n<p style="display: inline">I've read this sentence five times and still don't understand it. The idea is that you contribute by taking a program and modifying it (using the GPL as a license) and then licensing your modification.</p><p>\nOkay, but the "on which the Program is based" gets the sequence backwards. Shouldn't it be "based on the Program"?</p>\ngpl3.definitions.p3.s1\nsanjoy\nlogin to agree\n2\n2668\n\n\n\ngpl3.definitions.p3.s1\nmerijn\nlogin to agree\n0\n2682\n\n\nA contributor contributes by adding things to the work. The converse doesn't make sense.\ngpl3.definitions.p3.s1\nmerijn\nlogin to agree\n0\n2683\n\n\nLet P and Q be unrelated programs, J be a patent holder and I be a hypothetical self. I and J are active contributors to P, and P doesn't infringe upon any of J's patents.\nI'm working on Q, and it will infringe upon J's patents. I'd like to get some patent protection for myself and other future users of Q, but P is not willing to offer me a suitable license.\nSince P, as distributed by J, does not implement any of J's patents, there's no patent license as per section 11.\nHowever, the last paragraph of section 10 appears to provide for what I need, because I know P is critical to J's business.\nConsider that I take any code whatsoever from P, as distributed by J, and add it to Q, even though it plays no functional role at all. Then, if J were to sue anyone for patent infringement in Q, I'd put it on notice and threaten to terminate its license for P.\nI understand we'd love if software patents didn't affect Free Software in any way, but it appears to me that the consequences of the above, if I go it right, might be for patent holders to stay away from development and distribution of any GPLv3 programs whatsoever. This could be a strategic loss.\nSome questions:\nIs my reading correct?\nIs posting patches (perhaps under more permissive licenses) enough to avoid the risks above for the patent holder?\nShould we perhaps try to find some way to accommodate such concerns, such as by limiting the scope of the termination for litigation such that it applies only to patents implemented in A's version of P (distributed or not), rather than to any work whatsoever that contains code from P?\nIt could be a bit wider than the provisions for implied patent licensing, in that it could still cover versions not distributed by P, but I'm not sure it's fair, or even strategically useful, to enable such tricks to bring in patent protection to unrelated projects.\n\ngpl3.definitions.p3.s1\noliva\nlogin to agree\n0\n2743\n\n\nA better way of saying this is: A 'contributor' is a party who licenses under this License a work based on the Program.\ngpl3.definitions.p3.s1\njkoenig\nlogin to agree\n0\n2845\n\n\n<p style="display: inline">Rationale says "contributors include all the copyright holders for the Program, other than copyright holders of material originally licensed under non-GPL terms and later incorporated into a GPL-covered work."</p><p>\nIf so, current text is wrong, because the Program itself is dropped off. (The term "based on" is premised on modifying.)</p><p>\nI think, the above rationale's text is more appropriate.\n</p>\ngpl3.definitions.p3.s1\nyusuke\nlogin to agree\n1\n2947\n\n\n", "2669": "\n\n\n\n=> "This work is..."\ngpl3.definitions.p3.s2\nsanjoy\nlogin to agree\n0\n2669\n\n\n", "2669:2682:2683:2743:2810:2845:2947": "\n\n\n\n=> "This work is..."\ngpl3.definitions.p3.s2\nsanjoy\nlogin to agree\n0\n2669\n\n\n\ngpl3.definitions.p3.s1\nmerijn\nlogin to agree\n0\n2682\n\n\nA contributor contributes by adding things to the work. The converse doesn't make sense.\ngpl3.definitions.p3.s1\nmerijn\nlogin to agree\n0\n2683\n\n\nLet P and Q be unrelated programs, J be a patent holder and I be a hypothetical self. I and J are active contributors to P, and P doesn't infringe upon any of J's patents.\nI'm working on Q, and it will infringe upon J's patents. I'd like to get some patent protection for myself and other future users of Q, but P is not willing to offer me a suitable license.\nSince P, as distributed by J, does not implement any of J's patents, there's no patent license as per section 11.\nHowever, the last paragraph of section 10 appears to provide for what I need, because I know P is critical to J's business.\nConsider that I take any code whatsoever from P, as distributed by J, and add it to Q, even though it plays no functional role at all. Then, if J were to sue anyone for patent infringement in Q, I'd put it on notice and threaten to terminate its license for P.\nI understand we'd love if software patents didn't affect Free Software in any way, but it appears to me that the consequences of the above, if I go it right, might be for patent holders to stay away from development and distribution of any GPLv3 programs whatsoever. This could be a strategic loss.\nSome questions:\nIs my reading correct?\nIs posting patches (perhaps under more permissive licenses) enough to avoid the risks above for the patent holder?\nShould we perhaps try to find some way to accommodate such concerns, such as by limiting the scope of the termination for litigation such that it applies only to patents implemented in A's version of P (distributed or not), rather than to any work whatsoever that contains code from P?\nIt could be a bit wider than the provisions for implied patent licensing, in that it could still cover versions not distributed by P, but I'm not sure it's fair, or even strategically useful, to enable such tricks to bring in patent protection to unrelated projects.\n\ngpl3.definitions.p3.s1\noliva\nlogin to agree\n0\n2743\n\n\n<p style="display: inline">I think that, with respect to a particular covered work, the contribution should be just the part of the contributor's work that is actually present in the covered work.</p><p>\nThat is, if author A takes a library routine from open-source program X written by author B, and uses it in a new work Z, both A and B are contributors to the new work Z, but B's contribution to that work should be just the library routine, not all of X.</p><p>\nThis distinction becomes important in the clauses involving patents and automatic licenses.\n\n\n</p>\ngpl3.definitions.p3.s2\nsepreece\nlogin to agree\n0\n2810\n\n\nA better way of saying this is: A 'contributor' is a party who licenses under this License a work based on the Program.\ngpl3.definitions.p3.s1\njkoenig\nlogin to agree\n0\n2845\n\n\n<p style="display: inline">Rationale says "contributors include all the copyright holders for the Program, other than copyright holders of material originally licensed under non-GPL terms and later incorporated into a GPL-covered work."</p><p>\nIf so, current text is wrong, because the Program itself is dropped off. (The term "based on" is premised on modifying.)</p><p>\nI think, the above rationale's text is more appropriate.\n</p>\ngpl3.definitions.p3.s1\nyusuke\nlogin to agree\n1\n2947\n\n\n", "2754": "\n\n\n\nThe definitions of "propagate" and "convey" seem fairly clear.\nAgain, their linking to copyright law ensures that the license does not place restrictions on activities that do not require permission under applicable copyright law.\n\ngpl3.definitions.p4.s1\nfrx\nlogin to agree\n0\n2754\n\n\n", "2645:2754:2959": "\n\n\n\n<p style="display: inline">If "or cause others to do" is to be included, it should at least be "or *intentionally* cause others to do", in order to avoid punishing people for unintended consequences, or "the butterfly effect".</p><p>\nWikipedia's article on Unintended consequence includes this example: "it is often conjectured that if the Treaty of Versailles had not imposed such harsh conditions on Germany, World War II would not have occurred." Statements of this sort may well be true, but this is not the type of cause that ought to be covered by the license.</p><p>\nBut even holding people responsible for intentionally causing others to propogate a GPL covered work may be too broad. Say I order a copy of a program which infringes on a GPL license, e.g. I order a copy of StarOffice, and Sun has infringed on the GPL license for some code included in OpenOffice. In ordering the program, I would be intentionally causing Sun to propogate the work. Should I be held responsible for this propogation?</p><p>\nPerhaps it would be better to only hold people responsible for intentionally causing propogation if they know in advance that the propogation is infringing?</p>\ngpl3.definitions.p4.s1\njamesgnz\nlogin to agree\n0\n2645\n\n\nThe definitions of "propagate" and "convey" seem fairly clear.\nAgain, their linking to copyright law ensures that the license does not place restrictions on activities that do not require permission under applicable copyright law.\n\ngpl3.definitions.p4.s1\nfrx\nlogin to agree\n0\n2754\n\n\nJust causing others to do so does not require permission under applicable copyright law.\nThis sentence should be amended as below. \n\nTo "propagate" a work means to do anything with it that requires permission under applicable copyright law (including to cause others to do it and to be deemed as your own activity under applicable law).\n\n\ngpl3.definitions.p4.s1\nyusuke\nlogin to agree\n0\n2959\n\n\n", "2474:2754": "\n\n\n\nSo sync the above with this text\ngpl3.definitions.p4.s1\nmerijn\nlogin to agree\n0\n2474\n\n\nThe definitions of "propagate" and "convey" seem fairly clear.\nAgain, their linking to copyright law ensures that the license does not place restrictions on activities that do not require permission under applicable copyright law.\n\ngpl3.definitions.p4.s1\nfrx\nlogin to agree\n0\n2754\n\n\n", "2747:2754:3008": "\n\n\n\n<p style="display: inline">This should be extended to say "except executing on a single computer which is accessible only by a single user" in order to exclude those circumstances where a copy is placed on a shared computer which can then be accessed by any number of other people.</p><p>\nOne of the dictionary definitions of propagate is: "Extend the action or operation of; transmit in some direction or through some medium". Similar definitions are available at http://www.thefreedictionary.com/propagate</p><p>\nIf software which is derived from or a modification of a GPL'd work is placed on a server which can be accessed by others, then that software is in effect being shared with others. Surely it does not matter that a third party is not running that software from their local hard drive but from a remote server? They are still running the software, supplying input to it and getting responses from it, regardless of the distance.</p><p>\nPropagate means sharing with others regardless of the means. The only way to not propagate is to not share.</p>\ngpl3.definitions.p4.s1\ntony32\nlogin to agree\n1\n2747\n\n\nThe definitions of "propagate" and "convey" seem fairly clear.\nAgain, their linking to copyright law ensures that the license does not place restrictions on activities that do not require permission under applicable copyright law.\n\ngpl3.definitions.p4.s1\nfrx\nlogin to agree\n0\n2754\n\n\n<p style="display: inline">The major problem I see with this paragraph is that it is styled like a math test in set theory. You start out with a set R (requirements from copyright law) and by various operations you obtain the sets P(ropagate) and C(onvey).<br><br></p><p>\nSome of the complication arises from the cutting out of sublicensing, executing and making private copies from the definitions. In the case of "propagate", the "except" part seems to be an attempt to include implicitly in the definitions what instead should be explicit grants of the license, namely "You are allowed to execute the work on a computer." and "You are allowed to make modifications to the work. If you do not share your modifications with other parties, you have no obligations under this license."<br><br></p><p>\nI suggest that the definition of "propagate" be simplified to "anything that requires permission under copyright law. This includes executing it on a computer, copying,..." Section "2. Basic Permissions" already states that executing the program on a computer and making private modifications is allowed. If there are places where the broader definition of "propagate" would make a difference, those should be amended. This will serve overall clarity.<br><br></p>\ngpl3.definitions.p4.s1\nmux2005\nlogin to agree\n0\n3008\n\n\n", "2754:3008": "\n\n\n\nThe definitions of "propagate" and "convey" seem fairly clear.\nAgain, their linking to copyright law ensures that the license does not place restrictions on activities that do not require permission under applicable copyright law.\n\ngpl3.definitions.p4.s1\nfrx\nlogin to agree\n0\n2754\n\n\n<p style="display: inline">The major problem I see with this paragraph is that it is styled like a math test in set theory. You start out with a set R (requirements from copyright law) and by various operations you obtain the sets P(ropagate) and C(onvey).<br><br></p><p>\nSome of the complication arises from the cutting out of sublicensing, executing and making private copies from the definitions. In the case of "propagate", the "except" part seems to be an attempt to include implicitly in the definitions what instead should be explicit grants of the license, namely "You are allowed to execute the work on a computer." and "You are allowed to make modifications to the work. If you do not share your modifications with other parties, you have no obligations under this license."<br><br></p><p>\nI suggest that the definition of "propagate" be simplified to "anything that requires permission under copyright law. This includes executing it on a computer, copying,..." Section "2. Basic Permissions" already states that executing the program on a computer and making private modifications is allowed. If there are places where the broader definition of "propagate" would make a difference, those should be amended. This will serve overall clarity.<br><br></p>\ngpl3.definitions.p4.s1\nmux2005\nlogin to agree\n0\n3008\n\n\n", "2556:2754:3008": "\n\n\n\n<p style="display: inline">If someone takes a GPL'd work, modifies it and places the result on an internet server which he then charges third parties to use, is this classed as "sharing"? As neither the executable nor the source code ever leaves the server it is not "conveyed", but the third parties are still in effect running the executable, and paying for the privilege.</p><p>\n"Making a modification that you do not share" would imply that only the person who made the modification can run the modified executable. Making that modified executable available to others, whether it be giving them copies which they can run on their own servers, or allowing them to access the executable from your own server, would still be classed as "sharing" in my book.</p>\ngpl3.definitions.p4.s1\ntony32\nlogin to agree\n0\n2556\n\n\nThe definitions of "propagate" and "convey" seem fairly clear.\nAgain, their linking to copyright law ensures that the license does not place restrictions on activities that do not require permission under applicable copyright law.\n\ngpl3.definitions.p4.s1\nfrx\nlogin to agree\n0\n2754\n\n\n<p style="display: inline">The major problem I see with this paragraph is that it is styled like a math test in set theory. You start out with a set R (requirements from copyright law) and by various operations you obtain the sets P(ropagate) and C(onvey).<br><br></p><p>\nSome of the complication arises from the cutting out of sublicensing, executing and making private copies from the definitions. In the case of "propagate", the "except" part seems to be an attempt to include implicitly in the definitions what instead should be explicit grants of the license, namely "You are allowed to execute the work on a computer." and "You are allowed to make modifications to the work. If you do not share your modifications with other parties, you have no obligations under this license."<br><br></p><p>\nI suggest that the definition of "propagate" be simplified to "anything that requires permission under copyright law. This includes executing it on a computer, copying,..." Section "2. Basic Permissions" already states that executing the program on a computer and making private modifications is allowed. If there are places where the broader definition of "propagate" would make a difference, those should be amended. This will serve overall clarity.<br><br></p>\ngpl3.definitions.p4.s1\nmux2005\nlogin to agree\n0\n3008\n\n\n", "2654:3010": "\n\n\n\n<p style="display: inline">May I suggest this sentence to clarify that (IIRC) "convey" is a subset of "propagation":</p><p>\nPropagation includes copying, to "convey" (distribution with or without modification), making available to the public, and in some countries other activities as well.</p>\ngpl3.definitions.p4.s2\nknassar\nlogin to agree\n0\n2654\n\n\nYou should put this after the definitions of "propagate" and "convey" and expand it to become examples of what is propagation, what is conveyance and what is propagation but not conveyance. As I've said in another comment, these concepts are so important for the understanding of the license that they must be explained properly.\ngpl3.definitions.p4.s2\nmux2005\nlogin to agree\n0\n3010\n\n\n", "2654:2748:3010": "\n\n\n\n<p style="display: inline">May I suggest this sentence to clarify that (IIRC) "convey" is a subset of "propagation":</p><p>\nPropagation includes copying, to "convey" (distribution with or without modification), making available to the public, and in some countries other activities as well.</p>\ngpl3.definitions.p4.s2\nknassar\nlogin to agree\n0\n2654\n\n\n<p style="display: inline">Propagation includes sharing something with *any* others, which may include others within a closed community and not necessarily the general public, therefore the phrase "making available to the public" should be changed to "making available to others". This would then follow more closely the notions of "private" and "shared" which appear in other places in this document.</p><p>\n"Private" is when I keep something to myself, but if I share it with *any* others then it is "shared" and ceases to be "private". I should not be able to use the excuse that I only shared it with the people in my street and not the general public, therefore it is still "private".\n\n</p>\ngpl3.definitions.p4.s2\ntony32\nlogin to agree\n0\n2748\n\n\nYou should put this after the definitions of "propagate" and "convey" and expand it to become examples of what is propagation, what is conveyance and what is propagation but not conveyance. As I've said in another comment, these concepts are so important for the understanding of the license that they must be explained properly.\ngpl3.definitions.p4.s2\nmux2005\nlogin to agree\n0\n3010\n\n\n", "2994": "\n\n\n\nConveying, and the distinctions between propagation and conveying seems important enough that the definition of convey should get its own paragraph. \n\nIf I understand correctly, making copies of a program within an organization would be propagation (making copies requiring permission under copyright law), but not conveying (as other copies are not receiving copies). \ngpl3.definitions.p4.s3\nmole\nlogin to agree\n1\n2994\n\n\n", "2994:3007": "\n\n\n\nConveying, and the distinctions between propagation and conveying seems important enough that the definition of convey should get its own paragraph. \n\nIf I understand correctly, making copies of a program within an organization would be propagation (making copies requiring permission under copyright law), but not conveying (as other copies are not receiving copies). \ngpl3.definitions.p4.s3\nmole\nlogin to agree\n1\n2994\n\n\n<p style="display: inline">The word "convey" gave me a lot of trouble. When I read this paragraph for the first (and 2nd, and 3rd,...) time, I didn't understand the distinction between propagate and convey. I understood from the construction of the "convey" part that is was a subset of propagation but I didn't understand what was included and was not.<br><br></p><p>\nAlthough now in hindsight, the definition "any kind of propagation that enables other parties to make or receive copies" seems clear enough it was not until I had read the rationale document that I understood that "convey" is supposed to mean what I would usually call "distribute". Once I realized this, suddenly everywhere in the license things started to make sense that I had previously not understood.<br><br></p><p>\nI strongly urge you to revise this paragraph. You should split it in 2 (one for "propagate" and one for "convey"), rephrase it in short, simple(!!!) sentences and provide an example or 2. Yes, this will double its size, but it is necessary. The concepts "propagate" and "convey" are so essential to the license that it must be ensured that readers will understand them. This is not one of those parts (like the stuff about patents) that only company lawyers need to understand. The terms "propagate" and "convey" must be crystal clear even to hobbyist programmers whose native language is not English, because otherwise they cannot understand the rest of the license.</p>\ngpl3.definitions.p4.s3\nmux2005\nlogin to agree\n0\n3007\n\n\n", "3009": "\n\n\n\n<p style="display: inline">Cf. my other comments on why I think the whole "convey" and "propagate" definition is too complex and needs to be simplified. In the case of "convey" this "excluding sublicensing" is another factor that makes the whole thing more difficult to understand. In this case, aside from making the definition (mathematically) more complex, I also find it confusing, because it does not seem to make sense. I don't see how sublicensing can "enable other parties to make or receive copies". Putting the statement "I hereby allow you to do ... with the work X" on a website or into a contract (this is what "(sub)licensing" means, isn't it?) does not enable anyone to do anything with work X. To do anything with work X, one has to have a copy of work X. So the only way to enable a party to make or receive copies of work X is to transfer (convey:-) work X to that party. Sublicensing doesn't do that. So excluding it does not seem to have an effect. <br><br></p><p>\nFurthermore, what I said with respect to "propagate" applies here, too. If you want to explicitly forbid sublicensing, put that in an explicit statement (I think there already is one, although I don't find it at the moment), rather than sneak it in implicitly by excluding it from the definition of "convey".</p>\ngpl3.definitions.p4.s3\nmux2005\nlogin to agree\n1\n3009\n\n\n", "2750": "\n\n\n\n<p style="display: inline">It may not be conveying the software, but I would contend that it is propagating the use of that software. Whether you run software from your local hard drive or a networked server you are still running that software, and if you allow others to run that software you are sharing and therefore propagating the use of that software. The physical distance between the user and the hard drive, and the ownership of that hard drive, should be irrelevant.</p><p>\nSomething can only be said to be "not propagated" if it is also "not shared". An idea can be propagated if it is shared with others, therefore software can also be propagated if it is shared with others. It is either "private and not propagated", or "propagated and not private".</p><p>\nThe notion that you must give a third party a physical copy of you software in order to share it is fundamentally flawed. You can have just one copy on a single networked server which can be accessed by millions of people and you are immediately sharing it with and propagating it to those millions of people.</p>\ngpl3.definitions.p4.s4\ntony32\nlogin to agree\n1\n2750\n\n\n", "2994:3007:3008:3009:3010": "\n\n\n\nConveying, and the distinctions between propagation and conveying seems important enough that the definition of convey should get its own paragraph. \n\nIf I understand correctly, making copies of a program within an organization would be propagation (making copies requiring permission under copyright law), but not conveying (as other copies are not receiving copies). \ngpl3.definitions.p4.s3\nmole\nlogin to agree\n1\n2994\n\n\n<p style="display: inline">The word "convey" gave me a lot of trouble. When I read this paragraph for the first (and 2nd, and 3rd,...) time, I didn't understand the distinction between propagate and convey. I understood from the construction of the "convey" part that is was a subset of propagation but I didn't understand what was included and was not.<br><br></p><p>\nAlthough now in hindsight, the definition "any kind of propagation that enables other parties to make or receive copies" seems clear enough it was not until I had read the rationale document that I understood that "convey" is supposed to mean what I would usually call "distribute". Once I realized this, suddenly everywhere in the license things started to make sense that I had previously not understood.<br><br></p><p>\nI strongly urge you to revise this paragraph. You should split it in 2 (one for "propagate" and one for "convey"), rephrase it in short, simple(!!!) sentences and provide an example or 2. Yes, this will double its size, but it is necessary. The concepts "propagate" and "convey" are so essential to the license that it must be ensured that readers will understand them. This is not one of those parts (like the stuff about patents) that only company lawyers need to understand. The terms "propagate" and "convey" must be crystal clear even to hobbyist programmers whose native language is not English, because otherwise they cannot understand the rest of the license.</p>\ngpl3.definitions.p4.s3\nmux2005\nlogin to agree\n0\n3007\n\n\n<p style="display: inline">The major problem I see with this paragraph is that it is styled like a math test in set theory. You start out with a set R (requirements from copyright law) and by various operations you obtain the sets P(ropagate) and C(onvey).<br><br></p><p>\nSome of the complication arises from the cutting out of sublicensing, executing and making private copies from the definitions. In the case of "propagate", the "except" part seems to be an attempt to include implicitly in the definitions what instead should be explicit grants of the license, namely "You are allowed to execute the work on a computer." and "You are allowed to make modifications to the work. If you do not share your modifications with other parties, you have no obligations under this license."<br><br></p><p>\nI suggest that the definition of "propagate" be simplified to "anything that requires permission under copyright law. This includes executing it on a computer, copying,..." Section "2. Basic Permissions" already states that executing the program on a computer and making private modifications is allowed. If there are places where the broader definition of "propagate" would make a difference, those should be amended. This will serve overall clarity.<br><br></p>\ngpl3.definitions.p4.s1\nmux2005\nlogin to agree\n0\n3008\n\n\n<p style="display: inline">Cf. my other comments on why I think the whole "convey" and "propagate" definition is too complex and needs to be simplified. In the case of "convey" this "excluding sublicensing" is another factor that makes the whole thing more difficult to understand. In this case, aside from making the definition (mathematically) more complex, I also find it confusing, because it does not seem to make sense. I don't see how sublicensing can "enable other parties to make or receive copies". Putting the statement "I hereby allow you to do ... with the work X" on a website or into a contract (this is what "(sub)licensing" means, isn't it?) does not enable anyone to do anything with work X. To do anything with work X, one has to have a copy of work X. So the only way to enable a party to make or receive copies of work X is to transfer (convey:-) work X to that party. Sublicensing doesn't do that. So excluding it does not seem to have an effect. <br><br></p><p>\nFurthermore, what I said with respect to "propagate" applies here, too. If you want to explicitly forbid sublicensing, put that in an explicit statement (I think there already is one, although I don't find it at the moment), rather than sneak it in implicitly by excluding it from the definition of "convey".</p>\ngpl3.definitions.p4.s3\nmux2005\nlogin to agree\n1\n3009\n\n\nYou should put this after the definitions of "propagate" and "convey" and expand it to become examples of what is propagation, what is conveyance and what is propagation but not conveyance. As I've said in another comment, these concepts are so important for the understanding of the license that they must be explained properly.\ngpl3.definitions.p4.s2\nmux2005\nlogin to agree\n0\n3010\n\n\n", "2977": "\n\n\n\nAll the definitions in section 0 are used throughout the license - except for this one, which is only used in section 11. Like other section-specific definitions, it should live in the appropriate section.\ngpl3.definitions.p5.s1\ngerv\nlogin to agree\n0\n2977\n\n\n", "2551": "\n\n\n\n<p style="display: inline">I suggest: \n"... that would be infringed [in a manner permitted by This License, including using, modifying, conveying, or selling the work], but do not ..."</p><p>\nor, more simply:\n"... that would be infringed [by the Basic Permissions defined in section 2], but do not ..."</p><p>\nI do not understand what "making" the work means.</p>\ngpl3.definitions.p5.s1\nccady\nlogin to agree\n1\n2551\n\n\n", "2551:2977": "\n\n\n\n<p style="display: inline">I suggest: \n"... that would be infringed [in a manner permitted by This License, including using, modifying, conveying, or selling the work], but do not ..."</p><p>\nor, more simply:\n"... that would be infringed [by the Basic Permissions defined in section 2], but do not ..."</p><p>\nI do not understand what "making" the work means.</p>\ngpl3.definitions.p5.s1\nccady\nlogin to agree\n1\n2551\n\n\nAll the definitions in section 0 are used throughout the license - except for this one, which is only used in section 11. Like other section-specific definitions, it should live in the appropriate section.\ngpl3.definitions.p5.s1\ngerv\nlogin to agree\n0\n2977\n\n\n", "2548": "\n\n\n\nThis section seems to include definitions. Shouldn't "Source Code" and the other definitions ("Standard Interface", "System Libraries", "Corresponding Source") just be items under the Definitions section. It might also be nice to make each definition a subsection.\ngpl3.sourcecode.0.0\ndkrohn70\nlogin to agree\n3\n2548\n\n\n", "2548:2715:2755": "\n\n\n\nThis section seems to include definitions. Shouldn't "Source Code" and the other definitions ("Standard Interface", "System Libraries", "Corresponding Source") just be items under the Definitions section. It might also be nice to make each definition a subsection.\ngpl3.sourcecode.0.0\ndkrohn70\nlogin to agree\n3\n2548\n\n\nThese definitions need to be drastically simplified, even if that means moving back to the GPLv2 defs. (which which didn't really seem that problematic).\ngpl3.sourcecode.0.0\nflaschen\nlogin to agree\n0\n2715\n\n\nThis section seems OK to me.\ngpl3.sourcecode.0.0\nfrx\nlogin to agree\n1\n2755\n\n\n", "2629:2756": "\n\n\n\nMove to Definition section and use mixed case (eg 'Source Code')\ngpl3.sourcecode.p0.s1\nsamj\nlogin to agree\n0\n2629\n\n\nI like the definitions of "source code" and "Object code" as they are. Good to see that they were kept unaltered.\ngpl3.sourcecode.p0.s1\nfrx\nlogin to agree\n0\n2756\n\n\n", "2465:2629:2756": "\n\n\n\n<p style="display: inline">As I already mentioned when the 1st draft came out, the "prefered form" for making modifications to the work is a bit vague and depends on the context. Should it be "prefered" by the original author, by anyone who modifies or propagates the work, or by anyone who just receives the work?</p><p>\nConcrete example: someone creates a cross-platform application and distributes it. The source code includes the usual Makefiles and configure scripts for building it using free software tools. Later, someone else makes extensive changes and improvements to the application but since they are using proprietary tools for building it (say, Visual Studio running on Windows), they replace all the original Makefiles by project files that are specific to that proprietary tool. This is the "prefered form" for the second developer, but this excludes the original author who will not be able to use the modified code. To some extent, the new code has lost a part of its freedom.</p><p>\nCredits to tml (who ported GTK+ and GIMP to Windows) for discussing this issue with me last year.\n</p>\ngpl3.sourcecode.p0.s1\nraphael\nlogin to agree\n8\n2465\n\n\nMove to Definition section and use mixed case (eg 'Source Code')\ngpl3.sourcecode.p0.s1\nsamj\nlogin to agree\n0\n2629\n\n\nI like the definitions of "source code" and "Object code" as they are. Good to see that they were kept unaltered.\ngpl3.sourcecode.p0.s1\nfrx\nlogin to agree\n0\n2756\n\n\n", "2628": "\n\n\n\nMove to Definitions section and use mixed case (eg 'Object Code')\ngpl3.sourcecode.p0.s2\nsamj\nlogin to agree\n1\n2628\n\n\n", "2628:2629:2756": "\n\n\n\nMove to Definitions section and use mixed case (eg 'Object Code')\ngpl3.sourcecode.p0.s2\nsamj\nlogin to agree\n1\n2628\n\n\nMove to Definition section and use mixed case (eg 'Source Code')\ngpl3.sourcecode.p0.s1\nsamj\nlogin to agree\n0\n2629\n\n\nI like the definitions of "source code" and "Object code" as they are. Good to see that they were kept unaltered.\ngpl3.sourcecode.p0.s1\nfrx\nlogin to agree\n0\n2756\n\n\n", "2448": "\n\n\n\nThe "Standard Inferface" definition, like the "Major Component" definition, are used in this license only for defining what "System Libraries" are. However, "Standard Interface" is given its own paragraph, while "Major Component" is defined as a clarifying comment in the "System Libraries" definition paragraph. I suggest putting the "Standard Interface" and "Major Component" definitions consecutively.\ngpl3.sourcecode.p1.s1\ncyd\nlogin to agree\n4\n2448\n\n\n", "2448:3072": "\n\n\n\nThe "Standard Inferface" definition, like the "Major Component" definition, are used in this license only for defining what "System Libraries" are. However, "Standard Interface" is given its own paragraph, while "Major Component" is defined as a clarifying comment in the "System Libraries" definition paragraph. I suggest putting the "Standard Interface" and "Major Component" definitions consecutively.\ngpl3.sourcecode.p1.s1\ncyd\nlogin to agree\n4\n2448\n\n\nI feel "widely used among developers working in that language" is too unspecific, and could be redefined against its meaning by those looking to do so. If nothing else, you to emphasise that it is widely used amongst multiple independent developers\ngpl3.sourcecode.p1.s1\njamesj\nlogin to agree\n0\n3072\n\n\n", "2631:2675:3013": "\n\n\n\nMove to Definitions section.\ngpl3.sourcecode.p2.s1\nsamj\nlogin to agree\n1\n2631\n\n\nI've read this sentence many times and still don't understand it and cannot come up with an exmaple of a system library. In part (a), the "but which is not part of that Major component" phrase trips me up. And I cannot see why there's an "and (b) serves only...".\ngpl3.sourcecode.p2.s1\nsanjoy\nlogin to agree\n1\n2675\n\n\n<p style="display: inline">Maybe we can characterize System Libraries by typical properties, that communicate the intent of what is supposed to be covered. How's this for a start:<br><br></p><p>\n"System Libraries" of an executable work are components that are normally included in the distribution of the operating environment in which the work is to be executed. These components are characterized by several of the following aspects.<br><br> \n\n(a) they are not included in distributions of the executable work<br><br></p><p>\n(b) they are provided by a party not affiliated with the executable work <br><br></p><p>\n(c) they are used by programs other than the executable work and this includes programs from parties not affiliated with the executable work.<br><br></p><p>\n(d) they are not normally included in distributions of other executable works for the same operating environment, even if those works require them.<br><br></p><p>\nI think it's clear that something that fits one or more of the above criteria (in addition to the primary criterion that it is "normally included in the distribution of the operating environment") is a System Library for which people should be required to provide source. Defining it this way has the added benefit that it offers a clear list of easily verifiable conditions to check if something is a System Library or not. </p>\ngpl3.sourcecode.p2.s1\nmux2005\nlogin to agree\n0\n3013\n\n\n", "2631:2675:2864:3013": "\n\n\n\nMove to Definitions section.\ngpl3.sourcecode.p2.s1\nsamj\nlogin to agree\n1\n2631\n\n\nI've read this sentence many times and still don't understand it and cannot come up with an exmaple of a system library. In part (a), the "but which is not part of that Major component" phrase trips me up. And I cannot see why there's an "and (b) serves only...".\ngpl3.sourcecode.p2.s1\nsanjoy\nlogin to agree\n1\n2675\n\n\nThe definition of "System Libraries" itself uses two definitions; "Major Components" (defined after it) and "Standard Interface (defined before it). There should be a consistent rule that terms are defined before they are used, unless a circle has to be broken.\ngpl3.sourcecode.p2.s1\ngerv\nlogin to agree\n0\n2864\n\n\n<p style="display: inline">Maybe we can characterize System Libraries by typical properties, that communicate the intent of what is supposed to be covered. How's this for a start:<br><br></p><p>\n"System Libraries" of an executable work are components that are normally included in the distribution of the operating environment in which the work is to be executed. These components are characterized by several of the following aspects.<br><br> \n\n(a) they are not included in distributions of the executable work<br><br></p><p>\n(b) they are provided by a party not affiliated with the executable work <br><br></p><p>\n(c) they are used by programs other than the executable work and this includes programs from parties not affiliated with the executable work.<br><br></p><p>\n(d) they are not normally included in distributions of other executable works for the same operating environment, even if those works require them.<br><br></p><p>\nI think it's clear that something that fits one or more of the above criteria (in addition to the primary criterion that it is "normally included in the distribution of the operating environment") is a System Library for which people should be required to provide source. Defining it this way has the added benefit that it offers a clear list of easily verifiable conditions to check if something is a System Library or not. </p>\ngpl3.sourcecode.p2.s1\nmux2005\nlogin to agree\n0\n3013\n\n\n", "2631:2675:2714:3013": "\n\n\n\nMove to Definitions section.\ngpl3.sourcecode.p2.s1\nsamj\nlogin to agree\n1\n2631\n\n\nI've read this sentence many times and still don't understand it and cannot come up with an exmaple of a system library. In part (a), the "but which is not part of that Major component" phrase trips me up. And I cannot see why there's an "and (b) serves only...".\ngpl3.sourcecode.p2.s1\nsanjoy\nlogin to agree\n1\n2675\n\n\nThis should be replaced with object code, to avoid causing problems with interpreted code, or non-software works.\ngpl3.sourcecode.p2.s1\nflaschen\nlogin to agree\n0\n2714\n\n\n<p style="display: inline">Maybe we can characterize System Libraries by typical properties, that communicate the intent of what is supposed to be covered. How's this for a start:<br><br></p><p>\n"System Libraries" of an executable work are components that are normally included in the distribution of the operating environment in which the work is to be executed. These components are characterized by several of the following aspects.<br><br> \n\n(a) they are not included in distributions of the executable work<br><br></p><p>\n(b) they are provided by a party not affiliated with the executable work <br><br></p><p>\n(c) they are used by programs other than the executable work and this includes programs from parties not affiliated with the executable work.<br><br></p><p>\n(d) they are not normally included in distributions of other executable works for the same operating environment, even if those works require them.<br><br></p><p>\nI think it's clear that something that fits one or more of the above criteria (in addition to the primary criterion that it is "normally included in the distribution of the operating environment") is a System Library for which people should be required to provide source. Defining it this way has the added benefit that it offers a clear list of easily verifiable conditions to check if something is a System Library or not. </p>\ngpl3.sourcecode.p2.s1\nmux2005\nlogin to agree\n0\n3013\n\n\n", "2631:2675:3013:3061": "\n\n\n\nMove to Definitions section.\ngpl3.sourcecode.p2.s1\nsamj\nlogin to agree\n1\n2631\n\n\nI've read this sentence many times and still don't understand it and cannot come up with an exmaple of a system library. In part (a), the "but which is not part of that Major component" phrase trips me up. And I cannot see why there's an "and (b) serves only...".\ngpl3.sourcecode.p2.s1\nsanjoy\nlogin to agree\n1\n2675\n\n\n<p style="display: inline">Maybe we can characterize System Libraries by typical properties, that communicate the intent of what is supposed to be covered. How's this for a start:<br><br></p><p>\n"System Libraries" of an executable work are components that are normally included in the distribution of the operating environment in which the work is to be executed. These components are characterized by several of the following aspects.<br><br> \n\n(a) they are not included in distributions of the executable work<br><br></p><p>\n(b) they are provided by a party not affiliated with the executable work <br><br></p><p>\n(c) they are used by programs other than the executable work and this includes programs from parties not affiliated with the executable work.<br><br></p><p>\n(d) they are not normally included in distributions of other executable works for the same operating environment, even if those works require them.<br><br></p><p>\nI think it's clear that something that fits one or more of the above criteria (in addition to the primary criterion that it is "normally included in the distribution of the operating environment") is a System Library for which people should be required to provide source. Defining it this way has the added benefit that it offers a clear list of easily verifiable conditions to check if something is a System Library or not. </p>\ngpl3.sourcecode.p2.s1\nmux2005\nlogin to agree\n0\n3013\n\n\n<p style="display: inline">If the portion is other than the work as a whole, it is not part of the Corresponding Source literally.</p><p>\nI think this "System Libraries" definition is intended to show distinction whether the portion, which is part of the work as a whole under the section 5 of this License, is included in the Coorespong Source or not.</p><p>\nIs my understanding wrong?</p>\ngpl3.sourcecode.p2.s1\nyusuke\nlogin to agree\n0\n3061\n\n\n", "2631:2675:3011:3013": "\n\n\n\nMove to Definitions section.\ngpl3.sourcecode.p2.s1\nsamj\nlogin to agree\n1\n2631\n\n\nI've read this sentence many times and still don't understand it and cannot come up with an exmaple of a system library. In part (a), the "but which is not part of that Major component" phrase trips me up. And I cannot see why there's an "and (b) serves only...".\ngpl3.sourcecode.p2.s1\nsanjoy\nlogin to agree\n1\n2675\n\n\n<p style="display: inline">My understanding is that<br><br>\n1. glibc is included in the distribution of a Major Component (what this package is called depends on the GNU/Linux distribution you use)<br>\n2. glibc is a part of that Major Component<br>\n3. glibc is (at least to programs that use glibc-specific functions) *not* just an implementation of a Standard Interface.<br><br></p><p>\nSo as I understand it, glibc fails the definition of "System Library" on all counts. This cannot be intended. I think the whole "System Libraries" paragraph is unfixable and will need to be written again.<br><br>\n</p>\ngpl3.sourcecode.p2.s1\nmux2005\nlogin to agree\n0\n3011\n\n\n<p style="display: inline">Maybe we can characterize System Libraries by typical properties, that communicate the intent of what is supposed to be covered. How's this for a start:<br><br></p><p>\n"System Libraries" of an executable work are components that are normally included in the distribution of the operating environment in which the work is to be executed. These components are characterized by several of the following aspects.<br><br> \n\n(a) they are not included in distributions of the executable work<br><br></p><p>\n(b) they are provided by a party not affiliated with the executable work <br><br></p><p>\n(c) they are used by programs other than the executable work and this includes programs from parties not affiliated with the executable work.<br><br></p><p>\n(d) they are not normally included in distributions of other executable works for the same operating environment, even if those works require them.<br><br></p><p>\nI think it's clear that something that fits one or more of the above criteria (in addition to the primary criterion that it is "normally included in the distribution of the operating environment") is a System Library for which people should be required to provide source. Defining it this way has the added benefit that it offers a clear list of easily verifiable conditions to check if something is a System Library or not. </p>\ngpl3.sourcecode.p2.s1\nmux2005\nlogin to agree\n0\n3013\n\n\n", "2631:2675:3011": "\n\n\n\nMove to Definitions section.\ngpl3.sourcecode.p2.s1\nsamj\nlogin to agree\n1\n2631\n\n\nI've read this sentence many times and still don't understand it and cannot come up with an exmaple of a system library. In part (a), the "but which is not part of that Major component" phrase trips me up. And I cannot see why there's an "and (b) serves only...".\ngpl3.sourcecode.p2.s1\nsanjoy\nlogin to agree\n1\n2675\n\n\n<p style="display: inline">My understanding is that<br><br>\n1. glibc is included in the distribution of a Major Component (what this package is called depends on the GNU/Linux distribution you use)<br>\n2. glibc is a part of that Major Component<br>\n3. glibc is (at least to programs that use glibc-specific functions) *not* just an implementation of a Standard Interface.<br><br></p><p>\nSo as I understand it, glibc fails the definition of "System Library" on all counts. This cannot be intended. I think the whole "System Libraries" paragraph is unfixable and will need to be written again.<br><br>\n</p>\ngpl3.sourcecode.p2.s1\nmux2005\nlogin to agree\n0\n3011\n\n\n", "2631:2675:2689:3011": "\n\n\n\nMove to Definitions section.\ngpl3.sourcecode.p2.s1\nsamj\nlogin to agree\n1\n2631\n\n\nI've read this sentence many times and still don't understand it and cannot come up with an exmaple of a system library. In part (a), the "but which is not part of that Major component" phrase trips me up. And I cannot see why there's an "and (b) serves only...".\ngpl3.sourcecode.p2.s1\nsanjoy\nlogin to agree\n1\n2675\n\n\n<p style="display: inline">This section has gotten more convoluted, to no good end. How do you distinguish a major component from a minor component? Why should the definition of a system library be dependent on use by "major components" to beging with? How do you define "essential"?</p><p>\nI suggest replacing the whole paragraph with something along the lines of \n\nAn "enabler" of an executable work is any library or other component needed to use the work in confunction with another work, and that either is normally included with that other work or is a standard interface for which a source-code implementation is available to the public.</p>\ngpl3.sourcecode.p2.s1\nsepreece\nlogin to agree\n0\n2689\n\n\n<p style="display: inline">My understanding is that<br><br>\n1. glibc is included in the distribution of a Major Component (what this package is called depends on the GNU/Linux distribution you use)<br>\n2. glibc is a part of that Major Component<br>\n3. glibc is (at least to programs that use glibc-specific functions) *not* just an implementation of a Standard Interface.<br><br></p><p>\nSo as I understand it, glibc fails the definition of "System Library" on all counts. This cannot be intended. I think the whole "System Libraries" paragraph is unfixable and will need to be written again.<br><br>\n</p>\ngpl3.sourcecode.p2.s1\nmux2005\nlogin to agree\n0\n3011\n\n\n", "2631:2675": "\n\n\n\nMove to Definitions section.\ngpl3.sourcecode.p2.s1\nsamj\nlogin to agree\n1\n2631\n\n\nI've read this sentence many times and still don't understand it and cannot come up with an exmaple of a system library. In part (a), the "but which is not part of that Major component" phrase trips me up. And I cannot see why there's an "and (b) serves only...".\ngpl3.sourcecode.p2.s1\nsanjoy\nlogin to agree\n1\n2675\n\n\n", "3063": "\n\n\n\n<p style="display: inline">If the portion just implements a Standard Interface, we can make another implementation.\nTherefore, the requirement of existing another implementation in source code form is not needed.</p><p>\nI think this sentence, "for which an implementation is available to the public in source code form", may be deleted.</p>\ngpl3.sourcecode.p2.s1\nyusuke\nlogin to agree\n1\n3063\n\n\n", "2507": "\n\n\n\nDefinition of "Major Component", like the other definitions, should come before its first use in the text.\ngpl3.sourcecode.p2.s2\nraulir\nlogin to agree\n5\n2507\n\n\n", "2631:2632:2670:2675:2689:2714:2864:3011:3012:3013:3061:3063": "\n\n\n\nMove to Definitions section.\ngpl3.sourcecode.p2.s1\nsamj\nlogin to agree\n1\n2631\n\n\nMove to Definitions section.\ngpl3.sourcecode.p2.s2\nsamj\nlogin to agree\n0\n2632\n\n\n"Major Component" is referenced before this definition, so move the definition to the beginning of the paragraph (and delete "in this context"). Then define "System Libraries", perhaps in a separate paragraph.\ngpl3.sourcecode.p2.s2\nsanjoy\nlogin to agree\n0\n2670\n\n\nI've read this sentence many times and still don't understand it and cannot come up with an exmaple of a system library. In part (a), the "but which is not part of that Major component" phrase trips me up. And I cannot see why there's an "and (b) serves only...".\ngpl3.sourcecode.p2.s1\nsanjoy\nlogin to agree\n1\n2675\n\n\n<p style="display: inline">This section has gotten more convoluted, to no good end. How do you distinguish a major component from a minor component? Why should the definition of a system library be dependent on use by "major components" to beging with? How do you define "essential"?</p><p>\nI suggest replacing the whole paragraph with something along the lines of \n\nAn "enabler" of an executable work is any library or other component needed to use the work in confunction with another work, and that either is normally included with that other work or is a standard interface for which a source-code implementation is available to the public.</p>\ngpl3.sourcecode.p2.s1\nsepreece\nlogin to agree\n0\n2689\n\n\nThis should be replaced with object code, to avoid causing problems with interpreted code, or non-software works.\ngpl3.sourcecode.p2.s1\nflaschen\nlogin to agree\n0\n2714\n\n\nThe definition of "System Libraries" itself uses two definitions; "Major Components" (defined after it) and "Standard Interface (defined before it). There should be a consistent rule that terms are defined before they are used, unless a circle has to be broken.\ngpl3.sourcecode.p2.s1\ngerv\nlogin to agree\n0\n2864\n\n\n<p style="display: inline">My understanding is that<br><br>\n1. glibc is included in the distribution of a Major Component (what this package is called depends on the GNU/Linux distribution you use)<br>\n2. glibc is a part of that Major Component<br>\n3. glibc is (at least to programs that use glibc-specific functions) *not* just an implementation of a Standard Interface.<br><br></p><p>\nSo as I understand it, glibc fails the definition of "System Library" on all counts. This cannot be intended. I think the whole "System Libraries" paragraph is unfixable and will need to be written again.<br><br>\n</p>\ngpl3.sourcecode.p2.s1\nmux2005\nlogin to agree\n0\n3011\n\n\n<p style="display: inline">This definition sounds as if an OS itself is not a Major Component, because an OS cannot be a major essential component of itself. Hmm. I think that this means that GPLv3 programs cannot be ported to MS Windows. My reasoning is as follows:<br><br></p><p>\n1. An Operating System is not a Major Component<br><br></p><p>\n2. Windows is an Operating System (I know that different people have different definitions of "Operating System", but I think no one will disagree that it is not unlikely that a judge will find that Windows as a whole is an "Operating System")<br><br></p><p>\n3. Windows is not a Major Component. Therefore, none of its DLLs can be System Libraries.<br><br></p><p>\n4. Source to Windows DLLs has to be part of the Corresponding Source. Because it is not available, one cannot satisfy GPL's requirements and therefore is not allowed to convey a Windows port of a GPL'ed program.<br><br></p><p>\nBut even if this text is changed so that Windows qualifies as a Major Component, this doesn't help, because the "not part of that Major Component" language would not match those DLLs, because at least in my (and I think most other's) reading of "part", those DLLs are "part" of Windows.<br><br></p><p>\nFurthermore, a system library needs to "serve only to enable use of the work with that Major Component, or to implement...available to the public in source code form." The 2nd part (Standard Interface...implementation available in source code form) certainly doesn't apply to most Windows DLLs. And I think that the 1st part doesn't apply either, because Windows DLLs are used internally by Windows and thus don't (as I understand it) "serve *only* to enable use of the work with that Major Component".<br><br></p><p>\nNo matter how I look at it, I don't see how this language can be easily amended to cover Windows DLLs as System Libraries. As I've said in another comment, I think this whole part has to be scrapped and written again.</p>\ngpl3.sourcecode.p2.s2\nmux2005\nlogin to agree\n0\n3012\n\n\n<p style="display: inline">Maybe we can characterize System Libraries by typical properties, that communicate the intent of what is supposed to be covered. How's this for a start:<br><br></p><p>\n"System Libraries" of an executable work are components that are normally included in the distribution of the operating environment in which the work is to be executed. These components are characterized by several of the following aspects.<br><br> \n\n(a) they are not included in distributions of the executable work<br><br></p><p>\n(b) they are provided by a party not affiliated with the executable work <br><br></p><p>\n(c) they are used by programs other than the executable work and this includes programs from parties not affiliated with the executable work.<br><br></p><p>\n(d) they are not normally included in distributions of other executable works for the same operating environment, even if those works require them.<br><br></p><p>\nI think it's clear that something that fits one or more of the above criteria (in addition to the primary criterion that it is "normally included in the distribution of the operating environment") is a System Library for which people should be required to provide source. Defining it this way has the added benefit that it offers a clear list of easily verifiable conditions to check if something is a System Library or not. </p>\ngpl3.sourcecode.p2.s1\nmux2005\nlogin to agree\n0\n3013\n\n\n<p style="display: inline">If the portion is other than the work as a whole, it is not part of the Corresponding Source literally.</p><p>\nI think this "System Libraries" definition is intended to show distinction whether the portion, which is part of the work as a whole under the section 5 of this License, is included in the Coorespong Source or not.</p><p>\nIs my understanding wrong?</p>\ngpl3.sourcecode.p2.s1\nyusuke\nlogin to agree\n0\n3061\n\n\n<p style="display: inline">If the portion just implements a Standard Interface, we can make another implementation.\nTherefore, the requirement of existing another implementation in source code form is not needed.</p><p>\nI think this sentence, "for which an implementation is available to the public in source code form", may be deleted.</p>\ngpl3.sourcecode.p2.s1\nyusuke\nlogin to agree\n1\n3063\n\n\n", "2630": "\n\n\n\nMove to Definitions section.\ngpl3.sourcecode.p3.s1\nsamj\nlogin to agree\n0\n2630\n\n\n", "3064": "\n\n\n\n<p style="display: inline">If some portions are not part of the work, those portions are not included in the Corresponding Source literally.</p><p>\nIs this a new copyleft provision?\n(Not modifying or linking with the GPL Program but just using in performing those activities causes transmission of the GPL?)</p><p>\nThis part should be re-written as described below.</p><p>\n", or any tools which are used in performing those activities unless those tools themselves are part of the work."</p><p>\nOf course, Makefiles and/or configure scripts are part of the work.</p>\ngpl3.sourcecode.p3.s2\nyusuke\nlogin to agree\n0\n3064\n\n\n", "2671:2684:2734:2797:2938:3064": "\n\n\n\nDoes "free programs" mean "free software" in the sense of the FSF? Or (less likely!) free as in beer? Does the distinction need to be made clear here?\ngpl3.sourcecode.p3.s2\nsanjoy\nlogin to agree\n2\n2671\n\n\nIf the intention is really to restrict this coverage to "free as in freedom", then this should be more explicit. \ngpl3.sourcecode.p3.s2\nmerijn\nlogin to agree\n11\n2684\n\n\nWhat about non-free, not-general purpose tools? It is conceivable that a non-free program might be used to compile and install a free one.\ngpl3.sourcecode.p3.s2\nmkorman\nlogin to agree\n0\n2734\n\n\nThose of us reading this license, and with a sense of Freedom understand this means "libre". However, common (english) usage does not mean this. Since the term "Free" is not defined in this license, this phrase is ambiguous.\ngpl3.sourcecode.p3.s2\nkdean06\nlogin to agree\n1\n2797\n\n\nThe usage of this term should definitely be explained a bit better.\ngpl3.sourcecode.p3.s2\nmsikma\nlogin to agree\n2\n2938\n\n\n<p style="display: inline">If some portions are not part of the work, those portions are not included in the Corresponding Source literally.</p><p>\nIs this a new copyleft provision?\n(Not modifying or linking with the GPL Program but just using in performing those activities causes transmission of the GPL?)</p><p>\nThis part should be re-written as described below.</p><p>\n", or any tools which are used in performing those activities unless those tools themselves are part of the work."</p><p>\nOf course, Makefiles and/or configure scripts are part of the work.</p>\ngpl3.sourcecode.p3.s2\nyusuke\nlogin to agree\n0\n3064\n\n\n", "3014": "\n\n\n\nThis seems much too broad. Let's say I take a GPLv3 licensed GNU tar and modify it, so that it uses (and requires) some glibc-specific functionality, then certainly glibc is a "shared library...that the work is specifically designed to require". So I'd have to distribute the complete glibc source code with it, if I provided my modified tar in object code form. Is this intentional? There also should be a reiteration that this does not include system libraries. Otherwise this example seems to contradict the text that stands before it, because system libraries usually fit the description "shared libraries and dynamically linked subprograms that the work is specifically designed to require" and the statement earlier says that System Libraries are not part of the Corresponding Source.\n\ngpl3.sourcecode.p3.s3\nmux2005\nlogin to agree\n0\n3014\n\n\n", "2653": "\n\n\n\nMaybe the term "symbiosis" can be used to make this more precise. See http://en.wikipedia.org/wiki/Symbiosis .\ngpl3.sourcecode.p3.s3\nhkbst\nlogin to agree\n1\n2653\n\n\n", "2443": "\n\n\n\nShould read "among those subprograms" between is for interactions of just two subprograms.\ngpl3.sourcecode.p3.s3\nmaginnis\nlogin to agree\n1\n2443\n\n\n", "2671:2684:2734:2797:2938:3014:3064": "\n\n\n\nDoes "free programs" mean "free software" in the sense of the FSF? Or (less likely!) free as in beer? Does the distinction need to be made clear here?\ngpl3.sourcecode.p3.s2\nsanjoy\nlogin to agree\n2\n2671\n\n\nIf the intention is really to restrict this coverage to "free as in freedom", then this should be more explicit. \ngpl3.sourcecode.p3.s2\nmerijn\nlogin to agree\n11\n2684\n\n\nWhat about non-free, not-general purpose tools? It is conceivable that a non-free program might be used to compile and install a free one.\ngpl3.sourcecode.p3.s2\nmkorman\nlogin to agree\n0\n2734\n\n\nThose of us reading this license, and with a sense of Freedom understand this means "libre". However, common (english) usage does not mean this. Since the term "Free" is not defined in this license, this phrase is ambiguous.\ngpl3.sourcecode.p3.s2\nkdean06\nlogin to agree\n1\n2797\n\n\nThe usage of this term should definitely be explained a bit better.\ngpl3.sourcecode.p3.s2\nmsikma\nlogin to agree\n2\n2938\n\n\nThis seems much too broad. Let's say I take a GPLv3 licensed GNU tar and modify it, so that it uses (and requires) some glibc-specific functionality, then certainly glibc is a "shared library...that the work is specifically designed to require". So I'd have to distribute the complete glibc source code with it, if I provided my modified tar in object code form. Is this intentional? There also should be a reiteration that this does not include system libraries. Otherwise this example seems to contradict the text that stands before it, because system libraries usually fit the description "shared libraries and dynamically linked subprograms that the work is specifically designed to require" and the statement earlier says that System Libraries are not part of the Corresponding Source.\n\ngpl3.sourcecode.p3.s3\nmux2005\nlogin to agree\n0\n3014\n\n\n<p style="display: inline">If some portions are not part of the work, those portions are not included in the Corresponding Source literally.</p><p>\nIs this a new copyleft provision?\n(Not modifying or linking with the GPL Program but just using in performing those activities causes transmission of the GPL?)</p><p>\nThis part should be re-written as described below.</p><p>\n", or any tools which are used in performing those activities unless those tools themselves are part of the work."</p><p>\nOf course, Makefiles and/or configure scripts are part of the work.</p>\ngpl3.sourcecode.p3.s2\nyusuke\nlogin to agree\n0\n3064\n\n\n", "2652": "\n\n\n\n<p style="display: inline">What if the re-generation requires a proprietary tool that is $100,000 a copy ?</p><p>\n"automatically from parts of the Corresponding Source using....'\n</p>\ngpl3.sourcecode.p4.s1\nAlanCox\nlogin to agree\n2\n2652\n\n\n", "2662": "\n\n\n\nWhat about works with no "source" such as photographic images or sound samples? As distinct from a bitmap of a structured vector image (where the latter is clearly the source) or an OGG file of a structured sound sequencer composition (where the latter is clearly the source).\ngpl3.sourcecode.p5.s1\nrobmyers\nlogin to agree\n0\n2662\n\n\n", "3094": "\n\n\n\n<p style="display: inline">This section recognises two kinds of use: use of the unmodified work, which is exempt from termination; and use of a modified work, which is not exempt. This seems like a strange distinction. A person whose license has been terminated is not permitted to use the work in certain ways within the privacy of their own home, yet a company whose license has been terminated is permitted to use the work (albiet unmodified only) on a public server, as a core part of a commercial venture.</p><p>\nIt seems to me that, if anything, the right to use the work on a public server should be subject to termination, as, like conveying the work, it effects others, and can be used to generate income, whereas the right to private use should be sanctified, as it does not effect others, and determining whether or not it is occuring requires an invasion of privacy.</p>\ngpl3.basicperms.0.0\njamesgnz\nlogin to agree\n0\n3094\n\n\n", "2758:3094": "\n\n\n\nI'm glad that the patent retaliation clause has been dropped: it was overreaching IMHO.\n\ngpl3.basicperms.0.0\nfrx\nlogin to agree\n0\n2758\n\n\n<p style="display: inline">This section recognises two kinds of use: use of the unmodified work, which is exempt from termination; and use of a modified work, which is not exempt. This seems like a strange distinction. A person whose license has been terminated is not permitted to use the work in certain ways within the privacy of their own home, yet a company whose license has been terminated is permitted to use the work (albiet unmodified only) on a public server, as a core part of a commercial venture.</p><p>\nIt seems to me that, if anything, the right to use the work on a public server should be subject to termination, as, like conveying the work, it effects others, and can be used to generate income, whereas the right to private use should be sanctified, as it does not effect others, and determining whether or not it is occuring requires an invasion of privacy.</p>\ngpl3.basicperms.0.0\njamesgnz\nlogin to agree\n0\n3094\n\n\n", "2690": "\n\n\n\n<p style="display: inline">Either your irrevocable or you're not. These terms aren't, since you define in several places conditions under which they may be revoked. Make it:</p><p>\nAll rights granted under this License are granted for the term of copyright on the Program, so long as the stated conditions are met.</p>\ngpl3.basicperms.p0.s1\nsepreece\nlogin to agree\n7\n2690\n\n\n", "2847": "\n\n\n\n<p style="display: inline">This seems to be poor wording in that it "affirms" something that may not be clear to the uninitiated. In law, "affirm" has the following meanings: \tLaw.\na. to state something solemnly before a court or magistrate, but without oath.\nb. to ratify and accept a voidable transaction.\nc. (of an appellate court) to determine that the action of the lower court shall stand.</p><p>\nI believe the use of the word "affirm" is unnecessary and confusing. The four corners of the document should speak clearly to its effect. Using the word "affirm" to me falls way short. Why not just a standard grant?</p>\ngpl3.basicperms.p0.s2\njkoenig\nlogin to agree\n2\n2847\n\n\n", "2847:2993": "\n\n\n\n<p style="display: inline">This seems to be poor wording in that it "affirms" something that may not be clear to the uninitiated. In law, "affirm" has the following meanings: \tLaw.\na. to state something solemnly before a court or magistrate, but without oath.\nb. to ratify and accept a voidable transaction.\nc. (of an appellate court) to determine that the action of the lower court shall stand.</p><p>\nI believe the use of the word "affirm" is unnecessary and confusing. The four corners of the document should speak clearly to its effect. Using the word "affirm" to me falls way short. Why not just a standard grant?</p>\ngpl3.basicperms.p0.s2\njkoenig\nlogin to agree\n2\n2847\n\n\nDoes "run" include compiling the program from source? Perhaps s/run the unmodified Program/use the unmodified Program/\ngpl3.basicperms.p0.s2\nmole\nlogin to agree\n1\n2993\n\n\n", "2757": "\n\n\n\nGood to see that fair use and similar rights are explicitly mentioned. This is an improvement over GPLv2.\ngpl3.basicperms.p0.s4\nfrx\nlogin to agree\n1\n2757\n\n\n", "2757:3041": "\n\n\n\nGood to see that fair use and similar rights are explicitly mentioned. This is an improvement over GPLv2.\ngpl3.basicperms.p0.s4\nfrx\nlogin to agree\n1\n2757\n\n\n<p style="display: inline">Fair use is a only one of several exceptions and limitations of copyright monopoly recognized by law. And it is least useful for software development among them. \n\nWhy only fair use mentioned? Why not first sale/exhaustion of rigths? Why not transient copies exceptions? Why not software backup exception?</p><p>\nGPLv3 should acknowledge ALL exceptions an limitation of copyright, not just fair use.</p>\ngpl3.basicperms.p0.s4\nlqp\nlogin to agree\n0\n3041\n\n\n", "2847:2993:3041": "\n\n\n\n<p style="display: inline">This seems to be poor wording in that it "affirms" something that may not be clear to the uninitiated. In law, "affirm" has the following meanings: \tLaw.\na. to state something solemnly before a court or magistrate, but without oath.\nb. to ratify and accept a voidable transaction.\nc. (of an appellate court) to determine that the action of the lower court shall stand.</p><p>\nI believe the use of the word "affirm" is unnecessary and confusing. The four corners of the document should speak clearly to its effect. Using the word "affirm" to me falls way short. Why not just a standard grant?</p>\ngpl3.basicperms.p0.s2\njkoenig\nlogin to agree\n2\n2847\n\n\nDoes "run" include compiling the program from source? Perhaps s/run the unmodified Program/use the unmodified Program/\ngpl3.basicperms.p0.s2\nmole\nlogin to agree\n1\n2993\n\n\n<p style="display: inline">Fair use is a only one of several exceptions and limitations of copyright monopoly recognized by law. And it is least useful for software development among them. \n\nWhy only fair use mentioned? Why not first sale/exhaustion of rigths? Why not transient copies exceptions? Why not software backup exception?</p><p>\nGPLv3 should acknowledge ALL exceptions an limitation of copyright, not just fair use.</p>\ngpl3.basicperms.p0.s4\nlqp\nlogin to agree\n0\n3041\n\n\n", "3036": "\n\n\n\nAs I've stated in another comment, the terms "propagation" and "convey" are essential to understanding the license. I think it would be a good idea to add an example here of something that is propagation but not conveyance (e.g. copying the program to your own computers). \ngpl3.basicperms.p1.s1\nmux2005\nlogin to agree\n0\n3036\n\n\n", "2744": "\n\n\n\nIIUC it's not true that making private modifications is permitted without conditions. There are actions, such as initiating patent litigation against covered works, that may lead to termination, and that thus serve as conditions for making private modifications.\ngpl3.basicperms.p1.s1\noliva\nlogin to agree\n0\n2744\n\n\n", "2709:2744": "\n\n\n\n<p style="display: inline">Personal use should not be entirely without conditions. There should at least be the section 15 disclaimer (and possibly additional section 7a disclaimers).</p><p>\nThe clause "so long as your license otherwise remains in force" also looks suspiciously like a condition. (Is this stating that section 8 termination applies here? It's not clear to me.)</p><p>\nI suggest truncating this sentance, so it simply ends with "are permitted."</p>\ngpl3.basicperms.p1.s1\njamesgnz\nlogin to agree\n1\n2709\n\n\nIIUC it's not true that making private modifications is permitted without conditions. There are actions, such as initiating patent litigation against covered works, that may lead to termination, and that thus serve as conditions for making private modifications.\ngpl3.basicperms.p1.s1\noliva\nlogin to agree\n0\n2744\n\n\n", "2709:3037": "\n\n\n\n<p style="display: inline">Personal use should not be entirely without conditions. There should at least be the section 15 disclaimer (and possibly additional section 7a disclaimers).</p><p>\nThe clause "so long as your license otherwise remains in force" also looks suspiciously like a condition. (Is this stating that section 8 termination applies here? It's not clear to me.)</p><p>\nI suggest truncating this sentance, so it simply ends with "are permitted."</p>\ngpl3.basicperms.p1.s1\njamesgnz\nlogin to agree\n1\n2709\n\n\n<p style="display: inline">Rather than a vague hint at the possibility of termination, you should clearly crossreference the sections about termination, i.e. replace "so long as your license otherwise remains in force" with\n"so long as your license is not terminated according to section bla, section bla paragraph bla or section bla paragraphs bla and bla."</p><p>\nYou should not require that people read the license a dozen times while taking crossreferencing notes, because people (people != lawyers ;-) don't do that. A lot of people (unfortunately including many if not most non-corporate licensors and licensees) never read the GPL in their entirety. They instead skim the text and read selectively. Parts such as this that say "somewhere in this license there's something very important that you should know, but I'll let you find it on your own" are not friendly to those people.</p>\ngpl3.basicperms.p1.s1\nmux2005\nlogin to agree\n0\n3037\n\n\n", "2709:2744:3036:3037": "\n\n\n\n<p style="display: inline">Personal use should not be entirely without conditions. There should at least be the section 15 disclaimer (and possibly additional section 7a disclaimers).</p><p>\nThe clause "so long as your license otherwise remains in force" also looks suspiciously like a condition. (Is this stating that section 8 termination applies here? It's not clear to me.)</p><p>\nI suggest truncating this sentance, so it simply ends with "are permitted."</p>\ngpl3.basicperms.p1.s1\njamesgnz\nlogin to agree\n1\n2709\n\n\nIIUC it's not true that making private modifications is permitted without conditions. There are actions, such as initiating patent litigation against covered works, that may lead to termination, and that thus serve as conditions for making private modifications.\ngpl3.basicperms.p1.s1\noliva\nlogin to agree\n0\n2744\n\n\nAs I've stated in another comment, the terms "propagation" and "convey" are essential to understanding the license. I think it would be a good idea to add an example here of something that is propagation but not conveyance (e.g. copying the program to your own computers). \ngpl3.basicperms.p1.s1\nmux2005\nlogin to agree\n0\n3036\n\n\n<p style="display: inline">Rather than a vague hint at the possibility of termination, you should clearly crossreference the sections about termination, i.e. replace "so long as your license otherwise remains in force" with\n"so long as your license is not terminated according to section bla, section bla paragraph bla or section bla paragraphs bla and bla."</p><p>\nYou should not require that people read the license a dozen times while taking crossreferencing notes, because people (people != lawyers ;-) don't do that. A lot of people (unfortunately including many if not most non-corporate licensors and licensees) never read the GPL in their entirety. They instead skim the text and read selectively. Parts such as this that say "somewhere in this license there's something very important that you should know, but I'll let you find it on your own" are not friendly to those people.</p>\ngpl3.basicperms.p1.s1\nmux2005\nlogin to agree\n0\n3037\n\n\n", "2928": "\n\n\n\n<p style="display: inline">As I understand it the aim of this section is to prevent content from being locked up with technical protection measures thus potentially inhibiting access to content that should be legally or morally available to the user. I am in favour of this philosophy in principle.</p><p>\nHowever I worry that this section, as it is worded, would not allow the covered work to be used in applications where only rights information is being stored. Rights information that might justly need technical protection form unauthorized modification or abuse. For example, an application and database that stored author information and identification details of the works they have authored might well need technical protection. This technical measure has no impact on someone accessing the content itself but it serves an important social and copyright function of protecting the details and the rights of the author and other metadata about the content of the work. Waiving all legal protection for this type of rights management seems unnecessary and even counter productive.</p><p>\nBy all means protect user rights but lets protect author rights at the same time.</p>\ngpl3.drm.0.0\nnickb\nlogin to agree\n0\n2928\n\n\n", "2557:2661:2759": "\n\n\n\nAlthough it is much better (for a Dane) to quote an international law than a US-only law, I intensely dislike a reference to *any* specific law. (The reference to Magnuson-Moss has the same problem, BTW).\n\n\nIt weakens the license in my view and smells of defeat. The GPL is not merely a legal contract, it is a statement of intent.\n\n\nIf you want me to choose the GPL, you better leave me with the impression that I understand it. I understand there are nuances I do not understand (or are even unknown until established in court), but referencing a specific law just smack it in my face.\n\n\nSure I could find the law, but I probably could not understand it and feel I should be able to specify what people can do with my creation.\n\n\nWhat does this paragraph accomplish that the second paragraph does not cover? Or are there better ways?\ngpl3.drm.p0.s1\nhawk\nlogin to agree\n7\n2557\n\n\nThis is much better than the previous US/EU language. Thanks!\ngpl3.drm.p0.s1\nrobmyers\nlogin to agree\n0\n2661\n\n\nThis clause is improved (being now denationalized), but still problematic. It could be seen as an untrue statement in some cases. How can the licensor say that the covered work won't be judged as "part of an effective technological measure" under a given law? That is for the courts to decide. In some scenarios, GnuPG may actually be considered "part of an effective technological measure" and could be deemed so by a judge...\ngpl3.drm.p0.s1\nfrx\nlogin to agree\n2\n2759\n\n\n", "2557:2661:2759:2798:2968": "\n\n\n\nAlthough it is much better (for a Dane) to quote an international law than a US-only law, I intensely dislike a reference to *any* specific law. (The reference to Magnuson-Moss has the same problem, BTW).\n\n\nIt weakens the license in my view and smells of defeat. The GPL is not merely a legal contract, it is a statement of intent.\n\n\nIf you want me to choose the GPL, you better leave me with the impression that I understand it. I understand there are nuances I do not understand (or are even unknown until established in court), but referencing a specific law just smack it in my face.\n\n\nSure I could find the law, but I probably could not understand it and feel I should be able to specify what people can do with my creation.\n\n\nWhat does this paragraph accomplish that the second paragraph does not cover? Or are there better ways?\ngpl3.drm.p0.s1\nhawk\nlogin to agree\n7\n2557\n\n\nThis is much better than the previous US/EU language. Thanks!\ngpl3.drm.p0.s1\nrobmyers\nlogin to agree\n0\n2661\n\n\nThis clause is improved (being now denationalized), but still problematic. It could be seen as an untrue statement in some cases. How can the licensor say that the covered work won't be judged as "part of an effective technological measure" under a given law? That is for the courts to decide. In some scenarios, GnuPG may actually be considered "part of an effective technological measure" and could be deemed so by a judge...\ngpl3.drm.p0.s1\nfrx\nlogin to agree\n2\n2759\n\n\nLike one poster above, I feel that referencing a specific law inside of a license is problematic. \n\nFor instance, stating the obligation of such laws may also inadvertently pull in judicial opinions of such laws, invalidating or conflicting with the intent of the license. \ngpl3.drm.p0.s1\nkdean06\nlogin to agree\n0\n2798\n\n\n<p style="display: inline">The text of article 11 of the copyright treaty is ambiguous so this license should define its interpretation of the treaty before it refers to it. One interpretation (1.) would prevent a covered work from being used to protect rights metadata which I don't think is a good idea or necessary.</p><p>\nArticle 11 reads:</p><p>\n\u00c3\u00a2\u00c2\u0080\u00c2\u009cContracting Parties shall provide adequate legal protection and effective legal remedies against the circumvention of effective technological measures that are used by authors in connection with the exercise of their rights under this Treaty or the Berne Convention [AND] that restrict acts, in respect of their works, which are not authorized by the authors concerned or permitted by law.\u00c3\u00a2\u00c2\u0080\u00c2\u009d\n \nInterpretation 1.</p><p>\n\u00c3\u00a2\u00c2\u0080\u00c2\u009cContracting Parties shall provide adequate legal protection and effective legal remedies against the circumvention of effective technological measures that are used by authors in connection with the exercise of their rights which are not authorized by the authors concerned or permitted by law. [AND], Contracting Parties shall provide adequate legal protection and effective legal remedies against the circumvention of effective technological measures that are used by authors that restrict acts, in respect of their works, which are not authorized by the authors concerned or permitted by law.\u00c3\u00a2\u00c2\u0080\u00c2\u009d</p><p>\nInterpretation 2.</p><p>\n\u00c3\u00a2\u00c2\u0080\u00c2\u009cContracting Parties shall provide adequate legal protection and effective legal remedies against the circumvention of effective technological measures that are used by authors in connection with the exercise of their rights under this Treaty or the Berne Convention [IN THAT THEY] restrict acts, in respect of their works, which are not authorized by the authors concerned or permitted by law.\u00c3\u00a2\u00c2\u0080\u00c2\u009d \n\nInterpretation 3.</p><p>\nContracting Parties shall provide adequate legal protection and effective legal remedies against the circumvention of effective technological measures that are used by authors in connection with the exercise of their rights under this Treaty or the Berne Convention [BUT ONLY IN CASES WHERE THE EXERCISE OF THEIR RIGHTS] restrict acts in respect of their works which are not authorized by the authors concerned or permitted by law. </p>\ngpl3.drm.p0.s1\nnickb\nlogin to agree\n1\n2968\n\n\n", "2575:2735": "\n\n\n\n<p style="display: inline">Author releases with no DRM. Licensee tier #1 copies with no DRM. Licensee tier #2 slaps on some DRM. Licensee tier #3 has to crack the DRM to make further copies, but at least it is legal.</p><p>\nBut, a GPL program should be able to process copy-limiting DRM-ed data although the open-source requirement means it will likely be cracked. A GPL OS should be able to run closed-source programs that process copy-limiting DRM-ed data (DRM-ed AAC tunes through Itunes on Linux) also.</p><p>\nA closed-source driver running under a GPL OS should be able to process DRM data, also, as long as it doesn't have to be compiled into the kernal (which is how linux works, but Windoze doesn't do drivers this way).</p><p>\nA GPL program should NEVER have copy-limiting DRM directly on itself though.</p>\ngpl3.drm.p1.s1\ninca\nlogin to agree\n0\n2575\n\n\n<p style="display: inline">What if a work is installed on an encrypted file system? I will be in violation of the license, since I am asserting my legal right to the protection of my data, even though the GPL'd program never even factored into this.</p><p>\nOr, perhaps the phrase "legal rights" needs to be clarified.</p>\ngpl3.drm.p1.s1\nmkorman\nlogin to agree\n2\n2735\n\n\n", "2475:2575:2735:2819": "\n\n\n\nThe covered work, to remain consistent\n\ngpl3.drm.p1.s1\nmerijn\nlogin to agree\n1\n2475\n\n\n<p style="display: inline">Author releases with no DRM. Licensee tier #1 copies with no DRM. Licensee tier #2 slaps on some DRM. Licensee tier #3 has to crack the DRM to make further copies, but at least it is legal.</p><p>\nBut, a GPL program should be able to process copy-limiting DRM-ed data although the open-source requirement means it will likely be cracked. A GPL OS should be able to run closed-source programs that process copy-limiting DRM-ed data (DRM-ed AAC tunes through Itunes on Linux) also.</p><p>\nA closed-source driver running under a GPL OS should be able to process DRM data, also, as long as it doesn't have to be compiled into the kernal (which is how linux works, but Windoze doesn't do drivers this way).</p><p>\nA GPL program should NEVER have copy-limiting DRM directly on itself though.</p>\ngpl3.drm.p1.s1\ninca\nlogin to agree\n0\n2575\n\n\n<p style="display: inline">What if a work is installed on an encrypted file system? I will be in violation of the license, since I am asserting my legal right to the protection of my data, even though the GPL'd program never even factored into this.</p><p>\nOr, perhaps the phrase "legal rights" needs to be clarified.</p>\ngpl3.drm.p1.s1\nmkorman\nlogin to agree\n2\n2735\n\n\n<p style="display: inline">Quoting part of section 3:</p><p>\n"you waive any legal power to forbid circumvention of technical measures to the extent such circumvention is effected by exercising rights under this License with respect to the covered work," \n\nAside from a clause that makes every licensee agree to "disclaim any intention to limit operation or modification of the work as a means of enforcing, against the work's users, your or third parties' legal rights to forbid circumvention of technical measures," there is nothing else to actually STOP a person from actually limiting the exercise of rights of end-users under this license, through the actual inclusion of technical measures.</p><p>\nI don't feel that merely making a person agree not forbid circumvention of these measures, or to merely agree not have the intention "to limit operation or modification" of the program is adequate. Why merely limit the power to forbid circumvention of a technical measures? Why not ban any technical measures, altogether, that compromise rights under this license. I recommend adding this clause:</p><p>\n"You agree that you will not include, within any covered work you convey, any technical measures, which are clearly designed to limit, in whole, or in part, the exercise of any right under this license, with respect to the covered work."</p><p>\nI have a written a few programs of my own, and at present, they are licensed under the GNU GPL, version 2. I plan on re-licensing them under version 3, as soon as the official version of the license is published. I DO NOT believe things like product activation, copy or usage limiting or disabling features, or other forms of digital restrictions management should be in any software. I certainly don't want it in my programs and it seems to go against the mission of the FSF and free software as well. Therefore, I strongly recommend a provision that doesn't just prevent circumvention of these types of abusive technical measures, but that forbids, outright, any inclusion of them, into any covered work.</p>\ngpl3.drm.p1.s1\nthefuchian\nlogin to agree\n0\n2819\n\n\n", "2575:2735:2760:2819:2823:3038:3042": "\n\n\n\n<p style="display: inline">Author releases with no DRM. Licensee tier #1 copies with no DRM. Licensee tier #2 slaps on some DRM. Licensee tier #3 has to crack the DRM to make further copies, but at least it is legal.</p><p>\nBut, a GPL program should be able to process copy-limiting DRM-ed data although the open-source requirement means it will likely be cracked. A GPL OS should be able to run closed-source programs that process copy-limiting DRM-ed data (DRM-ed AAC tunes through Itunes on Linux) also.</p><p>\nA closed-source driver running under a GPL OS should be able to process DRM data, also, as long as it doesn't have to be compiled into the kernal (which is how linux works, but Windoze doesn't do drivers this way).</p><p>\nA GPL program should NEVER have copy-limiting DRM directly on itself though.</p>\ngpl3.drm.p1.s1\ninca\nlogin to agree\n0\n2575\n\n\n<p style="display: inline">What if a work is installed on an encrypted file system? I will be in violation of the license, since I am asserting my legal right to the protection of my data, even though the GPL'd program never even factored into this.</p><p>\nOr, perhaps the phrase "legal rights" needs to be clarified.</p>\ngpl3.drm.p1.s1\nmkorman\nlogin to agree\n2\n2735\n\n\n<p style="display: inline">This clause is clearer than in the previous draft, but still troublesome, as it seems to be overreaching. For instance, it could be interpreted as covering legal powers to forbid "computer crimes" such as unauthorized intrusion into computer systems.</p><p>\nE.g.: suppose that the covered work is a vulnerability scanner, or password cracker, or anyway a tool that could be used (among other things) to break into other people's computers. Using that tool in this manner is exercising a right "under this License" and is a circumvention of appropriate technical measures set to protect a computer system or network from unauthorized access. Gaining unauthorized access to a protected computer system or network is forbidden by law in several jurisdictions; do I waive such a legal protection, when I convey the covered work?</p><p>\nI suggest dropping the waiver entirely, thus leaving the following disclaimer only.</p>\ngpl3.drm.p1.s1\nfrx\nlogin to agree\n2\n2760\n\n\n<p style="display: inline">Quoting part of section 3:</p><p>\n"you waive any legal power to forbid circumvention of technical measures to the extent such circumvention is effected by exercising rights under this License with respect to the covered work," \n\nAside from a clause that makes every licensee agree to "disclaim any intention to limit operation or modification of the work as a means of enforcing, against the work's users, your or third parties' legal rights to forbid circumvention of technical measures," there is nothing else to actually STOP a person from actually limiting the exercise of rights of end-users under this license, through the actual inclusion of technical measures.</p><p>\nI don't feel that merely making a person agree not forbid circumvention of these measures, or to merely agree not have the intention "to limit operation or modification" of the program is adequate. Why merely limit the power to forbid circumvention of a technical measures? Why not ban any technical measures, altogether, that compromise rights under this license. I recommend adding this clause:</p><p>\n"You agree that you will not include, within any covered work you convey, any technical measures, which are clearly designed to limit, in whole, or in part, the exercise of any right under this license, with respect to the covered work."</p><p>\nI have a written a few programs of my own, and at present, they are licensed under the GNU GPL, version 2. I plan on re-licensing them under version 3, as soon as the official version of the license is published. I DO NOT believe things like product activation, copy or usage limiting or disabling features, or other forms of digital restrictions management should be in any software. I certainly don't want it in my programs and it seems to go against the mission of the FSF and free software as well. Therefore, I strongly recommend a provision that doesn't just prevent circumvention of these types of abusive technical measures, but that forbids, outright, any inclusion of them, into any covered work.</p>\ngpl3.drm.p1.s1\nthefuchian\nlogin to agree\n0\n2819\n\n\n<p style="display: inline">I apologize, there was a mistake in the very last sentence of my last post. In that sentence, I actually meant to say:</p><p>\nI strongly recommend a provision that doesn't just forbid licensees from preventing other people from circumventing these types of abusive technical measures, but one that also forbids, outright, ANY inclusion of such abusive technical measures, into any covered work.</p>\ngpl3.drm.p1.s1\nthefuchian\nlogin to agree\n0\n2823\n\n\nAs I wrote in a followup to a comment on the previous paragraph about the WIPO treaty, this section seems like I would waive the power to enforce even those anti-circumvention laws that are legitimate and morally accetable, such as laws forbidding the forging of authorship information. The right of the author to get proper attribution is recognized even by the GPLv3. Laws that forbid the circumvention of systems that protect authorship information are not in conflict with free software and waiving the power to enforce these laws against people who take credit for someone else's work is not acceptable.\ngpl3.drm.p1.s1\nmux2005\nlogin to agree\n0\n3038\n\n\nAdd a note that you can still exercise your legal powers under copyright law.\ngpl3.drm.p1.s1\nyuhong\nlogin to agree\n0\n3042\n\n\n", "2575:2735:3039": "\n\n\n\n<p style="display: inline">Author releases with no DRM. Licensee tier #1 copies with no DRM. Licensee tier #2 slaps on some DRM. Licensee tier #3 has to crack the DRM to make further copies, but at least it is legal.</p><p>\nBut, a GPL program should be able to process copy-limiting DRM-ed data although the open-source requirement means it will likely be cracked. A GPL OS should be able to run closed-source programs that process copy-limiting DRM-ed data (DRM-ed AAC tunes through Itunes on Linux) also.</p><p>\nA closed-source driver running under a GPL OS should be able to process DRM data, also, as long as it doesn't have to be compiled into the kernal (which is how linux works, but Windoze doesn't do drivers this way).</p><p>\nA GPL program should NEVER have copy-limiting DRM directly on itself though.</p>\ngpl3.drm.p1.s1\ninca\nlogin to agree\n0\n2575\n\n\n<p style="display: inline">What if a work is installed on an encrypted file system? I will be in violation of the license, since I am asserting my legal right to the protection of my data, even though the GPL'd program never even factored into this.</p><p>\nOr, perhaps the phrase "legal rights" needs to be clarified.</p>\ngpl3.drm.p1.s1\nmkorman\nlogin to agree\n2\n2735\n\n\nMake a period here and start a new sentence. Those overlong lawyer-speak sentences do not improve the GPL. The same applies to many other parts of the license. Wherever you can, split long sentences into multiple sentences. Yes, the result doesn't sound as "professional" anymore, but that's a GOOD THING. Non-native speakers have trouble enough with English legalese (in fact, native speakers do, too). Don't make it harder than it needs to be.\ngpl3.drm.p1.s1\nmux2005\nlogin to agree\n0\n3039\n\n\n", "3133": "\n\n\n\nNetwork Security Bugs have a way of causing corporate distributors of software to want to restrict the speech of those who discover vulnerabilities in software. This practice is intended to give distributors ample time to create and distribute a "fix" for the software before software can be written to exploit the vulnerability. I feel that users must agree to not restrict the speech of other users of the program relating to vulnerabilities in the code. \n\n-Joe Baker\nDigital Communications Research, Inc.\nBurlington, Wisconsin\ngpl3.drm.p1.s1\njoebaker\nlogin to agree\n0\n3133\n\n\n", "2475:2575:2735:2760:2819:2823:3038:3039:3042:3133": "\n\n\n\nThe covered work, to remain consistent\n\ngpl3.drm.p1.s1\nmerijn\nlogin to agree\n1\n2475\n\n\n<p style="display: inline">Author releases with no DRM. Licensee tier #1 copies with no DRM. Licensee tier #2 slaps on some DRM. Licensee tier #3 has to crack the DRM to make further copies, but at least it is legal.</p><p>\nBut, a GPL program should be able to process copy-limiting DRM-ed data although the open-source requirement means it will likely be cracked. A GPL OS should be able to run closed-source programs that process copy-limiting DRM-ed data (DRM-ed AAC tunes through Itunes on Linux) also.</p><p>\nA closed-source driver running under a GPL OS should be able to process DRM data, also, as long as it doesn't have to be compiled into the kernal (which is how linux works, but Windoze doesn't do drivers this way).</p><p>\nA GPL program should NEVER have copy-limiting DRM directly on itself though.</p>\ngpl3.drm.p1.s1\ninca\nlogin to agree\n0\n2575\n\n\n<p style="display: inline">What if a work is installed on an encrypted file system? I will be in violation of the license, since I am asserting my legal right to the protection of my data, even though the GPL'd program never even factored into this.</p><p>\nOr, perhaps the phrase "legal rights" needs to be clarified.</p>\ngpl3.drm.p1.s1\nmkorman\nlogin to agree\n2\n2735\n\n\n<p style="display: inline">This clause is clearer than in the previous draft, but still troublesome, as it seems to be overreaching. For instance, it could be interpreted as covering legal powers to forbid "computer crimes" such as unauthorized intrusion into computer systems.</p><p>\nE.g.: suppose that the covered work is a vulnerability scanner, or password cracker, or anyway a tool that could be used (among other things) to break into other people's computers. Using that tool in this manner is exercising a right "under this License" and is a circumvention of appropriate technical measures set to protect a computer system or network from unauthorized access. Gaining unauthorized access to a protected computer system or network is forbidden by law in several jurisdictions; do I waive such a legal protection, when I convey the covered work?</p><p>\nI suggest dropping the waiver entirely, thus leaving the following disclaimer only.</p>\ngpl3.drm.p1.s1\nfrx\nlogin to agree\n2\n2760\n\n\n<p style="display: inline">Quoting part of section 3:</p><p>\n"you waive any legal power to forbid circumvention of technical measures to the extent such circumvention is effected by exercising rights under this License with respect to the covered work," \n\nAside from a clause that makes every licensee agree to "disclaim any intention to limit operation or modification of the work as a means of enforcing, against the work's users, your or third parties' legal rights to forbid circumvention of technical measures," there is nothing else to actually STOP a person from actually limiting the exercise of rights of end-users under this license, through the actual inclusion of technical measures.</p><p>\nI don't feel that merely making a person agree not forbid circumvention of these measures, or to merely agree not have the intention "to limit operation or modification" of the program is adequate. Why merely limit the power to forbid circumvention of a technical measures? Why not ban any technical measures, altogether, that compromise rights under this license. I recommend adding this clause:</p><p>\n"You agree that you will not include, within any covered work you convey, any technical measures, which are clearly designed to limit, in whole, or in part, the exercise of any right under this license, with respect to the covered work."</p><p>\nI have a written a few programs of my own, and at present, they are licensed under the GNU GPL, version 2. I plan on re-licensing them under version 3, as soon as the official version of the license is published. I DO NOT believe things like product activation, copy or usage limiting or disabling features, or other forms of digital restrictions management should be in any software. I certainly don't want it in my programs and it seems to go against the mission of the FSF and free software as well. Therefore, I strongly recommend a provision that doesn't just prevent circumvention of these types of abusive technical measures, but that forbids, outright, any inclusion of them, into any covered work.</p>\ngpl3.drm.p1.s1\nthefuchian\nlogin to agree\n0\n2819\n\n\n<p style="display: inline">I apologize, there was a mistake in the very last sentence of my last post. In that sentence, I actually meant to say:</p><p>\nI strongly recommend a provision that doesn't just forbid licensees from preventing other people from circumventing these types of abusive technical measures, but one that also forbids, outright, ANY inclusion of such abusive technical measures, into any covered work.</p>\ngpl3.drm.p1.s1\nthefuchian\nlogin to agree\n0\n2823\n\n\nAs I wrote in a followup to a comment on the previous paragraph about the WIPO treaty, this section seems like I would waive the power to enforce even those anti-circumvention laws that are legitimate and morally accetable, such as laws forbidding the forging of authorship information. The right of the author to get proper attribution is recognized even by the GPLv3. Laws that forbid the circumvention of systems that protect authorship information are not in conflict with free software and waiving the power to enforce these laws against people who take credit for someone else's work is not acceptable.\ngpl3.drm.p1.s1\nmux2005\nlogin to agree\n0\n3038\n\n\nMake a period here and start a new sentence. Those overlong lawyer-speak sentences do not improve the GPL. The same applies to many other parts of the license. Wherever you can, split long sentences into multiple sentences. Yes, the result doesn't sound as "professional" anymore, but that's a GOOD THING. Non-native speakers have trouble enough with English legalese (in fact, native speakers do, too). Don't make it harder than it needs to be.\ngpl3.drm.p1.s1\nmux2005\nlogin to agree\n0\n3039\n\n\nAdd a note that you can still exercise your legal powers under copyright law.\ngpl3.drm.p1.s1\nyuhong\nlogin to agree\n0\n3042\n\n\nNetwork Security Bugs have a way of causing corporate distributors of software to want to restrict the speech of those who discover vulnerabilities in software. This practice is intended to give distributors ample time to create and distribute a "fix" for the software before software can be written to exploit the vulnerability. I feel that users must agree to not restrict the speech of other users of the program relating to vulnerabilities in the code. \n\n-Joe Baker\nDigital Communications Research, Inc.\nBurlington, Wisconsin\ngpl3.drm.p1.s1\njoebaker\nlogin to agree\n0\n3133\n\n\n", "2761:2978": "\n\n\n\nThis section seems to be reworded for clarity, but not much changed in its meaning with respect to previous drafts. It's not that dissimilar to the corresponding section in GPLv2 (apart from the references to section 7) and looks OK to me.\ngpl3.verbatimcopying.0.0\nfrx\nlogin to agree\n0\n2761\n\n\n<p style="display: inline">Sections 4 and 5 both apply to source code, and section 6 to object code. However, the title of 4 does not say so (whereas 5 and 6 do) and the title of 6 does not use the pre-defined term "object code". I suggest the following set of consistent and progressive titles:\n\n\n* 4] Conveying Verbatim Source Code</p><p>\n* 5] Conveying Modified Source Code</p><p>\n* 6] Conveying Object Code\n\n</p>\ngpl3.verbatimcopying.0.0\ngerv\nlogin to agree\n0\n2978\n\n\n", "2848": "\n\n\n\nI do not like the word "verbatim copies" since it assumes text (or software), a scope which is less than the intended scope of the license. My alternative word choice would be "clones".\ngpl3.verbatimcopying.p0.s1\njkoenig\nlogin to agree\n0\n2848\n\n\n", "2633": "\n\n\n\nUse mixed case for defined term 'Source Code'\ngpl3.verbatimcopying.p0.s1\nsamj\nlogin to agree\n0\n2633\n\n\n", "2736": "\n\n\n\nWhat if I wish to _provide_ a warranty for my recipients? In this case, it would make no sense to leave others' warranty disclaimers intact.\ngpl3.verbatimcopying.p0.s1\nmkorman\nlogin to agree\n0\n2736\n\n\n", "2633:2736:2848": "\n\n\n\nUse mixed case for defined term 'Source Code'\ngpl3.verbatimcopying.p0.s1\nsamj\nlogin to agree\n0\n2633\n\n\nWhat if I wish to _provide_ a warranty for my recipients? In this case, it would make no sense to leave others' warranty disclaimers intact.\ngpl3.verbatimcopying.p0.s1\nmkorman\nlogin to agree\n0\n2736\n\n\nI do not like the word "verbatim copies" since it assumes text (or software), a scope which is less than the intended scope of the license. My alternative word choice would be "clones".\ngpl3.verbatimcopying.p0.s1\njkoenig\nlogin to agree\n0\n2848\n\n\n", "2849:2872": "\n\n\n\nA better wording of this sentence would be: "You may charge any fee or no fee for each copy that you convey, and you may additionally offer for example, support, warranty or indemnification for a fee, or no fee.\ngpl3.verbatimcopying.p1.s1\njkoenig\nlogin to agree\n1\n2849\n\n\n-> "You may charge or elect not to charge a fee for each copy that you convey, and you may offer or elect not to offer support or warranty protection for a fee."\ngpl3.verbatimcopying.p1.s1\nskquinn\nlogin to agree\n0\n2872\n\n\n", "2737:2849:2872": "\n\n\n\nThe phrase "no price" is implied by "any price".\ngpl3.verbatimcopying.p1.s1\nmkorman\nlogin to agree\n1\n2737\n\n\nA better wording of this sentence would be: "You may charge any fee or no fee for each copy that you convey, and you may additionally offer for example, support, warranty or indemnification for a fee, or no fee.\ngpl3.verbatimcopying.p1.s1\njkoenig\nlogin to agree\n1\n2849\n\n\n-> "You may charge or elect not to charge a fee for each copy that you convey, and you may offer or elect not to offer support or warranty protection for a fee."\ngpl3.verbatimcopying.p1.s1\nskquinn\nlogin to agree\n0\n2872\n\n\n", "2625:2849:2872": "\n\n\n\nThe license should not dictate whether or not I do this 'for a fee', though perhaps 'for consideration' is relevant/necessary.\ngpl3.verbatimcopying.p1.s1\nsamj\nlogin to agree\n4\n2625\n\n\nA better wording of this sentence would be: "You may charge any fee or no fee for each copy that you convey, and you may additionally offer for example, support, warranty or indemnification for a fee, or no fee.\ngpl3.verbatimcopying.p1.s1\njkoenig\nlogin to agree\n1\n2849\n\n\n-> "You may charge or elect not to charge a fee for each copy that you convey, and you may offer or elect not to offer support or warranty protection for a fee."\ngpl3.verbatimcopying.p1.s1\nskquinn\nlogin to agree\n0\n2872\n\n\n", "2625:2737:2849:2872": "\n\n\n\nThe license should not dictate whether or not I do this 'for a fee', though perhaps 'for consideration' is relevant/necessary.\ngpl3.verbatimcopying.p1.s1\nsamj\nlogin to agree\n4\n2625\n\n\nThe phrase "no price" is implied by "any price".\ngpl3.verbatimcopying.p1.s1\nmkorman\nlogin to agree\n1\n2737\n\n\nA better wording of this sentence would be: "You may charge any fee or no fee for each copy that you convey, and you may additionally offer for example, support, warranty or indemnification for a fee, or no fee.\ngpl3.verbatimcopying.p1.s1\njkoenig\nlogin to agree\n1\n2849\n\n\n-> "You may charge or elect not to charge a fee for each copy that you convey, and you may offer or elect not to offer support or warranty protection for a fee."\ngpl3.verbatimcopying.p1.s1\nskquinn\nlogin to agree\n0\n2872\n\n\n", "3022": "\n\n\n\n<p style="display: inline">\nIn some countries, the author of a creative work can be legally entitled to certain moral rights. These moral rights can be separate from any copyright for a creative work. For more information, see http://en.wikipedia.org/wiki/Moral_rights (checked on April 29, 2007.) In particular, the article mentions the moral right for an author to have the integrity of a work preserved. It appears that this right can disallow others from altering a work if the alterations would hurt the honor and/or reputation of the author. When a work legally qualifies for moral rights, do those rights still apply if the work is licensed under the GPL? For such a work, would it be permissible for others to do something such as modifying the work in a manner that dishonors the author? Moral rights have come up in the case of the Creative Commons licenses. See http://wiki.creativecommons.org/Version_3#International_Harmonization_.E2.80.93_Moral_Rights (checked on May 6, 2007) for more information. In general, the Creative Commons policy was to preserve any moral right of integrity for the author of a work. An issue did come up with respect to the jurisdiction of Japan. The Japanese moral right of integrity was supposedly of such a nature that it could end up disallowing any alterations to a work. The solution was to limit the moral right of integrity for jurisdictions where it could completely disallow modifications to CC-licensed works. The unported (not specific to a certain jurisdiction) version of the Creative Commons Attribution-ShareAlike 3.0 license has a clause along these lines in its legal code. See section 4(d) at http://creativecommons.org/licenses/by-sa/3.0/legalcode for details. It might be useful for the GPL to do something similar.</p><p>\nThe Wikipedia article also mentions "artist resale rights" which may affect the resale of covered works, but it is not clear if such rights apply to the sale of a modified work or to the sale of a work that incorporates part of another work.</p><p>\nThe Creative Commons page regarding moral rights (http://wiki.creativecommons.org/Version_3#International_Harmonization_.E2.80.93_Moral_Rights) mentions the moral right of attribution. Newer versions of the Creative Commons licenses can require the attribution of authors. In the case of the GPLv3, item b of Section 7 (which can require the attribution of authors) may help to avoid problems with any moral right of attribution.\n\n</p>\ngpl3.distribmod.p0.s1\nmezzanine1\nlogin to agree\n0\n3022\n\n\n", "2477:3022": "\n\n\n\n\ngpl3.distribmod.p0.s1\nmerijn\nlogin to agree\n0\n2477\n\n\n<p style="display: inline">\nIn some countries, the author of a creative work can be legally entitled to certain moral rights. These moral rights can be separate from any copyright for a creative work. For more information, see http://en.wikipedia.org/wiki/Moral_rights (checked on April 29, 2007.) In particular, the article mentions the moral right for an author to have the integrity of a work preserved. It appears that this right can disallow others from altering a work if the alterations would hurt the honor and/or reputation of the author. When a work legally qualifies for moral rights, do those rights still apply if the work is licensed under the GPL? For such a work, would it be permissible for others to do something such as modifying the work in a manner that dishonors the author? Moral rights have come up in the case of the Creative Commons licenses. See http://wiki.creativecommons.org/Version_3#International_Harmonization_.E2.80.93_Moral_Rights (checked on May 6, 2007) for more information. In general, the Creative Commons policy was to preserve any moral right of integrity for the author of a work. An issue did come up with respect to the jurisdiction of Japan. The Japanese moral right of integrity was supposedly of such a nature that it could end up disallowing any alterations to a work. The solution was to limit the moral right of integrity for jurisdictions where it could completely disallow modifications to CC-licensed works. The unported (not specific to a certain jurisdiction) version of the Creative Commons Attribution-ShareAlike 3.0 license has a clause along these lines in its legal code. See section 4(d) at http://creativecommons.org/licenses/by-sa/3.0/legalcode for details. It might be useful for the GPL to do something similar.</p><p>\nThe Wikipedia article also mentions "artist resale rights" which may affect the resale of covered works, but it is not clear if such rights apply to the sale of a modified work or to the sale of a work that incorporates part of another work.</p><p>\nThe Creative Commons page regarding moral rights (http://wiki.creativecommons.org/Version_3#International_Harmonization_.E2.80.93_Moral_Rights) mentions the moral right of attribution. Newer versions of the Creative Commons licenses can require the attribution of authors. In the case of the GPLv3, item b of Section 7 (which can require the attribution of authors) may help to avoid problems with any moral right of attribution.\n\n</p>\ngpl3.distribmod.p0.s1\nmezzanine1\nlogin to agree\n0\n3022\n\n\n", "3043": "\n\n\n\n<p style="display: inline">This term is undefined and IMHO is not self-explaining. Does this include diff-patches for instance? <br><br></p><p>\nI think a "modification" at least in the ordinary sense (that a non-technie judge/lawyer might apply) is not something tangible that can be conveyed. You may convey instructions to perform a modification, programs that perform a modification, etc. but a "modification" is something abstract that cannot be conveyed. Programmers might call a diff-file a "modification", but more correctly a diff is a description of a modification that can be used by the patch program to perform said modification.<br><br></p><p>\nFurthermore I'm not sure how this interacts with the definition of source code as the preferred form for making modifications. This sentence requires that modifications be in the form of source code, so we're talking about the preferred form of making modifications to a modification. Does that make sense? <br><br></p><p>\nAnd is a diff the preferred form for making modifications to a modification? I don't think so. The preferred form is the modified source code that can be generated from applying the diff. Then, I'm not sure if the "prominent notices" requirement can be satisfied in a useful way on a diff. At least no diff I've ever seen had such prominent notices. So as I read this, diffs would be disallowed. Is that intended? If it is, do you believe that everyone understands the text that way? Do you believe that anybody cares?<br><br></p><p>\nThe only option I see to satisfy this condition is to convey modifications in the form of complete modified source files. If that is indeed the only way to have "modifications to produce it from the Program, in the form of source code", then you should simply say so. <br><br></p><p>\nIn any case this section needs to be amended to make clear what is a permissible way to convey a modification.<br><br></p><p>\nPlease don't take this as nitpicking. The Linux From Scratch project (and many other places) are at this very moment distributing patches to GNU programs, such as this one<br><br></p><p>\nhttp://www.linuxfromscratch.org/patches/downloads/tar/tar-1.13.25-y_compat_flag-1.patch\n<br><br>\nThis is certainly not the preferred form of working on the modified work. It also doesn't carry any "prominent notices stating that it is released under this License"<br><br></p><p>\nSo as you can see, the question what are and what aren't permissible ways of conveying modified source versions has concrete consequences on everyday life of hobbyist open-source programmers, so it needs to be made clear in the GPL in a way that non-lawyers can understand it right away. And hopefully, you can change the wording to make the common practice of distributing diff-patches legal, because I can tell you, people won't stop doing it.</p>\ngpl3.distribmod.p0.s1\nmux2005\nlogin to agree\n0\n3043\n\n\n", "2476:3043": "\n\n\n\nTo produce such modified work\ngpl3.distribmod.p0.s1\nmerijn\nlogin to agree\n0\n2476\n\n\n<p style="display: inline">This term is undefined and IMHO is not self-explaining. Does this include diff-patches for instance? <br><br></p><p>\nI think a "modification" at least in the ordinary sense (that a non-technie judge/lawyer might apply) is not something tangible that can be conveyed. You may convey instructions to perform a modification, programs that perform a modification, etc. but a "modification" is something abstract that cannot be conveyed. Programmers might call a diff-file a "modification", but more correctly a diff is a description of a modification that can be used by the patch program to perform said modification.<br><br></p><p>\nFurthermore I'm not sure how this interacts with the definition of source code as the preferred form for making modifications. This sentence requires that modifications be in the form of source code, so we're talking about the preferred form of making modifications to a modification. Does that make sense? <br><br></p><p>\nAnd is a diff the preferred form for making modifications to a modification? I don't think so. The preferred form is the modified source code that can be generated from applying the diff. Then, I'm not sure if the "prominent notices" requirement can be satisfied in a useful way on a diff. At least no diff I've ever seen had such prominent notices. So as I read this, diffs would be disallowed. Is that intended? If it is, do you believe that everyone understands the text that way? Do you believe that anybody cares?<br><br></p><p>\nThe only option I see to satisfy this condition is to convey modifications in the form of complete modified source files. If that is indeed the only way to have "modifications to produce it from the Program, in the form of source code", then you should simply say so. <br><br></p><p>\nIn any case this section needs to be amended to make clear what is a permissible way to convey a modification.<br><br></p><p>\nPlease don't take this as nitpicking. The Linux From Scratch project (and many other places) are at this very moment distributing patches to GNU programs, such as this one<br><br></p><p>\nhttp://www.linuxfromscratch.org/patches/downloads/tar/tar-1.13.25-y_compat_flag-1.patch\n<br><br>\nThis is certainly not the preferred form of working on the modified work. It also doesn't carry any "prominent notices stating that it is released under this License"<br><br></p><p>\nSo as you can see, the question what are and what aren't permissible ways of conveying modified source versions has concrete consequences on everyday life of hobbyist open-source programmers, so it needs to be made clear in the GPL in a way that non-lawyers can understand it right away. And hopefully, you can change the wording to make the common practice of distributing diff-patches legal, because I can tell you, people won't stop doing it.</p>\ngpl3.distribmod.p0.s1\nmux2005\nlogin to agree\n0\n3043\n\n\n", "2634": "\n\n\n\nUse mixed case for defined term 'Source Code'\ngpl3.distribmod.p0.s1\nsamj\nlogin to agree\n0\n2634\n\n\n", "2870": "\n\n\n\nThe word "above" is unnecessary. To keep the text thin, should stay off.\ngpl3.distribmod.p0.s1\nlcchueri\nlogin to agree\n2\n2870\n\n\n", "2476:2477:2634:2870:3022:3043": "\n\n\n\nTo produce such modified work\ngpl3.distribmod.p0.s1\nmerijn\nlogin to agree\n0\n2476\n\n\n\ngpl3.distribmod.p0.s1\nmerijn\nlogin to agree\n0\n2477\n\n\nUse mixed case for defined term 'Source Code'\ngpl3.distribmod.p0.s1\nsamj\nlogin to agree\n0\n2634\n\n\nThe word "above" is unnecessary. To keep the text thin, should stay off.\ngpl3.distribmod.p0.s1\nlcchueri\nlogin to agree\n2\n2870\n\n\n<p style="display: inline">\nIn some countries, the author of a creative work can be legally entitled to certain moral rights. These moral rights can be separate from any copyright for a creative work. For more information, see http://en.wikipedia.org/wiki/Moral_rights (checked on April 29, 2007.) In particular, the article mentions the moral right for an author to have the integrity of a work preserved. It appears that this right can disallow others from altering a work if the alterations would hurt the honor and/or reputation of the author. When a work legally qualifies for moral rights, do those rights still apply if the work is licensed under the GPL? For such a work, would it be permissible for others to do something such as modifying the work in a manner that dishonors the author? Moral rights have come up in the case of the Creative Commons licenses. See http://wiki.creativecommons.org/Version_3#International_Harmonization_.E2.80.93_Moral_Rights (checked on May 6, 2007) for more information. In general, the Creative Commons policy was to preserve any moral right of integrity for the author of a work. An issue did come up with respect to the jurisdiction of Japan. The Japanese moral right of integrity was supposedly of such a nature that it could end up disallowing any alterations to a work. The solution was to limit the moral right of integrity for jurisdictions where it could completely disallow modifications to CC-licensed works. The unported (not specific to a certain jurisdiction) version of the Creative Commons Attribution-ShareAlike 3.0 license has a clause along these lines in its legal code. See section 4(d) at http://creativecommons.org/licenses/by-sa/3.0/legalcode for details. It might be useful for the GPL to do something similar.</p><p>\nThe Wikipedia article also mentions "artist resale rights" which may affect the resale of covered works, but it is not clear if such rights apply to the sale of a modified work or to the sale of a work that incorporates part of another work.</p><p>\nThe Creative Commons page regarding moral rights (http://wiki.creativecommons.org/Version_3#International_Harmonization_.E2.80.93_Moral_Rights) mentions the moral right of attribution. Newer versions of the Creative Commons licenses can require the attribution of authors. In the case of the GPLv3, item b of Section 7 (which can require the attribution of authors) may help to avoid problems with any moral right of attribution.\n\n</p>\ngpl3.distribmod.p0.s1\nmezzanine1\nlogin to agree\n0\n3022\n\n\n<p style="display: inline">This term is undefined and IMHO is not self-explaining. Does this include diff-patches for instance? <br><br></p><p>\nI think a "modification" at least in the ordinary sense (that a non-technie judge/lawyer might apply) is not something tangible that can be conveyed. You may convey instructions to perform a modification, programs that perform a modification, etc. but a "modification" is something abstract that cannot be conveyed. Programmers might call a diff-file a "modification", but more correctly a diff is a description of a modification that can be used by the patch program to perform said modification.<br><br></p><p>\nFurthermore I'm not sure how this interacts with the definition of source code as the preferred form for making modifications. This sentence requires that modifications be in the form of source code, so we're talking about the preferred form of making modifications to a modification. Does that make sense? <br><br></p><p>\nAnd is a diff the preferred form for making modifications to a modification? I don't think so. The preferred form is the modified source code that can be generated from applying the diff. Then, I'm not sure if the "prominent notices" requirement can be satisfied in a useful way on a diff. At least no diff I've ever seen had such prominent notices. So as I read this, diffs would be disallowed. Is that intended? If it is, do you believe that everyone understands the text that way? Do you believe that anybody cares?<br><br></p><p>\nThe only option I see to satisfy this condition is to convey modifications in the form of complete modified source files. If that is indeed the only way to have "modifications to produce it from the Program, in the form of source code", then you should simply say so. <br><br></p><p>\nIn any case this section needs to be amended to make clear what is a permissible way to convey a modification.<br><br></p><p>\nPlease don't take this as nitpicking. The Linux From Scratch project (and many other places) are at this very moment distributing patches to GNU programs, such as this one<br><br></p><p>\nhttp://www.linuxfromscratch.org/patches/downloads/tar/tar-1.13.25-y_compat_flag-1.patch\n<br><br>\nThis is certainly not the preferred form of working on the modified work. It also doesn't carry any "prominent notices stating that it is released under this License"<br><br></p><p>\nSo as you can see, the question what are and what aren't permissible ways of conveying modified source versions has concrete consequences on everyday life of hobbyist open-source programmers, so it needs to be made clear in the GPL in a way that non-lawyers can understand it right away. And hopefully, you can change the wording to make the common practice of distributing diff-patches legal, because I can tell you, people won't stop doing it.</p>\ngpl3.distribmod.p0.s1\nmux2005\nlogin to agree\n0\n3043\n\n\n", "2762:3045": "\n\n\n\nThis subsection is clarified as to which date(s) must be given. This is an improvement, since many people were uncertain on how the corresponding language in GPLv2 should be interpreted.\ngpl3.distribmod.p1.s1\nfrx\nlogin to agree\n0\n2762\n\n\nPlease change "the work" back to "the modified work" as in draft 2. This makes it easier to read, especially when you read only selective parts of the license and skim the rest (which is the primary mode for many hobbyist programmers to consume the GPL).\ngpl3.distribmod.p1.s1\nmux2005\nlogin to agree\n0\n3045\n\n\n", "2762": "\n\n\n\nThis subsection is clarified as to which date(s) must be given. This is an improvement, since many people were uncertain on how the corresponding language in GPLv2 should be interpreted.\ngpl3.distribmod.p1.s1\nfrx\nlogin to agree\n0\n2762\n\n\n", "2635:2762": "\n\n\n\n'Prominent Notices' should be a defined term and as such should be mixed case.\ngpl3.distribmod.p1.s1\nsamj\nlogin to agree\n2\n2635\n\n\nThis subsection is clarified as to which date(s) must be given. This is an improvement, since many people were uncertain on how the corresponding language in GPLv2 should be interpreted.\ngpl3.distribmod.p1.s1\nfrx\nlogin to agree\n0\n2762\n\n\n", "2635:2762:3045": "\n\n\n\n'Prominent Notices' should be a defined term and as such should be mixed case.\ngpl3.distribmod.p1.s1\nsamj\nlogin to agree\n2\n2635\n\n\nThis subsection is clarified as to which date(s) must be given. This is an improvement, since many people were uncertain on how the corresponding language in GPLv2 should be interpreted.\ngpl3.distribmod.p1.s1\nfrx\nlogin to agree\n0\n2762\n\n\nPlease change "the work" back to "the modified work" as in draft 2. This makes it easier to read, especially when you read only selective parts of the license and skim the rest (which is the primary mode for many hobbyist programmers to consume the GPL).\ngpl3.distribmod.p1.s1\nmux2005\nlogin to agree\n0\n3045\n\n\n", "2811:2812": "\n\n\n\nShort version: point 5.[2] b) in GPL3 draft3 is not strong enough!\r\n\r\nINSTEAD OF: 5.[2] b) The work must carry prominent notices stating that it is released under [GPL3] and any conditions added under section 7. This requirement modifies the requirement in section 4 to "keep intact all notices",\r\n\r\nI SUGGEST: 5.[2] b) The work must carry prominent notices stating that it is released either under "GPL3" or under "GPL3+C" (the latter meaning "GPL3 + any Compatibility conditions added under section 7"). "GPL3+C" also stands for "GPL3 and some GPL3+C material, such as blah-blah", meaning that, even if only a small part of the work is "GPL3+C", the whole work must be advertized as "GPL3+C". The licensor must not under any circumstances label or advertize the work released under "(GPL3 and) GPL3+C" as merely "GPL3". This requirement modifies the requirement in section 4 to "keep intact all notices".\r\n\r\nRationale: When introducing a GPL3+conditions project in a news story or even on a project's official page, it's much more excuseable for the publishers to shorten "GPL3 and the following conditions: blah-blah" to "GPL3" than it is to shorten "GPL3+C" to "GPL3". Now, it's obviously impossible for the LEGAL enforcement of this "GPL3+C" labeling to cover wikis, support forums, or publications, BUT, because of its "ergonomics", this "GPL3+C" idea will be way much more "enforceable" at least as a matter of fairplay, reputation, and credibility. After all, nobody can ever find any respectable and innocent excuse to drop the "+C" part, even informally, on a forum! Also, just try to imagine, as an analogy, how weak GPL's copyleft would've become if instead of "LGPL" we'd allowed licensors to advertize their projects as "GPL [with additional permissions, as mentioned in section blah-blah]"!\r\n\r\nIn conclusion, when adding "additional terms", licensors should be forced by GPL3 to NEVER, NOWHERE, advertize their software as being GPL3 without adding something like this: "+C, namely: MPL / CDDL / etc. material[, as_follows: blah-blah / as_detailed_at: blah blah]". AND, in situations in which such long mentions would be unreasonable (like when repeated too often on the same page), THE ONLY VALID available alternative should be that such notices be shortened to "GPL3+C" (or "CGPL3"?). But never should the GPL3 implicitely allow them to simply shorten it to "GPL3" (as the current draft3 sadly does)! If GPL3 won't tell them what's the reasonable way to shorten that, they will be forced to find their own "reasonable" ways. Reasonable and... convenient.\r\n\r\nMoreover, some licensors will feel INVITED to add further "terms" (restrictions) to their work "just because they can" and because "it will still be GPL3, anyway", while they can still safely advertize it as GPL3, which is "approximately true, anyway" (and, by the way, "just sue us if you think it sounds misleading!"), and then they leave to us the "pleasure" to find out "what kind of" GPL3 their work actually is (yes, there will be more than one [kind of] GPL3!) -- usually only after we get the source code. So it will become very easy to use GPL3 as a buzzword. Well, we don't want such traps! My proposed GPL3 / GPL3+C distinction will "invite" licensors to abandon such deceitful labeling. This distinction is especially important (and hopefully enforceable even legally) in legal/technical (con)texts, on the project's official pages, and in the object code (not only source). Otherwise, not only that the current GPL3 (draft3) would effectively invite them to add restrictions, but GPL3's very definition would be weakened. GPL3 won't mean "GPL3" anymore, but rather something like "open source" (a general expression with no safe and precise meaning). Weakened because of its Section 7 ("additional terms"), unlike GPL2, GPL3 won't give us that feeling of uncompromized protection anymore! :(\ngpl3.distribmod.p2.s1\nvincentc\nlogin to agree\n0\n2811\n\n\n<br><br>\r\nShort version: point 5.[2] b) in GPL3 draft3 is not strong enough!<br><br>\r\n\r\nINSTEAD OF: 5.[2] b) The work must carry prominent notices stating that it is released under [GPL3] and any conditions added under section 7. This requirement modifies the requirement in section 4 to "keep intact all notices",<br><br>\r\n\r\nI SUGGEST: 5.[2] b) The work must carry prominent notices stating that it is released either under "GPL3" or under "GPL3+C" (the latter meaning "GPL3 + any Compatibility conditions added under section 7"). "GPL3+C" also stands for "GPL3 and some GPL3+C material, such as blah-blah", meaning that, even if only a small part of the work is "GPL3+C", the whole work must be advertized as "GPL3+C". The licensor must not under any circumstances label or advertize the work released under "(GPL3 and) GPL3+C" as merely "GPL3". This requirement modifies the requirement in section 4 to "keep intact all notices".<br><br>\r\n\r\nRationale: When introducing a GPL3+conditions project in a news story or even on a project's official page, it's much more excuseable for the publishers to shorten "GPL3 and the following conditions: blah-blah" to "GPL3" than it is to shorten "GPL3+C" to "GPL3". Now, it's obviously impossible for the LEGAL enforcement of this "GPL3+C" labeling to cover wikis, support forums, or publications, BUT, because of its "ergonomics", this "GPL3+C" idea will be way much more "enforceable" at least as a matter of fairplay, reputation, and credibility. After all, nobody can ever find any respectable and innocent excuse to drop the "+C" part, even informally, on a forum! Also, just try to imagine, as an analogy, how weak GPL's copyleft would've become if instead of "LGPL" we'd allowed licensors to advertize their projects as "GPL [with additional permissions, as mentioned in section blah-blah]"!<br><br>\r\n\r\nIn conclusion, when adding "additional terms", licensors should be forced by GPL3 to NEVER, NOWHERE, advertize their software as being GPL3 without adding something like this: "+C, namely: MPL / CDDL / etc. material[, as_follows: blah-blah / as_detailed_at: blah blah]". AND, in situations in which such long mentions would be unreasonable (like when repeated too often on the same page), THE ONLY VALID available alternative should be that such notices be shortened to "GPL3+C" (or "CGPL3"?). But never should the GPL3 implicitely allow them to simply shorten it to "GPL3" (as the current draft3 sadly does)! If GPL3 won't tell them what's the reasonable way to shorten that, they will be forced to find their own "reasonable" ways. Reasonable and... convenient.<br><br>\r\n\r\nMoreover, some licensors will feel INVITED to add further "terms" (restrictions) to their work "just because they can" and because "it will still be GPL3, anyway", while they can still safely advertize it as GPL3, which is "approximately true, anyway" (and, by the way, "just sue us if you think it sounds misleading!"), and then they leave to us the "pleasure" to find out "what kind of" GPL3 their work actually is (yes, there will be more than one [kind of] GPL3!) -- usually only after we get the source code. So it will become very easy to use GPL3 as a buzzword. Well, we don't want such traps! My proposed GPL3 / GPL3+C distinction will "invite" licensors to abandon such deceitful labeling. This distinction is especially important (and hopefully enforceable even legally) in legal/technical (con)texts, on the project's official pages, and in the object code (not only source). Otherwise, not only that the current GPL3 (draft3) would effectively invite them to add restrictions, but GPL3's very definition would be weakened. GPL3 won't mean "GPL3" anymore, but rather something like "open source" (a general expression with no safe and precise meaning). Weakened because of its Section 7 ("additional terms"), unlike GPL2, GPL3 won't give us that feeling of uncompromized protection anymore! :(<br><br>\ngpl3.distribmod.p2.s1\nvincentc\nlogin to agree\n3\n2812\n\n\n", "2636:2811:2812": "\n\n\n\n'Prominent Notices' should be a defined term and as such should be mixed case.\ngpl3.distribmod.p2.s1\nsamj\nlogin to agree\n1\n2636\n\n\nShort version: point 5.[2] b) in GPL3 draft3 is not strong enough!\r\n\r\nINSTEAD OF: 5.[2] b) The work must carry prominent notices stating that it is released under [GPL3] and any conditions added under section 7. This requirement modifies the requirement in section 4 to "keep intact all notices",\r\n\r\nI SUGGEST: 5.[2] b) The work must carry prominent notices stating that it is released either under "GPL3" or under "GPL3+C" (the latter meaning "GPL3 + any Compatibility conditions added under section 7"). "GPL3+C" also stands for "GPL3 and some GPL3+C material, such as blah-blah", meaning that, even if only a small part of the work is "GPL3+C", the whole work must be advertized as "GPL3+C". The licensor must not under any circumstances label or advertize the work released under "(GPL3 and) GPL3+C" as merely "GPL3". This requirement modifies the requirement in section 4 to "keep intact all notices".\r\n\r\nRationale: When introducing a GPL3+conditions project in a news story or even on a project's official page, it's much more excuseable for the publishers to shorten "GPL3 and the following conditions: blah-blah" to "GPL3" than it is to shorten "GPL3+C" to "GPL3". Now, it's obviously impossible for the LEGAL enforcement of this "GPL3+C" labeling to cover wikis, support forums, or publications, BUT, because of its "ergonomics", this "GPL3+C" idea will be way much more "enforceable" at least as a matter of fairplay, reputation, and credibility. After all, nobody can ever find any respectable and innocent excuse to drop the "+C" part, even informally, on a forum! Also, just try to imagine, as an analogy, how weak GPL's copyleft would've become if instead of "LGPL" we'd allowed licensors to advertize their projects as "GPL [with additional permissions, as mentioned in section blah-blah]"!\r\n\r\nIn conclusion, when adding "additional terms", licensors should be forced by GPL3 to NEVER, NOWHERE, advertize their software as being GPL3 without adding something like this: "+C, namely: MPL / CDDL / etc. material[, as_follows: blah-blah / as_detailed_at: blah blah]". AND, in situations in which such long mentions would be unreasonable (like when repeated too often on the same page), THE ONLY VALID available alternative should be that such notices be shortened to "GPL3+C" (or "CGPL3"?). But never should the GPL3 implicitely allow them to simply shorten it to "GPL3" (as the current draft3 sadly does)! If GPL3 won't tell them what's the reasonable way to shorten that, they will be forced to find their own "reasonable" ways. Reasonable and... convenient.\r\n\r\nMoreover, some licensors will feel INVITED to add further "terms" (restrictions) to their work "just because they can" and because "it will still be GPL3, anyway", while they can still safely advertize it as GPL3, which is "approximately true, anyway" (and, by the way, "just sue us if you think it sounds misleading!"), and then they leave to us the "pleasure" to find out "what kind of" GPL3 their work actually is (yes, there will be more than one [kind of] GPL3!) -- usually only after we get the source code. So it will become very easy to use GPL3 as a buzzword. Well, we don't want such traps! My proposed GPL3 / GPL3+C distinction will "invite" licensors to abandon such deceitful labeling. This distinction is especially important (and hopefully enforceable even legally) in legal/technical (con)texts, on the project's official pages, and in the object code (not only source). Otherwise, not only that the current GPL3 (draft3) would effectively invite them to add restrictions, but GPL3's very definition would be weakened. GPL3 won't mean "GPL3" anymore, but rather something like "open source" (a general expression with no safe and precise meaning). Weakened because of its Section 7 ("additional terms"), unlike GPL2, GPL3 won't give us that feeling of uncompromized protection anymore! :(\ngpl3.distribmod.p2.s1\nvincentc\nlogin to agree\n0\n2811\n\n\n<br><br>\r\nShort version: point 5.[2] b) in GPL3 draft3 is not strong enough!<br><br>\r\n\r\nINSTEAD OF: 5.[2] b) The work must carry prominent notices stating that it is released under [GPL3] and any conditions added under section 7. This requirement modifies the requirement in section 4 to "keep intact all notices",<br><br>\r\n\r\nI SUGGEST: 5.[2] b) The work must carry prominent notices stating that it is released either under "GPL3" or under "GPL3+C" (the latter meaning "GPL3 + any Compatibility conditions added under section 7"). "GPL3+C" also stands for "GPL3 and some GPL3+C material, such as blah-blah", meaning that, even if only a small part of the work is "GPL3+C", the whole work must be advertized as "GPL3+C". The licensor must not under any circumstances label or advertize the work released under "(GPL3 and) GPL3+C" as merely "GPL3". This requirement modifies the requirement in section 4 to "keep intact all notices".<br><br>\r\n\r\nRationale: When introducing a GPL3+conditions project in a news story or even on a project's official page, it's much more excuseable for the publishers to shorten "GPL3 and the following conditions: blah-blah" to "GPL3" than it is to shorten "GPL3+C" to "GPL3". Now, it's obviously impossible for the LEGAL enforcement of this "GPL3+C" labeling to cover wikis, support forums, or publications, BUT, because of its "ergonomics", this "GPL3+C" idea will be way much more "enforceable" at least as a matter of fairplay, reputation, and credibility. After all, nobody can ever find any respectable and innocent excuse to drop the "+C" part, even informally, on a forum! Also, just try to imagine, as an analogy, how weak GPL's copyleft would've become if instead of "LGPL" we'd allowed licensors to advertize their projects as "GPL [with additional permissions, as mentioned in section blah-blah]"!<br><br>\r\n\r\nIn conclusion, when adding "additional terms", licensors should be forced by GPL3 to NEVER, NOWHERE, advertize their software as being GPL3 without adding something like this: "+C, namely: MPL / CDDL / etc. material[, as_follows: blah-blah / as_detailed_at: blah blah]". AND, in situations in which such long mentions would be unreasonable (like when repeated too often on the same page), THE ONLY VALID available alternative should be that such notices be shortened to "GPL3+C" (or "CGPL3"?). But never should the GPL3 implicitely allow them to simply shorten it to "GPL3" (as the current draft3 sadly does)! If GPL3 won't tell them what's the reasonable way to shorten that, they will be forced to find their own "reasonable" ways. Reasonable and... convenient.<br><br>\r\n\r\nMoreover, some licensors will feel INVITED to add further "terms" (restrictions) to their work "just because they can" and because "it will still be GPL3, anyway", while they can still safely advertize it as GPL3, which is "approximately true, anyway" (and, by the way, "just sue us if you think it sounds misleading!"), and then they leave to us the "pleasure" to find out "what kind of" GPL3 their work actually is (yes, there will be more than one [kind of] GPL3!) -- usually only after we get the source code. So it will become very easy to use GPL3 as a buzzword. Well, we don't want such traps! My proposed GPL3 / GPL3+C distinction will "invite" licensors to abandon such deceitful labeling. This distinction is especially important (and hopefully enforceable even legally) in legal/technical (con)texts, on the project's official pages, and in the object code (not only source). Otherwise, not only that the current GPL3 (draft3) would effectively invite them to add restrictions, but GPL3's very definition would be weakened. GPL3 won't mean "GPL3" anymore, but rather something like "open source" (a general expression with no safe and precise meaning). Weakened because of its Section 7 ("additional terms"), unlike GPL2, GPL3 won't give us that feeling of uncompromized protection anymore! :(<br><br>\ngpl3.distribmod.p2.s1\nvincentc\nlogin to agree\n3\n2812\n\n\n", "2672:2811:2812": "\n\n\n\n=> "conditions and permissions"?\ngpl3.distribmod.p2.s1\nsanjoy\nlogin to agree\n0\n2672\n\n\nShort version: point 5.[2] b) in GPL3 draft3 is not strong enough!\r\n\r\nINSTEAD OF: 5.[2] b) The work must carry prominent notices stating that it is released under [GPL3] and any conditions added under section 7. This requirement modifies the requirement in section 4 to "keep intact all notices",\r\n\r\nI SUGGEST: 5.[2] b) The work must carry prominent notices stating that it is released either under "GPL3" or under "GPL3+C" (the latter meaning "GPL3 + any Compatibility conditions added under section 7"). "GPL3+C" also stands for "GPL3 and some GPL3+C material, such as blah-blah", meaning that, even if only a small part of the work is "GPL3+C", the whole work must be advertized as "GPL3+C". The licensor must not under any circumstances label or advertize the work released under "(GPL3 and) GPL3+C" as merely "GPL3". This requirement modifies the requirement in section 4 to "keep intact all notices".\r\n\r\nRationale: When introducing a GPL3+conditions project in a news story or even on a project's official page, it's much more excuseable for the publishers to shorten "GPL3 and the following conditions: blah-blah" to "GPL3" than it is to shorten "GPL3+C" to "GPL3". Now, it's obviously impossible for the LEGAL enforcement of this "GPL3+C" labeling to cover wikis, support forums, or publications, BUT, because of its "ergonomics", this "GPL3+C" idea will be way much more "enforceable" at least as a matter of fairplay, reputation, and credibility. After all, nobody can ever find any respectable and innocent excuse to drop the "+C" part, even informally, on a forum! Also, just try to imagine, as an analogy, how weak GPL's copyleft would've become if instead of "LGPL" we'd allowed licensors to advertize their projects as "GPL [with additional permissions, as mentioned in section blah-blah]"!\r\n\r\nIn conclusion, when adding "additional terms", licensors should be forced by GPL3 to NEVER, NOWHERE, advertize their software as being GPL3 without adding something like this: "+C, namely: MPL / CDDL / etc. material[, as_follows: blah-blah / as_detailed_at: blah blah]". AND, in situations in which such long mentions would be unreasonable (like when repeated too often on the same page), THE ONLY VALID available alternative should be that such notices be shortened to "GPL3+C" (or "CGPL3"?). But never should the GPL3 implicitely allow them to simply shorten it to "GPL3" (as the current draft3 sadly does)! If GPL3 won't tell them what's the reasonable way to shorten that, they will be forced to find their own "reasonable" ways. Reasonable and... convenient.\r\n\r\nMoreover, some licensors will feel INVITED to add further "terms" (restrictions) to their work "just because they can" and because "it will still be GPL3, anyway", while they can still safely advertize it as GPL3, which is "approximately true, anyway" (and, by the way, "just sue us if you think it sounds misleading!"), and then they leave to us the "pleasure" to find out "what kind of" GPL3 their work actually is (yes, there will be more than one [kind of] GPL3!) -- usually only after we get the source code. So it will become very easy to use GPL3 as a buzzword. Well, we don't want such traps! My proposed GPL3 / GPL3+C distinction will "invite" licensors to abandon such deceitful labeling. This distinction is especially important (and hopefully enforceable even legally) in legal/technical (con)texts, on the project's official pages, and in the object code (not only source). Otherwise, not only that the current GPL3 (draft3) would effectively invite them to add restrictions, but GPL3's very definition would be weakened. GPL3 won't mean "GPL3" anymore, but rather something like "open source" (a general expression with no safe and precise meaning). Weakened because of its Section 7 ("additional terms"), unlike GPL2, GPL3 won't give us that feeling of uncompromized protection anymore! :(\ngpl3.distribmod.p2.s1\nvincentc\nlogin to agree\n0\n2811\n\n\n<br><br>\r\nShort version: point 5.[2] b) in GPL3 draft3 is not strong enough!<br><br>\r\n\r\nINSTEAD OF: 5.[2] b) The work must carry prominent notices stating that it is released under [GPL3] and any conditions added under section 7. This requirement modifies the requirement in section 4 to "keep intact all notices",<br><br>\r\n\r\nI SUGGEST: 5.[2] b) The work must carry prominent notices stating that it is released either under "GPL3" or under "GPL3+C" (the latter meaning "GPL3 + any Compatibility conditions added under section 7"). "GPL3+C" also stands for "GPL3 and some GPL3+C material, such as blah-blah", meaning that, even if only a small part of the work is "GPL3+C", the whole work must be advertized as "GPL3+C". The licensor must not under any circumstances label or advertize the work released under "(GPL3 and) GPL3+C" as merely "GPL3". This requirement modifies the requirement in section 4 to "keep intact all notices".<br><br>\r\n\r\nRationale: When introducing a GPL3+conditions project in a news story or even on a project's official page, it's much more excuseable for the publishers to shorten "GPL3 and the following conditions: blah-blah" to "GPL3" than it is to shorten "GPL3+C" to "GPL3". Now, it's obviously impossible for the LEGAL enforcement of this "GPL3+C" labeling to cover wikis, support forums, or publications, BUT, because of its "ergonomics", this "GPL3+C" idea will be way much more "enforceable" at least as a matter of fairplay, reputation, and credibility. After all, nobody can ever find any respectable and innocent excuse to drop the "+C" part, even informally, on a forum! Also, just try to imagine, as an analogy, how weak GPL's copyleft would've become if instead of "LGPL" we'd allowed licensors to advertize their projects as "GPL [with additional permissions, as mentioned in section blah-blah]"!<br><br>\r\n\r\nIn conclusion, when adding "additional terms", licensors should be forced by GPL3 to NEVER, NOWHERE, advertize their software as being GPL3 without adding something like this: "+C, namely: MPL / CDDL / etc. material[, as_follows: blah-blah / as_detailed_at: blah blah]". AND, in situations in which such long mentions would be unreasonable (like when repeated too often on the same page), THE ONLY VALID available alternative should be that such notices be shortened to "GPL3+C" (or "CGPL3"?). But never should the GPL3 implicitely allow them to simply shorten it to "GPL3" (as the current draft3 sadly does)! If GPL3 won't tell them what's the reasonable way to shorten that, they will be forced to find their own "reasonable" ways. Reasonable and... convenient.<br><br>\r\n\r\nMoreover, some licensors will feel INVITED to add further "terms" (restrictions) to their work "just because they can" and because "it will still be GPL3, anyway", while they can still safely advertize it as GPL3, which is "approximately true, anyway" (and, by the way, "just sue us if you think it sounds misleading!"), and then they leave to us the "pleasure" to find out "what kind of" GPL3 their work actually is (yes, there will be more than one [kind of] GPL3!) -- usually only after we get the source code. So it will become very easy to use GPL3 as a buzzword. Well, we don't want such traps! My proposed GPL3 / GPL3+C distinction will "invite" licensors to abandon such deceitful labeling. This distinction is especially important (and hopefully enforceable even legally) in legal/technical (con)texts, on the project's official pages, and in the object code (not only source). Otherwise, not only that the current GPL3 (draft3) would effectively invite them to add restrictions, but GPL3's very definition would be weakened. GPL3 won't mean "GPL3" anymore, but rather something like "open source" (a general expression with no safe and precise meaning). Weakened because of its Section 7 ("additional terms"), unlike GPL2, GPL3 won't give us that feeling of uncompromized protection anymore! :(<br><br>\ngpl3.distribmod.p2.s1\nvincentc\nlogin to agree\n3\n2812\n\n\n", "3044": "\n\n\n\n<p style="display: inline">I don't understand how this modifies the requirement to keep intact all notices. Section 4 does not AFAICS preclude someone from adding more notices, so requiring the adding of a new notice does not in my view modify section 4's requirements. Instead it would be an addition to it. The only way I can see how this new requirement could *modify* the old requirement is by replacing it. But that's not what's meant, right? I'm not allowed to strip notices from a modified work, am I?<br><br></p><p>\nPlease change "modifies" to either "replaces" or "is an addition to", depending on what you want it to mean.</p>\ngpl3.distribmod.p2.s2\nmux2005\nlogin to agree\n0\n3044\n\n\n", "2541:2820": "\n\n\n\nI should think this would be limited to anyone who legitimately comes into possession of a copy. If I modify a GPL program for internal use never intending to convey or propagate, and someone steals a laptop with the binaries, why should I be obligated to provide a license? I should be allowed to reclaim the bits (along with the transfer medium).\ngpl3.distribmod.p3.s1\nmbreuer\nlogin to agree\n12\n2541\n\n\nThe earlier sections of this draft, taken as a whole, do not clearly define the scope of a covered work. There was language in GPL v2 stating, in effect, that independent parts of a work can be considered separate works and are not part of the covered work. The guidance on that issue ("intimate communication") migrated to the definition of Corresponding Source. The scope of a covered work remains vague and needs to be clarified.\ngpl3.distribmod.p3.s1\nsklein\nlogin to agree\n0\n2820\n\n\n", "2541": "\n\n\n\nI should think this would be limited to anyone who legitimately comes into possession of a copy. If I modify a GPL program for internal use never intending to convey or propagate, and someone steals a laptop with the binaries, why should I be obligated to provide a license? I should be allowed to reclaim the bits (along with the transfer medium).\ngpl3.distribmod.p3.s1\nmbreuer\nlogin to agree\n12\n2541\n\n\n", "2541:2626:2685": "\n\n\n\nI should think this would be limited to anyone who legitimately comes into possession of a copy. If I modify a GPL program for internal use never intending to convey or propagate, and someone steals a laptop with the binaries, why should I be obligated to provide a license? I should be allowed to reclaim the bits (along with the transfer medium).\ngpl3.distribmod.p3.s1\nmbreuer\nlogin to agree\n12\n2541\n\n\nIf the owner of the modification does not authorise the distribution they should be protected from unintended distribution (eg theft by an employee).\ngpl3.distribmod.p3.s1\nsamj\nlogin to agree\n8\n2626\n\n\nsamj: without "legitimately" people will argue the GPL legalizes such theft. it's better to be 100% explicit in licenses and not to rely on implicit invocation of other laws.\ngpl3.distribmod.p3.s1\nmerijn\nlogin to agree\n2\n2685\n\n\n", "2541:2626:2685:2995": "\n\n\n\nI should think this would be limited to anyone who legitimately comes into possession of a copy. If I modify a GPL program for internal use never intending to convey or propagate, and someone steals a laptop with the binaries, why should I be obligated to provide a license? I should be allowed to reclaim the bits (along with the transfer medium).\ngpl3.distribmod.p3.s1\nmbreuer\nlogin to agree\n12\n2541\n\n\nIf the owner of the modification does not authorise the distribution they should be protected from unintended distribution (eg theft by an employee).\ngpl3.distribmod.p3.s1\nsamj\nlogin to agree\n8\n2626\n\n\nsamj: without "legitimately" people will argue the GPL legalizes such theft. it's better to be 100% explicit in licenses and not to rely on implicit invocation of other laws.\ngpl3.distribmod.p3.s1\nmerijn\nlogin to agree\n2\n2685\n\n\ns/who comes into possession of a copy/ to whom the work is conveyed/\ngpl3.distribmod.p3.s1\nmole\nlogin to agree\n0\n2995\n\n\n", "3147": "\n\n\n\nThis section is the spiritual successor to GPLv2 2b.\nWould defining "the whole" to explicitly include patents strengthen the license or would further defining "the whole" imply that "the whole" might not include some things weakening the license?\ngpl3.distribmod.p3.s2\ndrcj\nlogin to agree\n0\n3147\n\n\n", "2763": "\n\n\n\n<p style="display: inline">I think that explicitly acknowledging the possibility of dual licensing (GPLv3 or OTHERLICENSE at the recipient option) is good, as it clarifies that the GPL does not try to take away permissions that are separately granted. This is an improvement over GPLv2: good to see that it was kept in GPLv3draft3 too.</p><p>\nSection 5 seems OK, so far (apart from the reference to Section 7, more on this in a later comment...)\n</p>\ngpl3.distribmod.p3.s3\nfrx\nlogin to agree\n0\n2763\n\n\n", "2763:2820:2995:3147": "\n\n\n\n<p style="display: inline">I think that explicitly acknowledging the possibility of dual licensing (GPLv3 or OTHERLICENSE at the recipient option) is good, as it clarifies that the GPL does not try to take away permissions that are separately granted. This is an improvement over GPLv2: good to see that it was kept in GPLv3draft3 too.</p><p>\nSection 5 seems OK, so far (apart from the reference to Section 7, more on this in a later comment...)\n</p>\ngpl3.distribmod.p3.s3\nfrx\nlogin to agree\n0\n2763\n\n\nThe earlier sections of this draft, taken as a whole, do not clearly define the scope of a covered work. There was language in GPL v2 stating, in effect, that independent parts of a work can be considered separate works and are not part of the covered work. The guidance on that issue ("intimate communication") migrated to the definition of Corresponding Source. The scope of a covered work remains vague and needs to be clarified.\ngpl3.distribmod.p3.s1\nsklein\nlogin to agree\n0\n2820\n\n\ns/who comes into possession of a copy/ to whom the work is conveyed/\ngpl3.distribmod.p3.s1\nmole\nlogin to agree\n0\n2995\n\n\nThis section is the spiritual successor to GPLv2 2b.\nWould defining "the whole" to explicitly include patents strengthen the license or would further defining "the whole" imply that "the whole" might not include some things weakening the license?\ngpl3.distribmod.p3.s2\ndrcj\nlogin to agree\n0\n3147\n\n\n", "2764:2799:2932:2972:2974:3082": "\n\n\n\n<p style="display: inline">Clause 5d in GPLv3draft3 is basically unchanged with respect to previous drafts. It's worse than the corresponding clause 2c in GPLv2... :-(</p><p>\nIt's an inconvenience and border-line with respect to freeness. Actually this clause restricts how I can modify what an interactive program does when run. It mandates a feature that I *must* implement in *any* interactive interface of my modified work. It's very close to place an unacceptable restriction on modification. What is more awkward is that it seems that when a non-interactive work is modified so that it becomes an interactive work, the modifier is *compelled* to implement these features in *any* newly created interactive interface...</p><p>\nI would like to see clause 5d dropped entirely.</p>\ngpl3.distribmod.p4.s1\nfrx\nlogin to agree\n3\n2764\n\n\n<p style="display: inline">I believe in some cases this conflicts directly. \n\nFirstly, I believe this steps on Freedom 1, dictating specifically how one must write or design a program.</p><p>\nIn many cases, optimization for space is important, specifically to embedded devices. Also, when designing web application, ease of navigation is important. This clause would restrict a user from removing such notice to reduce space or improve navigation, hence, limiting one's ability to improve a program.</p>\ngpl3.distribmod.p4.s1\nkdean06\nlogin to agree\n3\n2799\n\n\n<p style="display: inline">\nI feel uncomfortable with the provisions of clause 5d about user\ninterfaces. In a way it reminds me of the historically obnoxious BSD\nclause.\nWhat happens to miniature devices, whose interactive displays have no\nroom for the long messages required to comply with clause 5d?\n --- Omer</p><p>\n--\nDelay is the deadliest form of denial. C. Northcote Parkinson\nMy own blog is at http://tddpirate.livejournal.com/</p><p>\nMy opinions, as expressed in this E-mail message, are mine alone.\nThey do not represent the official policy of any organization with which\nI may be affiliated in any way.\nWARNING TO SPAMMERS: at http://www.zak.co.il/spamwarning.html\n\n\n</p>\ngpl3.distribmod.p4.s1\nw1\nlogin to agree\n1\n2932\n\n\n<p style="display: inline">As I understand it, the purpose of this section is to make sure that users of the program know their rights. However, I believe it has several practical and philosophical problems. In true lawyerly fashion, I will try multiple possible arguments (some of which have been made by others in this comments system):\n\n * 5d) is misplaced in section 5, given that it really relates to object code forms and yet 5 is a source code section.\n * 5d) is redundant, given the existence of section 5b) and the word "prominent" therein, and the existence of section 4, para 1 (which is applicable because of the inclusion of the terms of 4 into 5 - see 5 para 1).\n * Examples such as menus are bad because they will date the license and may force a particular inappropriate implementation of the underlying goal.\n * The current way 5d) is phrased is close to being (some would say, is) a restriction on modification, thereby violating freedom 1.\n * The current version of 5d) is more intrusive in that regard than the version in GPLv2, and applies in more circumstances.\n * 5d) should be written to focus on policy, not mechanism - as was done to turn the requirement to surrender keys into the requirement to provide Installation Instructions.</p><p>\nI believe 5d) should be scrapped entirely but, if that is not to be, perhaps something like the following wording would achieve the goal while alleviating the problems in the latter half of the list:\n\n d) If the work has interactive user interfaces, users of each interface must have an easily-discoverable mechanism to display an appropriate copyright notice, and to be told:\n\n - the extent of the warranty, or lack of it, for the work;\n - that licensees may convey the work under this License;\n - how to view a copy of this License.</p><p>\nHowever, if the work has interactive interfaces that do not comply with this subsection, your changes need not make them comply.</p><p>\nThis rephrasing removes most things that might be construed as a restriction on modification, and becomes solely an expression of policy - you must make users aware of their rights.\n</p>\ngpl3.distribmod.p4.s1\ngerv\nlogin to agree\n2\n2972\n\n\nAs I understand it, the purpose of this section is to make sure that users of the program know their rights. However, I believe it has several practical and philosophical problems. In true lawyerly fashion, I will try multiple possible arguments (some of which have been made by others in this comments system):\n\n\n * 5d) is misplaced in section 5, given that it really relates to object code forms and yet 5 is a source code section.\n\n\n * 5d) is redundant, given the existence of section 5b) and the word "prominent" therein, and the existence of section 4, para 1 (which is applicable because of the inclusion of the terms of 4 into 5 - see 5 para 1).\n\n\n * Examples such as menus are bad because they will date the license and may force a particular inappropriate implementation of the underlying goal.\n\n\n * The current way 5d) is phrased is close to being (some would say, is) a restriction on modification, thereby violating freedom 1.\n\n\n * The current version of 5d) is more intrusive in that regard than the version in GPLv2, and applies in more circumstances.\n\n\n * 5d) should be written to focus on policy, not mechanism - as was done to turn the requirement to surrender keys into the requirement to provide Installation Instructions.\n\n\nI believe 5d) should be scrapped entirely but, if that is not to be, perhaps something like the following wording would achieve the goal while alleviating the problems in the latter half of the list:\n\n\n d) If the work has interactive user interfaces, users of each interface must have an easily-discoverable mechanism to display an appropriate copyright notice, and to be told:\n\n\n - the extent of the warranty, or lack of it, for the work;\n\n\n - that licensees may convey the work \nunder this License;\n\n\n - how to view a copy of this License.\n\n\nHowever, if the work has interactive interfaces that do not comply with this subsection, your changes need not make them comply.\n\n\nThis rephrasing removes most things that might be construed as a restriction on modification, and becomes solely an expression of policy - you must make users aware of their rights.\ngpl3.distribmod.p4.s1\ngerv\nlogin to agree\n0\n2974\n\n\n<p style="display: inline">If you don't want to remove 5d entirely, make it a permissible additional term according to section 7.<br><br></p><p>\nAs stated in other comments, there are good reasons for removing this section entirely. If the FSF doesn't want to do this for whatever reasons, then a compromise would be to move this to section 7.Additional Terms. After all, this is closely related to "requiring preservation of ..." and "prohibiting misrepresentation of..."<br><br></p><p>\nThis way, licensors can decide if they want to impose this restriction on licensees. The statement "However, if the Program...need not make them comply" already acknowledges the current practice that some authors of free software (me, for instance) decide not to include a complying message (and as observed elsewhere, even some GNU software such as bash doesn't have this message). So this is really better placed in section 7 Additional Terms.</p>\ngpl3.distribmod.p4.s1\nmux2005\nlogin to agree\n0\n3082\n\n\n", "2764:2788:2799:2972:2974:3082": "\n\n\n\n<p style="display: inline">Clause 5d in GPLv3draft3 is basically unchanged with respect to previous drafts. It's worse than the corresponding clause 2c in GPLv2... :-(</p><p>\nIt's an inconvenience and border-line with respect to freeness. Actually this clause restricts how I can modify what an interactive program does when run. It mandates a feature that I *must* implement in *any* interactive interface of my modified work. It's very close to place an unacceptable restriction on modification. What is more awkward is that it seems that when a non-interactive work is modified so that it becomes an interactive work, the modifier is *compelled* to implement these features in *any* newly created interactive interface...</p><p>\nI would like to see clause 5d dropped entirely.</p>\ngpl3.distribmod.p4.s1\nfrx\nlogin to agree\n3\n2764\n\n\nIs it too strong to require that *each* interactive user interface include the feature? Perhaps instead "the main interface or menu must include..."?\ngpl3.distribmod.p4.s1\nsanjoy\nlogin to agree\n0\n2788\n\n\n<p style="display: inline">I believe in some cases this conflicts directly. \n\nFirstly, I believe this steps on Freedom 1, dictating specifically how one must write or design a program.</p><p>\nIn many cases, optimization for space is important, specifically to embedded devices. Also, when designing web application, ease of navigation is important. This clause would restrict a user from removing such notice to reduce space or improve navigation, hence, limiting one's ability to improve a program.</p>\ngpl3.distribmod.p4.s1\nkdean06\nlogin to agree\n3\n2799\n\n\n<p style="display: inline">As I understand it, the purpose of this section is to make sure that users of the program know their rights. However, I believe it has several practical and philosophical problems. In true lawyerly fashion, I will try multiple possible arguments (some of which have been made by others in this comments system):\n\n * 5d) is misplaced in section 5, given that it really relates to object code forms and yet 5 is a source code section.\n * 5d) is redundant, given the existence of section 5b) and the word "prominent" therein, and the existence of section 4, para 1 (which is applicable because of the inclusion of the terms of 4 into 5 - see 5 para 1).\n * Examples such as menus are bad because they will date the license and may force a particular inappropriate implementation of the underlying goal.\n * The current way 5d) is phrased is close to being (some would say, is) a restriction on modification, thereby violating freedom 1.\n * The current version of 5d) is more intrusive in that regard than the version in GPLv2, and applies in more circumstances.\n * 5d) should be written to focus on policy, not mechanism - as was done to turn the requirement to surrender keys into the requirement to provide Installation Instructions.</p><p>\nI believe 5d) should be scrapped entirely but, if that is not to be, perhaps something like the following wording would achieve the goal while alleviating the problems in the latter half of the list:\n\n d) If the work has interactive user interfaces, users of each interface must have an easily-discoverable mechanism to display an appropriate copyright notice, and to be told:\n\n - the extent of the warranty, or lack of it, for the work;\n - that licensees may convey the work under this License;\n - how to view a copy of this License.</p><p>\nHowever, if the work has interactive interfaces that do not comply with this subsection, your changes need not make them comply.</p><p>\nThis rephrasing removes most things that might be construed as a restriction on modification, and becomes solely an expression of policy - you must make users aware of their rights.\n</p>\ngpl3.distribmod.p4.s1\ngerv\nlogin to agree\n2\n2972\n\n\nAs I understand it, the purpose of this section is to make sure that users of the program know their rights. However, I believe it has several practical and philosophical problems. In true lawyerly fashion, I will try multiple possible arguments (some of which have been made by others in this comments system):\n\n\n * 5d) is misplaced in section 5, given that it really relates to object code forms and yet 5 is a source code section.\n\n\n * 5d) is redundant, given the existence of section 5b) and the word "prominent" therein, and the existence of section 4, para 1 (which is applicable because of the inclusion of the terms of 4 into 5 - see 5 para 1).\n\n\n * Examples such as menus are bad because they will date the license and may force a particular inappropriate implementation of the underlying goal.\n\n\n * The current way 5d) is phrased is close to being (some would say, is) a restriction on modification, thereby violating freedom 1.\n\n\n * The current version of 5d) is more intrusive in that regard than the version in GPLv2, and applies in more circumstances.\n\n\n * 5d) should be written to focus on policy, not mechanism - as was done to turn the requirement to surrender keys into the requirement to provide Installation Instructions.\n\n\nI believe 5d) should be scrapped entirely but, if that is not to be, perhaps something like the following wording would achieve the goal while alleviating the problems in the latter half of the list:\n\n\n d) If the work has interactive user interfaces, users of each interface must have an easily-discoverable mechanism to display an appropriate copyright notice, and to be told:\n\n\n - the extent of the warranty, or lack of it, for the work;\n\n\n - that licensees may convey the work \nunder this License;\n\n\n - how to view a copy of this License.\n\n\nHowever, if the work has interactive interfaces that do not comply with this subsection, your changes need not make them comply.\n\n\nThis rephrasing removes most things that might be construed as a restriction on modification, and becomes solely an expression of policy - you must make users aware of their rights.\ngpl3.distribmod.p4.s1\ngerv\nlogin to agree\n0\n2974\n\n\n<p style="display: inline">If you don't want to remove 5d entirely, make it a permissible additional term according to section 7.<br><br></p><p>\nAs stated in other comments, there are good reasons for removing this section entirely. If the FSF doesn't want to do this for whatever reasons, then a compromise would be to move this to section 7.Additional Terms. After all, this is closely related to "requiring preservation of ..." and "prohibiting misrepresentation of..."<br><br></p><p>\nThis way, licensors can decide if they want to impose this restriction on licensees. The statement "However, if the Program...need not make them comply" already acknowledges the current practice that some authors of free software (me, for instance) decide not to include a complying message (and as observed elsewhere, even some GNU software such as bash doesn't have this message). So this is really better placed in section 7 Additional Terms.</p>\ngpl3.distribmod.p4.s1\nmux2005\nlogin to agree\n0\n3082\n\n\n", "2764:2799:2972:2974:3082": "\n\n\n\n<p style="display: inline">Clause 5d in GPLv3draft3 is basically unchanged with respect to previous drafts. It's worse than the corresponding clause 2c in GPLv2... :-(</p><p>\nIt's an inconvenience and border-line with respect to freeness. Actually this clause restricts how I can modify what an interactive program does when run. It mandates a feature that I *must* implement in *any* interactive interface of my modified work. It's very close to place an unacceptable restriction on modification. What is more awkward is that it seems that when a non-interactive work is modified so that it becomes an interactive work, the modifier is *compelled* to implement these features in *any* newly created interactive interface...</p><p>\nI would like to see clause 5d dropped entirely.</p>\ngpl3.distribmod.p4.s1\nfrx\nlogin to agree\n3\n2764\n\n\n<p style="display: inline">I believe in some cases this conflicts directly. \n\nFirstly, I believe this steps on Freedom 1, dictating specifically how one must write or design a program.</p><p>\nIn many cases, optimization for space is important, specifically to embedded devices. Also, when designing web application, ease of navigation is important. This clause would restrict a user from removing such notice to reduce space or improve navigation, hence, limiting one's ability to improve a program.</p>\ngpl3.distribmod.p4.s1\nkdean06\nlogin to agree\n3\n2799\n\n\n<p style="display: inline">As I understand it, the purpose of this section is to make sure that users of the program know their rights. However, I believe it has several practical and philosophical problems. In true lawyerly fashion, I will try multiple possible arguments (some of which have been made by others in this comments system):\n\n * 5d) is misplaced in section 5, given that it really relates to object code forms and yet 5 is a source code section.\n * 5d) is redundant, given the existence of section 5b) and the word "prominent" therein, and the existence of section 4, para 1 (which is applicable because of the inclusion of the terms of 4 into 5 - see 5 para 1).\n * Examples such as menus are bad because they will date the license and may force a particular inappropriate implementation of the underlying goal.\n * The current way 5d) is phrased is close to being (some would say, is) a restriction on modification, thereby violating freedom 1.\n * The current version of 5d) is more intrusive in that regard than the version in GPLv2, and applies in more circumstances.\n * 5d) should be written to focus on policy, not mechanism - as was done to turn the requirement to surrender keys into the requirement to provide Installation Instructions.</p><p>\nI believe 5d) should be scrapped entirely but, if that is not to be, perhaps something like the following wording would achieve the goal while alleviating the problems in the latter half of the list:\n\n d) If the work has interactive user interfaces, users of each interface must have an easily-discoverable mechanism to display an appropriate copyright notice, and to be told:\n\n - the extent of the warranty, or lack of it, for the work;\n - that licensees may convey the work under this License;\n - how to view a copy of this License.</p><p>\nHowever, if the work has interactive interfaces that do not comply with this subsection, your changes need not make them comply.</p><p>\nThis rephrasing removes most things that might be construed as a restriction on modification, and becomes solely an expression of policy - you must make users aware of their rights.\n</p>\ngpl3.distribmod.p4.s1\ngerv\nlogin to agree\n2\n2972\n\n\nAs I understand it, the purpose of this section is to make sure that users of the program know their rights. However, I believe it has several practical and philosophical problems. In true lawyerly fashion, I will try multiple possible arguments (some of which have been made by others in this comments system):\n\n\n * 5d) is misplaced in section 5, given that it really relates to object code forms and yet 5 is a source code section.\n\n\n * 5d) is redundant, given the existence of section 5b) and the word "prominent" therein, and the existence of section 4, para 1 (which is applicable because of the inclusion of the terms of 4 into 5 - see 5 para 1).\n\n\n * Examples such as menus are bad because they will date the license and may force a particular inappropriate implementation of the underlying goal.\n\n\n * The current way 5d) is phrased is close to being (some would say, is) a restriction on modification, thereby violating freedom 1.\n\n\n * The current version of 5d) is more intrusive in that regard than the version in GPLv2, and applies in more circumstances.\n\n\n * 5d) should be written to focus on policy, not mechanism - as was done to turn the requirement to surrender keys into the requirement to provide Installation Instructions.\n\n\nI believe 5d) should be scrapped entirely but, if that is not to be, perhaps something like the following wording would achieve the goal while alleviating the problems in the latter half of the list:\n\n\n d) If the work has interactive user interfaces, users of each interface must have an easily-discoverable mechanism to display an appropriate copyright notice, and to be told:\n\n\n - the extent of the warranty, or lack of it, for the work;\n\n\n - that licensees may convey the work \nunder this License;\n\n\n - how to view a copy of this License.\n\n\nHowever, if the work has interactive interfaces that do not comply with this subsection, your changes need not make them comply.\n\n\nThis rephrasing removes most things that might be construed as a restriction on modification, and becomes solely an expression of policy - you must make users aware of their rights.\ngpl3.distribmod.p4.s1\ngerv\nlogin to agree\n0\n2974\n\n\n<p style="display: inline">If you don't want to remove 5d entirely, make it a permissible additional term according to section 7.<br><br></p><p>\nAs stated in other comments, there are good reasons for removing this section entirely. If the FSF doesn't want to do this for whatever reasons, then a compromise would be to move this to section 7.Additional Terms. After all, this is closely related to "requiring preservation of ..." and "prohibiting misrepresentation of..."<br><br></p><p>\nThis way, licensors can decide if they want to impose this restriction on licensees. The statement "However, if the Program...need not make them comply" already acknowledges the current practice that some authors of free software (me, for instance) decide not to include a complying message (and as observed elsewhere, even some GNU software such as bash doesn't have this message). So this is really better placed in section 7 Additional Terms.</p>\ngpl3.distribmod.p4.s1\nmux2005\nlogin to agree\n0\n3082\n\n\n", "2764:2972:2974": "\n\n\n\n<p style="display: inline">Clause 5d in GPLv3draft3 is basically unchanged with respect to previous drafts. It's worse than the corresponding clause 2c in GPLv2... :-(</p><p>\nIt's an inconvenience and border-line with respect to freeness. Actually this clause restricts how I can modify what an interactive program does when run. It mandates a feature that I *must* implement in *any* interactive interface of my modified work. It's very close to place an unacceptable restriction on modification. What is more awkward is that it seems that when a non-interactive work is modified so that it becomes an interactive work, the modifier is *compelled* to implement these features in *any* newly created interactive interface...</p><p>\nI would like to see clause 5d dropped entirely.</p>\ngpl3.distribmod.p4.s1\nfrx\nlogin to agree\n3\n2764\n\n\n<p style="display: inline">As I understand it, the purpose of this section is to make sure that users of the program know their rights. However, I believe it has several practical and philosophical problems. In true lawyerly fashion, I will try multiple possible arguments (some of which have been made by others in this comments system):\n\n * 5d) is misplaced in section 5, given that it really relates to object code forms and yet 5 is a source code section.\n * 5d) is redundant, given the existence of section 5b) and the word "prominent" therein, and the existence of section 4, para 1 (which is applicable because of the inclusion of the terms of 4 into 5 - see 5 para 1).\n * Examples such as menus are bad because they will date the license and may force a particular inappropriate implementation of the underlying goal.\n * The current way 5d) is phrased is close to being (some would say, is) a restriction on modification, thereby violating freedom 1.\n * The current version of 5d) is more intrusive in that regard than the version in GPLv2, and applies in more circumstances.\n * 5d) should be written to focus on policy, not mechanism - as was done to turn the requirement to surrender keys into the requirement to provide Installation Instructions.</p><p>\nI believe 5d) should be scrapped entirely but, if that is not to be, perhaps something like the following wording would achieve the goal while alleviating the problems in the latter half of the list:\n\n d) If the work has interactive user interfaces, users of each interface must have an easily-discoverable mechanism to display an appropriate copyright notice, and to be told:\n\n - the extent of the warranty, or lack of it, for the work;\n - that licensees may convey the work under this License;\n - how to view a copy of this License.</p><p>\nHowever, if the work has interactive interfaces that do not comply with this subsection, your changes need not make them comply.</p><p>\nThis rephrasing removes most things that might be construed as a restriction on modification, and becomes solely an expression of policy - you must make users aware of their rights.\n</p>\ngpl3.distribmod.p4.s1\ngerv\nlogin to agree\n2\n2972\n\n\nAs I understand it, the purpose of this section is to make sure that users of the program know their rights. However, I believe it has several practical and philosophical problems. In true lawyerly fashion, I will try multiple possible arguments (some of which have been made by others in this comments system):\n\n\n * 5d) is misplaced in section 5, given that it really relates to object code forms and yet 5 is a source code section.\n\n\n * 5d) is redundant, given the existence of section 5b) and the word "prominent" therein, and the existence of section 4, para 1 (which is applicable because of the inclusion of the terms of 4 into 5 - see 5 para 1).\n\n\n * Examples such as menus are bad because they will date the license and may force a particular inappropriate implementation of the underlying goal.\n\n\n * The current way 5d) is phrased is close to being (some would say, is) a restriction on modification, thereby violating freedom 1.\n\n\n * The current version of 5d) is more intrusive in that regard than the version in GPLv2, and applies in more circumstances.\n\n\n * 5d) should be written to focus on policy, not mechanism - as was done to turn the requirement to surrender keys into the requirement to provide Installation Instructions.\n\n\nI believe 5d) should be scrapped entirely but, if that is not to be, perhaps something like the following wording would achieve the goal while alleviating the problems in the latter half of the list:\n\n\n d) If the work has interactive user interfaces, users of each interface must have an easily-discoverable mechanism to display an appropriate copyright notice, and to be told:\n\n\n - the extent of the warranty, or lack of it, for the work;\n\n\n - that licensees may convey the work \nunder this License;\n\n\n - how to view a copy of this License.\n\n\nHowever, if the work has interactive interfaces that do not comply with this subsection, your changes need not make them comply.\n\n\nThis rephrasing removes most things that might be construed as a restriction on modification, and becomes solely an expression of policy - you must make users aware of their rights.\ngpl3.distribmod.p4.s1\ngerv\nlogin to agree\n0\n2974\n\n\n", "3073": "\n\n\n\n\ngpl3.distribmod.p4.s1\njamesj\nlogin to agree\n0\n3073\n\n\n", "2451": "\n\n\n\nIn the interest of space, the license should provide examples in this way only on important points, such as the precise definition of corresponding source. It is not necessary, and in fact distasteful, to dictate that the program should add "a list of user commands" and so forth.\ngpl3.distribmod.p4.s2\ncyd\nlogin to agree\n2\n2451\n\n\n", "2502:2503:2509:2800": "\n\n\n\n<p style="display: inline">This confuses me; either it contradicts the requirement or I failed to grasp the difference between "The Program" and "the work". Would it be too basic to define "The Work" in section 0?</p><p>\nI presume the idea is that you need not add the feature, and the original authors need not supply it (they do not rely on the GPL for anything). However, since 5d starts with a requirement, the language makes me wonder whether I missed something important.</p><p>\nSomehow the extension to "copyright-like" laws made by uneasy about this.</p>\ngpl3.distribmod.p4.s3\nhawk\nlogin to agree\n0\n2502\n\n\n<p style="display: inline">The rationale documents describe changes made to the language in 5d but they don't explain why the subsection is necessary. Isn't the language in 5b sufficient to ensure the user is informed of his rights?</p><p>\n5d requires licensees distributing certain types of works, but not others, to make functional modifications to the work over-and-above the necessary documentary modifications. This seems both onerous (the licensee may only want to modify the non-interface parts of the work) and arbitrary (licensees of certain works -- those whose interactive interfaces do not comply with 5d -- are exempt).</p>\ngpl3.distribmod.p4.s3\nfitzsim\nlogin to agree\n1\n2503\n\n\nSeems like this means that all new interactive interfaces must display such notices, whether or not the Program does so with its other interfaces (if any). Why not just say that such notices may not be removed from interactive interfaces that already have them?\ngpl3.distribmod.p4.s3\nraulir\nlogin to agree\n4\n2509\n\n\n<p style="display: inline">I simply don't get this. This subsection doesn't clearly explain how one has a compliant or non-compliant user interface. Interpreted at face value, it would seem that not providing such notifications would be non-compliance. If so, the last clause says "If you don't comply, you don't have to".</p><p>\nClarification here is needed.</p>\ngpl3.distribmod.p4.s3\nkdean06\nlogin to agree\n0\n2800\n\n\n", "2502:2503:2509": "\n\n\n\n<p style="display: inline">This confuses me; either it contradicts the requirement or I failed to grasp the difference between "The Program" and "the work". Would it be too basic to define "The Work" in section 0?</p><p>\nI presume the idea is that you need not add the feature, and the original authors need not supply it (they do not rely on the GPL for anything). However, since 5d starts with a requirement, the language makes me wonder whether I missed something important.</p><p>\nSomehow the extension to "copyright-like" laws made by uneasy about this.</p>\ngpl3.distribmod.p4.s3\nhawk\nlogin to agree\n0\n2502\n\n\n<p style="display: inline">The rationale documents describe changes made to the language in 5d but they don't explain why the subsection is necessary. Isn't the language in 5b sufficient to ensure the user is informed of his rights?</p><p>\n5d requires licensees distributing certain types of works, but not others, to make functional modifications to the work over-and-above the necessary documentary modifications. This seems both onerous (the licensee may only want to modify the non-interface parts of the work) and arbitrary (licensees of certain works -- those whose interactive interfaces do not comply with 5d -- are exempt).</p>\ngpl3.distribmod.p4.s3\nfitzsim\nlogin to agree\n1\n2503\n\n\nSeems like this means that all new interactive interfaces must display such notices, whether or not the Program does so with its other interfaces (if any). Why not just say that such notices may not be removed from interactive interfaces that already have them?\ngpl3.distribmod.p4.s3\nraulir\nlogin to agree\n4\n2509\n\n\n", "2502:2503:2509:2553": "\n\n\n\n<p style="display: inline">This confuses me; either it contradicts the requirement or I failed to grasp the difference between "The Program" and "the work". Would it be too basic to define "The Work" in section 0?</p><p>\nI presume the idea is that you need not add the feature, and the original authors need not supply it (they do not rely on the GPL for anything). However, since 5d starts with a requirement, the language makes me wonder whether I missed something important.</p><p>\nSomehow the extension to "copyright-like" laws made by uneasy about this.</p>\ngpl3.distribmod.p4.s3\nhawk\nlogin to agree\n0\n2502\n\n\n<p style="display: inline">The rationale documents describe changes made to the language in 5d but they don't explain why the subsection is necessary. Isn't the language in 5b sufficient to ensure the user is informed of his rights?</p><p>\n5d requires licensees distributing certain types of works, but not others, to make functional modifications to the work over-and-above the necessary documentary modifications. This seems both onerous (the licensee may only want to modify the non-interface parts of the work) and arbitrary (licensees of certain works -- those whose interactive interfaces do not comply with 5d -- are exempt).</p>\ngpl3.distribmod.p4.s3\nfitzsim\nlogin to agree\n1\n2503\n\n\nSeems like this means that all new interactive interfaces must display such notices, whether or not the Program does so with its other interfaces (if any). Why not just say that such notices may not be removed from interactive interfaces that already have them?\ngpl3.distribmod.p4.s3\nraulir\nlogin to agree\n4\n2509\n\n\nShould this be "changes in a modified work not make them comply." It sounds like the intent here is to not force people who modify code that doesn't comply with this section to add a compliant interface. The phrase "your work" seems to include the original author, and thus seems to make the section optional.\ngpl3.distribmod.p4.s3\nmole\nlogin to agree\n0\n2553\n\n\n", "2502:2503:2509:2553:2637:2764:2788:2799:2800:2850:2865:2932:2972:2974:3073:3082": "\n\n\n\n<p style="display: inline">This confuses me; either it contradicts the requirement or I failed to grasp the difference between "The Program" and "the work". Would it be too basic to define "The Work" in section 0?</p><p>\nI presume the idea is that you need not add the feature, and the original authors need not supply it (they do not rely on the GPL for anything). However, since 5d starts with a requirement, the language makes me wonder whether I missed something important.</p><p>\nSomehow the extension to "copyright-like" laws made by uneasy about this.</p>\ngpl3.distribmod.p4.s3\nhawk\nlogin to agree\n0\n2502\n\n\n<p style="display: inline">The rationale documents describe changes made to the language in 5d but they don't explain why the subsection is necessary. Isn't the language in 5b sufficient to ensure the user is informed of his rights?</p><p>\n5d requires licensees distributing certain types of works, but not others, to make functional modifications to the work over-and-above the necessary documentary modifications. This seems both onerous (the licensee may only want to modify the non-interface parts of the work) and arbitrary (licensees of certain works -- those whose interactive interfaces do not comply with 5d -- are exempt).</p>\ngpl3.distribmod.p4.s3\nfitzsim\nlogin to agree\n1\n2503\n\n\nSeems like this means that all new interactive interfaces must display such notices, whether or not the Program does so with its other interfaces (if any). Why not just say that such notices may not be removed from interactive interfaces that already have them?\ngpl3.distribmod.p4.s3\nraulir\nlogin to agree\n4\n2509\n\n\nShould this be "changes in a modified work not make them comply." It sounds like the intent here is to not force people who modify code that doesn't comply with this section to add a compliant interface. The phrase "your work" seems to include the original author, and thus seems to make the section optional.\ngpl3.distribmod.p4.s3\nmole\nlogin to agree\n0\n2553\n\n\nThis smells like an advertising clause (albeit for the license rather than the authors). I would prefer that users simply be forbidden from removing any such notice.\ngpl3.distribmod.p4.s1\nsamj\nlogin to agree\n0\n2637\n\n\n<p style="display: inline">Clause 5d in GPLv3draft3 is basically unchanged with respect to previous drafts. It's worse than the corresponding clause 2c in GPLv2... :-(</p><p>\nIt's an inconvenience and border-line with respect to freeness. Actually this clause restricts how I can modify what an interactive program does when run. It mandates a feature that I *must* implement in *any* interactive interface of my modified work. It's very close to place an unacceptable restriction on modification. What is more awkward is that it seems that when a non-interactive work is modified so that it becomes an interactive work, the modifier is *compelled* to implement these features in *any* newly created interactive interface...</p><p>\nI would like to see clause 5d dropped entirely.</p>\ngpl3.distribmod.p4.s1\nfrx\nlogin to agree\n3\n2764\n\n\nIs it too strong to require that *each* interactive user interface include the feature? Perhaps instead "the main interface or menu must include..."?\ngpl3.distribmod.p4.s1\nsanjoy\nlogin to agree\n0\n2788\n\n\n<p style="display: inline">I believe in some cases this conflicts directly. \n\nFirstly, I believe this steps on Freedom 1, dictating specifically how one must write or design a program.</p><p>\nIn many cases, optimization for space is important, specifically to embedded devices. Also, when designing web application, ease of navigation is important. This clause would restrict a user from removing such notice to reduce space or improve navigation, hence, limiting one's ability to improve a program.</p>\ngpl3.distribmod.p4.s1\nkdean06\nlogin to agree\n3\n2799\n\n\n<p style="display: inline">I simply don't get this. This subsection doesn't clearly explain how one has a compliant or non-compliant user interface. Interpreted at face value, it would seem that not providing such notifications would be non-compliance. If so, the last clause says "If you don't comply, you don't have to".</p><p>\nClarification here is needed.</p>\ngpl3.distribmod.p4.s3\nkdean06\nlogin to agree\n0\n2800\n\n\n<p style="display: inline">I recommend the following language as a replacement:</p><p>\nIf the work has interactive user interfaces, each must a) display an appropriate copyright notice, b) inform the user that there is no warranty for the work, c) provide instructions on how to view a copy of this License, and d) inform licensees that they may convey the work under this License.</p>\ngpl3.distribmod.p4.s1\njkoenig\nlogin to agree\n0\n2850\n\n\n<p style="display: inline">Suggestion: focus on results, not mechanism. Rewrite this section to define what the result must be, not how to achieve it. E.g.:</p><p>\nIf the work has interactive user interfaces a user must be conveniently able to view an appropriate copyright notice and a copy of this License, and to discover there is no warranty for the work (unless you provide one) and that licensees may convey the work under this License.</p><p>\nHowever, if the work has interactive interfaces that do not comply with this subsection, your changes need not make them comply."\n\n</p>\ngpl3.distribmod.p4.s1\ngerv\nlogin to agree\n0\n2865\n\n\n<p style="display: inline">\nI feel uncomfortable with the provisions of clause 5d about user\ninterfaces. In a way it reminds me of the historically obnoxious BSD\nclause.\nWhat happens to miniature devices, whose interactive displays have no\nroom for the long messages required to comply with clause 5d?\n --- Omer</p><p>\n--\nDelay is the deadliest form of denial. C. Northcote Parkinson\nMy own blog is at http://tddpirate.livejournal.com/</p><p>\nMy opinions, as expressed in this E-mail message, are mine alone.\nThey do not represent the official policy of any organization with which\nI may be affiliated in any way.\nWARNING TO SPAMMERS: at http://www.zak.co.il/spamwarning.html\n\n\n</p>\ngpl3.distribmod.p4.s1\nw1\nlogin to agree\n1\n2932\n\n\n<p style="display: inline">As I understand it, the purpose of this section is to make sure that users of the program know their rights. However, I believe it has several practical and philosophical problems. In true lawyerly fashion, I will try multiple possible arguments (some of which have been made by others in this comments system):\n\n * 5d) is misplaced in section 5, given that it really relates to object code forms and yet 5 is a source code section.\n * 5d) is redundant, given the existence of section 5b) and the word "prominent" therein, and the existence of section 4, para 1 (which is applicable because of the inclusion of the terms of 4 into 5 - see 5 para 1).\n * Examples such as menus are bad because they will date the license and may force a particular inappropriate implementation of the underlying goal.\n * The current way 5d) is phrased is close to being (some would say, is) a restriction on modification, thereby violating freedom 1.\n * The current version of 5d) is more intrusive in that regard than the version in GPLv2, and applies in more circumstances.\n * 5d) should be written to focus on policy, not mechanism - as was done to turn the requirement to surrender keys into the requirement to provide Installation Instructions.</p><p>\nI believe 5d) should be scrapped entirely but, if that is not to be, perhaps something like the following wording would achieve the goal while alleviating the problems in the latter half of the list:\n\n d) If the work has interactive user interfaces, users of each interface must have an easily-discoverable mechanism to display an appropriate copyright notice, and to be told:\n\n - the extent of the warranty, or lack of it, for the work;\n - that licensees may convey the work under this License;\n - how to view a copy of this License.</p><p>\nHowever, if the work has interactive interfaces that do not comply with this subsection, your changes need not make them comply.</p><p>\nThis rephrasing removes most things that might be construed as a restriction on modification, and becomes solely an expression of policy - you must make users aware of their rights.\n</p>\ngpl3.distribmod.p4.s1\ngerv\nlogin to agree\n2\n2972\n\n\nAs I understand it, the purpose of this section is to make sure that users of the program know their rights. However, I believe it has several practical and philosophical problems. In true lawyerly fashion, I will try multiple possible arguments (some of which have been made by others in this comments system):\n\n\n * 5d) is misplaced in section 5, given that it really relates to object code forms and yet 5 is a source code section.\n\n\n * 5d) is redundant, given the existence of section 5b) and the word "prominent" therein, and the existence of section 4, para 1 (which is applicable because of the inclusion of the terms of 4 into 5 - see 5 para 1).\n\n\n * Examples such as menus are bad because they will date the license and may force a particular inappropriate implementation of the underlying goal.\n\n\n * The current way 5d) is phrased is close to being (some would say, is) a restriction on modification, thereby violating freedom 1.\n\n\n * The current version of 5d) is more intrusive in that regard than the version in GPLv2, and applies in more circumstances.\n\n\n * 5d) should be written to focus on policy, not mechanism - as was done to turn the requirement to surrender keys into the requirement to provide Installation Instructions.\n\n\nI believe 5d) should be scrapped entirely but, if that is not to be, perhaps something like the following wording would achieve the goal while alleviating the problems in the latter half of the list:\n\n\n d) If the work has interactive user interfaces, users of each interface must have an easily-discoverable mechanism to display an appropriate copyright notice, and to be told:\n\n\n - the extent of the warranty, or lack of it, for the work;\n\n\n - that licensees may convey the work \nunder this License;\n\n\n - how to view a copy of this License.\n\n\nHowever, if the work has interactive interfaces that do not comply with this subsection, your changes need not make them comply.\n\n\nThis rephrasing removes most things that might be construed as a restriction on modification, and becomes solely an expression of policy - you must make users aware of their rights.\ngpl3.distribmod.p4.s1\ngerv\nlogin to agree\n0\n2974\n\n\n\ngpl3.distribmod.p4.s1\njamesj\nlogin to agree\n0\n3073\n\n\n<p style="display: inline">If you don't want to remove 5d entirely, make it a permissible additional term according to section 7.<br><br></p><p>\nAs stated in other comments, there are good reasons for removing this section entirely. If the FSF doesn't want to do this for whatever reasons, then a compromise would be to move this to section 7.Additional Terms. After all, this is closely related to "requiring preservation of ..." and "prohibiting misrepresentation of..."<br><br></p><p>\nThis way, licensors can decide if they want to impose this restriction on licensees. The statement "However, if the Program...need not make them comply" already acknowledges the current practice that some authors of free software (me, for instance) decide not to include a complying message (and as observed elsewhere, even some GNU software such as bash doesn't have this message). So this is really better placed in section 7 Additional Terms.</p>\ngpl3.distribmod.p4.s1\nmux2005\nlogin to agree\n0\n3082\n\n\n", "2980": "\n\n\n\nThis term is not defined in the rest of the license, and its meaning is not obvious. What constitutes an "extension" of the covered work? Is that like extensions for Firefox? Or something else?\ngpl3.distribmod.p5.s1\ngerv\nlogin to agree\n0\n2980\n\n\n", "2627": "\n\n\n\nSuggest removal as it is software specific and does not appear to add value (does an ISO on an FTP site qualify for example).\ngpl3.distribmod.p5.s1\nsamj\nlogin to agree\n2\n2627\n\n\n", "2979": "\n\n\n\nThe definition of Aggregate is inconsistent with the other definitions, in that it does not capitalise the word, and does not begin 'An "Aggregate" is...".\ngpl3.distribmod.p5.s1\ngerv\nlogin to agree\n0\n2979\n\n\n", "2997": "\n\n\n\n<p style="display: inline">This clause appears to add complexity to the license without adding any utility or protections. I suggest deleting it.</p><p>\nIf a Covered Work is Conveyed in a compilation with other works restrictions to the compilation can't restrict the Covered Work more tightly than the GPL (e.g. I bundle GPL software with proprietary software on a CD, I can't legitimately claim Copyright, All Rights Reserved on the entire compilation.) How I restrict other works on the compilation is irrelevant to this license, I just can't put a restriction on the entire compilation that violates the license terms for the Covered Work. All this seems to be explicit elsewhere in the GPL text, so this clause seems 1) irrelevant, and 2) to be attempting to restrict limitations that can be applied to other software distributed in the same compilation as GPL software. </p>\ngpl3.distribmod.p5.s1\nmole\nlogin to agree\n0\n2997\n\n\n", "2716:2997": "\n\n\n\nThis should be clarified a bit, I think. If GPL works and proprietary are aggregated, the compilation license can say that the compilation *as a whole* can't be redistributed; however, it can't restrict the GPL software from being redistributed/modified/etc. separately.\ngpl3.distribmod.p5.s1\nflaschen\nlogin to agree\n2\n2716\n\n\n<p style="display: inline">This clause appears to add complexity to the license without adding any utility or protections. I suggest deleting it.</p><p>\nIf a Covered Work is Conveyed in a compilation with other works restrictions to the compilation can't restrict the Covered Work more tightly than the GPL (e.g. I bundle GPL software with proprietary software on a CD, I can't legitimately claim Copyright, All Rights Reserved on the entire compilation.) How I restrict other works on the compilation is irrelevant to this license, I just can't put a restriction on the entire compilation that violates the license terms for the Covered Work. All this seems to be explicit elsewhere in the GPL text, so this clause seems 1) irrelevant, and 2) to be attempting to restrict limitations that can be applied to other software distributed in the same compilation as GPL software. </p>\ngpl3.distribmod.p5.s1\nmole\nlogin to agree\n0\n2997\n\n\n", "2967": "\n\n\n\nIn 5[2].d:Inclusion of a covered work in an aggregate does not cause this License to apply to the other parts of the aggregate.\r\n\r\nI suggest test be added to prohibit someone from placing restrictions on redistributing GPL'd source or binary, even if that someone elects to commingle their trademarked material with it. I intend this specifically to prohibit Red Hat from claiming that someone may not duplicate their CDs/DVDs because of their absurd "red hat" trademark.\r\n\r\nIMHO Red Hat has gone over to the dark side and their restriction, based on their claime of trademark infringement, violates both the spirit and the letter of copyright law w.r.t. the previous versions of the GPL. Possible wording might be:\r\n\r\n"Commingling GPL-protected material with non-GPL-protected material (whether non-free source, trademarked or patented material) on distribution medium implies permission for others to further distribute and duplicate said medium in accordance to the GPL and without further restrictions."\ngpl3.distribmod.p5.s2\nBobToxen\nlogin to agree\n1\n2967\n\n\n", "2967:2979:2980:2996:2997": "\n\n\n\nIn 5[2].d:Inclusion of a covered work in an aggregate does not cause this License to apply to the other parts of the aggregate.\r\n\r\nI suggest test be added to prohibit someone from placing restrictions on redistributing GPL'd source or binary, even if that someone elects to commingle their trademarked material with it. I intend this specifically to prohibit Red Hat from claiming that someone may not duplicate their CDs/DVDs because of their absurd "red hat" trademark.\r\n\r\nIMHO Red Hat has gone over to the dark side and their restriction, based on their claime of trademark infringement, violates both the spirit and the letter of copyright law w.r.t. the previous versions of the GPL. Possible wording might be:\r\n\r\n"Commingling GPL-protected material with non-GPL-protected material (whether non-free source, trademarked or patented material) on distribution medium implies permission for others to further distribute and duplicate said medium in accordance to the GPL and without further restrictions."\ngpl3.distribmod.p5.s2\nBobToxen\nlogin to agree\n1\n2967\n\n\nThe definition of Aggregate is inconsistent with the other definitions, in that it does not capitalise the word, and does not begin 'An "Aggregate" is...".\ngpl3.distribmod.p5.s1\ngerv\nlogin to agree\n0\n2979\n\n\nThis term is not defined in the rest of the license, and its meaning is not obvious. What constitutes an "extension" of the covered work? Is that like extensions for Firefox? Or something else?\ngpl3.distribmod.p5.s1\ngerv\nlogin to agree\n0\n2980\n\n\nRephrase to: "An Aggregate is a compilation...." \ngpl3.distribmod.p5.s1\nmole\nlogin to agree\n0\n2996\n\n\n<p style="display: inline">This clause appears to add complexity to the license without adding any utility or protections. I suggest deleting it.</p><p>\nIf a Covered Work is Conveyed in a compilation with other works restrictions to the compilation can't restrict the Covered Work more tightly than the GPL (e.g. I bundle GPL software with proprietary software on a CD, I can't legitimately claim Copyright, All Rights Reserved on the entire compilation.) How I restrict other works on the compilation is irrelevant to this license, I just can't put a restriction on the entire compilation that violates the license terms for the Covered Work. All this seems to be explicit elsewhere in the GPL text, so this clause seems 1) irrelevant, and 2) to be attempting to restrict limitations that can be applied to other software distributed in the same compilation as GPL software. </p>\ngpl3.distribmod.p5.s1\nmole\nlogin to agree\n0\n2997\n\n\n", "2544:2738:2927:2983": "\n\n\n\n<p style="display: inline">I'd like to give away copies of a freely copyable binary to a software user group in order to demonstrate that I'd produced the underlying source code to an enhancement that the group would probably pool funds to pay for.</p><p>\nIn other words, in order for the GPL to assist in rewarding authors for production and publication of GPL software shouldn't it enable authors to promote their as yet unreleased work by allowing them to publish a free demo in binary form?</p><p>\n<em>f) Convey the object code at no charge for demonstration purposes as long as the corresponding source has not yet been conveyed to any other party (outside yourself or your organisation).</em></p><p>\nSo, authors can give away binaries as long as they haven't conveyed the source. Once they have conveyed the source (presumably given payment), then they can no longer convey binaries via this method.</p><p>\nI'd also say that the demo binaries should clearly identify themselves as demos of GPL software, that they are freely copyable and, precisely because their purpose is to demonstrate and promote the sale of unreleased source code, are exempt from the requirement for accompanying source code.\n\n</p>\ngpl3.nonsource.0.0\ncrosbie\nlogin to agree\n1\n2544\n\n\nUnless I'm misreading this, it seems to preclude the possibility of providing object code over a network, and source code by mail. What is wrong with this situation? Maybe (c) and (e) allow this somewhat, but those appear to be limited special cases.\ngpl3.nonsource.0.0\nmkorman\nlogin to agree\n0\n2738\n\n\n<p style="display: inline">In so far as prohibiting Tivoisation, this section has a big loophole: It only applies to non-source forms.</p><p>\nTivo (or anyone) can circumvent this section simply by conveying their tivoised code only in source form. Section 5 does not make any provision against tivoisation.</p>\ngpl3.nonsource.0.0\njmd\nlogin to agree\n1\n2927\n\n\n<p style="display: inline">Different rules for user products is not the way to deal with this issue.</p><p>\nThe annotated version of the license says that some business users want the option of having hardware that will only run particular versions of the software.</p><p>\nI could guess why some users, business and home users, would want hardware that would not allow changes to be made.</p><p>\nAssuming that there is indeed an advantage to having hardware lockouts shouldn't home users have the option of the advantage?</p><p>\nA clause that deals with locked hardware should deal with the reasons some people and businesses would want locked hardware.</p><p>\nDealing with these issues directly would be much more eloquent.</p>\ngpl3.nonsource.0.0\ndogshed\nlogin to agree\n1\n2983\n\n\n", "3083": "\n\n\n\nThe introduction says "convey the [...] Source [...] in one of these ways:" which I understand to mean that the following paragraphs are ways for conveying the source, but they aren't. They are ways for conveying object code and source. Please rephrase. I suggest to simply remove "in one of these ways:". Or replace it with ". You may:"\ngpl3.nonsource.p0.s1\nmux2005\nlogin to agree\n0\n3083\n\n\n", "2441:2478:2697:3083": "\n\n\n\nShouldn't this instead be 'preferred form for modification'? A machine can read a book using OCR (optical characters recognition) or read a DRM restricted file with the correct key, but neither is suitable for easy modification.\ngpl3.nonsource.p0.s1\ntcort\nlogin to agree\n7\n2441\n\n\nNo need to add the "machine-readable". A printout of the C source is not the preferred form for modification so it's not Corresponding Source\ngpl3.nonsource.p0.s1\nmerijn\nlogin to agree\n5\n2478\n\n\n"machine-readable" seems a bit vague. What if I claim I have built a special machine which can read my source code (and no other machine can read it). Or what if I claim it is "machine readable", but you need a special program (which only I have a copy of) in order to read it ?\ngpl3.nonsource.p0.s1\nsalsaman\nlogin to agree\n1\n2697\n\n\nThe introduction says "convey the [...] Source [...] in one of these ways:" which I understand to mean that the following paragraphs are ways for conveying the source, but they aren't. They are ways for conveying object code and source. Please rephrase. I suggest to simply remove "in one of these ways:". Or replace it with ". You may:"\ngpl3.nonsource.p0.s1\nmux2005\nlogin to agree\n0\n3083\n\n\n", "2441:2478:3083": "\n\n\n\nShouldn't this instead be 'preferred form for modification'? A machine can read a book using OCR (optical characters recognition) or read a DRM restricted file with the correct key, but neither is suitable for easy modification.\ngpl3.nonsource.p0.s1\ntcort\nlogin to agree\n7\n2441\n\n\nNo need to add the "machine-readable". A printout of the C source is not the preferred form for modification so it's not Corresponding Source\ngpl3.nonsource.p0.s1\nmerijn\nlogin to agree\n5\n2478\n\n\nThe introduction says "convey the [...] Source [...] in one of these ways:" which I understand to mean that the following paragraphs are ways for conveying the source, but they aren't. They are ways for conveying object code and source. Please rephrase. I suggest to simply remove "in one of these ways:". Or replace it with ". You may:"\ngpl3.nonsource.p0.s1\nmux2005\nlogin to agree\n0\n3083\n\n\n", "2638": "\n\n\n\nUse mixed case for defined term 'Object Code'.\ngpl3.nonsource.p1.s1\nsamj\nlogin to agree\n0\n2638\n\n\n", "2494:3084:3088": "\n\n\n\nWe all know that original Emacs used the magnetic tape route. Is that a "customarily used" medium? Probably not anymore. Are floppies? There it's a bit gray'er. Perhaps a more explicit list is wise here with a method for adding to the list (???). The danger is that one can choose a significantly difficult medium (e.g. DEC TK-70 format) that sort of ensures the source never gets read.\ngpl3.nonsource.p1.s1\ncjcox\nlogin to agree\n1\n2494\n\n\nThe term "durable" is not well-defined and should be removed entirely. What about CD-Rs or DVD-Rs? Are those durable? I've seen reports about failure after just 1 or 2 years of storage (under bad conditions). I think the word "durable" should be removed. I don't see any benefits of it. The important part is that the medium be "customarily used for software interchange". This is the requirement that makes sure recipients will be able to read the Source. As long as what I get is something my computer can read, I don't care about durability, because I can always make a backup copy. If some environmentally friendly form of CD-R is ever invented that decomposes after 6 months I think it's perfectly okay for someone to use that to convey the Corresponding Source, even though it's not durable.\ngpl3.nonsource.p1.s1\nmux2005\nlogin to agree\n1\n3084\n\n\n<p style="display: inline">The intent of "durable" and "used for software interchange" are pretty loose. Would the intent of the license be satisfied by including the source code for an embedded device in the flash memory on the device (with a means to allow copying it from the device)?</p><p>\nI tend to think it should be, but is it really "a medium customarily used for software interchange"?\n</p>\ngpl3.nonsource.p1.s1\nsepreece\nlogin to agree\n0\n3088\n\n\n", "2494:3088": "\n\n\n\nWe all know that original Emacs used the magnetic tape route. Is that a "customarily used" medium? Probably not anymore. Are floppies? There it's a bit gray'er. Perhaps a more explicit list is wise here with a method for adding to the list (???). The danger is that one can choose a significantly difficult medium (e.g. DEC TK-70 format) that sort of ensures the source never gets read.\ngpl3.nonsource.p1.s1\ncjcox\nlogin to agree\n1\n2494\n\n\n<p style="display: inline">The intent of "durable" and "used for software interchange" are pretty loose. Would the intent of the license be satisfied by including the source code for an embedded device in the flash memory on the device (with a means to allow copying it from the device)?</p><p>\nI tend to think it should be, but is it really "a medium customarily used for software interchange"?\n</p>\ngpl3.nonsource.p1.s1\nsepreece\nlogin to agree\n0\n3088\n\n\n", "2494:2638:2639:3084:3088": "\n\n\n\nWe all know that original Emacs used the magnetic tape route. Is that a "customarily used" medium? Probably not anymore. Are floppies? There it's a bit gray'er. Perhaps a more explicit list is wise here with a method for adding to the list (???). The danger is that one can choose a significantly difficult medium (e.g. DEC TK-70 format) that sort of ensures the source never gets read.\ngpl3.nonsource.p1.s1\ncjcox\nlogin to agree\n1\n2494\n\n\nUse mixed case for defined term 'Object Code'.\ngpl3.nonsource.p1.s1\nsamj\nlogin to agree\n0\n2638\n\n\nDefine the term 'Physical Product' and use mixed case throughout.\ngpl3.nonsource.p1.s1\nsamj\nlogin to agree\n0\n2639\n\n\nThe term "durable" is not well-defined and should be removed entirely. What about CD-Rs or DVD-Rs? Are those durable? I've seen reports about failure after just 1 or 2 years of storage (under bad conditions). I think the word "durable" should be removed. I don't see any benefits of it. The important part is that the medium be "customarily used for software interchange". This is the requirement that makes sure recipients will be able to read the Source. As long as what I get is something my computer can read, I don't care about durability, because I can always make a backup copy. If some environmentally friendly form of CD-R is ever invented that decomposes after 6 months I think it's perfectly okay for someone to use that to convey the Corresponding Source, even though it's not durable.\ngpl3.nonsource.p1.s1\nmux2005\nlogin to agree\n1\n3084\n\n\n<p style="display: inline">The intent of "durable" and "used for software interchange" are pretty loose. Would the intent of the license be satisfied by including the source code for an embedded device in the flash memory on the device (with a means to allow copying it from the device)?</p><p>\nI tend to think it should be, but is it really "a medium customarily used for software interchange"?\n</p>\ngpl3.nonsource.p1.s1\nsepreece\nlogin to agree\n0\n3088\n\n\n", "3085": "\n\n\n\n<p style="display: inline">Is this whole paragraph just one long convoluted sentence? I can't find any period in it. Please dumb this down for those who don't have a master's degree in English. One simplification would be a structure like this:<br><br></p><p>\nConvey .. accompanied by a written offer to provide the Corresponding Source in one of the following ways (you need not offer both):<br></p><p>\n(1) on a physical medium customarily used for software interchange, for a price no more than your reasonable cost of physically performing this conveying of source.<br></p><p>\n(2) by providing access to copy the Corresponding Source from a network server at no charge.<br><br></p><p>\nThe offer must extend to anyone who possesses the object code. It has to cover the Corresponding Source for all the software in the product that is covered by this License. The offer must be valid for at least three years. If you offer spare parts or customer support for that product model for more than 3 years, the offer must be valid for that longer period of time.</p>\ngpl3.nonsource.p2.s1\nmux2005\nlogin to agree\n0\n3085\n\n\n", "2640": "\n\n\n\nUse mixed case for defined term 'Object Code'\ngpl3.nonsource.p2.s1\nsamj\nlogin to agree\n0\n2640\n\n\n", "2912:2922": "\n\n\n\n<p style="display: inline">Does the offer have to be written by you, or (say, if you are a distributor or retailer), can you pass on a written offer that you received? Does the offer have to be valid for 3 years from the time the product is manufactured, the time it is ordered, or the time it is delivered?</p><p>\nDo distributors or retailers even have to agree to the license?</p>\ngpl3.nonsource.p2.s1\njamesgnz\nlogin to agree\n0\n2912\n\n\n<p style="display: inline">Reword this so that distributors must:</p><p>\ni) offer the source</p><p>\nii) provide that offer in writing</p><p>\nOtherwise it could lead to legal shenanigans if you can prove that they distributed you a device, but you can't find the written offer.</p>\ngpl3.nonsource.p2.s1\njamesgnz\nlogin to agree\n0\n2922\n\n\n", "2912:2922:3093": "\n\n\n\n<p style="display: inline">Does the offer have to be written by you, or (say, if you are a distributor or retailer), can you pass on a written offer that you received? Does the offer have to be valid for 3 years from the time the product is manufactured, the time it is ordered, or the time it is delivered?</p><p>\nDo distributors or retailers even have to agree to the license?</p>\ngpl3.nonsource.p2.s1\njamesgnz\nlogin to agree\n0\n2912\n\n\n<p style="display: inline">Reword this so that distributors must:</p><p>\ni) offer the source</p><p>\nii) provide that offer in writing</p><p>\nOtherwise it could lead to legal shenanigans if you can prove that they distributed you a device, but you can't find the written offer.</p>\ngpl3.nonsource.p2.s1\njamesgnz\nlogin to agree\n0\n2922\n\n\n<p style="display: inline">Is it acceptable to limit the offer to making a single copy per recipient?</p><p>\nI don't see a problem with such a restriction, but the language should probably make it clear, with respect to all such offers, whether they have to be open-ended or can be limited to one copy.\n</p>\ngpl3.nonsource.p2.s1\nsepreece\nlogin to agree\n1\n3093\n\n\n", "2437:2479:2912": "\n\n\n\nThis statement is in the form: A and B or C. Maybe it could be reworded to "valid for at least three years. After the initial three years, the offer must remain valid for as long as you offer spare parts or customer support for that product model"\ngpl3.nonsource.p2.s1\ntcort\nlogin to agree\n8\n2437\n\n\n\ngpl3.nonsource.p2.s1\nmerijn\nlogin to agree\n0\n2479\n\n\n<p style="display: inline">Does the offer have to be written by you, or (say, if you are a distributor or retailer), can you pass on a written offer that you received? Does the offer have to be valid for 3 years from the time the product is manufactured, the time it is ordered, or the time it is delivered?</p><p>\nDo distributors or retailers even have to agree to the license?</p>\ngpl3.nonsource.p2.s1\njamesgnz\nlogin to agree\n0\n2912\n\n\n", "2437:2479:2866:2912": "\n\n\n\nThis statement is in the form: A and B or C. Maybe it could be reworded to "valid for at least three years. After the initial three years, the offer must remain valid for as long as you offer spare parts or customer support for that product model"\ngpl3.nonsource.p2.s1\ntcort\nlogin to agree\n8\n2437\n\n\n\ngpl3.nonsource.p2.s1\nmerijn\nlogin to agree\n0\n2479\n\n\n<p style="display: inline">When the GPL was written, software evolved at a slower pace than it does now. GNOME makes an entire new release every six months.</p><p>\nGiven the new stipulation that source has to be available for as long as support is, I think we could reduce the absolute time. </p>\ngpl3.nonsource.p2.s1\ngerv\nlogin to agree\n1\n2866\n\n\n<p style="display: inline">Does the offer have to be written by you, or (say, if you are a distributor or retailer), can you pass on a written offer that you received? Does the offer have to be valid for 3 years from the time the product is manufactured, the time it is ordered, or the time it is delivered?</p><p>\nDo distributors or retailers even have to agree to the license?</p>\ngpl3.nonsource.p2.s1\njamesgnz\nlogin to agree\n0\n2912\n\n\n", "2437:2479": "\n\n\n\nThis statement is in the form: A and B or C. Maybe it could be reworded to "valid for at least three years. After the initial three years, the offer must remain valid for as long as you offer spare parts or customer support for that product model"\ngpl3.nonsource.p2.s1\ntcort\nlogin to agree\n8\n2437\n\n\n\ngpl3.nonsource.p2.s1\nmerijn\nlogin to agree\n0\n2479\n\n\n", "2437": "\n\n\n\nThis statement is in the form: A and B or C. Maybe it could be reworded to "valid for at least three years. After the initial three years, the offer must remain valid for as long as you offer spare parts or customer support for that product model"\ngpl3.nonsource.p2.s1\ntcort\nlogin to agree\n8\n2437\n\n\n", "2924": "\n\n\n\nI order GPL CDs from someone here in New Zealand. The person I get the CDs from says they don't make a lot of money, and I believe them. I think the CDs are reasonably priced. I wonder, if he was required to sell source CDs at cost, whether it would be worth his while distributing GPL software at all. I think distribution should be allowed at twice the cost of performing it.\ngpl3.nonsource.p2.s1\njamesgnz\nlogin to agree\n0\n2924\n\n\n", "3032": "\n\n\n\nIf the object code is being distributed on a physical medium, so should the source. Otherwise, a great many people that don't have access to the Internet wouldn't be able to get the source code. This is especially true of larger programs, where even just having dial-up may prevent the practical acquisition of source code for larger files. \n\nThe source code should be provided in whatever form the object code was. That is if you distribute a binary online, you can allow downloads for the source code and be covered. But if you give the object code in a physical form, you should be obligated to do the same for the source.\ngpl3.nonsource.p2.s1\nLovasco\nlogin to agree\n0\n3032\n\n\n", "2437:2462:2479:2480:2511:2546:2640:2674:2717:2718:2721:2722:2786:2831:2851:2866:2911:2912:2913:2922:2924:2926:2943:2944:2963:2981:3032:3034:3085:3086:3087:3093:3109": "\n\n\n\nThis statement is in the form: A and B or C. Maybe it could be reworded to "valid for at least three years. After the initial three years, the offer must remain valid for as long as you offer spare parts or customer support for that product model"\ngpl3.nonsource.p2.s1\ntcort\nlogin to agree\n8\n2437\n\n\nThere is no mention of who should operate the network server in b) (2), contrary to option d) that requires the distributor to ensure that the server remains available. It could be tempting for some distributors to point to a third-party server that they do not control (and do not pay for). I am not convinced that the requirement for the offer to be "valid for at least three years..." would be sufficient to discourage some distributors from pointing to a third-party server and just hope that it will stay around long enough, without contacting the owners of that server.\ngpl3.nonsource.p2.s1\nraphael\nlogin to agree\n0\n2462\n\n\n\ngpl3.nonsource.p2.s1\nmerijn\nlogin to agree\n0\n2479\n\n\n<p style="display: inline">Does this really mean, you must offer to provide access to download? No benefit in that compared to sending out a disk.</p><p>\nIt would make more sense to say "(1) an offer to send on durable medium, valid for 3 years, or (2) providing access for three years".</p>\ngpl3.nonsource.p2.s1\nmerijn\nlogin to agree\n0\n2480\n\n\nMost of the license is not restricted to software. Why here, then? There may be other relevant free content besides software, too.\ngpl3.nonsource.p2.s1\nraulir\nlogin to agree\n0\n2511\n\n\n<p style="display: inline">There is no easy way to verify\nif somebody possesses the objectcode.</p><p>\nSo in reality there either will be a fallback to the rule in gpl 2,\nor some clever person reqally requires\na prove about possesion of object code,\nbefore sendig source, \nwhich would make the whole thing much more \ncomplex for both sides.</p>\ngpl3.nonsource.p2.s1\nerge\nlogin to agree\n0\n2546\n\n\nUse mixed case for defined term 'Object Code'\ngpl3.nonsource.p2.s1\nsamj\nlogin to agree\n0\n2640\n\n\nI'm very happy that this convenient distribution option has remained in the draft.\ngpl3.nonsource.p2.s1\nsanjoy\nlogin to agree\n0\n2674\n\n\nPeople commented that this was undesirable during the previous draft. I think I agree; not everyone that buys a product with GPL code will have access to the internet.\ngpl3.nonsource.p2.s1\nflaschen\nlogin to agree\n0\n2717\n\n\nIt is conceivable that I might want the source code exactly because I want to rebuild the object code that has become damaged or lost. With this limitation (which does not exist in GPLv2) one would lose all the rights he had in GPLv2.\ngpl3.nonsource.p2.s1\nmnalis\nlogin to agree\n0\n2718\n\n\nThis was mentioned before, but not everyone will have access to a network server. This may be a mistake.\ngpl3.nonsource.p2.s1\nflaschen\nlogin to agree\n0\n2721\n\n\nNot everyone has network access.\ngpl3.nonsource.p2.s1\nflaschen\nlogin to agree\n0\n2722\n\n\nGreat! This network-distribution option is very useful and takes account of the spread of Internet distribution (I can't remember the last time I loaded software via durable physical media).\ngpl3.nonsource.p2.s1\nsanjoy\nlogin to agree\n0\n2786\n\n\nIt is conceivable that I might want the source code exactly because I want to rebuild the object code that has become damaged or lost. With this limitation (which does not exist in GPLv2) one would lose all the rights he had in GPLv2.\ngpl3.nonsource.p2.s1\nmnalis\nlogin to agree\n0\n2831\n\n\nMy interpretation of the GPLv2 was that it gave the whole world rights to source, whether or not a particular individual had an object code version in possession. As long as a single distribution tool place, the whole world could enforce the right to obtain source code. That was my interpretation. The GPLv3 departs by now requiring that the individual "possess" object code before demanding source code from the modifying party. I believe this is undesirable and creates the possibility of contrived technical and economic barriers to sharing source code. For instance, a modifying party could charge tens of thousands of dollars for a device containing the object code. It would constitute a distribution. However, most individuals would have no ability to obtain an object version, and thus would have no access to source, even though a modified distribution clearly took place. \ngpl3.nonsource.p2.s1\njkoenig\nlogin to agree\n0\n2851\n\n\n<p style="display: inline">When the GPL was written, software evolved at a slower pace than it does now. GNOME makes an entire new release every six months.</p><p>\nGiven the new stipulation that source has to be available for as long as support is, I think we could reduce the absolute time. </p>\ngpl3.nonsource.p2.s1\ngerv\nlogin to agree\n1\n2866\n\n\n<p style="display: inline">Shouldn't there be a requirement that the networked server be\naccessable from remote locations connected to the Internet,\nexcept for remote locations that themselves block access?</p><p>\nIn other words. A networked server available only to\nthose physically present within Guantanamno Bay is not acceptable.</p><p>\nThere are other places where the language uses the\nwords "networked server". Shouldn't there be\nsomething in the Definition section for this phrase?</p>\ngpl3.nonsource.p2.s1\nkop\nlogin to agree\n0\n2911\n\n\n<p style="display: inline">Does the offer have to be written by you, or (say, if you are a distributor or retailer), can you pass on a written offer that you received? Does the offer have to be valid for 3 years from the time the product is manufactured, the time it is ordered, or the time it is delivered?</p><p>\nDo distributors or retailers even have to agree to the license?</p>\ngpl3.nonsource.p2.s1\njamesgnz\nlogin to agree\n0\n2912\n\n\nSo, if I get someone to download and burn a CD with binary GPL code on it, then I can demand that they spend the next 3 years of their life downloading copies of the source code, at cost, for anyone who wants it?\ngpl3.nonsource.p2.s1\njamesgnz\nlogin to agree\n0\n2913\n\n\n<p style="display: inline">Reword this so that distributors must:</p><p>\ni) offer the source</p><p>\nii) provide that offer in writing</p><p>\nOtherwise it could lead to legal shenanigans if you can prove that they distributed you a device, but you can't find the written offer.</p>\ngpl3.nonsource.p2.s1\njamesgnz\nlogin to agree\n0\n2922\n\n\nI order GPL CDs from someone here in New Zealand. The person I get the CDs from says they don't make a lot of money, and I believe them. I think the CDs are reasonably priced. I wonder, if he was required to sell source CDs at cost, whether it would be worth his while distributing GPL software at all. I think distribution should be allowed at twice the cost of performing it.\ngpl3.nonsource.p2.s1\njamesgnz\nlogin to agree\n0\n2924\n\n\nThe FSF must know how much it costs to run a server. I would not begrudge them for charging a minimal fee for downloads. Allow twice the cost of bandwidth. This would be a pittance for individuals downloading a file, but it would add up, and help to pay for running the server.\ngpl3.nonsource.p2.s1\njamesgnz\nlogin to agree\n0\n2926\n\n\n<p style="display: inline">The parenthetical clause can be read two ways. As it is written, I would read it as meaning "a physical product that includes a physical distribution medium" - that is, treating the flash or other persistent storage as the medium.</p><p>\nThe other reading is that it means to say that a physical distribution medium is an example of a physical product.</p><p>\nI think the latter reading is what is intended.</p><p>\nThis could be handled by rewriting (perhaps making "A physical distribution medium is considered such a product." a separate sentence) or by adding a separate definition of "physical product" that covers all the appropriate cases.\n</p>\ngpl3.nonsource.p2.s1\nsepreece\nlogin to agree\n0\n2943\n\n\nPerhaps stipulate a reasonable time within which the source code must be produced? "Allow 28 days for delivery" comes to mind. Without a specified time, it is not clear at what point a party has failed to deliver on this requirement.\ngpl3.nonsource.p2.s1\njamesgnz\nlogin to agree\n0\n2944\n\n\n<p style="display: inline">The Corresponding Source should be available without the need for excessive personal information, or any information that could compromise the user. This should include, but is not limited to, the users SSN or gov't id number, a credit card number, date of birth, etc.</p><p>\nThis should apply anytime conveyance occurs over a network or peer-to-peer connection.</p>\ngpl3.nonsource.p2.s1\nghuber\nlogin to agree\n0\n2963\n\n\nIt should not be required to give everyone who possesses the object code access to the source, but only those who received the object code directly or indirectly from 'you' through 6b or 6c. I further suggest that distributors be allowed to require that people provide evidence, such as an id key from their written offer.\ngpl3.nonsource.p2.s1\njamesgnz\nlogin to agree\n1\n2981\n\n\nIf the object code is being distributed on a physical medium, so should the source. Otherwise, a great many people that don't have access to the Internet wouldn't be able to get the source code. This is especially true of larger programs, where even just having dial-up may prevent the practical acquisition of source code for larger files. \n\nThe source code should be provided in whatever form the object code was. That is if you distribute a binary online, you can allow downloads for the source code and be covered. But if you give the object code in a physical form, you should be obligated to do the same for the source.\ngpl3.nonsource.p2.s1\nLovasco\nlogin to agree\n0\n3032\n\n\nOffer to provide access to whom? Is this still only required to be an offer to anyone who possesses the object code? To anyone who possesses a copy of the offer? An original copy of the offer? Or just to anyone?\n\ngpl3.nonsource.p2.s1\nsepreece\nlogin to agree\n0\n3034\n\n\n<p style="display: inline">Is this whole paragraph just one long convoluted sentence? I can't find any period in it. Please dumb this down for those who don't have a master's degree in English. One simplification would be a structure like this:<br><br></p><p>\nConvey .. accompanied by a written offer to provide the Corresponding Source in one of the following ways (you need not offer both):<br></p><p>\n(1) on a physical medium customarily used for software interchange, for a price no more than your reasonable cost of physically performing this conveying of source.<br></p><p>\n(2) by providing access to copy the Corresponding Source from a network server at no charge.<br><br></p><p>\nThe offer must extend to anyone who possesses the object code. It has to cover the Corresponding Source for all the software in the product that is covered by this License. The offer must be valid for at least three years. If you offer spare parts or customer support for that product model for more than 3 years, the offer must be valid for that longer period of time.</p>\ngpl3.nonsource.p2.s1\nmux2005\nlogin to agree\n0\n3085\n\n\nWhy is the "for all the software in the product that is covered by this License" only in (1)? This way it seems as if it doesn't apply to (2). You should move this out of the (1) block so that it applies to both. On 2nd thought, why is it there in the 1st place? Isn't this in the definition of Corresponding Source already? After all that definition says "ALL the source code needed..."\ngpl3.nonsource.p2.s1\nmux2005\nlogin to agree\n0\n3086\n\n\nsee my other comment. I think "durable" should be removed everywhere.\ngpl3.nonsource.p2.s1\nmux2005\nlogin to agree\n0\n3087\n\n\n<p style="display: inline">Is it acceptable to limit the offer to making a single copy per recipient?</p><p>\nI don't see a problem with such a restriction, but the language should probably make it clear, with respect to all such offers, whether they have to be open-ended or can be limited to one copy.\n</p>\ngpl3.nonsource.p2.s1\nsepreece\nlogin to agree\n1\n3093\n\n\nDoes this mean you have to actually provide it on a server you run yourself, or can you just direct them to a server run by someone else where they can download the same source code at no charge (such as the place you originally downloaded it from)?\ngpl3.nonsource.p2.s1\njohnston\nlogin to agree\n0\n3109\n\n\n", "2921": "\n\n\n\n<p style="display: inline">What if the offer has already expired? Or what if it only has 1 week left to run?</p><p>\nI think 6c should be dropped.</p>\ngpl3.nonsource.p3.s1\njamesgnz\nlogin to agree\n0\n2921\n\n\n", "2438:2971:3030:3056": "\n\n\n\nWhat is this alternative allowed or not allowed? What is the exact meaning of 'only occasionally'?\ngpl3.nonsource.p3.s2\ntcort\nlogin to agree\n6\n2438\n\n\n<p style="display: inline">Someone reselling hardware accompanied by a section 6b) written offer is in trouble. They cannot pass on the offer, because that's allowed only "occasionally and noncommercially". So they must either make their own offer (and set up facilities to fulfil it) or download the source and burn it to a CD and put it in the box!</p><p>\nDetails: Let's imagine I'm in the business of buying consumer hardware in bulk and reselling it - a common business model. If I buy a pallet of routers which contain GPLv3 software and come with a 6b) written offer, can I resell them as-is? My reading of the definition of "propagate" and "convey" leads me to believe that this reselling is "conveying", which I may only do under one of sub-clauses a) to e) of section 6.</p><p>\nI am not covered by 6b) because it's not my offer, and I'm not offering support or spare parts. What about 6c)? The current wording suggests no, because it permits such conveying only "occasionally and non commercially", and my conveying is regular and commercial. Do I have to download the source, burn to a CD and pop it in the box (thereby being covered by 6a) before I can resell it? Surely the license must make it possible for me to merely pass on the boxes as I receive them.</p><p>\nMy suggestion is add another exception to the "occasionally and non commercially" one, to permit passing on copies you received. One might phrase a modified 6c) as follows:\n\n c) Convey individual copies of the object code with a copy of the written offer to provide the Corresponding Source. This alternative is allowed only if you received the object code with such an offer (in accord with subsection 6b) and either:\n\n - you are conveying copies that were conveyed to you; or\n\n - your conveying is occasional and non-commercial.</p><p>\nSo the extra exception permits a high-volume and commercial business model reselling items, yet one cannot use it to increase the number of copies the offer covers. The danger of a large increase of this type and the consequent unfair effect on the person making the written offer is, as I understand it, the reasoning behind the "occasional and non-commercial" wording.\n </p>\ngpl3.nonsource.p3.s2\ngerv\nlogin to agree\n2\n2971\n\n\nWhat does "only occasionally" mean? I'm not sure it is needed, anyway, given the noncommercial restriction.\ngpl3.nonsource.p3.s2\nLovasco\nlogin to agree\n1\n3030\n\n\nI think the "only occasionally" point is much to vague. I think it should be removed altogether, so that option is allowed for all non-commercial copying.\ngpl3.nonsource.p3.s2\njohnston\nlogin to agree\n0\n3056\n\n\n", "2971": "\n\n\n\n<p style="display: inline">Someone reselling hardware accompanied by a section 6b) written offer is in trouble. They cannot pass on the offer, because that's allowed only "occasionally and noncommercially". So they must either make their own offer (and set up facilities to fulfil it) or download the source and burn it to a CD and put it in the box!</p><p>\nDetails: Let's imagine I'm in the business of buying consumer hardware in bulk and reselling it - a common business model. If I buy a pallet of routers which contain GPLv3 software and come with a 6b) written offer, can I resell them as-is? My reading of the definition of "propagate" and "convey" leads me to believe that this reselling is "conveying", which I may only do under one of sub-clauses a) to e) of section 6.</p><p>\nI am not covered by 6b) because it's not my offer, and I'm not offering support or spare parts. What about 6c)? The current wording suggests no, because it permits such conveying only "occasionally and non commercially", and my conveying is regular and commercial. Do I have to download the source, burn to a CD and pop it in the box (thereby being covered by 6a) before I can resell it? Surely the license must make it possible for me to merely pass on the boxes as I receive them.</p><p>\nMy suggestion is add another exception to the "occasionally and non commercially" one, to permit passing on copies you received. One might phrase a modified 6c) as follows:\n\n c) Convey individual copies of the object code with a copy of the written offer to provide the Corresponding Source. This alternative is allowed only if you received the object code with such an offer (in accord with subsection 6b) and either:\n\n - you are conveying copies that were conveyed to you; or\n\n - your conveying is occasional and non-commercial.</p><p>\nSo the extra exception permits a high-volume and commercial business model reselling items, yet one cannot use it to increase the number of copies the offer covers. The danger of a large increase of this type and the consequent unfair effect on the person making the written offer is, as I understand it, the reasoning behind the "occasional and non-commercial" wording.\n </p>\ngpl3.nonsource.p3.s2\ngerv\nlogin to agree\n2\n2971\n\n\n", "2439:2463:2971": "\n\n\n\n<p style="display: inline">What does 'noncommercially' mean here?</p><p>\nDoes it mean... The source code is conveyed without a fee? The object code is conveyed without a fee? Both are conveyed without a fee? A bussiness isn't allowed to use this method?</p>\ngpl3.nonsource.p3.s2\ntcort\nlogin to agree\n3\n2439\n\n\nIf this word would be removed, GPLv3 wouldn't discriminate between commercial and noncommercial uses, at all. I think we would be better off if it would be that way.\ngpl3.nonsource.p3.s2\nkaol\nlogin to agree\n8\n2463\n\n\n<p style="display: inline">Someone reselling hardware accompanied by a section 6b) written offer is in trouble. They cannot pass on the offer, because that's allowed only "occasionally and noncommercially". So they must either make their own offer (and set up facilities to fulfil it) or download the source and burn it to a CD and put it in the box!</p><p>\nDetails: Let's imagine I'm in the business of buying consumer hardware in bulk and reselling it - a common business model. If I buy a pallet of routers which contain GPLv3 software and come with a 6b) written offer, can I resell them as-is? My reading of the definition of "propagate" and "convey" leads me to believe that this reselling is "conveying", which I may only do under one of sub-clauses a) to e) of section 6.</p><p>\nI am not covered by 6b) because it's not my offer, and I'm not offering support or spare parts. What about 6c)? The current wording suggests no, because it permits such conveying only "occasionally and non commercially", and my conveying is regular and commercial. Do I have to download the source, burn to a CD and pop it in the box (thereby being covered by 6a) before I can resell it? Surely the license must make it possible for me to merely pass on the boxes as I receive them.</p><p>\nMy suggestion is add another exception to the "occasionally and non commercially" one, to permit passing on copies you received. One might phrase a modified 6c) as follows:\n\n c) Convey individual copies of the object code with a copy of the written offer to provide the Corresponding Source. This alternative is allowed only if you received the object code with such an offer (in accord with subsection 6b) and either:\n\n - you are conveying copies that were conveyed to you; or\n\n - your conveying is occasional and non-commercial.</p><p>\nSo the extra exception permits a high-volume and commercial business model reselling items, yet one cannot use it to increase the number of copies the offer covers. The danger of a large increase of this type and the consequent unfair effect on the person making the written offer is, as I understand it, the reasoning behind the "occasional and non-commercial" wording.\n </p>\ngpl3.nonsource.p3.s2\ngerv\nlogin to agree\n2\n2971\n\n\n", "2538:2962": "\n\n\n\n6c implies that if you recive the object code under the conditions of 6c, then you cannot redistribute yourself, should be edited to "in accordance with subsection 6b or 6c"\ngpl3.nonsource.p3.s2\nilly\nlogin to agree\n0\n2538\n\n\nif 6c must be left in, make it only available if you received your copy *commercially*.\ngpl3.nonsource.p3.s2\njamesgnz\nlogin to agree\n0\n2962\n\n\n", "2538:2962:2998": "\n\n\n\n6c implies that if you recive the object code under the conditions of 6c, then you cannot redistribute yourself, should be edited to "in accordance with subsection 6b or 6c"\ngpl3.nonsource.p3.s2\nilly\nlogin to agree\n0\n2538\n\n\nif 6c must be left in, make it only available if you received your copy *commercially*.\ngpl3.nonsource.p3.s2\njamesgnz\nlogin to agree\n0\n2962\n\n\n<p style="display: inline">s/the object code with such an offer/only the object code and a written offer/</p><p>\nThe intent of this section appears to be to allow transfer of object code + written offer they received to another party without having to obtain a copy of the source code themselves. If that is the intent, the "occasionally and noncommercially" seems very unclear.</p>\ngpl3.nonsource.p3.s2\nmole\nlogin to agree\n0\n2998\n\n\n", "2921:2962:2971:2998:3030:3056": "\n\n\n\n<p style="display: inline">What if the offer has already expired? Or what if it only has 1 week left to run?</p><p>\nI think 6c should be dropped.</p>\ngpl3.nonsource.p3.s1\njamesgnz\nlogin to agree\n0\n2921\n\n\nif 6c must be left in, make it only available if you received your copy *commercially*.\ngpl3.nonsource.p3.s2\njamesgnz\nlogin to agree\n0\n2962\n\n\n<p style="display: inline">Someone reselling hardware accompanied by a section 6b) written offer is in trouble. They cannot pass on the offer, because that's allowed only "occasionally and noncommercially". So they must either make their own offer (and set up facilities to fulfil it) or download the source and burn it to a CD and put it in the box!</p><p>\nDetails: Let's imagine I'm in the business of buying consumer hardware in bulk and reselling it - a common business model. If I buy a pallet of routers which contain GPLv3 software and come with a 6b) written offer, can I resell them as-is? My reading of the definition of "propagate" and "convey" leads me to believe that this reselling is "conveying", which I may only do under one of sub-clauses a) to e) of section 6.</p><p>\nI am not covered by 6b) because it's not my offer, and I'm not offering support or spare parts. What about 6c)? The current wording suggests no, because it permits such conveying only "occasionally and non commercially", and my conveying is regular and commercial. Do I have to download the source, burn to a CD and pop it in the box (thereby being covered by 6a) before I can resell it? Surely the license must make it possible for me to merely pass on the boxes as I receive them.</p><p>\nMy suggestion is add another exception to the "occasionally and non commercially" one, to permit passing on copies you received. One might phrase a modified 6c) as follows:\n\n c) Convey individual copies of the object code with a copy of the written offer to provide the Corresponding Source. This alternative is allowed only if you received the object code with such an offer (in accord with subsection 6b) and either:\n\n - you are conveying copies that were conveyed to you; or\n\n - your conveying is occasional and non-commercial.</p><p>\nSo the extra exception permits a high-volume and commercial business model reselling items, yet one cannot use it to increase the number of copies the offer covers. The danger of a large increase of this type and the consequent unfair effect on the person making the written offer is, as I understand it, the reasoning behind the "occasional and non-commercial" wording.\n </p>\ngpl3.nonsource.p3.s2\ngerv\nlogin to agree\n2\n2971\n\n\n<p style="display: inline">s/the object code with such an offer/only the object code and a written offer/</p><p>\nThe intent of this section appears to be to allow transfer of object code + written offer they received to another party without having to obtain a copy of the source code themselves. If that is the intent, the "occasionally and noncommercially" seems very unclear.</p>\ngpl3.nonsource.p3.s2\nmole\nlogin to agree\n0\n2998\n\n\nWhat does "only occasionally" mean? I'm not sure it is needed, anyway, given the noncommercial restriction.\ngpl3.nonsource.p3.s2\nLovasco\nlogin to agree\n1\n3030\n\n\nI think the "only occasionally" point is much to vague. I think it should be removed altogether, so that option is allowed for all non-commercial copying.\ngpl3.nonsource.p3.s2\njohnston\nlogin to agree\n0\n3056\n\n\n", "2651": "\n\n\n\nWhat is access - does this need to use the "via a method normally used for the delivery of ..' type text like the physical media\n\ngpl3.nonsource.p4.s1\nAlanCox\nlogin to agree\n0\n2651\n\n\n", "2452": "\n\n\n\nIt is not clear, until further in this paragraph, that "place" means (typically) a network server. I suggest "...from a designated place, such as a network server, ..."\ngpl3.nonsource.p4.s1\ncyd\nlogin to agree\n2\n2452\n\n\n", "2987": "\n\n\n\nI think "additional" would be slightly clearer than "further"...\ngpl3.nonsource.p4.s1\nsepreece\nlogin to agree\n0\n2987\n\n\n", "3035": "\n\n\n\nMust the offer of access to download the source code be made to anyone, or only to those who have already downloaded the object code?\n\ngpl3.nonsource.p4.s2\nsepreece\nlogin to agree\n0\n3035\n\n\n", "2975:3035": "\n\n\n\n<p style="display: inline">A common method of distributing free software is the handing out of CDs, for example at trade shows or conferences. With some sets of software, such as the Debian GNU/Linux distribution, there are a large number of CDs of source to go with the binaries. 6a) requires the targets of such distribution to compulsorily take source; 6b) may be impractical and 3 years is a long commitment to take on for a volunteer.</p><p>\n6d) is almost perfect, but it is focussed on network distribution. The generalisation of a single word in 6d) - "copy" -> "obtain" - makes it much easier to distribute free software in this way.</p><p>\nThe CDs could then be distributed under 6d), with the "designated place" being the distribution location. There would be a pile of binary CDs and one of source; people could take the source CDs or not, as they chose.</p>\ngpl3.nonsource.p4.s2\ngerv\nlogin to agree\n1\n2975\n\n\nMust the offer of access to download the source code be made to anyone, or only to those who have already downloaded the object code?\n\ngpl3.nonsource.p4.s2\nsepreece\nlogin to agree\n0\n3035\n\n\n", "3031": "\n\n\n\nI would include some kind of statement to the effect of "the server hosting the Corresponding Source must provide equal to or greater than the bandwidth used to provide the object code" Otherwise, someone could throttle the server such that downloading the source wouldn't be practical. For example, a person offers a 650mb iso containing binary packages. The server allows downloaded to acquire the file at 300kbps. Without a restriction, that person could then set bandwidth limits such that the source code for those packages could only be obtained at 1kbps. Since they would still be providing the source (technically), they'd be covered, but the for all practical purposes no one would be able to get it. By adding a statement such as the one above, the said person would have to provide access to the source at at least the 300kbps allowed with the binaries. \ngpl3.nonsource.p4.s4\nLovasco\nlogin to agree\n0\n3031\n\n\n", "2975:2985:2987:3031:3035": "\n\n\n\n<p style="display: inline">A common method of distributing free software is the handing out of CDs, for example at trade shows or conferences. With some sets of software, such as the Debian GNU/Linux distribution, there are a large number of CDs of source to go with the binaries. 6a) requires the targets of such distribution to compulsorily take source; 6b) may be impractical and 3 years is a long commitment to take on for a volunteer.</p><p>\n6d) is almost perfect, but it is focussed on network distribution. The generalisation of a single word in 6d) - "copy" -> "obtain" - makes it much easier to distribute free software in this way.</p><p>\nThe CDs could then be distributed under 6d), with the "designated place" being the distribution location. There would be a pile of binary CDs and one of source; people could take the source CDs or not, as they chose.</p>\ngpl3.nonsource.p4.s2\ngerv\nlogin to agree\n1\n2975\n\n\nI think this needs to state that it is not only disallowed to charge more for source in this situation, but it is also disallowed to do anything that makes obtaining the source difficult, time-consuming, incur obligations, or disclose information.\ngpl3.nonsource.p4.s1\njamesgnz\nlogin to agree\n0\n2985\n\n\nI think "additional" would be slightly clearer than "further"...\ngpl3.nonsource.p4.s1\nsepreece\nlogin to agree\n0\n2987\n\n\nI would include some kind of statement to the effect of "the server hosting the Corresponding Source must provide equal to or greater than the bandwidth used to provide the object code" Otherwise, someone could throttle the server such that downloading the source wouldn't be practical. For example, a person offers a 650mb iso containing binary packages. The server allows downloaded to acquire the file at 300kbps. Without a restriction, that person could then set bandwidth limits such that the source code for those packages could only be obtained at 1kbps. Since they would still be providing the source (technically), they'd be covered, but the for all practical purposes no one would be able to get it. By adding a statement such as the one above, the said person would have to provide access to the source at at least the 300kbps allowed with the binaries. \ngpl3.nonsource.p4.s4\nLovasco\nlogin to agree\n0\n3031\n\n\nMust the offer of access to download the source code be made to anyone, or only to those who have already downloaded the object code?\n\ngpl3.nonsource.p4.s2\nsepreece\nlogin to agree\n0\n3035\n\n\n", "2650:2982:2999:3057:3058": "\n\n\n\nThis needs a clause to match designated place. Peer to peer is not implicitly free. I can make it available for $1M on a private patent protected peer to peer network otherwise\n\ngpl3.nonsource.p5.s1\nAlanCox\nlogin to agree\n0\n2650\n\n\nAs written, this would never allow me to download object code using bittorrent, as I (the downloader) do not personally inform other peers where the corresponding source is available (even if the tracker's website does, say). I suggest replacing "provided you inform other peers" with "provided other peers are informed"\ngpl3.nonsource.p5.s1\ndivec\nlogin to agree\n1\n2982\n\n\nRather than treating peer-to-peer networks as a special case with their own clause merge this into 6d by defining designated place so as to include peer-to-peer network distribution, that is to treat 6d as describing how to convey object + source over computer networks. \ngpl3.nonsource.p5.s1\nmole\nlogin to agree\n0\n2999\n\n\nI agree with divec. I don't see how this would work. How can each person downloading it using Bittorrent (for example) realistically inform every peer they're uploading it to? This would only work if whoever provides the torrent file (e.g. on a webpage) has to give this information with the file, or include it in the filename/comment field/whatever.\ngpl3.nonsource.p5.s1\njohnston\nlogin to agree\n0\n3057\n\n\nI just read divec's other comment and realised the problem I mentioned doesn't exist because of section 9. Maybe that should be made more clear. It's fairly obvious that simply having and running a copy isn't affected by a copyright licence, but not that "ancillary propagation" through peer-to-peer is exempt.\ngpl3.nonsource.p5.s1\njohnston\nlogin to agree\n0\n3058\n\n\n", "2650:2982:2999:3057:3058:3074": "\n\n\n\nThis needs a clause to match designated place. Peer to peer is not implicitly free. I can make it available for $1M on a private patent protected peer to peer network otherwise\n\ngpl3.nonsource.p5.s1\nAlanCox\nlogin to agree\n0\n2650\n\n\nAs written, this would never allow me to download object code using bittorrent, as I (the downloader) do not personally inform other peers where the corresponding source is available (even if the tracker's website does, say). I suggest replacing "provided you inform other peers" with "provided other peers are informed"\ngpl3.nonsource.p5.s1\ndivec\nlogin to agree\n1\n2982\n\n\nRather than treating peer-to-peer networks as a special case with their own clause merge this into 6d by defining designated place so as to include peer-to-peer network distribution, that is to treat 6d as describing how to convey object + source over computer networks. \ngpl3.nonsource.p5.s1\nmole\nlogin to agree\n0\n2999\n\n\nI agree with divec. I don't see how this would work. How can each person downloading it using Bittorrent (for example) realistically inform every peer they're uploading it to? This would only work if whoever provides the torrent file (e.g. on a webpage) has to give this information with the file, or include it in the filename/comment field/whatever.\ngpl3.nonsource.p5.s1\njohnston\nlogin to agree\n0\n3057\n\n\nI just read divec's other comment and realised the problem I mentioned doesn't exist because of section 9. Maybe that should be made more clear. It's fairly obvious that simply having and running a copy isn't affected by a copyright licence, but not that "ancillary propagation" through peer-to-peer is exempt.\ngpl3.nonsource.p5.s1\njohnston\nlogin to agree\n0\n3058\n\n\n<p style="display: inline">"Inform other peers" perhaps should be replaced with "a reasonable attempt has been made by you to inform other peers, present or current" or somesuch.</p><p>\nIt's not realistic to inform all peers all the time by the nature of P2P.</p>\ngpl3.nonsource.p5.s1\njamesj\nlogin to agree\n0\n3074\n\n\n", "2765:3099": "\n\n\n\nSection 6 seems so far to be basically similar to section 3 of GPLv2, just more explicit and clear in some respects and a little more permissive in other ones. Good.\ngpl3.nonsource.p6.s1\nfrx\nlogin to agree\n0\n2765\n\n\n<p style="display: inline">I don't understand this. The GPL has a lot of requirements with respect to the SOURCE CODE that I must convey, but I'm not aware of any requirements that would force me to convey some OBJECT CODE. So as I understand this sentence it is completely unnecessary. And worse than that: Because it appears to relax requirements on what object code I have to convey it leads me to believe that there are such requirements, which is confusing, since I can't find them.<br><br></p><p>\nIf there are indeed such requirements, you have some work to do to make that clear, because in all the years I've been using the GPL, I've never noticed such requirements. I always thought that an object code work could be anything and in any form, complete, incomplete, with libraries, without libraries, statically linked, dynamically linked, whatever, so long as you provide the Corresponding Source. I've now read draft 3 several times and I still don't see a place that would force me to include certain parts in my object code work. So what's the point in telling me I don't need to include system libraries in my object code work? Of course I don't need to.</p>\ngpl3.nonsource.p6.s1\nmux2005\nlogin to agree\n0\n3099\n\n\n", "2765:3099:3100": "\n\n\n\nSection 6 seems so far to be basically similar to section 3 of GPLv2, just more explicit and clear in some respects and a little more permissive in other ones. Good.\ngpl3.nonsource.p6.s1\nfrx\nlogin to agree\n0\n2765\n\n\n<p style="display: inline">I don't understand this. The GPL has a lot of requirements with respect to the SOURCE CODE that I must convey, but I'm not aware of any requirements that would force me to convey some OBJECT CODE. So as I understand this sentence it is completely unnecessary. And worse than that: Because it appears to relax requirements on what object code I have to convey it leads me to believe that there are such requirements, which is confusing, since I can't find them.<br><br></p><p>\nIf there are indeed such requirements, you have some work to do to make that clear, because in all the years I've been using the GPL, I've never noticed such requirements. I always thought that an object code work could be anything and in any form, complete, incomplete, with libraries, without libraries, statically linked, dynamically linked, whatever, so long as you provide the Corresponding Source. I've now read draft 3 several times and I still don't see a place that would force me to include certain parts in my object code work. So what's the point in telling me I don't need to include system libraries in my object code work? Of course I don't need to.</p>\ngpl3.nonsource.p6.s1\nmux2005\nlogin to agree\n0\n3099\n\n\nAfter posting my comment that I don't understand the purpose of this sentence it dawned on me that maybe this should be "must not". That would make some sense since it would emphasize that system libraries are supposed to already be on the system and if you have to ship them with your product, they're probably not system libraries. Although I don't think I would agree with this argument, since there are often different versions of system libraries and shipping them with a product is a common way of avoiding incompatibilities. Prohibiting this is undesirable.\ngpl3.nonsource.p6.s1\nmux2005\nlogin to agree\n0\n3100\n\n\n", "2455:2487:2656:2935:3101:3103": "\n\n\n\nThis definition, and the definition of "Installation information", should be merged into a single paragraph and moved to the Definitions section.\ngpl3.nonsource.p7.s1\ncyd\nlogin to agree\n6\n2455\n\n\n<p style="display: inline">This introduction of user and consumer products seems overly specific.</p><p>\nWhy not encompass all and any device or medium able to contain or represent a copy or derivative of the licensed work?</p><p>\nThere are software containers, which when ownership is transferred, perforce transfer ownership of the software therein.</p><p>\nThere are machines able to execute the software. Some machines are software containers, some interact with software containers.</p><p>\nSo why not define 'software container' and 'software executing machine' - or whatever terms are considered more appropriate, e.g. media, computer, etc.</p><p>\nWhy the heck are we introducing yet another term to subclassify members of the public.</p><p>\nThere are human beings, licensees, software (and containers thereof), and computers. That's about it.</p><p>\nShow me a consumer product that executes software that cannot be classed more generically as a software executing machine.</p><p>\nFocussing on Consumer Products seems to accede to those who wish to produce medical or military devices that are exempt from some of the GPL's constraints.</p><p>\nAnd how would 'professional products' be classified? Are they 'consumer products'?\n\n</p>\ngpl3.nonsource.p7.s1\ncrosbie\nlogin to agree\n6\n2487\n\n\nThe purpose of this distinction is to allow DRM in corporate/government setting where support functions and responsiblity are outsourced. The condition that the customer can't get the keys to fiddle/tinker can be \nspecified in the support contract; if the customer really wants, he can obtain the keys via the GPL \nprovisions but then he loses all support. Why complicate the software license instead of laying it out \nin support contracts where it belongs?\ngpl3.nonsource.p7.s1\nalexbk\nlogin to agree\n5\n2656\n\n\n<p style="display: inline">This definition seems to leave a loophole, though I am not a lawyer and am not familiar with Magnuson-Moss. Many components of products are not sold as consumer products, but instead sold in bulk to other manufacturers. If some such component is then included in a User Product, it is unclear if this section would apply to it. Clause (2) seems to cover this case for components of a house, but what about components of a car, or of a PVR?</p><p>\nThis may be covered by the relevant case law, but as has been pointed out, referring to that may cause problems outside the US, and is also just confusing. I think it would be better to explicitly include a clause saying that "User Products" include components that are designed or sold to be incorporated into other "User Products".</p>\ngpl3.nonsource.p7.s1\nunlambda\nlogin to agree\n3\n2935\n\n\nThis definition leaves too many devices on which I may want to install new firmware uncovered. This section seems to be written with the idea that there is only the consumer on one side and big organizations on the other side. But that is wrong. In the middle, there are a lot people who run their own small one-man-shops or enthusiasts who obtain professional hardware. The freedoms of free software need to be available to them, too. But with this definition, a company is free to put DRM-protected GPLed software into diagnostic equipment and other professional tools. These are definitely not consumer products, so Installation Information is not required. Think about the car industry. AFAIK they try everything they can to prevent independent repair shops (let alone hobbyists) from getting the same access to the car's internals that affiliated shops get. Does the FSF want to support this? Just because something is a professional tool rather than a consumer product doesn't mean that it's okay to lock it down.\ngpl3.nonsource.p7.s1\nmux2005\nlogin to agree\n1\n3101\n\n\n<p style="display: inline">I think the whole User Products stuff needs its own section (if it stays in, which I hope it won't). It takes up as much space as the rest of 6. although it applies only to a special case. If you put it into its own section you can also revise the structure. I think a structure similar to 6 would make it easier to understand, i.e. something like this:<br><br></p><p>\nA "User Product" is ...\n"Installation Information" is.\nYou may convey a covered work incorporated in a user product in one of the following ways:<br>\na)...<br>\nb)...<br><br></p><p>\nThat way it's much easier to see how compliant conveyance can be achieved.</p>\ngpl3.nonsource.p7.s1\nmux2005\nlogin to agree\n0\n3103\n\n\n", "2455:2487:2543:2656:2836:2935:3101:3103": "\n\n\n\nThis definition, and the definition of "Installation information", should be merged into a single paragraph and moved to the Definitions section.\ngpl3.nonsource.p7.s1\ncyd\nlogin to agree\n6\n2455\n\n\n<p style="display: inline">This introduction of user and consumer products seems overly specific.</p><p>\nWhy not encompass all and any device or medium able to contain or represent a copy or derivative of the licensed work?</p><p>\nThere are software containers, which when ownership is transferred, perforce transfer ownership of the software therein.</p><p>\nThere are machines able to execute the software. Some machines are software containers, some interact with software containers.</p><p>\nSo why not define 'software container' and 'software executing machine' - or whatever terms are considered more appropriate, e.g. media, computer, etc.</p><p>\nWhy the heck are we introducing yet another term to subclassify members of the public.</p><p>\nThere are human beings, licensees, software (and containers thereof), and computers. That's about it.</p><p>\nShow me a consumer product that executes software that cannot be classed more generically as a software executing machine.</p><p>\nFocussing on Consumer Products seems to accede to those who wish to produce medical or military devices that are exempt from some of the GPL's constraints.</p><p>\nAnd how would 'professional products' be classified? Are they 'consumer products'?\n\n</p>\ngpl3.nonsource.p7.s1\ncrosbie\nlogin to agree\n6\n2487\n\n\n<p style="display: inline">I'm going to quote freedom 0 at you:</p><p>\nThe freedom to run the program, for any purpose.</p><p>\nDefining a subset of uses that get extra rights seems to be directly opposed to freedom 0. \nI have read the reasoning for this, and I disagree with the need to make this an exception. The distinction between owners and users rights should be enough to protect big buisness (users don't get any rights forced by the GPL, only the owners do), they can rent their hardware if they don't want control, let's call a spade a spade amd make it clear who has control in that situation. </p>\ngpl3.nonsource.p7.s1\nilly\nlogin to agree\n5\n2543\n\n\nThe purpose of this distinction is to allow DRM in corporate/government setting where support functions and responsiblity are outsourced. The condition that the customer can't get the keys to fiddle/tinker can be \nspecified in the support contract; if the customer really wants, he can obtain the keys via the GPL \nprovisions but then he loses all support. Why complicate the software license instead of laying it out \nin support contracts where it belongs?\ngpl3.nonsource.p7.s1\nalexbk\nlogin to agree\n5\n2656\n\n\nI'm uncomfortable with the concept of "User Product". Similar concepts have too often been used to arbitrarily limit end user freedoms. Couldn't the purpose of this be accomplished by other rephrasing that doesn't divide products into essentially arbitrary classes?\n\ngpl3.nonsource.p7.s1\ncharlesh\nlogin to agree\n0\n2836\n\n\n<p style="display: inline">This definition seems to leave a loophole, though I am not a lawyer and am not familiar with Magnuson-Moss. Many components of products are not sold as consumer products, but instead sold in bulk to other manufacturers. If some such component is then included in a User Product, it is unclear if this section would apply to it. Clause (2) seems to cover this case for components of a house, but what about components of a car, or of a PVR?</p><p>\nThis may be covered by the relevant case law, but as has been pointed out, referring to that may cause problems outside the US, and is also just confusing. I think it would be better to explicitly include a clause saying that "User Products" include components that are designed or sold to be incorporated into other "User Products".</p>\ngpl3.nonsource.p7.s1\nunlambda\nlogin to agree\n3\n2935\n\n\nThis definition leaves too many devices on which I may want to install new firmware uncovered. This section seems to be written with the idea that there is only the consumer on one side and big organizations on the other side. But that is wrong. In the middle, there are a lot people who run their own small one-man-shops or enthusiasts who obtain professional hardware. The freedoms of free software need to be available to them, too. But with this definition, a company is free to put DRM-protected GPLed software into diagnostic equipment and other professional tools. These are definitely not consumer products, so Installation Information is not required. Think about the car industry. AFAIK they try everything they can to prevent independent repair shops (let alone hobbyists) from getting the same access to the car's internals that affiliated shops get. Does the FSF want to support this? Just because something is a professional tool rather than a consumer product doesn't mean that it's okay to lock it down.\ngpl3.nonsource.p7.s1\nmux2005\nlogin to agree\n1\n3101\n\n\n<p style="display: inline">I think the whole User Products stuff needs its own section (if it stays in, which I hope it won't). It takes up as much space as the rest of 6. although it applies only to a special case. If you put it into its own section you can also revise the structure. I think a structure similar to 6 would make it easier to understand, i.e. something like this:<br><br></p><p>\nA "User Product" is ...\n"Installation Information" is.\nYou may convey a covered work incorporated in a user product in one of the following ways:<br>\na)...<br>\nb)...<br><br></p><p>\nThat way it's much easier to see how compliant conveyance can be achieved.</p>\ngpl3.nonsource.p7.s1\nmux2005\nlogin to agree\n0\n3103\n\n\n", "2656:2935:3101:3103": "\n\n\n\nThe purpose of this distinction is to allow DRM in corporate/government setting where support functions and responsiblity are outsourced. The condition that the customer can't get the keys to fiddle/tinker can be \nspecified in the support contract; if the customer really wants, he can obtain the keys via the GPL \nprovisions but then he loses all support. Why complicate the software license instead of laying it out \nin support contracts where it belongs?\ngpl3.nonsource.p7.s1\nalexbk\nlogin to agree\n5\n2656\n\n\n<p style="display: inline">This definition seems to leave a loophole, though I am not a lawyer and am not familiar with Magnuson-Moss. Many components of products are not sold as consumer products, but instead sold in bulk to other manufacturers. If some such component is then included in a User Product, it is unclear if this section would apply to it. Clause (2) seems to cover this case for components of a house, but what about components of a car, or of a PVR?</p><p>\nThis may be covered by the relevant case law, but as has been pointed out, referring to that may cause problems outside the US, and is also just confusing. I think it would be better to explicitly include a clause saying that "User Products" include components that are designed or sold to be incorporated into other "User Products".</p>\ngpl3.nonsource.p7.s1\nunlambda\nlogin to agree\n3\n2935\n\n\nThis definition leaves too many devices on which I may want to install new firmware uncovered. This section seems to be written with the idea that there is only the consumer on one side and big organizations on the other side. But that is wrong. In the middle, there are a lot people who run their own small one-man-shops or enthusiasts who obtain professional hardware. The freedoms of free software need to be available to them, too. But with this definition, a company is free to put DRM-protected GPLed software into diagnostic equipment and other professional tools. These are definitely not consumer products, so Installation Information is not required. Think about the car industry. AFAIK they try everything they can to prevent independent repair shops (let alone hobbyists) from getting the same access to the car's internals that affiliated shops get. Does the FSF want to support this? Just because something is a professional tool rather than a consumer product doesn't mean that it's okay to lock it down.\ngpl3.nonsource.p7.s1\nmux2005\nlogin to agree\n1\n3101\n\n\n<p style="display: inline">I think the whole User Products stuff needs its own section (if it stays in, which I hope it won't). It takes up as much space as the rest of 6. although it applies only to a special case. If you put it into its own section you can also revise the structure. I think a structure similar to 6 would make it easier to understand, i.e. something like this:<br><br></p><p>\nA "User Product" is ...\n"Installation Information" is.\nYou may convey a covered work incorporated in a user product in one of the following ways:<br>\na)...<br>\nb)...<br><br></p><p>\nThat way it's much easier to see how compliant conveyance can be achieved.</p>\ngpl3.nonsource.p7.s1\nmux2005\nlogin to agree\n0\n3103\n\n\n", "2656:2935:3101": "\n\n\n\nThe purpose of this distinction is to allow DRM in corporate/government setting where support functions and responsiblity are outsourced. The condition that the customer can't get the keys to fiddle/tinker can be \nspecified in the support contract; if the customer really wants, he can obtain the keys via the GPL \nprovisions but then he loses all support. Why complicate the software license instead of laying it out \nin support contracts where it belongs?\ngpl3.nonsource.p7.s1\nalexbk\nlogin to agree\n5\n2656\n\n\n<p style="display: inline">This definition seems to leave a loophole, though I am not a lawyer and am not familiar with Magnuson-Moss. Many components of products are not sold as consumer products, but instead sold in bulk to other manufacturers. If some such component is then included in a User Product, it is unclear if this section would apply to it. Clause (2) seems to cover this case for components of a house, but what about components of a car, or of a PVR?</p><p>\nThis may be covered by the relevant case law, but as has been pointed out, referring to that may cause problems outside the US, and is also just confusing. I think it would be better to explicitly include a clause saying that "User Products" include components that are designed or sold to be incorporated into other "User Products".</p>\ngpl3.nonsource.p7.s1\nunlambda\nlogin to agree\n3\n2935\n\n\nThis definition leaves too many devices on which I may want to install new firmware uncovered. This section seems to be written with the idea that there is only the consumer on one side and big organizations on the other side. But that is wrong. In the middle, there are a lot people who run their own small one-man-shops or enthusiasts who obtain professional hardware. The freedoms of free software need to be available to them, too. But with this definition, a company is free to put DRM-protected GPLed software into diagnostic equipment and other professional tools. These are definitely not consumer products, so Installation Information is not required. Think about the car industry. AFAIK they try everything they can to prevent independent repair shops (let alone hobbyists) from getting the same access to the car's internals that affiliated shops get. Does the FSF want to support this? Just because something is a professional tool rather than a consumer product doesn't mean that it's okay to lock it down.\ngpl3.nonsource.p7.s1\nmux2005\nlogin to agree\n1\n3101\n\n\n", "2656:2676:2935:3101": "\n\n\n\nThe purpose of this distinction is to allow DRM in corporate/government setting where support functions and responsiblity are outsourced. The condition that the customer can't get the keys to fiddle/tinker can be \nspecified in the support contract; if the customer really wants, he can obtain the keys via the GPL \nprovisions but then he loses all support. Why complicate the software license instead of laying it out \nin support contracts where it belongs?\ngpl3.nonsource.p7.s1\nalexbk\nlogin to agree\n5\n2656\n\n\nWhat does housing have to do with a copyright license? Maybe it does, but it seems like a hack.\ngpl3.nonsource.p7.s1\nsanjoy\nlogin to agree\n1\n2676\n\n\n<p style="display: inline">This definition seems to leave a loophole, though I am not a lawyer and am not familiar with Magnuson-Moss. Many components of products are not sold as consumer products, but instead sold in bulk to other manufacturers. If some such component is then included in a User Product, it is unclear if this section would apply to it. Clause (2) seems to cover this case for components of a house, but what about components of a car, or of a PVR?</p><p>\nThis may be covered by the relevant case law, but as has been pointed out, referring to that may cause problems outside the US, and is also just confusing. I think it would be better to explicitly include a clause saying that "User Products" include components that are designed or sold to be incorporated into other "User Products".</p>\ngpl3.nonsource.p7.s1\nunlambda\nlogin to agree\n3\n2935\n\n\nThis definition leaves too many devices on which I may want to install new firmware uncovered. This section seems to be written with the idea that there is only the consumer on one side and big organizations on the other side. But that is wrong. In the middle, there are a lot people who run their own small one-man-shops or enthusiasts who obtain professional hardware. The freedoms of free software need to be available to them, too. But with this definition, a company is free to put DRM-protected GPLed software into diagnostic equipment and other professional tools. These are definitely not consumer products, so Installation Information is not required. Think about the car industry. AFAIK they try everything they can to prevent independent repair shops (let alone hobbyists) from getting the same access to the car's internals that affiliated shops get. Does the FSF want to support this? Just because something is a professional tool rather than a consumer product doesn't mean that it's okay to lock it down.\ngpl3.nonsource.p7.s1\nmux2005\nlogin to agree\n1\n3101\n\n\n", "2570:2766:3102": "\n\n\n\nWhat do the square brackets mean?\ngpl3.nonsource.p7.s2\nadhemar\nlogin to agree\n0\n2570\n\n\n<p style="display: inline">This part between squared brackets looks too U.S.-centric and awkward to see in a license text. I feel uneasy in reading such specific references in a license...</p><p>\nI suggest dropping this sentence entirely.</p>\ngpl3.nonsource.p7.s2\nfrx\nlogin to agree\n4\n2766\n\n\n<p style="display: inline">If you leave this in the GPL, I will postpone relicensing my (currently GPL 2 only) software indefinitely. The reason is simple: I'm from Germany and I have no idea how a German judge will react to this, if he'll read this Act, if he'll understand it, whether it'll give him the feeling the GPL is very US centric and how that might influence his interpretation of the rest of the GPL.<br><br></p><p>\nThink about it for a moment. You have a German judge with limited English skills. Because he's thorough he reads this MMWA. How much time will he spend reading this Act and how much time will he spend reading the GPL? I have no idea how large this Act is, but I assume it's not shorter than the GPL itself. So not only does this force my poor judge to wade through twice the amount of legalese in a foreign language, it also distorts his perception, because the time he spends reading this MMWA is out of proportion with the rest of the GPL. <br><br></p><p>\nAlthough I'm not a lawyer I have an idea of how a German judge will interpret the GPL standing alone. That's because the GPL has been written to be understood by laypersons. I think that the GPL standing alone can be understood even if English is not your native language (although there certainly are a lot of passages that could be simplified further). I feel confident that the GPL standing alone works here in Germany exactly because it was written to be understood by laypersons. But this reference to the MMWA threatens this confidence of mine. Please remove it. (Actually, I would rather see the whole User Products restriction be removed. See my other comments/votes. This comment is just in case you don't do that.)\n</p>\ngpl3.nonsource.p7.s2\nmux2005\nlogin to agree\n0\n3102\n\n\n", "2766:3102": "\n\n\n\n<p style="display: inline">This part between squared brackets looks too U.S.-centric and awkward to see in a license text. I feel uneasy in reading such specific references in a license...</p><p>\nI suggest dropping this sentence entirely.</p>\ngpl3.nonsource.p7.s2\nfrx\nlogin to agree\n4\n2766\n\n\n<p style="display: inline">If you leave this in the GPL, I will postpone relicensing my (currently GPL 2 only) software indefinitely. The reason is simple: I'm from Germany and I have no idea how a German judge will react to this, if he'll read this Act, if he'll understand it, whether it'll give him the feeling the GPL is very US centric and how that might influence his interpretation of the rest of the GPL.<br><br></p><p>\nThink about it for a moment. You have a German judge with limited English skills. Because he's thorough he reads this MMWA. How much time will he spend reading this Act and how much time will he spend reading the GPL? I have no idea how large this Act is, but I assume it's not shorter than the GPL itself. So not only does this force my poor judge to wade through twice the amount of legalese in a foreign language, it also distorts his perception, because the time he spends reading this MMWA is out of proportion with the rest of the GPL. <br><br></p><p>\nAlthough I'm not a lawyer I have an idea of how a German judge will interpret the GPL standing alone. That's because the GPL has been written to be understood by laypersons. I think that the GPL standing alone can be understood even if English is not your native language (although there certainly are a lot of passages that could be simplified further). I feel confident that the GPL standing alone works here in Germany exactly because it was written to be understood by laypersons. But this reference to the MMWA threatens this confidence of mine. Please remove it. (Actually, I would rather see the whole User Products restriction be removed. See my other comments/votes. This comment is just in case you don't do that.)\n</p>\ngpl3.nonsource.p7.s2\nmux2005\nlogin to agree\n0\n3102\n\n\n", "2719:2766:2801:2933:3102": "\n\n\n\nDoesn't that risk foreign governments invalidating the license? It seems to hurt internationalization, which many other parts of the license (e.g. definitions, referring to the DRM treaty instead of DMCA) are designed to help.\ngpl3.nonsource.p7.s2\nflaschen\nlogin to agree\n3\n2719\n\n\n<p style="display: inline">This part between squared brackets looks too U.S.-centric and awkward to see in a license text. I feel uneasy in reading such specific references in a license...</p><p>\nI suggest dropping this sentence entirely.</p>\ngpl3.nonsource.p7.s2\nfrx\nlogin to agree\n4\n2766\n\n\nAgain, I feel a reference to a specific law is potentially dangerous, opening it to judicial interpretation and altering the terms of this license.\ngpl3.nonsource.p7.s2\nkdean06\nlogin to agree\n2\n2801\n\n\n<p style="display: inline">\nDefinition of "user product" in section 6:\n- Reliance upon a particular USA law is very bad idea. It may not cover\nadequately future technologies or situations, which are not foreseen\ntoday.\n(I was going to comment that "Congress might one day re-enact this act,\nunder lobbying by Microsoft or another anti-GPL company, to read that\nthe law provisions do not apply as intended for GPL-like licenses",\nhowever you commented that there are plenty of case law interpretations,\nwhich clarify the meaning of "user product", and which would survive any\nlegal changes by the Congress; you may want to emphasize that you do not\nrely upon the law, but upon the interpretations.)\n- Protection is needed also for products used by small businesses, even\nif they are not covered by "user product" or "consumer product" as\ncurrently defined.</p><p>\nAnother idea: to find some legal language with the intention that\none-off big custom projects may be excused from the requirements of\n"user product". One further condition may be that the customer is as\neconomically powerful (or more powerful) than the provider.</p><p>\nSuch big projects also have the property that the only conveyance of\nsoftware occurs from a single provider (the contractor) to a single\nclient (the big customer). This is a borderline case between private\ndistribution (software developed/modified in-company and distributed to\nits branches and offices, not outside, so the company does not have to\nrelease source code outside) and conveyance to the world.\n --- Omer\n--\nDelay is the deadliest form of denial. C. Northcote Parkinson\nMy own blog is at http://tddpirate.livejournal.com/</p><p>\nMy opinions, as expressed in this E-mail message, are mine alone.\nThey do not represent the official policy of any organization with which\nI may be affiliated in any way.\nWARNING TO SPAMMERS: at http://www.zak.co.il/spamwarning.html\n\n\n</p>\ngpl3.nonsource.p7.s2\nw1\nlogin to agree\n0\n2933\n\n\n<p style="display: inline">If you leave this in the GPL, I will postpone relicensing my (currently GPL 2 only) software indefinitely. The reason is simple: I'm from Germany and I have no idea how a German judge will react to this, if he'll read this Act, if he'll understand it, whether it'll give him the feeling the GPL is very US centric and how that might influence his interpretation of the rest of the GPL.<br><br></p><p>\nThink about it for a moment. You have a German judge with limited English skills. Because he's thorough he reads this MMWA. How much time will he spend reading this Act and how much time will he spend reading the GPL? I have no idea how large this Act is, but I assume it's not shorter than the GPL itself. So not only does this force my poor judge to wade through twice the amount of legalese in a foreign language, it also distorts his perception, because the time he spends reading this MMWA is out of proportion with the rest of the GPL. <br><br></p><p>\nAlthough I'm not a lawyer I have an idea of how a German judge will interpret the GPL standing alone. That's because the GPL has been written to be understood by laypersons. I think that the GPL standing alone can be understood even if English is not your native language (although there certainly are a lot of passages that could be simplified further). I feel confident that the GPL standing alone works here in Germany exactly because it was written to be understood by laypersons. But this reference to the MMWA threatens this confidence of mine. Please remove it. (Actually, I would rather see the whole User Products restriction be removed. See my other comments/votes. This comment is just in case you don't do that.)\n</p>\ngpl3.nonsource.p7.s2\nmux2005\nlogin to agree\n0\n3102\n\n\n", "2497:2564:2657:2719:2766:2801:2933:3102": "\n\n\n\nI think that if you can't make the point without referring to external documents, it should be rethought.\ngpl3.nonsource.p7.s2\nkaol\nlogin to agree\n9\n2497\n\n\nThis reference to specific US case(s) is perhaps not relevant elsewhere (eg Europe) where there is also much activity in the Open Source world.\ngpl3.nonsource.p7.s2\nsamj\nlogin to agree\n8\n2564\n\n\nJust a test.\ngpl3.nonsource.p7.s2\nfontana\nlogin to agree\n1\n2657\n\n\nDoesn't that risk foreign governments invalidating the license? It seems to hurt internationalization, which many other parts of the license (e.g. definitions, referring to the DRM treaty instead of DMCA) are designed to help.\ngpl3.nonsource.p7.s2\nflaschen\nlogin to agree\n3\n2719\n\n\n<p style="display: inline">This part between squared brackets looks too U.S.-centric and awkward to see in a license text. I feel uneasy in reading such specific references in a license...</p><p>\nI suggest dropping this sentence entirely.</p>\ngpl3.nonsource.p7.s2\nfrx\nlogin to agree\n4\n2766\n\n\nAgain, I feel a reference to a specific law is potentially dangerous, opening it to judicial interpretation and altering the terms of this license.\ngpl3.nonsource.p7.s2\nkdean06\nlogin to agree\n2\n2801\n\n\n<p style="display: inline">\nDefinition of "user product" in section 6:\n- Reliance upon a particular USA law is very bad idea. It may not cover\nadequately future technologies or situations, which are not foreseen\ntoday.\n(I was going to comment that "Congress might one day re-enact this act,\nunder lobbying by Microsoft or another anti-GPL company, to read that\nthe law provisions do not apply as intended for GPL-like licenses",\nhowever you commented that there are plenty of case law interpretations,\nwhich clarify the meaning of "user product", and which would survive any\nlegal changes by the Congress; you may want to emphasize that you do not\nrely upon the law, but upon the interpretations.)\n- Protection is needed also for products used by small businesses, even\nif they are not covered by "user product" or "consumer product" as\ncurrently defined.</p><p>\nAnother idea: to find some legal language with the intention that\none-off big custom projects may be excused from the requirements of\n"user product". One further condition may be that the customer is as\neconomically powerful (or more powerful) than the provider.</p><p>\nSuch big projects also have the property that the only conveyance of\nsoftware occurs from a single provider (the contractor) to a single\nclient (the big customer). This is a borderline case between private\ndistribution (software developed/modified in-company and distributed to\nits branches and offices, not outside, so the company does not have to\nrelease source code outside) and conveyance to the world.\n --- Omer\n--\nDelay is the deadliest form of denial. C. Northcote Parkinson\nMy own blog is at http://tddpirate.livejournal.com/</p><p>\nMy opinions, as expressed in this E-mail message, are mine alone.\nThey do not represent the official policy of any organization with which\nI may be affiliated in any way.\nWARNING TO SPAMMERS: at http://www.zak.co.il/spamwarning.html\n\n\n</p>\ngpl3.nonsource.p7.s2\nw1\nlogin to agree\n0\n2933\n\n\n<p style="display: inline">If you leave this in the GPL, I will postpone relicensing my (currently GPL 2 only) software indefinitely. The reason is simple: I'm from Germany and I have no idea how a German judge will react to this, if he'll read this Act, if he'll understand it, whether it'll give him the feeling the GPL is very US centric and how that might influence his interpretation of the rest of the GPL.<br><br></p><p>\nThink about it for a moment. You have a German judge with limited English skills. Because he's thorough he reads this MMWA. How much time will he spend reading this Act and how much time will he spend reading the GPL? I have no idea how large this Act is, but I assume it's not shorter than the GPL itself. So not only does this force my poor judge to wade through twice the amount of legalese in a foreign language, it also distorts his perception, because the time he spends reading this MMWA is out of proportion with the rest of the GPL. <br><br></p><p>\nAlthough I'm not a lawyer I have an idea of how a German judge will interpret the GPL standing alone. That's because the GPL has been written to be understood by laypersons. I think that the GPL standing alone can be understood even if English is not your native language (although there certainly are a lot of passages that could be simplified further). I feel confident that the GPL standing alone works here in Germany exactly because it was written to be understood by laypersons. But this reference to the MMWA threatens this confidence of mine. Please remove it. (Actually, I would rather see the whole User Products restriction be removed. See my other comments/votes. This comment is just in case you don't do that.)\n</p>\ngpl3.nonsource.p7.s2\nmux2005\nlogin to agree\n0\n3102\n\n\n", "2719:2766:2801:3102": "\n\n\n\nDoesn't that risk foreign governments invalidating the license? It seems to hurt internationalization, which many other parts of the license (e.g. definitions, referring to the DRM treaty instead of DMCA) are designed to help.\ngpl3.nonsource.p7.s2\nflaschen\nlogin to agree\n3\n2719\n\n\n<p style="display: inline">This part between squared brackets looks too U.S.-centric and awkward to see in a license text. I feel uneasy in reading such specific references in a license...</p><p>\nI suggest dropping this sentence entirely.</p>\ngpl3.nonsource.p7.s2\nfrx\nlogin to agree\n4\n2766\n\n\nAgain, I feel a reference to a specific law is potentially dangerous, opening it to judicial interpretation and altering the terms of this license.\ngpl3.nonsource.p7.s2\nkdean06\nlogin to agree\n2\n2801\n\n\n<p style="display: inline">If you leave this in the GPL, I will postpone relicensing my (currently GPL 2 only) software indefinitely. The reason is simple: I'm from Germany and I have no idea how a German judge will react to this, if he'll read this Act, if he'll understand it, whether it'll give him the feeling the GPL is very US centric and how that might influence his interpretation of the rest of the GPL.<br><br></p><p>\nThink about it for a moment. You have a German judge with limited English skills. Because he's thorough he reads this MMWA. How much time will he spend reading this Act and how much time will he spend reading the GPL? I have no idea how large this Act is, but I assume it's not shorter than the GPL itself. So not only does this force my poor judge to wade through twice the amount of legalese in a foreign language, it also distorts his perception, because the time he spends reading this MMWA is out of proportion with the rest of the GPL. <br><br></p><p>\nAlthough I'm not a lawyer I have an idea of how a German judge will interpret the GPL standing alone. That's because the GPL has been written to be understood by laypersons. I think that the GPL standing alone can be understood even if English is not your native language (although there certainly are a lot of passages that could be simplified further). I feel confident that the GPL standing alone works here in Germany exactly because it was written to be understood by laypersons. But this reference to the MMWA threatens this confidence of mine. Please remove it. (Actually, I would rather see the whole User Products restriction be removed. See my other comments/votes. This comment is just in case you don't do that.)\n</p>\ngpl3.nonsource.p7.s2\nmux2005\nlogin to agree\n0\n3102\n\n\n", "2497:2543:2564:2570:2656:2657:2676:2719:2766:2801:2836:2933:2935:3101:3102:3103": "\n\n\n\nI think that if you can't make the point without referring to external documents, it should be rethought.\ngpl3.nonsource.p7.s2\nkaol\nlogin to agree\n9\n2497\n\n\n<p style="display: inline">I'm going to quote freedom 0 at you:</p><p>\nThe freedom to run the program, for any purpose.</p><p>\nDefining a subset of uses that get extra rights seems to be directly opposed to freedom 0. \nI have read the reasoning for this, and I disagree with the need to make this an exception. The distinction between owners and users rights should be enough to protect big buisness (users don't get any rights forced by the GPL, only the owners do), they can rent their hardware if they don't want control, let's call a spade a spade amd make it clear who has control in that situation. </p>\ngpl3.nonsource.p7.s1\nilly\nlogin to agree\n5\n2543\n\n\nThis reference to specific US case(s) is perhaps not relevant elsewhere (eg Europe) where there is also much activity in the Open Source world.\ngpl3.nonsource.p7.s2\nsamj\nlogin to agree\n8\n2564\n\n\nWhat do the square brackets mean?\ngpl3.nonsource.p7.s2\nadhemar\nlogin to agree\n0\n2570\n\n\nThe purpose of this distinction is to allow DRM in corporate/government setting where support functions and responsiblity are outsourced. The condition that the customer can't get the keys to fiddle/tinker can be \nspecified in the support contract; if the customer really wants, he can obtain the keys via the GPL \nprovisions but then he loses all support. Why complicate the software license instead of laying it out \nin support contracts where it belongs?\ngpl3.nonsource.p7.s1\nalexbk\nlogin to agree\n5\n2656\n\n\nJust a test.\ngpl3.nonsource.p7.s2\nfontana\nlogin to agree\n1\n2657\n\n\nWhat does housing have to do with a copyright license? Maybe it does, but it seems like a hack.\ngpl3.nonsource.p7.s1\nsanjoy\nlogin to agree\n1\n2676\n\n\nDoesn't that risk foreign governments invalidating the license? It seems to hurt internationalization, which many other parts of the license (e.g. definitions, referring to the DRM treaty instead of DMCA) are designed to help.\ngpl3.nonsource.p7.s2\nflaschen\nlogin to agree\n3\n2719\n\n\n<p style="display: inline">This part between squared brackets looks too U.S.-centric and awkward to see in a license text. I feel uneasy in reading such specific references in a license...</p><p>\nI suggest dropping this sentence entirely.</p>\ngpl3.nonsource.p7.s2\nfrx\nlogin to agree\n4\n2766\n\n\nAgain, I feel a reference to a specific law is potentially dangerous, opening it to judicial interpretation and altering the terms of this license.\ngpl3.nonsource.p7.s2\nkdean06\nlogin to agree\n2\n2801\n\n\nI'm uncomfortable with the concept of "User Product". Similar concepts have too often been used to arbitrarily limit end user freedoms. Couldn't the purpose of this be accomplished by other rephrasing that doesn't divide products into essentially arbitrary classes?\n\ngpl3.nonsource.p7.s1\ncharlesh\nlogin to agree\n0\n2836\n\n\n<p style="display: inline">\nDefinition of "user product" in section 6:\n- Reliance upon a particular USA law is very bad idea. It may not cover\nadequately future technologies or situations, which are not foreseen\ntoday.\n(I was going to comment that "Congress might one day re-enact this act,\nunder lobbying by Microsoft or another anti-GPL company, to read that\nthe law provisions do not apply as intended for GPL-like licenses",\nhowever you commented that there are plenty of case law interpretations,\nwhich clarify the meaning of "user product", and which would survive any\nlegal changes by the Congress; you may want to emphasize that you do not\nrely upon the law, but upon the interpretations.)\n- Protection is needed also for products used by small businesses, even\nif they are not covered by "user product" or "consumer product" as\ncurrently defined.</p><p>\nAnother idea: to find some legal language with the intention that\none-off big custom projects may be excused from the requirements of\n"user product". One further condition may be that the customer is as\neconomically powerful (or more powerful) than the provider.</p><p>\nSuch big projects also have the property that the only conveyance of\nsoftware occurs from a single provider (the contractor) to a single\nclient (the big customer). This is a borderline case between private\ndistribution (software developed/modified in-company and distributed to\nits branches and offices, not outside, so the company does not have to\nrelease source code outside) and conveyance to the world.\n --- Omer\n--\nDelay is the deadliest form of denial. C. Northcote Parkinson\nMy own blog is at http://tddpirate.livejournal.com/</p><p>\nMy opinions, as expressed in this E-mail message, are mine alone.\nThey do not represent the official policy of any organization with which\nI may be affiliated in any way.\nWARNING TO SPAMMERS: at http://www.zak.co.il/spamwarning.html\n\n\n</p>\ngpl3.nonsource.p7.s2\nw1\nlogin to agree\n0\n2933\n\n\n<p style="display: inline">This definition seems to leave a loophole, though I am not a lawyer and am not familiar with Magnuson-Moss. Many components of products are not sold as consumer products, but instead sold in bulk to other manufacturers. If some such component is then included in a User Product, it is unclear if this section would apply to it. Clause (2) seems to cover this case for components of a house, but what about components of a car, or of a PVR?</p><p>\nThis may be covered by the relevant case law, but as has been pointed out, referring to that may cause problems outside the US, and is also just confusing. I think it would be better to explicitly include a clause saying that "User Products" include components that are designed or sold to be incorporated into other "User Products".</p>\ngpl3.nonsource.p7.s1\nunlambda\nlogin to agree\n3\n2935\n\n\nThis definition leaves too many devices on which I may want to install new firmware uncovered. This section seems to be written with the idea that there is only the consumer on one side and big organizations on the other side. But that is wrong. In the middle, there are a lot people who run their own small one-man-shops or enthusiasts who obtain professional hardware. The freedoms of free software need to be available to them, too. But with this definition, a company is free to put DRM-protected GPLed software into diagnostic equipment and other professional tools. These are definitely not consumer products, so Installation Information is not required. Think about the car industry. AFAIK they try everything they can to prevent independent repair shops (let alone hobbyists) from getting the same access to the car's internals that affiliated shops get. Does the FSF want to support this? Just because something is a professional tool rather than a consumer product doesn't mean that it's okay to lock it down.\ngpl3.nonsource.p7.s1\nmux2005\nlogin to agree\n1\n3101\n\n\n<p style="display: inline">If you leave this in the GPL, I will postpone relicensing my (currently GPL 2 only) software indefinitely. The reason is simple: I'm from Germany and I have no idea how a German judge will react to this, if he'll read this Act, if he'll understand it, whether it'll give him the feeling the GPL is very US centric and how that might influence his interpretation of the rest of the GPL.<br><br></p><p>\nThink about it for a moment. You have a German judge with limited English skills. Because he's thorough he reads this MMWA. How much time will he spend reading this Act and how much time will he spend reading the GPL? I have no idea how large this Act is, but I assume it's not shorter than the GPL itself. So not only does this force my poor judge to wade through twice the amount of legalese in a foreign language, it also distorts his perception, because the time he spends reading this MMWA is out of proportion with the rest of the GPL. <br><br></p><p>\nAlthough I'm not a lawyer I have an idea of how a German judge will interpret the GPL standing alone. That's because the GPL has been written to be understood by laypersons. I think that the GPL standing alone can be understood even if English is not your native language (although there certainly are a lot of passages that could be simplified further). I feel confident that the GPL standing alone works here in Germany exactly because it was written to be understood by laypersons. But this reference to the MMWA threatens this confidence of mine. Please remove it. (Actually, I would rather see the whole User Products restriction be removed. See my other comments/votes. This comment is just in case you don't do that.)\n</p>\ngpl3.nonsource.p7.s2\nmux2005\nlogin to agree\n0\n3102\n\n\n<p style="display: inline">I think the whole User Products stuff needs its own section (if it stays in, which I hope it won't). It takes up as much space as the rest of 6. although it applies only to a special case. If you put it into its own section you can also revise the structure. I think a structure similar to 6 would make it easier to understand, i.e. something like this:<br><br></p><p>\nA "User Product" is ...\n"Installation Information" is.\nYou may convey a covered work incorporated in a user product in one of the following ways:<br>\na)...<br>\nb)...<br><br></p><p>\nThat way it's much easier to see how compliant conveyance can be achieved.</p>\ngpl3.nonsource.p7.s1\nmux2005\nlogin to agree\n0\n3103\n\n\n", "2529": "\n\n\n\nThis should be an "and". \ngpl3.nonsource.p8.s1\njring\nlogin to agree\n0\n2529\n\n\n", "2488": "\n\n\n\nWouldn't this language permit a distributor to describe a procedure that involves say emailing the desired modified source code to the distributor who would then apply their own private encryption key to the resulting binary and send back a binary which will execute. IANAL, but this language doesn't seem to force the distributor to give out the necessary key for me to make my own installable binaries, and it doesn't specify any restrictions on what those methods or procedures might involve or what rights or money I might have to give them in order to follow them.\ngpl3.nonsource.p8.s1\nanjrew\nlogin to agree\n2\n2488\n\n\n", "2966": "\n\n\n\nThis seems to overlook the situation where the distributor has a master key that allows installation on any device of its kind and distributes his binary signed with that key, but a recipeint only gets a limited key that allows installation only on a specific subset of the device pool (e.g. only on the device that the recipient owns). Should the license make it clear that such limitations of user rights are not allowed?\ngpl3.nonsource.p8.s1\nalexbk\nlogin to agree\n0\n2966\n\n\n", "2482:2966": "\n\n\n\n<p style="display: inline">I can imagine some electronics company allowing installation of modified binaries which subsequently don't get the decryption keys for the recorded movies. Is that execution?</p><p>\nIs execution in a sandbox from which the software can't do anything "execution"?</p>\ngpl3.nonsource.p8.s1\nmerijn\nlogin to agree\n0\n2482\n\n\nThis seems to overlook the situation where the distributor has a master key that allows installation on any device of its kind and distributes his binary signed with that key, but a recipeint only gets a limited key that allows installation only on a specific subset of the device pool (e.g. only on the device that the recipient owns). Should the license make it clear that such limitations of user rights are not allowed?\ngpl3.nonsource.p8.s1\nalexbk\nlogin to agree\n0\n2966\n\n\n", "2791": "\n\n\n\n<p style="display: inline">The point isn't that the object code will still function, the important part is that the entire device will. If the device contains several parts, one of which is running the GPL code, then it would be possible to make the device completely unusable with modified code, solely because it was modified, by letting the other parts of the device fail. This wouldn't be caught by this clause, because the code itself runs fine, only the rest of the device stops working.</p><p>\nTherefore I think it's better to change "the modified object code" into "the User Product [including the modified object code]". (I'm not sure if that last part is needed.)</p>\ngpl3.nonsource.p8.s2\nwijnen\nlogin to agree\n1\n2791\n\n\n", "2745": "\n\n\n\n<p style="display: inline">Today, many users are exposed to softwares that have been tampered with malicious code, or with bugs that may be used by external parties that want to interfere with the user installed software; if the software is digitally signed by its author or distributor, to allow users knowing if a software was modified or if it is the original tht was created by the original author or distributor, then, using the GPL rights, users are permitted to modify the software, thus canelling the digital signature.</p><p>\nThe effect is that the modification will effectively interfere with the digital signature mehanism, that may signal the existence of a modification (which may or may not be maliious). This kind of interference should be allowed, provided that it does not block the user to use the software as intended with the modifications (in other words, providing an extra information to the user that this is not an original, and letting the user choose whever he accepts to run the modifications or not should be left to the user, but not be prohibited arbitrarily by the implementation of a GPL-covered system).</p><p>\nDigital signatures will be useful to verify that the distributor of a GPL-covered software actually accepted the conditions of the GPL, and that he only was liable for the content of the software that he made himself, but that he is NOT liable of further modifications made by unknown third-parties to whom he licenced or sublicenced the software.</p><p>\nIn other words, each "Contributor" is responsible of his own modifications, and if that contributor cannot be identified, then he has abused the licence by not providing the source of his modifications along with the modified software, or by not including the required modification notices in the history, as required by the GPL.</p><p>\nBut if the modifications were made by the final licenced user himself, nothing should prohibit him to run the modified (patched, corrected, secured...) version that he made himself, even if those modifications cancelled the digital signature of the author or distributor.</p><p>\nIf the modified software is part of a more general system (such as a GNU/Linux system, or a Cygwin toolkit distribution, that includes an update mechanism), the system should continue to accept the modification, and should not automatically block it or revert it to the previous digitally signed version, because it would interfere with the legitimate user rights allowing such modifications.</p>\ngpl3.nonsource.p8.s2\nverdyph\nlogin to agree\n0\n2745\n\n\n", "2565": "\n\n\n\nPossibly too specific. Suggest 'The information must suffice to ensure that the continued functioning of the modified object code is in no case prevented or interfered with *even if* modification has been made.'\ngpl3.nonsource.p8.s2\nsamj\nlogin to agree\n0\n2565\n\n\n", "2565:2745:2791:2966": "\n\n\n\nPossibly too specific. Suggest 'The information must suffice to ensure that the continued functioning of the modified object code is in no case prevented or interfered with *even if* modification has been made.'\ngpl3.nonsource.p8.s2\nsamj\nlogin to agree\n0\n2565\n\n\n<p style="display: inline">Today, many users are exposed to softwares that have been tampered with malicious code, or with bugs that may be used by external parties that want to interfere with the user installed software; if the software is digitally signed by its author or distributor, to allow users knowing if a software was modified or if it is the original tht was created by the original author or distributor, then, using the GPL rights, users are permitted to modify the software, thus canelling the digital signature.</p><p>\nThe effect is that the modification will effectively interfere with the digital signature mehanism, that may signal the existence of a modification (which may or may not be maliious). This kind of interference should be allowed, provided that it does not block the user to use the software as intended with the modifications (in other words, providing an extra information to the user that this is not an original, and letting the user choose whever he accepts to run the modifications or not should be left to the user, but not be prohibited arbitrarily by the implementation of a GPL-covered system).</p><p>\nDigital signatures will be useful to verify that the distributor of a GPL-covered software actually accepted the conditions of the GPL, and that he only was liable for the content of the software that he made himself, but that he is NOT liable of further modifications made by unknown third-parties to whom he licenced or sublicenced the software.</p><p>\nIn other words, each "Contributor" is responsible of his own modifications, and if that contributor cannot be identified, then he has abused the licence by not providing the source of his modifications along with the modified software, or by not including the required modification notices in the history, as required by the GPL.</p><p>\nBut if the modifications were made by the final licenced user himself, nothing should prohibit him to run the modified (patched, corrected, secured...) version that he made himself, even if those modifications cancelled the digital signature of the author or distributor.</p><p>\nIf the modified software is part of a more general system (such as a GNU/Linux system, or a Cygwin toolkit distribution, that includes an update mechanism), the system should continue to accept the modification, and should not automatically block it or revert it to the previous digitally signed version, because it would interfere with the legitimate user rights allowing such modifications.</p>\ngpl3.nonsource.p8.s2\nverdyph\nlogin to agree\n0\n2745\n\n\n<p style="display: inline">The point isn't that the object code will still function, the important part is that the entire device will. If the device contains several parts, one of which is running the GPL code, then it would be possible to make the device completely unusable with modified code, solely because it was modified, by letting the other parts of the device fail. This wouldn't be caught by this clause, because the code itself runs fine, only the rest of the device stops working.</p><p>\nTherefore I think it's better to change "the modified object code" into "the User Product [including the modified object code]". (I'm not sure if that last part is needed.)</p>\ngpl3.nonsource.p8.s2\nwijnen\nlogin to agree\n1\n2791\n\n\nThis seems to overlook the situation where the distributor has a master key that allows installation on any device of its kind and distributes his binary signed with that key, but a recipeint only gets a limited key that allows installation only on a specific subset of the device pool (e.g. only on the device that the recipient owns). Should the license make it clear that such limitations of user rights are not allowed?\ngpl3.nonsource.p8.s1\nalexbk\nlogin to agree\n0\n2966\n\n\n", "2481:2720": "\n\n\n\n\ngpl3.nonsource.p9.s1\nmerijn\nlogin to agree\n0\n2481\n\n\nI really disagree with the decision to apply this provision only to User Products; it should apply to everything. The distinction means some users will have more freedom than others. Distributors will also have a difficult time determining whether their product is a "User Product". Finally, this makes it likely that another GPL revision to extend the applicability will eventually be needed.\ngpl3.nonsource.p9.s1\nflaschen\nlogin to agree\n1\n2720\n\n\n", "2481:2720:2923": "\n\n\n\n\ngpl3.nonsource.p9.s1\nmerijn\nlogin to agree\n0\n2481\n\n\nI really disagree with the decision to apply this provision only to User Products; it should apply to everything. The distinction means some users will have more freedom than others. Distributors will also have a difficult time determining whether their product is a "User Product". Finally, this makes it likely that another GPL revision to extend the applicability will eventually be needed.\ngpl3.nonsource.p9.s1\nflaschen\nlogin to agree\n1\n2720\n\n\n<p style="display: inline">Voluntary tivoization and eliminating "user product"</p><p>\nThe "user product" scope limitation to anti-tivoization seems very controversial. I'd like to argue that it's not really needed for addressing the concern about enterprise/government preferences. If this is deemed an acceptable resolution to the relevant parties, the "user product" limitation can be eliminated.</p><p>\nHere's how voluntary tivoization can be done in a GPL3-compliant way even under an anti-tivoization clause.\n\n\nThe vendor of the device adds a "vendor lock" status to the pruduct. The vendor lock has one of three values: UNLOCKED, LOCKED and BROKEN. The logic is as follows:</p><p>\n* The vendor lock starts as UNLOCKED.</p><p>\n* Iff the vendor lock is UNLOCKED or BROKEN, the Installation Instructions will work and will have the side effect of setting the status to BROKEN.</p><p>\n* Iff the vendor lock is UNLOCKED, the user can voluntarily and at its own discretion change it to LOCKED.</p><p>\n* Iff the vendor lock is LOCKED, the user can demonstrate this to third parties.</p><p>\n* The vendor lock has no other effect.\n\n\nThis is compatible with anti-tivoization. An enterprise or government agency which wishes for the device not to be under its control will simply lock the device as soon as it comes out of the shipping crate. Anyone else keeps the freedom of modification. With a simple adaptation, the existing language in the penultimate paragraph of section 6 can ensure that the vendor won't require a LOCKED status as a condition for functioning or service.\n</p>\ngpl3.nonsource.p9.s1\nblinken\nlogin to agree\n0\n2923\n\n\n", "2720:2923": "\n\n\n\nI really disagree with the decision to apply this provision only to User Products; it should apply to everything. The distinction means some users will have more freedom than others. Distributors will also have a difficult time determining whether their product is a "User Product". Finally, this makes it likely that another GPL revision to extend the applicability will eventually be needed.\ngpl3.nonsource.p9.s1\nflaschen\nlogin to agree\n1\n2720\n\n\n<p style="display: inline">Voluntary tivoization and eliminating "user product"</p><p>\nThe "user product" scope limitation to anti-tivoization seems very controversial. I'd like to argue that it's not really needed for addressing the concern about enterprise/government preferences. If this is deemed an acceptable resolution to the relevant parties, the "user product" limitation can be eliminated.</p><p>\nHere's how voluntary tivoization can be done in a GPL3-compliant way even under an anti-tivoization clause.\n\n\nThe vendor of the device adds a "vendor lock" status to the pruduct. The vendor lock has one of three values: UNLOCKED, LOCKED and BROKEN. The logic is as follows:</p><p>\n* The vendor lock starts as UNLOCKED.</p><p>\n* Iff the vendor lock is UNLOCKED or BROKEN, the Installation Instructions will work and will have the side effect of setting the status to BROKEN.</p><p>\n* Iff the vendor lock is UNLOCKED, the user can voluntarily and at its own discretion change it to LOCKED.</p><p>\n* Iff the vendor lock is LOCKED, the user can demonstrate this to third parties.</p><p>\n* The vendor lock has no other effect.\n\n\nThis is compatible with anti-tivoization. An enterprise or government agency which wishes for the device not to be under its control will simply lock the device as soon as it comes out of the shipping crate. Anyone else keeps the freedom of modification. With a simple adaptation, the existing language in the penultimate paragraph of section 6 can ensure that the vendor won't require a LOCKED status as a condition for functioning or service.\n</p>\ngpl3.nonsource.p9.s1\nblinken\nlogin to agree\n0\n2923\n\n\n", "2720:2833:2923": "\n\n\n\nI really disagree with the decision to apply this provision only to User Products; it should apply to everything. The distinction means some users will have more freedom than others. Distributors will also have a difficult time determining whether their product is a "User Product". Finally, this makes it likely that another GPL revision to extend the applicability will eventually be needed.\ngpl3.nonsource.p9.s1\nflaschen\nlogin to agree\n1\n2720\n\n\nMost of the early development of GNU software has been developed on machines, which I would not think of as user products. Today GNU/Linux software is used on IBM zSeries systems, which certainly are not user products. A small business might be interested to change the software on these systems to provide for instance improved security functions. Should we not ask the vendors to provide always a license under non-prohibitive financial terms which allows access to the installation information.\ngpl3.nonsource.p9.s1\nkune\nlogin to agree\n1\n2833\n\n\n<p style="display: inline">Voluntary tivoization and eliminating "user product"</p><p>\nThe "user product" scope limitation to anti-tivoization seems very controversial. I'd like to argue that it's not really needed for addressing the concern about enterprise/government preferences. If this is deemed an acceptable resolution to the relevant parties, the "user product" limitation can be eliminated.</p><p>\nHere's how voluntary tivoization can be done in a GPL3-compliant way even under an anti-tivoization clause.\n\n\nThe vendor of the device adds a "vendor lock" status to the pruduct. The vendor lock has one of three values: UNLOCKED, LOCKED and BROKEN. The logic is as follows:</p><p>\n* The vendor lock starts as UNLOCKED.</p><p>\n* Iff the vendor lock is UNLOCKED or BROKEN, the Installation Instructions will work and will have the side effect of setting the status to BROKEN.</p><p>\n* Iff the vendor lock is UNLOCKED, the user can voluntarily and at its own discretion change it to LOCKED.</p><p>\n* Iff the vendor lock is LOCKED, the user can demonstrate this to third parties.</p><p>\n* The vendor lock has no other effect.\n\n\nThis is compatible with anti-tivoization. An enterprise or government agency which wishes for the device not to be under its control will simply lock the device as soon as it comes out of the shipping crate. Anyone else keeps the freedom of modification. With a simple adaptation, the existing language in the penultimate paragraph of section 6 can ensure that the vendor won't require a LOCKED status as a condition for functioning or service.\n</p>\ngpl3.nonsource.p9.s1\nblinken\nlogin to agree\n0\n2923\n\n\n", "2490:2522:2644:2700": "\n\n\n\n<p style="display: inline">Let's keep it clean.</p><p>\nWhen a purchaser buys a 'User Product', its ownership is transferred to them. It is only ownership of the software that is in doubt - and hence fully addressed by the GPL.</p><p>\nThe software container or software executing machine is either physical (in which case ownership transfer is pretty clear cut) or digital (in which case it is a derivative work).</p><p>\nEither it's my mobile phone (and I am a GPL licensee) or its someone else's mobile phone that I'm borrowing or hiring (and I am not a GPL licensee).</p><p>\n'Consumers' can cope with the distinction between owning something and not owning it. If you say you only enjoy the GPL license if you own a copy of the software (or the container thereof) then that's fair. Why should anyone be a licensee simply because they get to use someone else's device (thus without owning the copy therein)?</p>\ngpl3.nonsource.p9.s1\ncrosbie\nlogin to agree\n3\n2490\n\n\nMy ADSL provider insists that the cable modem they installed is their property and that I am not allowed to modify it in any way. I can see TiVo saying they still own the devices they distribute, that they simply let you have some of their hardware in your house... \ngpl3.nonsource.p9.s1\njring\nlogin to agree\n0\n2522\n\n\n<p style="display: inline">I have to agree with the original wording and disagree with crosbie. If we distinguish between sale and rental, then we create a loophole that would allow Tivo etc to rent you a box for $500 for the first 100 years. To the average consumer it will look exactly like a sale, but legally it will be different.</p><p>\nI realize that applying GPL to rentals will be a pain for those offering short term rental of cellphones, PCs or the like, but with internet software distribution it will not be that hard to offer the source, and flashing tools.\n</p>\ngpl3.nonsource.p9.s1\nchresto\nlogin to agree\n1\n2644\n\n\n"right of possession and use" is too complicated - possession, in particular, has a very specific set of meanings in law. Say "ownership".\ngpl3.nonsource.p9.s1\nsepreece\nlogin to agree\n1\n2700\n\n\n", "2490:2522:2644": "\n\n\n\n<p style="display: inline">Let's keep it clean.</p><p>\nWhen a purchaser buys a 'User Product', its ownership is transferred to them. It is only ownership of the software that is in doubt - and hence fully addressed by the GPL.</p><p>\nThe software container or software executing machine is either physical (in which case ownership transfer is pretty clear cut) or digital (in which case it is a derivative work).</p><p>\nEither it's my mobile phone (and I am a GPL licensee) or its someone else's mobile phone that I'm borrowing or hiring (and I am not a GPL licensee).</p><p>\n'Consumers' can cope with the distinction between owning something and not owning it. If you say you only enjoy the GPL license if you own a copy of the software (or the container thereof) then that's fair. Why should anyone be a licensee simply because they get to use someone else's device (thus without owning the copy therein)?</p>\ngpl3.nonsource.p9.s1\ncrosbie\nlogin to agree\n3\n2490\n\n\nMy ADSL provider insists that the cable modem they installed is their property and that I am not allowed to modify it in any way. I can see TiVo saying they still own the devices they distribute, that they simply let you have some of their hardware in your house... \ngpl3.nonsource.p9.s1\njring\nlogin to agree\n0\n2522\n\n\n<p style="display: inline">I have to agree with the original wording and disagree with crosbie. If we distinguish between sale and rental, then we create a loophole that would allow Tivo etc to rent you a box for $500 for the first 100 years. To the average consumer it will look exactly like a sale, but legally it will be different.</p><p>\nI realize that applying GPL to rentals will be a pain for those offering short term rental of cellphones, PCs or the like, but with internet software distribution it will not be that hard to offer the source, and flashing tools.\n</p>\ngpl3.nonsource.p9.s1\nchresto\nlogin to agree\n1\n2644\n\n\n", "2490:2522:2568:2644": "\n\n\n\n<p style="display: inline">Let's keep it clean.</p><p>\nWhen a purchaser buys a 'User Product', its ownership is transferred to them. It is only ownership of the software that is in doubt - and hence fully addressed by the GPL.</p><p>\nThe software container or software executing machine is either physical (in which case ownership transfer is pretty clear cut) or digital (in which case it is a derivative work).</p><p>\nEither it's my mobile phone (and I am a GPL licensee) or its someone else's mobile phone that I'm borrowing or hiring (and I am not a GPL licensee).</p><p>\n'Consumers' can cope with the distinction between owning something and not owning it. If you say you only enjoy the GPL license if you own a copy of the software (or the container thereof) then that's fair. Why should anyone be a licensee simply because they get to use someone else's device (thus without owning the copy therein)?</p>\ngpl3.nonsource.p9.s1\ncrosbie\nlogin to agree\n3\n2490\n\n\nMy ADSL provider insists that the cable modem they installed is their property and that I am not allowed to modify it in any way. I can see TiVo saying they still own the devices they distribute, that they simply let you have some of their hardware in your house... \ngpl3.nonsource.p9.s1\njring\nlogin to agree\n0\n2522\n\n\nPerpetuity and fixed terms, is that a complete partition of timeframes? What about not-(completely)-fixed (potentially-)finite terms?\ngpl3.nonsource.p9.s1\nadhemar\nlogin to agree\n1\n2568\n\n\n<p style="display: inline">I have to agree with the original wording and disagree with crosbie. If we distinguish between sale and rental, then we create a loophole that would allow Tivo etc to rent you a box for $500 for the first 100 years. To the average consumer it will look exactly like a sale, but legally it will be different.</p><p>\nI realize that applying GPL to rentals will be a pain for those offering short term rental of cellphones, PCs or the like, but with internet software distribution it will not be that hard to offer the source, and flashing tools.\n</p>\ngpl3.nonsource.p9.s1\nchresto\nlogin to agree\n1\n2644\n\n\n", "3139": "\n\n\n\n<p style="display: inline">This seems to prevent TiVoisation, but what about blue ray and the like?</p><p>\nThey can still use GPLv3 code to implement a video player that refuses to process certain discs, and uses DRM technology to prevent altered versions of the program to disable this policy?</p><p>\nWhat about software that implements hardware-based Treacherous Computing architecture? It can still use GPLv3 code as long as they provide a way to replace the software (rendering the system crippled, of course).\n</p>\ngpl3.nonsource.p9.s1\nrmh\nlogin to agree\n0\n3139\n\n\n", "2492:2751:2930": "\n\n\n\n<p style="display: inline">Please remove this sentence. Exempting hard-wired code (ROM) means that the GPLv3 is useless for open hardware projects.</p><p>\nThe Open Graphics project (http://OpenGraphics.org) wanted to release the hardware schemas for the open-source graphics cards under the GPLv3 so that other companies that base products on the ones produced by the Open Graphics project would also have to provide the hardware schemas for their modified products. But if this exemption is in place, it looks like the Open Graphics project and other open hardware projects will have to find another license.</p><p>\nBut you can still fix this by removing the ROM exemption (highlighted in this comment).</p>\ngpl3.nonsource.p9.s2\nvladc\nlogin to agree\n5\n2492\n\n\n<p style="display: inline">This path does make it possible for vendors to use a flash memory in such a way that it can be only (uniquely for one customer) be written to. As those devices are not generally updated this is no loss to the manufacturer, but a loss to the customer. Similarly to write once read many memmory.</p><p>\nThe flash case might be invalid because the manufacturer can physically reprogram the flash (only when the encryption key used is retained). </p>\ngpl3.nonsource.p9.s2\npauldv\nlogin to agree\n1\n2751\n\n\n<p style="display: inline">software in product</p><p>\nProduct modificability, section 6:\n"But this requirement does not apply\nif neither you nor any third party retains the ability to install\nmodified object code on the User Product (for example, the work has\nbeen installed in ROM)."</p><p>\nThere may be cases in which ability to modify the product is limited to\ncertain parties (for example, cellular phone re-flashing); sometimes,\ndue to good reasons (such as implanted medical devices). I think that\nthe logical way to preserve freedoms is to require that any party, which\ncould (technically/legally) modify the object code, be provided - at\nrequest - with the corresponding source code and installation\ninformation.\n --- Omer\n\n\n</p>\ngpl3.nonsource.p9.s2\nw1\nlogin to agree\n0\n2930\n\n\n", "2492:2603:2751:2930": "\n\n\n\n<p style="display: inline">Please remove this sentence. Exempting hard-wired code (ROM) means that the GPLv3 is useless for open hardware projects.</p><p>\nThe Open Graphics project (http://OpenGraphics.org) wanted to release the hardware schemas for the open-source graphics cards under the GPLv3 so that other companies that base products on the ones produced by the Open Graphics project would also have to provide the hardware schemas for their modified products. But if this exemption is in place, it looks like the Open Graphics project and other open hardware projects will have to find another license.</p><p>\nBut you can still fix this by removing the ROM exemption (highlighted in this comment).</p>\ngpl3.nonsource.p9.s2\nvladc\nlogin to agree\n5\n2492\n\n\n-> "this requirement (to accompany the Corresponding Source with the Installation Information)"\ngpl3.nonsource.p9.s2\nskquinn\nlogin to agree\n1\n2603\n\n\n<p style="display: inline">This path does make it possible for vendors to use a flash memory in such a way that it can be only (uniquely for one customer) be written to. As those devices are not generally updated this is no loss to the manufacturer, but a loss to the customer. Similarly to write once read many memmory.</p><p>\nThe flash case might be invalid because the manufacturer can physically reprogram the flash (only when the encryption key used is retained). </p>\ngpl3.nonsource.p9.s2\npauldv\nlogin to agree\n1\n2751\n\n\n<p style="display: inline">software in product</p><p>\nProduct modificability, section 6:\n"But this requirement does not apply\nif neither you nor any third party retains the ability to install\nmodified object code on the User Product (for example, the work has\nbeen installed in ROM)."</p><p>\nThere may be cases in which ability to modify the product is limited to\ncertain parties (for example, cellular phone re-flashing); sometimes,\ndue to good reasons (such as implanted medical devices). I think that\nthe logical way to preserve freedoms is to require that any party, which\ncould (technically/legally) modify the object code, be provided - at\nrequest - with the corresponding source code and installation\ninformation.\n --- Omer\n\n\n</p>\ngpl3.nonsource.p9.s2\nw1\nlogin to agree\n0\n2930\n\n\n", "2492:2649:2699:2751:2930": "\n\n\n\n<p style="display: inline">Please remove this sentence. Exempting hard-wired code (ROM) means that the GPLv3 is useless for open hardware projects.</p><p>\nThe Open Graphics project (http://OpenGraphics.org) wanted to release the hardware schemas for the open-source graphics cards under the GPLv3 so that other companies that base products on the ones produced by the Open Graphics project would also have to provide the hardware schemas for their modified products. But if this exemption is in place, it looks like the Open Graphics project and other open hardware projects will have to find another license.</p><p>\nBut you can still fix this by removing the ROM exemption (highlighted in this comment).</p>\ngpl3.nonsource.p9.s2\nvladc\nlogin to agree\n5\n2492\n\n\nIf the software is on ROM I can unsolder the ROM and replace it. The ROM. Thus a third party often does have the power always even in cases the license seeks to specify\n\ngpl3.nonsource.p9.s2\nAlanCox\nlogin to agree\n3\n2649\n\n\n<p style="display: inline">Would it not suffice to say "if it is physically impossible to install modified versions of the work in the Product."</p><p>\nThe current language is ambiguous. What does "retains the ability to install" mean? The legal right? Having the right tools? The right technical knowledge and skill? Access to the device to make such replacement possible?\n\n</p>\ngpl3.nonsource.p9.s2\nsepreece\nlogin to agree\n3\n2699\n\n\n<p style="display: inline">This path does make it possible for vendors to use a flash memory in such a way that it can be only (uniquely for one customer) be written to. As those devices are not generally updated this is no loss to the manufacturer, but a loss to the customer. Similarly to write once read many memmory.</p><p>\nThe flash case might be invalid because the manufacturer can physically reprogram the flash (only when the encryption key used is retained). </p>\ngpl3.nonsource.p9.s2\npauldv\nlogin to agree\n1\n2751\n\n\n<p style="display: inline">software in product</p><p>\nProduct modificability, section 6:\n"But this requirement does not apply\nif neither you nor any third party retains the ability to install\nmodified object code on the User Product (for example, the work has\nbeen installed in ROM)."</p><p>\nThere may be cases in which ability to modify the product is limited to\ncertain parties (for example, cellular phone re-flashing); sometimes,\ndue to good reasons (such as implanted medical devices). I think that\nthe logical way to preserve freedoms is to require that any party, which\ncould (technically/legally) modify the object code, be provided - at\nrequest - with the corresponding source code and installation\ninformation.\n --- Omer\n\n\n</p>\ngpl3.nonsource.p9.s2\nw1\nlogin to agree\n0\n2930\n\n\n", "2492:2699:2751:2930": "\n\n\n\n<p style="display: inline">Please remove this sentence. Exempting hard-wired code (ROM) means that the GPLv3 is useless for open hardware projects.</p><p>\nThe Open Graphics project (http://OpenGraphics.org) wanted to release the hardware schemas for the open-source graphics cards under the GPLv3 so that other companies that base products on the ones produced by the Open Graphics project would also have to provide the hardware schemas for their modified products. But if this exemption is in place, it looks like the Open Graphics project and other open hardware projects will have to find another license.</p><p>\nBut you can still fix this by removing the ROM exemption (highlighted in this comment).</p>\ngpl3.nonsource.p9.s2\nvladc\nlogin to agree\n5\n2492\n\n\n<p style="display: inline">Would it not suffice to say "if it is physically impossible to install modified versions of the work in the Product."</p><p>\nThe current language is ambiguous. What does "retains the ability to install" mean? The legal right? Having the right tools? The right technical knowledge and skill? Access to the device to make such replacement possible?\n\n</p>\ngpl3.nonsource.p9.s2\nsepreece\nlogin to agree\n3\n2699\n\n\n<p style="display: inline">This path does make it possible for vendors to use a flash memory in such a way that it can be only (uniquely for one customer) be written to. As those devices are not generally updated this is no loss to the manufacturer, but a loss to the customer. Similarly to write once read many memmory.</p><p>\nThe flash case might be invalid because the manufacturer can physically reprogram the flash (only when the encryption key used is retained). </p>\ngpl3.nonsource.p9.s2\npauldv\nlogin to agree\n1\n2751\n\n\n<p style="display: inline">software in product</p><p>\nProduct modificability, section 6:\n"But this requirement does not apply\nif neither you nor any third party retains the ability to install\nmodified object code on the User Product (for example, the work has\nbeen installed in ROM)."</p><p>\nThere may be cases in which ability to modify the product is limited to\ncertain parties (for example, cellular phone re-flashing); sometimes,\ndue to good reasons (such as implanted medical devices). I think that\nthe logical way to preserve freedoms is to require that any party, which\ncould (technically/legally) modify the object code, be provided - at\nrequest - with the corresponding source code and installation\ninformation.\n --- Omer\n\n\n</p>\ngpl3.nonsource.p9.s2\nw1\nlogin to agree\n0\n2930\n\n\n", "2457:2492:2699:2751:2930:3112": "\n\n\n\nThis should be: "for example, *if* the work..."\ngpl3.nonsource.p9.s2\ncyd\nlogin to agree\n1\n2457\n\n\n<p style="display: inline">Please remove this sentence. Exempting hard-wired code (ROM) means that the GPLv3 is useless for open hardware projects.</p><p>\nThe Open Graphics project (http://OpenGraphics.org) wanted to release the hardware schemas for the open-source graphics cards under the GPLv3 so that other companies that base products on the ones produced by the Open Graphics project would also have to provide the hardware schemas for their modified products. But if this exemption is in place, it looks like the Open Graphics project and other open hardware projects will have to find another license.</p><p>\nBut you can still fix this by removing the ROM exemption (highlighted in this comment).</p>\ngpl3.nonsource.p9.s2\nvladc\nlogin to agree\n5\n2492\n\n\n<p style="display: inline">Would it not suffice to say "if it is physically impossible to install modified versions of the work in the Product."</p><p>\nThe current language is ambiguous. What does "retains the ability to install" mean? The legal right? Having the right tools? The right technical knowledge and skill? Access to the device to make such replacement possible?\n\n</p>\ngpl3.nonsource.p9.s2\nsepreece\nlogin to agree\n3\n2699\n\n\n<p style="display: inline">This path does make it possible for vendors to use a flash memory in such a way that it can be only (uniquely for one customer) be written to. As those devices are not generally updated this is no loss to the manufacturer, but a loss to the customer. Similarly to write once read many memmory.</p><p>\nThe flash case might be invalid because the manufacturer can physically reprogram the flash (only when the encryption key used is retained). </p>\ngpl3.nonsource.p9.s2\npauldv\nlogin to agree\n1\n2751\n\n\n<p style="display: inline">software in product</p><p>\nProduct modificability, section 6:\n"But this requirement does not apply\nif neither you nor any third party retains the ability to install\nmodified object code on the User Product (for example, the work has\nbeen installed in ROM)."</p><p>\nThere may be cases in which ability to modify the product is limited to\ncertain parties (for example, cellular phone re-flashing); sometimes,\ndue to good reasons (such as implanted medical devices). I think that\nthe logical way to preserve freedoms is to require that any party, which\ncould (technically/legally) modify the object code, be provided - at\nrequest - with the corresponding source code and installation\ninformation.\n --- Omer\n\n\n</p>\ngpl3.nonsource.p9.s2\nw1\nlogin to agree\n0\n2930\n\n\n<p style="display: inline">This is a bad example, because "ROM" as it is commonly used applies to a lot of technologies which give the vendor the ability to install modified object code. Flash memory for instance IS a type of ROM (EEP ROM).</p><p>\nhttp://en.wikipedia.org/wiki/Flash_rom</p>\ngpl3.nonsource.p9.s2\nmux2005\nlogin to agree\n0\n3112\n\n\n", "2457:2492:2699:2751:2930:3111:3112": "\n\n\n\nThis should be: "for example, *if* the work..."\ngpl3.nonsource.p9.s2\ncyd\nlogin to agree\n1\n2457\n\n\n<p style="display: inline">Please remove this sentence. Exempting hard-wired code (ROM) means that the GPLv3 is useless for open hardware projects.</p><p>\nThe Open Graphics project (http://OpenGraphics.org) wanted to release the hardware schemas for the open-source graphics cards under the GPLv3 so that other companies that base products on the ones produced by the Open Graphics project would also have to provide the hardware schemas for their modified products. But if this exemption is in place, it looks like the Open Graphics project and other open hardware projects will have to find another license.</p><p>\nBut you can still fix this by removing the ROM exemption (highlighted in this comment).</p>\ngpl3.nonsource.p9.s2\nvladc\nlogin to agree\n5\n2492\n\n\n<p style="display: inline">Would it not suffice to say "if it is physically impossible to install modified versions of the work in the Product."</p><p>\nThe current language is ambiguous. What does "retains the ability to install" mean? The legal right? Having the right tools? The right technical knowledge and skill? Access to the device to make such replacement possible?\n\n</p>\ngpl3.nonsource.p9.s2\nsepreece\nlogin to agree\n3\n2699\n\n\n<p style="display: inline">This path does make it possible for vendors to use a flash memory in such a way that it can be only (uniquely for one customer) be written to. As those devices are not generally updated this is no loss to the manufacturer, but a loss to the customer. Similarly to write once read many memmory.</p><p>\nThe flash case might be invalid because the manufacturer can physically reprogram the flash (only when the encryption key used is retained). </p>\ngpl3.nonsource.p9.s2\npauldv\nlogin to agree\n1\n2751\n\n\n<p style="display: inline">software in product</p><p>\nProduct modificability, section 6:\n"But this requirement does not apply\nif neither you nor any third party retains the ability to install\nmodified object code on the User Product (for example, the work has\nbeen installed in ROM)."</p><p>\nThere may be cases in which ability to modify the product is limited to\ncertain parties (for example, cellular phone re-flashing); sometimes,\ndue to good reasons (such as implanted medical devices). I think that\nthe logical way to preserve freedoms is to require that any party, which\ncould (technically/legally) modify the object code, be provided - at\nrequest - with the corresponding source code and installation\ninformation.\n --- Omer\n\n\n</p>\ngpl3.nonsource.p9.s2\nw1\nlogin to agree\n0\n2930\n\n\nWhile ROM may be a well-known acronym to geeks, you can hardly assume every lawyer and judge to know it. Even though it's just an example in parentheses, it's better not to rely on technical background like this. It doesn't hurt to write "read-only memory". This may even help with the problem that "ROM" as it is commonly used applies to a lot of technologies that are NOT read-only (see other comment)\ngpl3.nonsource.p9.s2\nmux2005\nlogin to agree\n0\n3111\n\n\n<p style="display: inline">This is a bad example, because "ROM" as it is commonly used applies to a lot of technologies which give the vendor the ability to install modified object code. Flash memory for instance IS a type of ROM (EEP ROM).</p><p>\nhttp://en.wikipedia.org/wiki/Flash_rom</p>\ngpl3.nonsource.p9.s2\nmux2005\nlogin to agree\n0\n3112\n\n\n", "2481:2490:2492:2522:2568:2603:2644:2649:2699:2700:2720:2751:2833:2923:2930:3111:3112:3139": "\n\n\n\n\ngpl3.nonsource.p9.s1\nmerijn\nlogin to agree\n0\n2481\n\n\n<p style="display: inline">Let's keep it clean.</p><p>\nWhen a purchaser buys a 'User Product', its ownership is transferred to them. It is only ownership of the software that is in doubt - and hence fully addressed by the GPL.</p><p>\nThe software container or software executing machine is either physical (in which case ownership transfer is pretty clear cut) or digital (in which case it is a derivative work).</p><p>\nEither it's my mobile phone (and I am a GPL licensee) or its someone else's mobile phone that I'm borrowing or hiring (and I am not a GPL licensee).</p><p>\n'Consumers' can cope with the distinction between owning something and not owning it. If you say you only enjoy the GPL license if you own a copy of the software (or the container thereof) then that's fair. Why should anyone be a licensee simply because they get to use someone else's device (thus without owning the copy therein)?</p>\ngpl3.nonsource.p9.s1\ncrosbie\nlogin to agree\n3\n2490\n\n\n<p style="display: inline">Please remove this sentence. Exempting hard-wired code (ROM) means that the GPLv3 is useless for open hardware projects.</p><p>\nThe Open Graphics project (http://OpenGraphics.org) wanted to release the hardware schemas for the open-source graphics cards under the GPLv3 so that other companies that base products on the ones produced by the Open Graphics project would also have to provide the hardware schemas for their modified products. But if this exemption is in place, it looks like the Open Graphics project and other open hardware projects will have to find another license.</p><p>\nBut you can still fix this by removing the ROM exemption (highlighted in this comment).</p>\ngpl3.nonsource.p9.s2\nvladc\nlogin to agree\n5\n2492\n\n\nMy ADSL provider insists that the cable modem they installed is their property and that I am not allowed to modify it in any way. I can see TiVo saying they still own the devices they distribute, that they simply let you have some of their hardware in your house... \ngpl3.nonsource.p9.s1\njring\nlogin to agree\n0\n2522\n\n\nPerpetuity and fixed terms, is that a complete partition of timeframes? What about not-(completely)-fixed (potentially-)finite terms?\ngpl3.nonsource.p9.s1\nadhemar\nlogin to agree\n1\n2568\n\n\n-> "this requirement (to accompany the Corresponding Source with the Installation Information)"\ngpl3.nonsource.p9.s2\nskquinn\nlogin to agree\n1\n2603\n\n\n<p style="display: inline">I have to agree with the original wording and disagree with crosbie. If we distinguish between sale and rental, then we create a loophole that would allow Tivo etc to rent you a box for $500 for the first 100 years. To the average consumer it will look exactly like a sale, but legally it will be different.</p><p>\nI realize that applying GPL to rentals will be a pain for those offering short term rental of cellphones, PCs or the like, but with internet software distribution it will not be that hard to offer the source, and flashing tools.\n</p>\ngpl3.nonsource.p9.s1\nchresto\nlogin to agree\n1\n2644\n\n\nIf the software is on ROM I can unsolder the ROM and replace it. The ROM. Thus a third party often does have the power always even in cases the license seeks to specify\n\ngpl3.nonsource.p9.s2\nAlanCox\nlogin to agree\n3\n2649\n\n\n<p style="display: inline">Would it not suffice to say "if it is physically impossible to install modified versions of the work in the Product."</p><p>\nThe current language is ambiguous. What does "retains the ability to install" mean? The legal right? Having the right tools? The right technical knowledge and skill? Access to the device to make such replacement possible?\n\n</p>\ngpl3.nonsource.p9.s2\nsepreece\nlogin to agree\n3\n2699\n\n\n"right of possession and use" is too complicated - possession, in particular, has a very specific set of meanings in law. Say "ownership".\ngpl3.nonsource.p9.s1\nsepreece\nlogin to agree\n1\n2700\n\n\nI really disagree with the decision to apply this provision only to User Products; it should apply to everything. The distinction means some users will have more freedom than others. Distributors will also have a difficult time determining whether their product is a "User Product". Finally, this makes it likely that another GPL revision to extend the applicability will eventually be needed.\ngpl3.nonsource.p9.s1\nflaschen\nlogin to agree\n1\n2720\n\n\n<p style="display: inline">This path does make it possible for vendors to use a flash memory in such a way that it can be only (uniquely for one customer) be written to. As those devices are not generally updated this is no loss to the manufacturer, but a loss to the customer. Similarly to write once read many memmory.</p><p>\nThe flash case might be invalid because the manufacturer can physically reprogram the flash (only when the encryption key used is retained). </p>\ngpl3.nonsource.p9.s2\npauldv\nlogin to agree\n1\n2751\n\n\nMost of the early development of GNU software has been developed on machines, which I would not think of as user products. Today GNU/Linux software is used on IBM zSeries systems, which certainly are not user products. A small business might be interested to change the software on these systems to provide for instance improved security functions. Should we not ask the vendors to provide always a license under non-prohibitive financial terms which allows access to the installation information.\ngpl3.nonsource.p9.s1\nkune\nlogin to agree\n1\n2833\n\n\n<p style="display: inline">Voluntary tivoization and eliminating "user product"</p><p>\nThe "user product" scope limitation to anti-tivoization seems very controversial. I'd like to argue that it's not really needed for addressing the concern about enterprise/government preferences. If this is deemed an acceptable resolution to the relevant parties, the "user product" limitation can be eliminated.</p><p>\nHere's how voluntary tivoization can be done in a GPL3-compliant way even under an anti-tivoization clause.\n\n\nThe vendor of the device adds a "vendor lock" status to the pruduct. The vendor lock has one of three values: UNLOCKED, LOCKED and BROKEN. The logic is as follows:</p><p>\n* The vendor lock starts as UNLOCKED.</p><p>\n* Iff the vendor lock is UNLOCKED or BROKEN, the Installation Instructions will work and will have the side effect of setting the status to BROKEN.</p><p>\n* Iff the vendor lock is UNLOCKED, the user can voluntarily and at its own discretion change it to LOCKED.</p><p>\n* Iff the vendor lock is LOCKED, the user can demonstrate this to third parties.</p><p>\n* The vendor lock has no other effect.\n\n\nThis is compatible with anti-tivoization. An enterprise or government agency which wishes for the device not to be under its control will simply lock the device as soon as it comes out of the shipping crate. Anyone else keeps the freedom of modification. With a simple adaptation, the existing language in the penultimate paragraph of section 6 can ensure that the vendor won't require a LOCKED status as a condition for functioning or service.\n</p>\ngpl3.nonsource.p9.s1\nblinken\nlogin to agree\n0\n2923\n\n\n<p style="display: inline">software in product</p><p>\nProduct modificability, section 6:\n"But this requirement does not apply\nif neither you nor any third party retains the ability to install\nmodified object code on the User Product (for example, the work has\nbeen installed in ROM)."</p><p>\nThere may be cases in which ability to modify the product is limited to\ncertain parties (for example, cellular phone re-flashing); sometimes,\ndue to good reasons (such as implanted medical devices). I think that\nthe logical way to preserve freedoms is to require that any party, which\ncould (technically/legally) modify the object code, be provided - at\nrequest - with the corresponding source code and installation\ninformation.\n --- Omer\n\n\n</p>\ngpl3.nonsource.p9.s2\nw1\nlogin to agree\n0\n2930\n\n\nWhile ROM may be a well-known acronym to geeks, you can hardly assume every lawyer and judge to know it. Even though it's just an example in parentheses, it's better not to rely on technical background like this. It doesn't hurt to write "read-only memory". This may even help with the problem that "ROM" as it is commonly used applies to a lot of technologies that are NOT read-only (see other comment)\ngpl3.nonsource.p9.s2\nmux2005\nlogin to agree\n0\n3111\n\n\n<p style="display: inline">This is a bad example, because "ROM" as it is commonly used applies to a lot of technologies which give the vendor the ability to install modified object code. Flash memory for instance IS a type of ROM (EEP ROM).</p><p>\nhttp://en.wikipedia.org/wiki/Flash_rom</p>\ngpl3.nonsource.p9.s2\nmux2005\nlogin to agree\n0\n3112\n\n\n<p style="display: inline">This seems to prevent TiVoisation, but what about blue ray and the like?</p><p>\nThey can still use GPLv3 code to implement a video player that refuses to process certain discs, and uses DRM technology to prevent altered versions of the program to disable this policy?</p><p>\nWhat about software that implements hardware-based Treacherous Computing architecture? It can still use GPLv3 code as long as they provide a way to replace the software (rendering the system crippled, of course).\n</p>\ngpl3.nonsource.p9.s1\nrmh\nlogin to agree\n0\n3139\n\n\n", "2552:2767": "\n\n\n\n<p style="display: inline">According to the rationale, the reason for the "User Product" language was that certain large enterprise customers prefer not to have the DRM keys, because they want the vendor to control their system updates.</p><p>\nHowever, according to this text, can't the vendor simply say that they will no longer update or warrantee the system if the user employs the DRM keys to modify it? It seems that this should remove the need for the controversial "User Product" language.</p>\ngpl3.nonsource.p10.s1\nstevenj\nlogin to agree\n3\n2552\n\n\nThis requirement to provide Installation Information is a reworded and clarified version of the requirement to include installation keys in the Corresponding Source (as the draft rationale explains). The current language seems clearer and better. Good.\ngpl3.nonsource.p10.s1\nfrx\nlogin to agree\n0\n2767\n\n\n", "2552:2925": "\n\n\n\n<p style="display: inline">According to the rationale, the reason for the "User Product" language was that certain large enterprise customers prefer not to have the DRM keys, because they want the vendor to control their system updates.</p><p>\nHowever, according to this text, can't the vendor simply say that they will no longer update or warrantee the system if the user employs the DRM keys to modify it? It seems that this should remove the need for the controversial "User Product" language.</p>\ngpl3.nonsource.p10.s1\nstevenj\nlogin to agree\n3\n2552\n\n\n<p style="display: inline">One way to circumvent the anti-tivoization is as follows: the device's vendor allows installation of modified works by default, but has a "vendor lock" flag that can be enabled by the user and prevents such modifications. Service is then conditioned upon the user "voluntarily" setting the flag. See my comment 2923 for a full description.</p><p>\nThis is a potentially devastating loophole: the user gets the choice of a free but useless device, or a serviced but forever-unmodifiable device.</p><p>\nTo close this loophole, I suggest adding a sentence to this paragraph, right after the first sentence:</p><p>\n"However, support service, warranty, and updates must not be conditioned upon acceptance of a constraint or obligation to avoid future modifications or installation of modified works."</p>\ngpl3.nonsource.p10.s1\nblinken\nlogin to agree\n0\n2925\n\n\n", "2552:2694:2925": "\n\n\n\n<p style="display: inline">According to the rationale, the reason for the "User Product" language was that certain large enterprise customers prefer not to have the DRM keys, because they want the vendor to control their system updates.</p><p>\nHowever, according to this text, can't the vendor simply say that they will no longer update or warrantee the system if the user employs the DRM keys to modify it? It seems that this should remove the need for the controversial "User Product" language.</p>\ngpl3.nonsource.p10.s1\nstevenj\nlogin to agree\n3\n2552\n\n\nThis language seems to allow a vendor to lock down a device by having a continuing update process (giving the new software a new key at frequent intervals). This language says that updates can be stopped if the software is modified, which means that the keys would stop being updated, and corresponding service could stop accepting the client on that basis.\n\ngpl3.nonsource.p10.s1\nsepreece\nlogin to agree\n2\n2694\n\n\n<p style="display: inline">One way to circumvent the anti-tivoization is as follows: the device's vendor allows installation of modified works by default, but has a "vendor lock" flag that can be enabled by the user and prevents such modifications. Service is then conditioned upon the user "voluntarily" setting the flag. See my comment 2923 for a full description.</p><p>\nThis is a potentially devastating loophole: the user gets the choice of a free but useless device, or a serviced but forever-unmodifiable device.</p><p>\nTo close this loophole, I suggest adding a sentence to this paragraph, right after the first sentence:</p><p>\n"However, support service, warranty, and updates must not be conditioned upon acceptance of a constraint or obligation to avoid future modifications or installation of modified works."</p>\ngpl3.nonsource.p10.s1\nblinken\nlogin to agree\n0\n2925\n\n\n", "2496:2523:2559:2578:2698:2827:3089": "\n\n\n\n<p style="display: inline">Please remove the sentence highlighted in this comment. First of all, it is unclear who is doing the denying of network access: the client's hardware or the server. Imagine a Tivo-like appliance designed so that the network card commits suicide if the md5sum of the loaded operating system has changed. The GPLv3 in its current form would allow this type of DRM because of the loophole introduced by the sentence highlighted in this comment.</p><p>\nIn combination with the "ROM exemption" (http://gplv3.fsf.org/comments/rt/readsay.html?filename=gplv3-draft-3&id=2492), this sentence forms an even greater loophole: GPL'ed but unmodifiable software stored in ROM would be the one performing the md5sum and killing the network card.</p><p>\nAs for "rules and protocols", this is an open invitation for DRM to be included in them. Terrible idea to even mention them. And it might give credence to those that are trying to kill Net Neutrality.</p><p>\nAgain, please delete this whole sentence.</p>\ngpl3.nonsource.p10.s2\nvladc\nlogin to agree\n7\n2496\n\n\nClarify that this is only allowed if the modification actually does alter network behavior and that simply suggesting that any modified code could possibly have security vulnerabilities is not enough. As it stands it is far to easy for companies to say "well we are afraid what might happen if we let everybody on" and the block any modified software. \ngpl3.nonsource.p10.s2\njring\nlogin to agree\n7\n2523\n\n\n<p style="display: inline">We need to distinguish between rules and protocols that involve the externally visible actions of the client, and those that are purely client-side. The mere fact of modification must not be sufficient to trigger denial. Nor must network operators be allowed to approve and disapprove of modifications on the basis of their purely client-side behavior. Any time a network operator can deny access to a client which is externally indistinguishable from one of its own, it gains an unacceptable amount of control over client operators.</p><p>\nThus remote attestation must be forbidden. The previous language did this. The new language does not.</p>\ngpl3.nonsource.p10.s2\nnovalis\nlogin to agree\n5\n2559\n\n\nIt is entirely unclear what network access is being discussed here. If you mean that a modified product may be denied interoperability with unmodified products, say so - although it is unclear whether that is good or bad. It might be bad if done spitefully to discourage mods, or good if a modified version might cause wide malfunctions.\ngpl3.nonsource.p10.s2\nrhouse\nlogin to agree\n3\n2578\n\n\n<p style="display: inline">"network access" is too narrow; the exception should apply to any service the software interacts with, whether accessed over a network or within the same device.</p><p>\nThat is, if the modified software misbehaves, there should be no expectation that other software in the device or that the device communicates with will ignore that misbehavior.</p><p>\nI agree that there is room for abuse by pretending that simple replacement is misbehavior, but I think the principle stated above is fundamental and that a court would recognize such abuse.\n </p>\ngpl3.nonsource.p10.s2\nsepreece\nlogin to agree\n1\n2698\n\n\nThis provision seems very specific to me. It looks as though the drafters have aimed it at a very few specific instances (DRM and remote attestation), which does not have wide applications, particularly in the free world. Given that, I wonder if we could move this into the list of additional terms, where authors of software that runs in circumstance where this come us as an issue could specify the appropriate conditions. \ngpl3.nonsource.p10.s2\nlocutas\nlogin to agree\n0\n2827\n\n\n<p style="display: inline">It's not immediately clear that the intent is that network access should not be denied based on modification unless there's a good reason. I'd suggest rewording to something like:</p><p>\n"However, you may not deny network access solely on the basis of a modification unless that modification itself materially and adversely affects the operation of the network or violates the rules and protocols for communication across the network."</p>\ngpl3.nonsource.p10.s2\ntfelker\nlogin to agree\n0\n3089\n\n\n", "2496:2523:2559:2578:2827:3089": "\n\n\n\n<p style="display: inline">Please remove the sentence highlighted in this comment. First of all, it is unclear who is doing the denying of network access: the client's hardware or the server. Imagine a Tivo-like appliance designed so that the network card commits suicide if the md5sum of the loaded operating system has changed. The GPLv3 in its current form would allow this type of DRM because of the loophole introduced by the sentence highlighted in this comment.</p><p>\nIn combination with the "ROM exemption" (http://gplv3.fsf.org/comments/rt/readsay.html?filename=gplv3-draft-3&id=2492), this sentence forms an even greater loophole: GPL'ed but unmodifiable software stored in ROM would be the one performing the md5sum and killing the network card.</p><p>\nAs for "rules and protocols", this is an open invitation for DRM to be included in them. Terrible idea to even mention them. And it might give credence to those that are trying to kill Net Neutrality.</p><p>\nAgain, please delete this whole sentence.</p>\ngpl3.nonsource.p10.s2\nvladc\nlogin to agree\n7\n2496\n\n\nClarify that this is only allowed if the modification actually does alter network behavior and that simply suggesting that any modified code could possibly have security vulnerabilities is not enough. As it stands it is far to easy for companies to say "well we are afraid what might happen if we let everybody on" and the block any modified software. \ngpl3.nonsource.p10.s2\njring\nlogin to agree\n7\n2523\n\n\n<p style="display: inline">We need to distinguish between rules and protocols that involve the externally visible actions of the client, and those that are purely client-side. The mere fact of modification must not be sufficient to trigger denial. Nor must network operators be allowed to approve and disapprove of modifications on the basis of their purely client-side behavior. Any time a network operator can deny access to a client which is externally indistinguishable from one of its own, it gains an unacceptable amount of control over client operators.</p><p>\nThus remote attestation must be forbidden. The previous language did this. The new language does not.</p>\ngpl3.nonsource.p10.s2\nnovalis\nlogin to agree\n5\n2559\n\n\nIt is entirely unclear what network access is being discussed here. If you mean that a modified product may be denied interoperability with unmodified products, say so - although it is unclear whether that is good or bad. It might be bad if done spitefully to discourage mods, or good if a modified version might cause wide malfunctions.\ngpl3.nonsource.p10.s2\nrhouse\nlogin to agree\n3\n2578\n\n\nThis provision seems very specific to me. It looks as though the drafters have aimed it at a very few specific instances (DRM and remote attestation), which does not have wide applications, particularly in the free world. Given that, I wonder if we could move this into the list of additional terms, where authors of software that runs in circumstance where this come us as an issue could specify the appropriate conditions. \ngpl3.nonsource.p10.s2\nlocutas\nlogin to agree\n0\n2827\n\n\n<p style="display: inline">It's not immediately clear that the intent is that network access should not be denied based on modification unless there's a good reason. I'd suggest rewording to something like:</p><p>\n"However, you may not deny network access solely on the basis of a modification unless that modification itself materially and adversely affects the operation of the network or violates the rules and protocols for communication across the network."</p>\ngpl3.nonsource.p10.s2\ntfelker\nlogin to agree\n0\n3089\n\n\n", "2458:2483:2491:2496:2523:2559:2578:2827:3089": "\n\n\n\nTivo, for instance, might argue that modified copies "materially" affects the operation of its network in the sense that it reduces their profits accruing from operating the network.\ngpl3.nonsource.p10.s2\ncyd\nlogin to agree\n4\n2458\n\n\n<p style="display: inline">FCC regulations for instance. So "when the modification itself violates a law or regulation applicable to the User Product, or violates the rules and protocols of a network".</p><p>\nThat way you cover things like FCC broadcast rules.</p>\ngpl3.nonsource.p10.s2\nmerijn\nlogin to agree\n1\n2483\n\n\n<p style="display: inline">Tivo can easily claim that everything except their encrypted protocol and DRM adversely affects the operation of the network. This wording is as weak as water!</p><p>\nKeep the wording from Draft 2:</p><p>\nIf the work communicates with an online service, it must be possible for modified versions to communicate with the same online service in the same way such that the service cannot distinguish.)</p>\ngpl3.nonsource.p10.s2\nvladc\nlogin to agree\n14\n2491\n\n\n<p style="display: inline">Please remove the sentence highlighted in this comment. First of all, it is unclear who is doing the denying of network access: the client's hardware or the server. Imagine a Tivo-like appliance designed so that the network card commits suicide if the md5sum of the loaded operating system has changed. The GPLv3 in its current form would allow this type of DRM because of the loophole introduced by the sentence highlighted in this comment.</p><p>\nIn combination with the "ROM exemption" (http://gplv3.fsf.org/comments/rt/readsay.html?filename=gplv3-draft-3&id=2492), this sentence forms an even greater loophole: GPL'ed but unmodifiable software stored in ROM would be the one performing the md5sum and killing the network card.</p><p>\nAs for "rules and protocols", this is an open invitation for DRM to be included in them. Terrible idea to even mention them. And it might give credence to those that are trying to kill Net Neutrality.</p><p>\nAgain, please delete this whole sentence.</p>\ngpl3.nonsource.p10.s2\nvladc\nlogin to agree\n7\n2496\n\n\nClarify that this is only allowed if the modification actually does alter network behavior and that simply suggesting that any modified code could possibly have security vulnerabilities is not enough. As it stands it is far to easy for companies to say "well we are afraid what might happen if we let everybody on" and the block any modified software. \ngpl3.nonsource.p10.s2\njring\nlogin to agree\n7\n2523\n\n\n<p style="display: inline">We need to distinguish between rules and protocols that involve the externally visible actions of the client, and those that are purely client-side. The mere fact of modification must not be sufficient to trigger denial. Nor must network operators be allowed to approve and disapprove of modifications on the basis of their purely client-side behavior. Any time a network operator can deny access to a client which is externally indistinguishable from one of its own, it gains an unacceptable amount of control over client operators.</p><p>\nThus remote attestation must be forbidden. The previous language did this. The new language does not.</p>\ngpl3.nonsource.p10.s2\nnovalis\nlogin to agree\n5\n2559\n\n\nIt is entirely unclear what network access is being discussed here. If you mean that a modified product may be denied interoperability with unmodified products, say so - although it is unclear whether that is good or bad. It might be bad if done spitefully to discourage mods, or good if a modified version might cause wide malfunctions.\ngpl3.nonsource.p10.s2\nrhouse\nlogin to agree\n3\n2578\n\n\nThis provision seems very specific to me. It looks as though the drafters have aimed it at a very few specific instances (DRM and remote attestation), which does not have wide applications, particularly in the free world. Given that, I wonder if we could move this into the list of additional terms, where authors of software that runs in circumstance where this come us as an issue could specify the appropriate conditions. \ngpl3.nonsource.p10.s2\nlocutas\nlogin to agree\n0\n2827\n\n\n<p style="display: inline">It's not immediately clear that the intent is that network access should not be denied based on modification unless there's a good reason. I'd suggest rewording to something like:</p><p>\n"However, you may not deny network access solely on the basis of a modification unless that modification itself materially and adversely affects the operation of the network or violates the rules and protocols for communication across the network."</p>\ngpl3.nonsource.p10.s2\ntfelker\nlogin to agree\n0\n3089\n\n\n", "2496:2523:2559:2578:2787:2827:3089": "\n\n\n\n<p style="display: inline">Please remove the sentence highlighted in this comment. First of all, it is unclear who is doing the denying of network access: the client's hardware or the server. Imagine a Tivo-like appliance designed so that the network card commits suicide if the md5sum of the loaded operating system has changed. The GPLv3 in its current form would allow this type of DRM because of the loophole introduced by the sentence highlighted in this comment.</p><p>\nIn combination with the "ROM exemption" (http://gplv3.fsf.org/comments/rt/readsay.html?filename=gplv3-draft-3&id=2492), this sentence forms an even greater loophole: GPL'ed but unmodifiable software stored in ROM would be the one performing the md5sum and killing the network card.</p><p>\nAs for "rules and protocols", this is an open invitation for DRM to be included in them. Terrible idea to even mention them. And it might give credence to those that are trying to kill Net Neutrality.</p><p>\nAgain, please delete this whole sentence.</p>\ngpl3.nonsource.p10.s2\nvladc\nlogin to agree\n7\n2496\n\n\nClarify that this is only allowed if the modification actually does alter network behavior and that simply suggesting that any modified code could possibly have security vulnerabilities is not enough. As it stands it is far to easy for companies to say "well we are afraid what might happen if we let everybody on" and the block any modified software. \ngpl3.nonsource.p10.s2\njring\nlogin to agree\n7\n2523\n\n\n<p style="display: inline">We need to distinguish between rules and protocols that involve the externally visible actions of the client, and those that are purely client-side. The mere fact of modification must not be sufficient to trigger denial. Nor must network operators be allowed to approve and disapprove of modifications on the basis of their purely client-side behavior. Any time a network operator can deny access to a client which is externally indistinguishable from one of its own, it gains an unacceptable amount of control over client operators.</p><p>\nThus remote attestation must be forbidden. The previous language did this. The new language does not.</p>\ngpl3.nonsource.p10.s2\nnovalis\nlogin to agree\n5\n2559\n\n\nIt is entirely unclear what network access is being discussed here. If you mean that a modified product may be denied interoperability with unmodified products, say so - although it is unclear whether that is good or bad. It might be bad if done spitefully to discourage mods, or good if a modified version might cause wide malfunctions.\ngpl3.nonsource.p10.s2\nrhouse\nlogin to agree\n3\n2578\n\n\n<p style="display: inline">So... the TOS for the service is "no modified clients", and so any modified client violates the "rules... for communication across the network".</p><p>\nIf you put this here, it basically explicitly allows the distributor to prevent modification of any network client software.</p>\ngpl3.nonsource.p10.s2\nchandon\nlogin to agree\n2\n2787\n\n\nThis provision seems very specific to me. It looks as though the drafters have aimed it at a very few specific instances (DRM and remote attestation), which does not have wide applications, particularly in the free world. Given that, I wonder if we could move this into the list of additional terms, where authors of software that runs in circumstance where this come us as an issue could specify the appropriate conditions. \ngpl3.nonsource.p10.s2\nlocutas\nlogin to agree\n0\n2827\n\n\n<p style="display: inline">It's not immediately clear that the intent is that network access should not be denied based on modification unless there's a good reason. I'd suggest rewording to something like:</p><p>\n"However, you may not deny network access solely on the basis of a modification unless that modification itself materially and adversely affects the operation of the network or violates the rules and protocols for communication across the network."</p>\ngpl3.nonsource.p10.s2\ntfelker\nlogin to agree\n0\n3089\n\n\n", "2496:2523:2559:2578:2787:2827:3089:3113": "\n\n\n\n<p style="display: inline">Please remove the sentence highlighted in this comment. First of all, it is unclear who is doing the denying of network access: the client's hardware or the server. Imagine a Tivo-like appliance designed so that the network card commits suicide if the md5sum of the loaded operating system has changed. The GPLv3 in its current form would allow this type of DRM because of the loophole introduced by the sentence highlighted in this comment.</p><p>\nIn combination with the "ROM exemption" (http://gplv3.fsf.org/comments/rt/readsay.html?filename=gplv3-draft-3&id=2492), this sentence forms an even greater loophole: GPL'ed but unmodifiable software stored in ROM would be the one performing the md5sum and killing the network card.</p><p>\nAs for "rules and protocols", this is an open invitation for DRM to be included in them. Terrible idea to even mention them. And it might give credence to those that are trying to kill Net Neutrality.</p><p>\nAgain, please delete this whole sentence.</p>\ngpl3.nonsource.p10.s2\nvladc\nlogin to agree\n7\n2496\n\n\nClarify that this is only allowed if the modification actually does alter network behavior and that simply suggesting that any modified code could possibly have security vulnerabilities is not enough. As it stands it is far to easy for companies to say "well we are afraid what might happen if we let everybody on" and the block any modified software. \ngpl3.nonsource.p10.s2\njring\nlogin to agree\n7\n2523\n\n\n<p style="display: inline">We need to distinguish between rules and protocols that involve the externally visible actions of the client, and those that are purely client-side. The mere fact of modification must not be sufficient to trigger denial. Nor must network operators be allowed to approve and disapprove of modifications on the basis of their purely client-side behavior. Any time a network operator can deny access to a client which is externally indistinguishable from one of its own, it gains an unacceptable amount of control over client operators.</p><p>\nThus remote attestation must be forbidden. The previous language did this. The new language does not.</p>\ngpl3.nonsource.p10.s2\nnovalis\nlogin to agree\n5\n2559\n\n\nIt is entirely unclear what network access is being discussed here. If you mean that a modified product may be denied interoperability with unmodified products, say so - although it is unclear whether that is good or bad. It might be bad if done spitefully to discourage mods, or good if a modified version might cause wide malfunctions.\ngpl3.nonsource.p10.s2\nrhouse\nlogin to agree\n3\n2578\n\n\n<p style="display: inline">So... the TOS for the service is "no modified clients", and so any modified client violates the "rules... for communication across the network".</p><p>\nIf you put this here, it basically explicitly allows the distributor to prevent modification of any network client software.</p>\ngpl3.nonsource.p10.s2\nchandon\nlogin to agree\n2\n2787\n\n\nThis provision seems very specific to me. It looks as though the drafters have aimed it at a very few specific instances (DRM and remote attestation), which does not have wide applications, particularly in the free world. Given that, I wonder if we could move this into the list of additional terms, where authors of software that runs in circumstance where this come us as an issue could specify the appropriate conditions. \ngpl3.nonsource.p10.s2\nlocutas\nlogin to agree\n0\n2827\n\n\n<p style="display: inline">It's not immediately clear that the intent is that network access should not be denied based on modification unless there's a good reason. I'd suggest rewording to something like:</p><p>\n"However, you may not deny network access solely on the basis of a modification unless that modification itself materially and adversely affects the operation of the network or violates the rules and protocols for communication across the network."</p>\ngpl3.nonsource.p10.s2\ntfelker\nlogin to agree\n0\n3089\n\n\n"the rules and protocols"? Whose rules and protocols? Wouldn't those usually be the rules and protocols of the very licensee who is tryint to circumvent the GPL? I don't think this is going to work. The whole concept of allowing the GPL licensee to terminate network access for "good cause" is flawed, because they'll just define the rules and protocols as they need them. So as I see it, this is equivalent to saying "network access may be denied to modified devices".\ngpl3.nonsource.p10.s2\nmux2005\nlogin to agree\n0\n3113\n\n\n", "2552:2559:2578:2694:2698:2767:2787:2827:2925:3089:3113": "\n\n\n\n<p style="display: inline">According to the rationale, the reason for the "User Product" language was that certain large enterprise customers prefer not to have the DRM keys, because they want the vendor to control their system updates.</p><p>\nHowever, according to this text, can't the vendor simply say that they will no longer update or warrantee the system if the user employs the DRM keys to modify it? It seems that this should remove the need for the controversial "User Product" language.</p>\ngpl3.nonsource.p10.s1\nstevenj\nlogin to agree\n3\n2552\n\n\n<p style="display: inline">We need to distinguish between rules and protocols that involve the externally visible actions of the client, and those that are purely client-side. The mere fact of modification must not be sufficient to trigger denial. Nor must network operators be allowed to approve and disapprove of modifications on the basis of their purely client-side behavior. Any time a network operator can deny access to a client which is externally indistinguishable from one of its own, it gains an unacceptable amount of control over client operators.</p><p>\nThus remote attestation must be forbidden. The previous language did this. The new language does not.</p>\ngpl3.nonsource.p10.s2\nnovalis\nlogin to agree\n5\n2559\n\n\nIt is entirely unclear what network access is being discussed here. If you mean that a modified product may be denied interoperability with unmodified products, say so - although it is unclear whether that is good or bad. It might be bad if done spitefully to discourage mods, or good if a modified version might cause wide malfunctions.\ngpl3.nonsource.p10.s2\nrhouse\nlogin to agree\n3\n2578\n\n\nThis language seems to allow a vendor to lock down a device by having a continuing update process (giving the new software a new key at frequent intervals). This language says that updates can be stopped if the software is modified, which means that the keys would stop being updated, and corresponding service could stop accepting the client on that basis.\n\ngpl3.nonsource.p10.s1\nsepreece\nlogin to agree\n2\n2694\n\n\n<p style="display: inline">"network access" is too narrow; the exception should apply to any service the software interacts with, whether accessed over a network or within the same device.</p><p>\nThat is, if the modified software misbehaves, there should be no expectation that other software in the device or that the device communicates with will ignore that misbehavior.</p><p>\nI agree that there is room for abuse by pretending that simple replacement is misbehavior, but I think the principle stated above is fundamental and that a court would recognize such abuse.\n </p>\ngpl3.nonsource.p10.s2\nsepreece\nlogin to agree\n1\n2698\n\n\nThis requirement to provide Installation Information is a reworded and clarified version of the requirement to include installation keys in the Corresponding Source (as the draft rationale explains). The current language seems clearer and better. Good.\ngpl3.nonsource.p10.s1\nfrx\nlogin to agree\n0\n2767\n\n\n<p style="display: inline">So... the TOS for the service is "no modified clients", and so any modified client violates the "rules... for communication across the network".</p><p>\nIf you put this here, it basically explicitly allows the distributor to prevent modification of any network client software.</p>\ngpl3.nonsource.p10.s2\nchandon\nlogin to agree\n2\n2787\n\n\nThis provision seems very specific to me. It looks as though the drafters have aimed it at a very few specific instances (DRM and remote attestation), which does not have wide applications, particularly in the free world. Given that, I wonder if we could move this into the list of additional terms, where authors of software that runs in circumstance where this come us as an issue could specify the appropriate conditions. \ngpl3.nonsource.p10.s2\nlocutas\nlogin to agree\n0\n2827\n\n\n<p style="display: inline">One way to circumvent the anti-tivoization is as follows: the device's vendor allows installation of modified works by default, but has a "vendor lock" flag that can be enabled by the user and prevents such modifications. Service is then conditioned upon the user "voluntarily" setting the flag. See my comment 2923 for a full description.</p><p>\nThis is a potentially devastating loophole: the user gets the choice of a free but useless device, or a serviced but forever-unmodifiable device.</p><p>\nTo close this loophole, I suggest adding a sentence to this paragraph, right after the first sentence:</p><p>\n"However, support service, warranty, and updates must not be conditioned upon acceptance of a constraint or obligation to avoid future modifications or installation of modified works."</p>\ngpl3.nonsource.p10.s1\nblinken\nlogin to agree\n0\n2925\n\n\n<p style="display: inline">It's not immediately clear that the intent is that network access should not be denied based on modification unless there's a good reason. I'd suggest rewording to something like:</p><p>\n"However, you may not deny network access solely on the basis of a modification unless that modification itself materially and adversely affects the operation of the network or violates the rules and protocols for communication across the network."</p>\ngpl3.nonsource.p10.s2\ntfelker\nlogin to agree\n0\n3089\n\n\n"the rules and protocols"? Whose rules and protocols? Wouldn't those usually be the rules and protocols of the very licensee who is tryint to circumvent the GPL? I don't think this is going to work. The whole concept of allowing the GPL licensee to terminate network access for "good cause" is flawed, because they'll just define the rules and protocols as they need them. So as I see it, this is equivalent to saying "network access may be denied to modified devices".\ngpl3.nonsource.p10.s2\nmux2005\nlogin to agree\n0\n3113\n\n\n", "2484": "\n\n\n\nAlready Corr Source must be in preferred form for modification. So move this text to the definition of Corresponding SOurcde.\ngpl3.nonsource.p11.s1\nmerijn\nlogin to agree\n3\n2484\n\n\n", "2484:3006": "\n\n\n\nAlready Corr Source must be in preferred form for modification. So move this text to the definition of Corresponding SOurcde.\ngpl3.nonsource.p11.s1\nmerijn\nlogin to agree\n3\n2484\n\n\n<p style="display: inline">Can this phrase ("publicly documented") stop code obfuscation? For example, would the obfuscated nv* driver be conveyed under GPLv3?</p><p>\n* http://cvsweb.xfree86.org/cvsweb/xc/programs/Xserver/hw/xfree86/vga256/drivers/nv/?hideattic=0&only_with_tag=xf-3_3_3</p>\ngpl3.nonsource.p11.s1\ndaa\nlogin to agree\n0\n3006\n\n\n", "2602": "\n\n\n\n-> "must be unpackable, readable, and copyable by normal means, specifically, a form requiring no special password, key, code, or other similar access control data to perform these tasks."\ngpl3.nonsource.p11.s1\nskquinn\nlogin to agree\n1\n2602\n\n\n", "2602:2695": "\n\n\n\n-> "must be unpackable, readable, and copyable by normal means, specifically, a form requiring no special password, key, code, or other similar access control data to perform these tasks."\ngpl3.nonsource.p11.s1\nskquinn\nlogin to agree\n1\n2602\n\n\nIt seems like it would be OK to require a key or password, so long as the needed key or password is provided or available on request.\n\ngpl3.nonsource.p11.s1\nsepreece\nlogin to agree\n1\n2695\n\n\n", "2531:2602": "\n\n\n\nmaybe also for "compiling and executing", because without this, the software could be checked by hardware before execution, which should go against GPL IMHO.\ngpl3.nonsource.p11.s1\nmichals\nlogin to agree\n2\n2531\n\n\n-> "must be unpackable, readable, and copyable by normal means, specifically, a form requiring no special password, key, code, or other similar access control data to perform these tasks."\ngpl3.nonsource.p11.s1\nskquinn\nlogin to agree\n1\n2602\n\n\n", "2484:2531:2602:2695:3006": "\n\n\n\nAlready Corr Source must be in preferred form for modification. So move this text to the definition of Corresponding SOurcde.\ngpl3.nonsource.p11.s1\nmerijn\nlogin to agree\n3\n2484\n\n\nmaybe also for "compiling and executing", because without this, the software could be checked by hardware before execution, which should go against GPL IMHO.\ngpl3.nonsource.p11.s1\nmichals\nlogin to agree\n2\n2531\n\n\n-> "must be unpackable, readable, and copyable by normal means, specifically, a form requiring no special password, key, code, or other similar access control data to perform these tasks."\ngpl3.nonsource.p11.s1\nskquinn\nlogin to agree\n1\n2602\n\n\nIt seems like it would be OK to require a key or password, so long as the needed key or password is provided or available on request.\n\ngpl3.nonsource.p11.s1\nsepreece\nlogin to agree\n1\n2695\n\n\n<p style="display: inline">Can this phrase ("publicly documented") stop code obfuscation? For example, would the obfuscated nv* driver be conveyed under GPLv3?</p><p>\n* http://cvsweb.xfree86.org/cvsweb/xc/programs/Xserver/hw/xfree86/vga256/drivers/nv/?hideattic=0&only_with_tag=xf-3_3_3</p>\ngpl3.nonsource.p11.s1\ndaa\nlogin to agree\n0\n3006\n\n\n", "2665:2768:2792:2846:2960:2965": "\n\n\n\nDepending on how you read it this section seems to allow either any kind of exception or to restrict you to only restrictions of kind a-d under local law. Please can you clarify this section.\ngpl3.licensecompat.0.0\nrobmyers\nlogin to agree\n0\n2665\n\n\nThis section is improved with respect to the previous draft, especially because two additional requirements can no longer be added to a covered work (forced distribution of source to network users, and strong patent retaliation clause). Unfortunately more improvements are needed for this section to be considered OK. Further details in later comments.\ngpl3.licensecompat.0.0\nfrx\nlogin to agree\n0\n2768\n\n\n<p style="display: inline">If I read this section correctly, it allows two things:</p><p>\n- additional permissions by making exceptions to sections\n- additional restrictions, in the categories a-d.</p><p>\nI may however be misreading it, which is exactly the problem. Please clarify this section. If it does indeed mean what I think it does, subheadings would probably help.</p>\ngpl3.licensecompat.0.0\nwijnen\nlogin to agree\n0\n2792\n\n\nThe previous section 7 made GPLv3 compatibility with the Apache license. This one doesn't. That's a huge bug.\ngpl3.licensecompat.0.0\nnovalis\nlogin to agree\n2\n2846\n\n\nI suggest a new additional requirement that licensors may require licensees to disclose security patches to a specified e-mail address (which must be at a domain that hosts the official website of a covered work) within, say 48 hours of that security patch being either conveyed or used on a public server.\ngpl3.licensecompat.0.0\njamesgnz\nlogin to agree\n0\n2960\n\n\n<p style="display: inline">In my eyes, the whole concept of how additional restrictions are invented here is broken. It clutters the meaning of the GPL in a cancer-like way. The code labeled GPL must stay GPL, so we know what it means.</p><p>\nThe GPL is complex enough on its own, inventing fine-printed variations of it does no good.</p><p>\nBut I agree that compatibility with other free licenses is a good thing.</p><p>\nIMHO, the solution is add an exception that allows linking / combining with code that comes with some additional restrictions (7a-7d, apache patent clause, affero web app clause, maybe even the BSD advertising clause), but is compatible to the GPL otherwise.</p><p>\nIt is much easier to manage a source project that has some files under GPL, some under BSD and some under Apache license, than having some files under plain GPL, some under GPL + foo, and others under GPL+bar.</p>\ngpl3.licensecompat.0.0\nschabi\nlogin to agree\n1\n2965\n\n\n", "2664": "\n\n\n\nCan the LGPL still just be an additional permission? If so, under which part of this section? It doesn't seem to fit a-d or be a fix for local law.\n\ngpl3.licensecompat.p0.s1\nrobmyers\nlogin to agree\n0\n2664\n\n\n", "3081": "\n\n\n\n\nIf a copy of a work is dual-licensed under the GPL and a different license, does the licensing of the work under the different license fall in the category of "Additional permissions"? When should other parties be allowed to remove additional permissions that have been specified by the copyright holder of a work? Should the removal of additional permissions always be permitted? If the copyright holder for a work gives permission for others to exploit the work in a certain manner, are there circumstances where other parties should or should not be allowed to remove that permission?\n\n\ngpl3.licensecompat.p1.s1\nmezzanine1\nlogin to agree\n0\n3081\n\n\n", "2677:3081": "\n\n\n\n=> "may"\ngpl3.licensecompat.p1.s1\nsanjoy\nlogin to agree\n1\n2677\n\n\n\nIf a copy of a work is dual-licensed under the GPL and a different license, does the licensing of the work under the different license fall in the category of "Additional permissions"? When should other parties be allowed to remove additional permissions that have been specified by the copyright holder of a work? Should the removal of additional permissions always be permitted? If the copyright holder for a work gives permission for others to exploit the work in a certain manner, are there circumstances where other parties should or should not be allowed to remove that permission?\n\n\ngpl3.licensecompat.p1.s1\nmezzanine1\nlogin to agree\n0\n3081\n\n\n", "2663:2742": "\n\n\n\n<p style="display: inline">I don't understand how this statement relates to the following clauses a-d.</p><p>\nCan you add terms of kind a-d that have been drafted to be international (as the GPL is an international license)? If so, why does this statement appear to say that they must be localized?</p><p>\nCan you add terms different from a-d if they are for local law only? If so, why do clauses a-d follow?</p><p>\nI think the colon at the end of this statement is the problem. :-)</p>\ngpl3.licensecompat.p2.s1\nrobmyers\nlogin to agree\n1\n2663\n\n\nThis paragraph should be (emphasized added text):\r\n\r\n"Notwithstanding any other provision of this License, AND ONLY IF REQUIRED BY LOCAL LAW, you may supplement the terms of this License with terms effective ONLY under local law:"\ngpl3.licensecompat.p2.s1\nvincentc\nlogin to agree\n2\n2742\n\n\n", "2663:2742:2769:3114": "\n\n\n\n<p style="display: inline">I don't understand how this statement relates to the following clauses a-d.</p><p>\nCan you add terms of kind a-d that have been drafted to be international (as the GPL is an international license)? If so, why does this statement appear to say that they must be localized?</p><p>\nCan you add terms different from a-d if they are for local law only? If so, why do clauses a-d follow?</p><p>\nI think the colon at the end of this statement is the problem. :-)</p>\ngpl3.licensecompat.p2.s1\nrobmyers\nlogin to agree\n1\n2663\n\n\nThis paragraph should be (emphasized added text):\r\n\r\n"Notwithstanding any other provision of this License, AND ONLY IF REQUIRED BY LOCAL LAW, you may supplement the terms of this License with terms effective ONLY under local law:"\ngpl3.licensecompat.p2.s1\nvincentc\nlogin to agree\n2\n2742\n\n\n<p style="display: inline">I strongly *dislike* the entire concept of allowing a limited set of additional requirements to be added.</p><p>\nThat is *against* the spirit of the GPLv2 (where the FSF promised that new versions would "be similar in spirit to the present version", see GPLv2, section 9.) and greatly weakens the copyleft.</p><p>\nYes, it enhances license compatibility (as the rationale of GPLv3draft1 explains), but the GPL never was a champion of compatibility. If the goal were maximizing license compatibility, the best choice would be adopting the 2-clause BSD license (http://www.gnu.org/licenses/info/BSD_2Clause.html) as the new GPL... It would be compatible with pretty everything and clearly free, even though non-copyleft. I'm *not* against non-copyleft free licenses, but they definitely aren't "similar in spirit to" GPLv2...</p><p>\nI would be happy to see all these "permissions to add restrictions" entirely dropped from Section 7.</p>\ngpl3.licensecompat.p2.s1\nfrx\nlogin to agree\n0\n2769\n\n\n<p style="display: inline">Like other commentors I have trouble understanding what the "...local law" is supposed to achieve. Is this meant to restrict what additional terms I can add? I can see how that would make sense with respect to a., so that people can add special disclaimers that are ONLY valid in certain jurisdictions. But I don't see how that would make sense with respect to d. Limiting the use of trademarks is a restriction that I'm sure a lot of corporate licensors will want to apply independent of jurisdiction. Let's take an example to make this clear.<br><br></p><p>\nThe company Gnomovision(R) takes The Gimp and improves it. Are they allowed to add the following additional terms to their improved Gimp "You may not use the trademark Gnomovision(R) in advertising materials for products based on or including this software." ?<br><br></p><p>\nThis is an additional term that Gnomovision would like to apply universally, everywhere in the world. So it's certainly not "drafted for compatibility with, local law". But, depending on what "local" means, it may be "effective under local law". But that is true for any terms of significance. After all, if a term is not effective under any local law, then it is not effective at all and therefore meaningless. So that brings me back to my initial question. Is the "...local law" part supposed to be a restriction of what kind of terms I can add? Because due to the "effective under" language, I don't think it has any restricting effect.</p>\ngpl3.licensecompat.p2.s1\nmux2005\nlogin to agree\n0\n3114\n\n\n", "2663:2742:2769:3114:3115": "\n\n\n\n<p style="display: inline">I don't understand how this statement relates to the following clauses a-d.</p><p>\nCan you add terms of kind a-d that have been drafted to be international (as the GPL is an international license)? If so, why does this statement appear to say that they must be localized?</p><p>\nCan you add terms different from a-d if they are for local law only? If so, why do clauses a-d follow?</p><p>\nI think the colon at the end of this statement is the problem. :-)</p>\ngpl3.licensecompat.p2.s1\nrobmyers\nlogin to agree\n1\n2663\n\n\nThis paragraph should be (emphasized added text):\r\n\r\n"Notwithstanding any other provision of this License, AND ONLY IF REQUIRED BY LOCAL LAW, you may supplement the terms of this License with terms effective ONLY under local law:"\ngpl3.licensecompat.p2.s1\nvincentc\nlogin to agree\n2\n2742\n\n\n<p style="display: inline">I strongly *dislike* the entire concept of allowing a limited set of additional requirements to be added.</p><p>\nThat is *against* the spirit of the GPLv2 (where the FSF promised that new versions would "be similar in spirit to the present version", see GPLv2, section 9.) and greatly weakens the copyleft.</p><p>\nYes, it enhances license compatibility (as the rationale of GPLv3draft1 explains), but the GPL never was a champion of compatibility. If the goal were maximizing license compatibility, the best choice would be adopting the 2-clause BSD license (http://www.gnu.org/licenses/info/BSD_2Clause.html) as the new GPL... It would be compatible with pretty everything and clearly free, even though non-copyleft. I'm *not* against non-copyleft free licenses, but they definitely aren't "similar in spirit to" GPLv2...</p><p>\nI would be happy to see all these "permissions to add restrictions" entirely dropped from Section 7.</p>\ngpl3.licensecompat.p2.s1\nfrx\nlogin to agree\n0\n2769\n\n\n<p style="display: inline">Like other commentors I have trouble understanding what the "...local law" is supposed to achieve. Is this meant to restrict what additional terms I can add? I can see how that would make sense with respect to a., so that people can add special disclaimers that are ONLY valid in certain jurisdictions. But I don't see how that would make sense with respect to d. Limiting the use of trademarks is a restriction that I'm sure a lot of corporate licensors will want to apply independent of jurisdiction. Let's take an example to make this clear.<br><br></p><p>\nThe company Gnomovision(R) takes The Gimp and improves it. Are they allowed to add the following additional terms to their improved Gimp "You may not use the trademark Gnomovision(R) in advertising materials for products based on or including this software." ?<br><br></p><p>\nThis is an additional term that Gnomovision would like to apply universally, everywhere in the world. So it's certainly not "drafted for compatibility with, local law". But, depending on what "local" means, it may be "effective under local law". But that is true for any terms of significance. After all, if a term is not effective under any local law, then it is not effective at all and therefore meaningless. So that brings me back to my initial question. Is the "...local law" part supposed to be a restriction of what kind of terms I can add? Because due to the "effective under" language, I don't think it has any restricting effect.</p>\ngpl3.licensecompat.p2.s1\nmux2005\nlogin to agree\n0\n3114\n\n\nWhat is "local law"? In a world of multi-national corporations that doesn't seem to make much sense. Take IBM for instance. What kinds of additional terms are they allowed to add? What is "local law" to them? Furthermore, law is usually nested. Most commonly there is state law and federal law, but there may be more layers. For a company based in New York, does "local law" refer to "New York state law" (which is "local" with respect to the US). Does it refer to "US federal law" (which is "local" with respect to the world)?\ngpl3.licensecompat.p2.s1\nmux2005\nlogin to agree\n0\n3115\n\n\n", "3144": "\n\n\n\n<p style="display: inline">Consider language commonly used among professional software pirates:\n... we make no claims and provide no assurances that this software does not infringe on the rights of third parties... even if we have been informed of the possibility of such infringement... the customer bears all responsibility for such infringement, agrees not to hold us liable for such infringement, agrees not to prosecute us or aid in the prosecution of us, and the customer agrees to pay any court cost we may bear resulting from such prosecution...\n[to paraphrase vague memories of the MS EULA]</p><p>\nImagine a scenario where Novell decided to utilize the potential GPLv3 section 7 part a to adopt such language, where the Novell/MS agreement dissipates, and MS (directly to customers) says "L may infringe x,y, or z; pay us and adopt our very restrictive licensing terms or else"</p><p>\nSuch would be a drastic change from the current situation where Novell must uphold the purity of Linux and must portray any such agreement as meaningless and just a way to avoid frivolous lawsuits.</p>\ngpl3.licensecompat.p3.s1\ndrcj\nlogin to agree\n0\n3144\n\n\n", "2770": "\n\n\n\nWhat exactly is a "reasonable legal notice"? What exactly is an "author attribution"? It seems that these terms are not defined anywhere in the license. I'm concerned that they could be interpreted in a broad sense and allow people to take a GPLv3'd work and add some sort of invariant long text that nobody will ever be able to remove or modify... This option could make a work include unmodifiable & unremovable parts and thus fail to fully grant the freedom to modify. I would rather avoid introducing such options in the GPLv3!\ngpl3.licensecompat.p4.s1\nfrx\nlogin to agree\n0\n2770\n\n\n", "2770:3140": "\n\n\n\nWhat exactly is a "reasonable legal notice"? What exactly is an "author attribution"? It seems that these terms are not defined anywhere in the license. I'm concerned that they could be interpreted in a broad sense and allow people to take a GPLv3'd work and add some sort of invariant long text that nobody will ever be able to remove or modify... This option could make a work include unmodifiable & unremovable parts and thus fail to fully grant the freedom to modify. I would rather avoid introducing such options in the GPLv3!\ngpl3.licensecompat.p4.s1\nfrx\nlogin to agree\n0\n2770\n\n\n<p style="display: inline">I noticed this language problem in section 7 b immediately upon publication of GPL V3 DD3 and should have commented earlier but I did not comment quickly because of my preoccupation with researching unrelated challenges to the GPL. I hope I correct for tardiness in commenting by thoroughness in treatment.</p><p>\nThe GPL is about protecting the rights of all software users and not merely software developer/users. "the GNU General Public License is intended to guarantee your freedom to share and change free software--to make sure the software is free for all its users."</p><p>\nWhen the an optional requirement in accordance with section 7 b is exercised, DD3, new wording seems as if it may allow removal of specifically required author attributions or reasonable legal notices from an object code mechanism to place them in the source code. The effect of moving information from object code to source code would be to render an attribution or legal notice all but invisible to the vast majority of software users.</p><p>\n1. GPL V3 DD3 SECTION 7 B LANGUAGE.</p><p>\n7 b. "requiring preservation of specified reasonable legal notices or author attributions in source or object code forms of material added by you to a covered work; or"</p><p>\nAdding the language "in source or object code forms" introduces a problem. While the effect is less obvious than if the problematic wording had been placed earlier, such as "requiring preservation in source or object code form", but there is no grammatical difference and the interpretation may be no different. The problematic interpretation might be understood as 'requiring the preservation in source or object code form at the propagator's discretion'.</p><p>\n2. EFFECT OF THIS LANGUAGE.</p><p>\nThis issue may be insufficiently well appreciated by software developers because developers do not generally think of source code as something scary and difficult to understand. However, most users do not think of source code a possible source of anything informative to them. Most users are more likely to think of source code as a scary foreign language which could only be informative to the specially initiated. The idea that valuable information in plain text form may accompany source code is outside the experience of most users and not something users are likely to discover merely with the knowledge that source code is available.</p><p>\nWhy is it important to have an optional requirement for preserving author attributions and reasonable legal notices if we have something similar in the requirement of section 5 d? While author attribution and some legal notices may seem to be covered in section 5 d there is a degree of under inclusion in the section 5 d requirement to have some function in object code form if one had been present to display copyright notices; lack of \nwarranty notice (unless warranty provided); right of conveyance under the GPL; and how to view the GPL.</p><p>\nAn author and an entity holding the copyright of a work are not necessarily the same or even closely related. Author assignment of copyright is an important element for ensuring that a single entity can act decisively to protect the copyright and hence the GPL. An organisation holding copyright may cease to function for all the reasons that happen in the world or may be acquired by another entity without an interest in some particular code leaving a work essentially orphaned. Similarly an individual person holding an assignment may become incapacitated or die.</p><p>\nReasonable legal notices are certainly not limited to a statement of lack of warranty and the license terms. I have no examples to offer as additional legal notices are outside my scope of interest at the present time.</p><p>\n3. AUTHOR ATTRIBUTIONS.</p><p>\n3.1 VALUE OF AUTHOR ATTRIBUTIONS.</p><p>\nRemember that this is about preserving the users' rights including the right to modify the source code. If code were automatically fully self documenting, then we would all have an easier time. However, often even the author may have difficulty remembering how best to make some changes to his own code. Other developers are likely to have a more difficult time making changes which take best advantage of another developer's code. Mere users have the most difficulty determining how best to make changes. If an author wishes to make himself known by including an attribution requirement in object code, the user has the maximum opportunity for learning how to best modify the source code even when the copyright holder no longer exists. An author attribution hidden only in the source code for software which at least has a copyright statement included as part of an object code mechanism will leave most users expecting that only the copyright holding entity may provide any special insight into the code.</p><p>\nAccess to the source code alone makes modification possible. Yet, maximising access to the people who might most easily provide information about how to make successful changes or who might best implement such changes themselves is important for making more free software available than would otherwise be the case.</p><p>\nThe use of an option to require author attribution in object code form so that users are most likely to see the attribution will help provide additional motivation for some authors to create free software. More free software is certainly better for user freedom than less free software.</p><p>\n3.2. AUTHOR ATTRIBUTION DIFFERENCE FROM ORIGINAL BSD ADVERTISING CLAUSE.</p><p>\nIs an optional object code attribution requirement similar to the original BSD license advertising clause? No, an object code attribution requirement only has an effect when invoked upon copies of the object code, and unlike the original BSD license would have no effect on any advertisement.</p><p>\n3.3. PROBLEMATIC STATEMENTS INCLUDED ALONG WITH ATTRIBUTIONS.</p><p>\nSome have identified a risk of statements described as author attributions but which cross the line from mere attribution to something else. Perhaps the problematic phrase in section 7 b is weak attempt to ameliorate the risk. The problematic phrase is absent from the related parts a, c, and d of section 7.</p><p>\nAuthor attribution can be easily understood without any need for formal definition. We all know that author attribution should not be understood to be anything more than an identifying name, functional description for the attributed part of the corresponding source code; along with the author's name; other information for identifying the author, such as affiliation, etc.; and means to contact the author.</p><p>\nAnything more than attribution is better located on the author's website. The URL for the author's website may be included as part of an author attribution in information about how to contact the author. If statements which are more than attributions are improperly designated as author attributions, then they might simply be separated from the actual attribution elements and removed. Extra statements should only survive modification at the discretion of the modifier as a courtesy to the original author.</p><p>\n4. REASONABLE LEGAL NOTICES.</p><p>\nI leave it for others to comment about what reasonable legal notices are important to them as may be found in some other software licenses which will become compatible with GPL V3.</p><p>\n5. ALLOWING REASONABLE OBJECT CODE METHODOLOGY.</p><p>\nThe language of section 7 b itself should allow a document external to GPL V3 which applies the GPL V3 and optional GPL V3 restrictions using methodology similar to GPL V3 section 5 d in the external document for applying optional restrictions of GPL V3. An external document might specify something like the following for modifications in accordance with a corrected 7 b allowing such an application.</p><p>\nIf the work has interactive user interfaces, each must include a convenient feature that displays the [specified author attributions / reasonable legal notices]. Specifically, if the interface presents a list of user commands or options, such as a menu, a command to display this information must be prominent in the list; otherwise, the work must display this information at startup.</p><p>\nNothing within reason should prevent software modification from changing the name of the menu item or command line switch or substituting a different information display function. The important point is that the display function is prominent in any list, irrespective of how the function actually works.</p><p>\nOf course, nothing should prevent removing code which has optional restrictions attached along with all code derived from that code to avoid optional restrictions.</p><p>\n6. POSSIBLE LANGUAGE CORRECTION.</p><p>\n6.1. CHANGING 'OR' TO 'AND'.</p><p>\nIf the wording of section 7 b changed 'or' to 'and' as "in source and object code forms" the problem would be mitigated, but would seem to then add an unnecessary requirement. A modifier might interpret a requirement to add the other form if either source or object form had been included but not both for the attribution or notice when the modifier had received a copy of the code.</p><p>\n6.2. DROPPING THE PROBLEMATIC PHRASE.</p><p>\nDropping the problematic phrase from section 7 b altogether would leave the following in conformity with the related parts of section 7.</p><p>\n"requiring preservation of specified reasonable legal notices or author attributions of material added by you to a covered work; or"</p><p>\nIf an inappropriate use risk is considered too troublesome even after my answer above, demonstrating the ease of determining what is inappropriate and the ease of removing the inappropriate parts; then a definition of author attribution and/or reasonable legal notices should be given. I offered the elements of a common understanding of author attribution above along with my assertion that a formal definition was unnecessary.\n</p>\ngpl3.licensecompat.p4.s1\nthomasd\nlogin to agree\n0\n3140\n\n\n", "2821": "\n\n\n\nSuch prohibition and requirement is already part of this draft of the GPLv3: section 5. Why would one need supplemental terms?\ngpl3.licensecompat.p5.s1\nadhemar\nlogin to agree\n0\n2821\n\n\n", "2821:2986": "\n\n\n\nSuch prohibition and requirement is already part of this draft of the GPLv3: section 5. Why would one need supplemental terms?\ngpl3.licensecompat.p5.s1\nadhemar\nlogin to agree\n0\n2821\n\n\n<p style="display: inline">What does it mean in saying that modified versions must be marked? Does this mean modified versions of X could be required to contain a certain string, such as, eg. "This work is not X." If Y is based on X, then could Y further require the inclusion of a string "This work is not Y."? Is it reasonable to require that these messages be displayed on startup?</p><p>\nSurely it is enough to not say that a work is X or Y (i.e. avoid X and Y's names and trademarks, etc.), rather than having to specifically say that a work is not X or Y?</p>\ngpl3.licensecompat.p5.s1\njamesgnz\nlogin to agree\n0\n2986\n\n\n", "2871": "\n\n\n\n<p style="display: inline">Doesn't relying on local law in this way mean that if there are countries where trademark law is overly harsh in some way, the GPL will make them more so?</p><p>\nThe section is presumably intended to prevent misleading use of trademarks, etc. (such as implying association or endorsement where none exists, mischaracterising association or endorsement, or presenting the work as another work).</p><p>\nI think the section should state some limit on the extent that the GPL can be used to enforce trademarks, etc., rather than enforcing anything that local law enforces.</p><p>\nPerhaps simply that *misleading* use may be limited to the extent permitted by law.\n</p>\ngpl3.licensecompat.p6.s1\njamesgnz\nlogin to agree\n0\n2871\n\n\n", "2746:2871": "\n\n\n\ndoes "service marks" include certifications and labels, that may be used shown for publicity on specific versions conveyed by one distributor?\nIt seems that such certifications and quality labels are reasonably efficient only if they conver unmodified copies, and it seems reasonnable to include additional terms that require the removal of such certifications or labels that were only asserted on unmodified versions, but do not extend to unknown modified versions. such terms may be useful to allow users to check if the versions they receive are effectively covered by applicable quality labels, that have not been tampered by their immediate distributor. Without such inclusion of additional terms, certification and labels could be kept in modified versions by the distributor, who may lie on the effective quality, because such labels are not as easy to protect as digital signatures that authenticate the origin of the certified software (and requiring the use of verifiable digital signatures, and the purchase of digital certificates, may be too restrictive for licencees that want to distribute their modified versions).\nFor this reason, a licence should be allowed to include additional terms that require removing all references to the quality or certifications from all modified versions that will be distributed.\ngpl3.licensecompat.p6.s1\nverdyph\nlogin to agree\n0\n2746\n\n\n<p style="display: inline">Doesn't relying on local law in this way mean that if there are countries where trademark law is overly harsh in some way, the GPL will make them more so?</p><p>\nThe section is presumably intended to prevent misleading use of trademarks, etc. (such as implying association or endorsement where none exists, mischaracterising association or endorsement, or presenting the work as another work).</p><p>\nI think the section should state some limit on the extent that the GPL can be used to enforce trademarks, etc., rather than enforcing anything that local law enforces.</p><p>\nPerhaps simply that *misleading* use may be limited to the extent permitted by law.\n</p>\ngpl3.licensecompat.p6.s1\njamesgnz\nlogin to agree\n0\n2871\n\n\n", "2829:2852:3116": "\n\n\n\nDoes the 'additional permission' option give me, as a user of the license, the ability to add language similar Section 13 that says "notwithstanding anything else in this license, you may link with a work licensed under the Apache License?" Or does the incompatible downstream patent licensing restriction in the Apache license that is causing the current Apache license to be GPL-incompatible render such an 'additional permission' ineffective? I understand this 'additional permission' would not make other people's GPL'ed code Apache-compatible (unless they also added such a permission), but suppose I must link my own code with apache.org's code -- but I do not need to make the derived work compatible with other GPL'ed code. Obviously this is much less satisfactory than having Apache be a GPLv3-compatible license, but is it effective at all as a work-around until that is achieved?\ngpl3.licensecompat.p7.s1\nsamizdat\nlogin to agree\n0\n2829\n\n\nI believe the word "conditions" should be used here instead of "terms" for purposes of consistency with section 5.2.b and 12.7, where "conditions" is used and not "terms".\ngpl3.licensecompat.p7.s1\njkoenig\nlogin to agree\n0\n2852\n\n\n<p style="display: inline">I think the term "non-permissive" is hard to understand, especially to non-native speakers. "Restriction" is much clearer So how about rephrasing as<br><br></p><p>\n"Terms as detailed above are the only restrictions you may add to this License. All other "further restrictions" are disallowed (see section 10)."</p>\ngpl3.licensecompat.p7.s1\nmux2005\nlogin to agree\n0\n3116\n\n\n", "2579:2771:2853": "\n\n\n\nYou seem to be banning someone saying things like "I release this under GPL3, but I insist you give me credit." If such an addition may be removed, it is purposeless; you would be better to say that licences with added restrictions are incompatible with GPL3, or listing restrictions you will accept and not permit people to cancel. The way you have done it, you have effectively prevented people from easily deriving licences that maintain most of the good work you have done whilst deviating on some minor point. The hypothetical licence above cannot be issued (and enforced), and so someone wanting such an added clause would have to write out their entire licence from scratch. They might overlook something you have catered for, their software might become extremely popular, and then a Novel or a Microsoft might use the author's oversight to cause havoc. I pointed out to you in the previous draft that removable restrictions are worthless. Please take more care to rethink this matter this time.\ngpl3.licensecompat.p7.s2\nrhouse\nlogin to agree\n2\n2579\n\n\nI'm glad to see that this is explicitly stated: every attempt to license a work under the terms of GPLv3 with further restrictions is equivalent to licensing under the plain GPLv3. This is good, since there are unfortunately many people that license works in inconsistent manners (such as GPLv2 + additional restrictions); creating a rule that resolves this kind of inconsistency for the better is a good thing to do.\ngpl3.licensecompat.p7.s2\nfrx\nlogin to agree\n5\n2771\n\n\nI believe the word "condition" should be used here instead of "term" for purposes of consistency with section 5.2.b and 12.7, where "conditions" is used and not "terms".\ngpl3.licensecompat.p7.s2\njkoenig\nlogin to agree\n0\n2853\n\n\n", "2854": "\n\n\n\nI believe the word "conditions" should be used here instead of "terms" for purposes of consistency with section 5.2.b and 12.7, where "conditions" is used and not "terms".\ngpl3.licensecompat.p7.s3\njkoenig\nlogin to agree\n0\n2854\n\n\n", "2854:3116": "\n\n\n\nI believe the word "conditions" should be used here instead of "terms" for purposes of consistency with section 5.2.b and 12.7, where "conditions" is used and not "terms".\ngpl3.licensecompat.p7.s3\njkoenig\nlogin to agree\n0\n2854\n\n\n<p style="display: inline">I think the term "non-permissive" is hard to understand, especially to non-native speakers. "Restriction" is much clearer So how about rephrasing as<br><br></p><p>\n"Terms as detailed above are the only restrictions you may add to this License. All other "further restrictions" are disallowed (see section 10)."</p>\ngpl3.licensecompat.p7.s1\nmux2005\nlogin to agree\n0\n3116\n\n\n", "2855": "\n\n\n\nI believe the word "conditions" should be used here instead of "terms" for purposes of consistency with section 5.2.b and 12.7, where "conditions" is used and not "terms".\ngpl3.licensecompat.p8.s1\njkoenig\nlogin to agree\n0\n2855\n\n\n", "2809:3075": "\n\n\n\n<p style="display: inline">This section is too harsh as it stands. It is presumably intended to be applicable to situations such as small non-profit organisations enforcing the GPL against large companies that have intentionally seriously breached the GPL to pervert copyleft, but it could just as easily be used by large companies to pervert copyleft by intimidating or destroying small non-profit organisations that have unintentionally breached the GPL in some minor way. It could potentially enable rather than prevent the next SCO.</p><p>\nConsider the sorts of things that can be GPL breaches:</p><p>\nMEPIS maintainer Warren Woodford distributes source code for binary packages he has modified, but not for binary packages that he distributes verbatim. I am sure this was not malicious.</p><p>\nCases where previously unenforced trademarks start being enforced: Linux, RedHat, Mozilla (against Debian). And the DCC (formerly Debian Core Consortium) ran into problems for using the trademark Debian. There are many potential violations here that would not be malicious but could be breaches of the GPLv3. In order to correct these breaches, it may be necessary to change domain names, and redo artwork, etc.</p><p>\nConsider that Linux distributions often involve thousands of packages for several platforms, and sometimes backport fixes to stable packages. It is entirely possible that some author attribution or licensing clause could be accidentally lost. In order to correct these breaches, it may be necessary to permanently take down archives, and entirely cease distribution until a new release is ready.</p><p>\nTwo breaches are required to make GPLv3d3 termination permanent, but this could be two uses of a trademark in artwork, or two downloads of a binary file for which source is not immediately available, or for which attribution is slightly incorrect. Even temporary termination could cause serious damage over a minor breach.</p><p>\nEssentially this clause is the equivalent of the death penalty for parking infringements. In order to protect freedom, it is necessary to adhere to the ideas of "innocent until proven guilty", "first do no harm" and "the punishment fits the crime". The most important thing to keep in mind for this clause is to ensure that it can not be used to deny the freedom that it is supposed to protect.</p><p>\nIt is not (I think) difficult to imagine that a former Linux company bought out or paid off could use this clause to devastating effect. Think of SCO, but whereas it seems that SCO never had a real case, the termination clause might well provide the next SCO with a real case.</p><p>\nMost GPL breaches do not require such a harsh penalty. It would suffice in the majority of cases to simply sue (or threaten to do so) for the breaches themselves, with the license still in place, thus ensuring that the punishment fits the crime. Anyone who thinks this would not be enough ought to be pushing for harsher copyright laws, rather than seeking the harshest possible punishment (termination) for even the smallest possible offence.</p><p>\nThere is only one class of breaches for which it is necessary (or desirable) to charge in with guns blazing, that being when someone brings a lawsuit in breach of the GPL. This is the only type of breach which could cause immediate devastating damage, so it is the only instance in which it is necessary to be able to respond with immediate devastating damage.</p><p>\nBy a happy co-incidence, this is also the only kind of breach which we can reasonably assume to be malicious, and the only kind of breach which we can be reasonably sure there is a quick and easy way for the party in violation to fix (i.e. by dropping the lawsuit or refunding damages).</p><p>\nThat is to say, this is the only kind of breach for which it is necessary to be able to terminate rights in order to prevent the next SCO, and it is also the only kind of breach for which it is reasonably safe to allow termination without risking enabling the next SCO.</p><p>\nFurther, even in instances where it is desirable to bring in the big guns and punish an aggressor by preventing them from benefiting from GPL code, it is both unnecessary and undesirable to also prevent them from offering a public service. And it is perverse to prevent them from fulfilling their GPL obligations. To highlight how absurd the situation is at the moment: Someone can have their license terminated for not providing source code, then be sued for providing source code, or they could have their license terminated for distributing incorrectly attributed works, then be sued for making the corrections. Think about this for a minute: The GPL being used to prevent someone from distributing source code in accordance with, and as required by, the GPL--this is lunacy.</p><p>\nIn the case of a commercial Linux distributor that sues on the grounds of a patent in GPL code, it would make sense to terminate their right to sell their distribution, but not to terminate their right to provide source code for their distribution, or provide a public mirror of the distribution that theirs is based on. (They would have to remove any allegedly patent infringing feature from source code they provided anyway, since conveying GPL works that depend on a patent is disallowed, but they may as well be allowed to provide the source code /without/ that feature, allowing them to mostly fulfill their obligations, even while their right to commercially distribute the work is entirely terminated, with or without the feature.)</p><p>\nTerminating the right to commercial use of the GPL work is all that is needed to hit the aggressor where it hurts (in the pocket). In attempting to pervert copyleft, they are likely to be after a financial gain, so a financial disincentive is appropriate. If this doesn't work (say the aggressor is a software company that doesn't rely on the GPL work), then also terminating their right to offer a public service isn't going to help at all. If it is desirable to make the punishment any harsher than this, then it could be required that they distribute source code (rather than just an offer for it) with every non-commercially distributed binary, and/or that they must offer source code for covered works run on a public server (the AGPL condition that was seen as to restrictive for GPL use might be appropriate in this context).</p><p>\nFinally, the primary goal of the termination clause should be, first and foremost, to prevent damage to innocent parties, or if damage has already been done, to undo it, in so far as is possible. The primary goal should not be to destroy the aggressor, no matter how appealing this may sound. It would actually be better for everyone if they started playing nicely. To this end, it makes sense to allow immediate and automatic reinstatement of the terminated rights as soon as the GPL breaches are corrected, regardless of whether this is after 2 breaches, or a month, or whatever. If the aggressor is ever put in the position where they have no way of escaping termination, then there is no incentive for them to play nice. This will simply ensure that whatever damage they have done will become permanent. It may even force them into making a business model out of suing the users of GPL works, which is exactly what this section should be trying to avoid.</p>\ngpl3.termination.0.0\njamesgnz\nlogin to agree\n1\n2809\n\n\nSection 4 of the GPL2 covers this better, without time limits or let-off clauses. This is too complex and too mild to be realistic to enforce.\ngpl3.termination.0.0\njamesj\nlogin to agree\n0\n3075\n\n\n", "2772:2809:3075": "\n\n\n\nSection 8 seems to be more or less similar to section 4 of GPLv2, except that it's more forgiving to first-time violators. This is good. However, as in GPLv2, it fails to mention fair use and similar rights and possible separate licenses (as in disjunctive dual-licensing...). It should be improved to acknowledge these possibilities.\ngpl3.termination.0.0\nfrx\nlogin to agree\n0\n2772\n\n\n<p style="display: inline">This section is too harsh as it stands. It is presumably intended to be applicable to situations such as small non-profit organisations enforcing the GPL against large companies that have intentionally seriously breached the GPL to pervert copyleft, but it could just as easily be used by large companies to pervert copyleft by intimidating or destroying small non-profit organisations that have unintentionally breached the GPL in some minor way. It could potentially enable rather than prevent the next SCO.</p><p>\nConsider the sorts of things that can be GPL breaches:</p><p>\nMEPIS maintainer Warren Woodford distributes source code for binary packages he has modified, but not for binary packages that he distributes verbatim. I am sure this was not malicious.</p><p>\nCases where previously unenforced trademarks start being enforced: Linux, RedHat, Mozilla (against Debian). And the DCC (formerly Debian Core Consortium) ran into problems for using the trademark Debian. There are many potential violations here that would not be malicious but could be breaches of the GPLv3. In order to correct these breaches, it may be necessary to change domain names, and redo artwork, etc.</p><p>\nConsider that Linux distributions often involve thousands of packages for several platforms, and sometimes backport fixes to stable packages. It is entirely possible that some author attribution or licensing clause could be accidentally lost. In order to correct these breaches, it may be necessary to permanently take down archives, and entirely cease distribution until a new release is ready.</p><p>\nTwo breaches are required to make GPLv3d3 termination permanent, but this could be two uses of a trademark in artwork, or two downloads of a binary file for which source is not immediately available, or for which attribution is slightly incorrect. Even temporary termination could cause serious damage over a minor breach.</p><p>\nEssentially this clause is the equivalent of the death penalty for parking infringements. In order to protect freedom, it is necessary to adhere to the ideas of "innocent until proven guilty", "first do no harm" and "the punishment fits the crime". The most important thing to keep in mind for this clause is to ensure that it can not be used to deny the freedom that it is supposed to protect.</p><p>\nIt is not (I think) difficult to imagine that a former Linux company bought out or paid off could use this clause to devastating effect. Think of SCO, but whereas it seems that SCO never had a real case, the termination clause might well provide the next SCO with a real case.</p><p>\nMost GPL breaches do not require such a harsh penalty. It would suffice in the majority of cases to simply sue (or threaten to do so) for the breaches themselves, with the license still in place, thus ensuring that the punishment fits the crime. Anyone who thinks this would not be enough ought to be pushing for harsher copyright laws, rather than seeking the harshest possible punishment (termination) for even the smallest possible offence.</p><p>\nThere is only one class of breaches for which it is necessary (or desirable) to charge in with guns blazing, that being when someone brings a lawsuit in breach of the GPL. This is the only type of breach which could cause immediate devastating damage, so it is the only instance in which it is necessary to be able to respond with immediate devastating damage.</p><p>\nBy a happy co-incidence, this is also the only kind of breach which we can reasonably assume to be malicious, and the only kind of breach which we can be reasonably sure there is a quick and easy way for the party in violation to fix (i.e. by dropping the lawsuit or refunding damages).</p><p>\nThat is to say, this is the only kind of breach for which it is necessary to be able to terminate rights in order to prevent the next SCO, and it is also the only kind of breach for which it is reasonably safe to allow termination without risking enabling the next SCO.</p><p>\nFurther, even in instances where it is desirable to bring in the big guns and punish an aggressor by preventing them from benefiting from GPL code, it is both unnecessary and undesirable to also prevent them from offering a public service. And it is perverse to prevent them from fulfilling their GPL obligations. To highlight how absurd the situation is at the moment: Someone can have their license terminated for not providing source code, then be sued for providing source code, or they could have their license terminated for distributing incorrectly attributed works, then be sued for making the corrections. Think about this for a minute: The GPL being used to prevent someone from distributing source code in accordance with, and as required by, the GPL--this is lunacy.</p><p>\nIn the case of a commercial Linux distributor that sues on the grounds of a patent in GPL code, it would make sense to terminate their right to sell their distribution, but not to terminate their right to provide source code for their distribution, or provide a public mirror of the distribution that theirs is based on. (They would have to remove any allegedly patent infringing feature from source code they provided anyway, since conveying GPL works that depend on a patent is disallowed, but they may as well be allowed to provide the source code /without/ that feature, allowing them to mostly fulfill their obligations, even while their right to commercially distribute the work is entirely terminated, with or without the feature.)</p><p>\nTerminating the right to commercial use of the GPL work is all that is needed to hit the aggressor where it hurts (in the pocket). In attempting to pervert copyleft, they are likely to be after a financial gain, so a financial disincentive is appropriate. If this doesn't work (say the aggressor is a software company that doesn't rely on the GPL work), then also terminating their right to offer a public service isn't going to help at all. If it is desirable to make the punishment any harsher than this, then it could be required that they distribute source code (rather than just an offer for it) with every non-commercially distributed binary, and/or that they must offer source code for covered works run on a public server (the AGPL condition that was seen as to restrictive for GPL use might be appropriate in this context).</p><p>\nFinally, the primary goal of the termination clause should be, first and foremost, to prevent damage to innocent parties, or if damage has already been done, to undo it, in so far as is possible. The primary goal should not be to destroy the aggressor, no matter how appealing this may sound. It would actually be better for everyone if they started playing nicely. To this end, it makes sense to allow immediate and automatic reinstatement of the terminated rights as soon as the GPL breaches are corrected, regardless of whether this is after 2 breaches, or a month, or whatever. If the aggressor is ever put in the position where they have no way of escaping termination, then there is no incentive for them to play nice. This will simply ensure that whatever damage they have done will become permanent. It may even force them into making a business model out of suing the users of GPL works, which is exactly what this section should be trying to avoid.</p>\ngpl3.termination.0.0\njamesgnz\nlogin to agree\n1\n2809\n\n\nSection 4 of the GPL2 covers this better, without time limits or let-off clauses. This is too complex and too mild to be realistic to enforce.\ngpl3.termination.0.0\njamesj\nlogin to agree\n0\n3075\n\n\n", "2485": "\n\n\n\n"an attempt" cannot be void.\nit can invalidate the license, rendering it void.\n\ngpl3.termination.p0.s2\nsds57\nlogin to agree\n15\n2485\n\n\n", "2485:2989": "\n\n\n\n"an attempt" cannot be void.\nit can invalidate the license, rendering it void.\n\ngpl3.termination.p0.s2\nsds57\nlogin to agree\n15\n2485\n\n\ns/propagate or modify it is void/propagate or modify the covered work voids your License/\ngpl3.termination.p0.s2\nmole\nlogin to agree\n1\n2989\n\n\n", "3005": "\n\n\n\nTermination should be automatic, no notice required. I have read Brett Smith's explanation of this non-automatic termination clause on Groklaw [http://www.groklaw.net/article.php?story=20070501092619462] \nand I don't buy it. Termination of the license should be automatic on violation - compliance with the license is the only thing that gives you the right to convey a covered work. Fixing the problem is straightforward - convey your modifications to the source or stop conveying. There are simply too many individual copyright holders in various GPL licensed code bases with too little ability for those individuals to examine all proprietary code for violations of their copyright. Providing notice and negotiating may work for a few large corporate entities that wish to use the GPL, and for the FSF for the GNU code base, but it this scheme makes no sense for the individual copyright holder who wishes to use the GPL for their code. I strongly urge a much simpler scheme of automatic termination and automatic reinsatement. Thus delete the two sentences "If you violate this license, any copyright .... granted to you under this Licence."\ngpl3.termination.p0.s3\nmole\nlogin to agree\n1\n3005\n\n\n", "2835:3005": "\n\n\n\nIs this term vague enough to permit a shadowy operation to claim that they have not been notified up until the point where a court affirms a judgement of violation? \n\nSuggestion: Possibly make this less ambiguous what serves as notification.\ngpl3.termination.p0.s3\ndmichaud\nlogin to agree\n0\n2835\n\n\nTermination should be automatic, no notice required. I have read Brett Smith's explanation of this non-automatic termination clause on Groklaw [http://www.groklaw.net/article.php?story=20070501092619462] \nand I don't buy it. Termination of the license should be automatic on violation - compliance with the license is the only thing that gives you the right to convey a covered work. Fixing the problem is straightforward - convey your modifications to the source or stop conveying. There are simply too many individual copyright holders in various GPL licensed code bases with too little ability for those individuals to examine all proprietary code for violations of their copyright. Providing notice and negotiating may work for a few large corporate entities that wish to use the GPL, and for the FSF for the GNU code base, but it this scheme makes no sense for the individual copyright holder who wishes to use the GPL for their code. I strongly urge a much simpler scheme of automatic termination and automatic reinsatement. Thus delete the two sentences "If you violate this license, any copyright .... granted to you under this Licence."\ngpl3.termination.p0.s3\nmole\nlogin to agree\n1\n3005\n\n\n", "2450:2835:3005": "\n\n\n\n<p style="display: inline">Please drop this section of the sentence. The 60 day rule is a "get out of jail free" card for unscrupulous interests who violate the GPL in the hope that no-one will immediately notice. What if court proceeding take more than 60 days? By the time the verdict is given out the guilty party won't be in violation of the license anymore and therefore the judge will be forced to rule in the violators' favor.</p><p>\nI've watched http://GPL-Violations.org for some time and realize that it takes months for a GPL violation to be noticed, months before developers who care are contacted, and months before those developers can hire a lawyer and mount a realistic lawsuit. Not to mention the exorbitant legal costs involved in asserting the FOSS developer's copyright.</p><p>\nAsk yourselves: would the RIAA/MPAA refrain from suing people if they downloaded a song/movie more than 60 days ago? Yeah right! So why should FOSS developers be handcuffed in this way?</p><p>\nFinally, judges are well prepared to decide whether to prosecute "old" violations, where "old" is defined by the jurisdiction's copyright laws and statute of limitations.</p><p>\nAgain, please don't shortchange GPL enforcers. If the GPL is unenforceable, all of other provisions are worthless!</p><p>\nRemoving this part of the sentence would give GPL enforcers however long they need to notify the infringer, and it will make GPL enforcement easier. As a developer who releases code under the GPL, I'd hate it if my own license started imposing deadlines on me!</p>\ngpl3.termination.p0.s3\nvladc\nlogin to agree\n23\n2450\n\n\nIs this term vague enough to permit a shadowy operation to claim that they have not been notified up until the point where a court affirms a judgement of violation? \n\nSuggestion: Possibly make this less ambiguous what serves as notification.\ngpl3.termination.p0.s3\ndmichaud\nlogin to agree\n0\n2835\n\n\nTermination should be automatic, no notice required. I have read Brett Smith's explanation of this non-automatic termination clause on Groklaw [http://www.groklaw.net/article.php?story=20070501092619462] \nand I don't buy it. Termination of the license should be automatic on violation - compliance with the license is the only thing that gives you the right to convey a covered work. Fixing the problem is straightforward - convey your modifications to the source or stop conveying. There are simply too many individual copyright holders in various GPL licensed code bases with too little ability for those individuals to examine all proprietary code for violations of their copyright. Providing notice and negotiating may work for a few large corporate entities that wish to use the GPL, and for the FSF for the GNU code base, but it this scheme makes no sense for the individual copyright holder who wishes to use the GPL for their code. I strongly urge a much simpler scheme of automatic termination and automatic reinsatement. Thus delete the two sentences "If you violate this license, any copyright .... granted to you under this Licence."\ngpl3.termination.p0.s3\nmole\nlogin to agree\n1\n3005\n\n\n", "2450:2542:2560:2834:2856:3005": "\n\n\n\n<p style="display: inline">Please drop this section of the sentence. The 60 day rule is a "get out of jail free" card for unscrupulous interests who violate the GPL in the hope that no-one will immediately notice. What if court proceeding take more than 60 days? By the time the verdict is given out the guilty party won't be in violation of the license anymore and therefore the judge will be forced to rule in the violators' favor.</p><p>\nI've watched http://GPL-Violations.org for some time and realize that it takes months for a GPL violation to be noticed, months before developers who care are contacted, and months before those developers can hire a lawyer and mount a realistic lawsuit. Not to mention the exorbitant legal costs involved in asserting the FOSS developer's copyright.</p><p>\nAsk yourselves: would the RIAA/MPAA refrain from suing people if they downloaded a song/movie more than 60 days ago? Yeah right! So why should FOSS developers be handcuffed in this way?</p><p>\nFinally, judges are well prepared to decide whether to prosecute "old" violations, where "old" is defined by the jurisdiction's copyright laws and statute of limitations.</p><p>\nAgain, please don't shortchange GPL enforcers. If the GPL is unenforceable, all of other provisions are worthless!</p><p>\nRemoving this part of the sentence would give GPL enforcers however long they need to notify the infringer, and it will make GPL enforcement easier. As a developer who releases code under the GPL, I'd hate it if my own license started imposing deadlines on me!</p>\ngpl3.termination.p0.s3\nvladc\nlogin to agree\n23\n2450\n\n\n60 days is an unreasonably short timeframe in the case of violations where it cannot easily be verified whether the violations continue or not.\ngpl3.termination.p0.s3\nnb\nlogin to agree\n8\n2542\n\n\nYou mean "since the most recent violation was cured".\ngpl3.termination.p0.s3\nnovalis\nlogin to agree\n1\n2560\n\n\n<p style="display: inline">I am wondering if this well-intended clause can be attacked by a planted insider. For example:</p><p>\n1. EvilCo pays their employee Phil to trivially contribute to GPL project GnuAge, thus becoming a copyright holder.</p><p>\n2. EvilCo violates the GPL in a way that is difficult to detect. \n\n3. Evilco tells Phil to quietly put the company on notice for the violation, and then forgive the company.</p><p>\n4. After 60 days, Evilco ramps up their GPL violation of GnuAge, and when the legitimate copyright owners notice and want to take action, they are prevented because the 60 day window has expired.</p><p>\nSuggestion: Allow any copyright holder in the violated work to put the violator on notice at any time, without prejudice.</p>\ngpl3.termination.p0.s3\ndmichaud\nlogin to agree\n0\n2834\n\n\nThis statement is incomplete in that the alternative is not stated. What happens if 60 days has passed?\ngpl3.termination.p0.s3\njkoenig\nlogin to agree\n0\n2856\n\n\nTermination should be automatic, no notice required. I have read Brett Smith's explanation of this non-automatic termination clause on Groklaw [http://www.groklaw.net/article.php?story=20070501092619462] \nand I don't buy it. Termination of the license should be automatic on violation - compliance with the license is the only thing that gives you the right to convey a covered work. Fixing the problem is straightforward - convey your modifications to the source or stop conveying. There are simply too many individual copyright holders in various GPL licensed code bases with too little ability for those individuals to examine all proprietary code for violations of their copyright. Providing notice and negotiating may work for a few large corporate entities that wish to use the GPL, and for the FSF for the GNU code base, but it this scheme makes no sense for the individual copyright holder who wishes to use the GPL for their code. I strongly urge a much simpler scheme of automatic termination and automatic reinsatement. Thus delete the two sentences "If you violate this license, any copyright .... granted to you under this Licence."\ngpl3.termination.p0.s3\nmole\nlogin to agree\n1\n3005\n\n\n", "2450:2542:2560:2834:3005": "\n\n\n\n<p style="display: inline">Please drop this section of the sentence. The 60 day rule is a "get out of jail free" card for unscrupulous interests who violate the GPL in the hope that no-one will immediately notice. What if court proceeding take more than 60 days? By the time the verdict is given out the guilty party won't be in violation of the license anymore and therefore the judge will be forced to rule in the violators' favor.</p><p>\nI've watched http://GPL-Violations.org for some time and realize that it takes months for a GPL violation to be noticed, months before developers who care are contacted, and months before those developers can hire a lawyer and mount a realistic lawsuit. Not to mention the exorbitant legal costs involved in asserting the FOSS developer's copyright.</p><p>\nAsk yourselves: would the RIAA/MPAA refrain from suing people if they downloaded a song/movie more than 60 days ago? Yeah right! So why should FOSS developers be handcuffed in this way?</p><p>\nFinally, judges are well prepared to decide whether to prosecute "old" violations, where "old" is defined by the jurisdiction's copyright laws and statute of limitations.</p><p>\nAgain, please don't shortchange GPL enforcers. If the GPL is unenforceable, all of other provisions are worthless!</p><p>\nRemoving this part of the sentence would give GPL enforcers however long they need to notify the infringer, and it will make GPL enforcement easier. As a developer who releases code under the GPL, I'd hate it if my own license started imposing deadlines on me!</p>\ngpl3.termination.p0.s3\nvladc\nlogin to agree\n23\n2450\n\n\n60 days is an unreasonably short timeframe in the case of violations where it cannot easily be verified whether the violations continue or not.\ngpl3.termination.p0.s3\nnb\nlogin to agree\n8\n2542\n\n\nYou mean "since the most recent violation was cured".\ngpl3.termination.p0.s3\nnovalis\nlogin to agree\n1\n2560\n\n\n<p style="display: inline">I am wondering if this well-intended clause can be attacked by a planted insider. For example:</p><p>\n1. EvilCo pays their employee Phil to trivially contribute to GPL project GnuAge, thus becoming a copyright holder.</p><p>\n2. EvilCo violates the GPL in a way that is difficult to detect. \n\n3. Evilco tells Phil to quietly put the company on notice for the violation, and then forgive the company.</p><p>\n4. After 60 days, Evilco ramps up their GPL violation of GnuAge, and when the legitimate copyright owners notice and want to take action, they are prevented because the 60 day window has expired.</p><p>\nSuggestion: Allow any copyright holder in the violated work to put the violator on notice at any time, without prejudice.</p>\ngpl3.termination.p0.s3\ndmichaud\nlogin to agree\n0\n2834\n\n\nTermination should be automatic, no notice required. I have read Brett Smith's explanation of this non-automatic termination clause on Groklaw [http://www.groklaw.net/article.php?story=20070501092619462] \nand I don't buy it. Termination of the license should be automatic on violation - compliance with the license is the only thing that gives you the right to convey a covered work. Fixing the problem is straightforward - convey your modifications to the source or stop conveying. There are simply too many individual copyright holders in various GPL licensed code bases with too little ability for those individuals to examine all proprietary code for violations of their copyright. Providing notice and negotiating may work for a few large corporate entities that wish to use the GPL, and for the FSF for the GNU code base, but it this scheme makes no sense for the individual copyright holder who wishes to use the GPL for their code. I strongly urge a much simpler scheme of automatic termination and automatic reinsatement. Thus delete the two sentences "If you violate this license, any copyright .... granted to you under this Licence."\ngpl3.termination.p0.s3\nmole\nlogin to agree\n1\n3005\n\n\n", "2545": "\n\n\n\nThe violator needs to be given the opportunity to cure the violation before termination. Absent this, large businesses will not use GPL V3 software in any business-critical application. Large companies have many employees, and they do sometimes make mistakes. I'd drop the 60 day provision (above) and the 30 day (below) and replace with a time limit after notice of violation to cure the violation. The possible cures probably need to be stated somewhere as well - releasing code, or re-writing without GPL components.\ngpl3.termination.p0.s4\nmbreuer\nlogin to agree\n1\n2545\n\n\n", "2545:2560:2834:2835:2856:2989:3005": "\n\n\n\nThe violator needs to be given the opportunity to cure the violation before termination. Absent this, large businesses will not use GPL V3 software in any business-critical application. Large companies have many employees, and they do sometimes make mistakes. I'd drop the 60 day provision (above) and the 30 day (below) and replace with a time limit after notice of violation to cure the violation. The possible cures probably need to be stated somewhere as well - releasing code, or re-writing without GPL components.\ngpl3.termination.p0.s4\nmbreuer\nlogin to agree\n1\n2545\n\n\nYou mean "since the most recent violation was cured".\ngpl3.termination.p0.s3\nnovalis\nlogin to agree\n1\n2560\n\n\n<p style="display: inline">I am wondering if this well-intended clause can be attacked by a planted insider. For example:</p><p>\n1. EvilCo pays their employee Phil to trivially contribute to GPL project GnuAge, thus becoming a copyright holder.</p><p>\n2. EvilCo violates the GPL in a way that is difficult to detect. \n\n3. Evilco tells Phil to quietly put the company on notice for the violation, and then forgive the company.</p><p>\n4. After 60 days, Evilco ramps up their GPL violation of GnuAge, and when the legitimate copyright owners notice and want to take action, they are prevented because the 60 day window has expired.</p><p>\nSuggestion: Allow any copyright holder in the violated work to put the violator on notice at any time, without prejudice.</p>\ngpl3.termination.p0.s3\ndmichaud\nlogin to agree\n0\n2834\n\n\nIs this term vague enough to permit a shadowy operation to claim that they have not been notified up until the point where a court affirms a judgement of violation? \n\nSuggestion: Possibly make this less ambiguous what serves as notification.\ngpl3.termination.p0.s3\ndmichaud\nlogin to agree\n0\n2835\n\n\nThis statement is incomplete in that the alternative is not stated. What happens if 60 days has passed?\ngpl3.termination.p0.s3\njkoenig\nlogin to agree\n0\n2856\n\n\ns/propagate or modify it is void/propagate or modify the covered work voids your License/\ngpl3.termination.p0.s2\nmole\nlogin to agree\n1\n2989\n\n\nTermination should be automatic, no notice required. I have read Brett Smith's explanation of this non-automatic termination clause on Groklaw [http://www.groklaw.net/article.php?story=20070501092619462] \nand I don't buy it. Termination of the license should be automatic on violation - compliance with the license is the only thing that gives you the right to convey a covered work. Fixing the problem is straightforward - convey your modifications to the source or stop conveying. There are simply too many individual copyright holders in various GPL licensed code bases with too little ability for those individuals to examine all proprietary code for violations of their copyright. Providing notice and negotiating may work for a few large corporate entities that wish to use the GPL, and for the FSF for the GNU code base, but it this scheme makes no sense for the individual copyright holder who wishes to use the GPL for their code. I strongly urge a much simpler scheme of automatic termination and automatic reinsatement. Thus delete the two sentences "If you violate this license, any copyright .... granted to you under this Licence."\ngpl3.termination.p0.s3\nmole\nlogin to agree\n1\n3005\n\n\n", "2466:3004": "\n\n\n\n<p style="display: inline">Please drop this whole paragraph. Don\u00c3\u00a2\u00c2\u0080\u00c2\u0099t give first-time offenders special privileges, because it is impossible to determine whether or not a legal entity is a first-time offender. If a Microsoft employee violated the GPL, does that violation count toward his employer or towards the individual? Imagine Microsoft taking advantage of this loophole by making each of their employees take turns violating the GPL. Or Microsoft might decide to pay non-employees for violating the GPL on their behalf.</p><p>\nAnd it\u00c3\u00a2\u00c2\u0080\u00c2\u0099s not clear what counts as an official violation. Is it being accused of violating? Is it being sued by the copyright holder? Is it losing a court case? How about settling out of court?</p><p>\nAll this paragraph does is suck up to powerful infringers and makes enforcing the GPL harder for the copyright holders. Please remove it!</p>\ngpl3.termination.p1.s1\nvladc\nlogin to agree\n9\n2466\n\n\n<p style="display: inline">How does reinstating work otherwise? What if the copyright holder is dead? It makes much more sense to me for there to be a single simple test - if you are in compliance with the terms of the license then you may propagate and convey, if you aren't in compliance then you can't. \n\nAn underlying goal of the GPL is ensuring that source code can't be locked up into a modified proprietary fork - by ensuring that we can build higher by standing on each others shoulders. Illicit distribution of a modified fork of gpl code without the source should be able to be easily cured by simply following the basic contract of the GPL - convey your modifications to the source and you are in compliance. \n\nI would suggest replacing this paragraph with a much simpler one: "If you are in violation of this license for a particular covered work, your license is automatically reinstated by your compliance with the terms of this license."</p><p>\nLikewise, remove the complex 60 day notification clause with an automatic termination clause - the only thing that grants you rights to propagate or convey a covered work is compliance with the license - violate the license and your rights to propagate and convey automatically terminate. </p>\ngpl3.termination.p1.s1\nmole\nlogin to agree\n0\n3004\n\n\n", "2466:2519:3004": "\n\n\n\n<p style="display: inline">Please drop this whole paragraph. Don\u00c3\u00a2\u00c2\u0080\u00c2\u0099t give first-time offenders special privileges, because it is impossible to determine whether or not a legal entity is a first-time offender. If a Microsoft employee violated the GPL, does that violation count toward his employer or towards the individual? Imagine Microsoft taking advantage of this loophole by making each of their employees take turns violating the GPL. Or Microsoft might decide to pay non-employees for violating the GPL on their behalf.</p><p>\nAnd it\u00c3\u00a2\u00c2\u0080\u00c2\u0099s not clear what counts as an official violation. Is it being accused of violating? Is it being sued by the copyright holder? Is it losing a court case? How about settling out of court?</p><p>\nAll this paragraph does is suck up to powerful infringers and makes enforcing the GPL harder for the copyright holders. Please remove it!</p>\ngpl3.termination.p1.s1\nvladc\nlogin to agree\n9\n2466\n\n\nMake it the first violation of the license, period. Otherwise it could be systematically abused against one copyright holder after another.\ngpl3.termination.p1.s1\njring\nlogin to agree\n2\n2519\n\n\n<p style="display: inline">How does reinstating work otherwise? What if the copyright holder is dead? It makes much more sense to me for there to be a single simple test - if you are in compliance with the terms of the license then you may propagate and convey, if you aren't in compliance then you can't. \n\nAn underlying goal of the GPL is ensuring that source code can't be locked up into a modified proprietary fork - by ensuring that we can build higher by standing on each others shoulders. Illicit distribution of a modified fork of gpl code without the source should be able to be easily cured by simply following the basic contract of the GPL - convey your modifications to the source and you are in compliance. \n\nI would suggest replacing this paragraph with a much simpler one: "If you are in violation of this license for a particular covered work, your license is automatically reinstated by your compliance with the terms of this license."</p><p>\nLikewise, remove the complex 60 day notification clause with an automatic termination clause - the only thing that grants you rights to propagate or convey a covered work is compliance with the license - violate the license and your rights to propagate and convey automatically terminate. </p>\ngpl3.termination.p1.s1\nmole\nlogin to agree\n0\n3004\n\n\n", "2466:2857:3004": "\n\n\n\n<p style="display: inline">Please drop this whole paragraph. Don\u00c3\u00a2\u00c2\u0080\u00c2\u0099t give first-time offenders special privileges, because it is impossible to determine whether or not a legal entity is a first-time offender. If a Microsoft employee violated the GPL, does that violation count toward his employer or towards the individual? Imagine Microsoft taking advantage of this loophole by making each of their employees take turns violating the GPL. Or Microsoft might decide to pay non-employees for violating the GPL on their behalf.</p><p>\nAnd it\u00c3\u00a2\u00c2\u0080\u00c2\u0099s not clear what counts as an official violation. Is it being accused of violating? Is it being sued by the copyright holder? Is it losing a court case? How about settling out of court?</p><p>\nAll this paragraph does is suck up to powerful infringers and makes enforcing the GPL harder for the copyright holders. Please remove it!</p>\ngpl3.termination.p1.s1\nvladc\nlogin to agree\n9\n2466\n\n\nThere is no guidance here as to what constitutes a "cure". I think that makes the paragraph useless.\ngpl3.termination.p1.s1\njkoenig\nlogin to agree\n0\n2857\n\n\n<p style="display: inline">How does reinstating work otherwise? What if the copyright holder is dead? It makes much more sense to me for there to be a single simple test - if you are in compliance with the terms of the license then you may propagate and convey, if you aren't in compliance then you can't. \n\nAn underlying goal of the GPL is ensuring that source code can't be locked up into a modified proprietary fork - by ensuring that we can build higher by standing on each others shoulders. Illicit distribution of a modified fork of gpl code without the source should be able to be easily cured by simply following the basic contract of the GPL - convey your modifications to the source and you are in compliance. \n\nI would suggest replacing this paragraph with a much simpler one: "If you are in violation of this license for a particular covered work, your license is automatically reinstated by your compliance with the terms of this license."</p><p>\nLikewise, remove the complex 60 day notification clause with an automatic termination clause - the only thing that grants you rights to propagate or convey a covered work is compliance with the license - violate the license and your rights to propagate and convey automatically terminate. </p>\ngpl3.termination.p1.s1\nmole\nlogin to agree\n0\n3004\n\n\n", "2466:2530:3004": "\n\n\n\n<p style="display: inline">Please drop this whole paragraph. Don\u00c3\u00a2\u00c2\u0080\u00c2\u0099t give first-time offenders special privileges, because it is impossible to determine whether or not a legal entity is a first-time offender. If a Microsoft employee violated the GPL, does that violation count toward his employer or towards the individual? Imagine Microsoft taking advantage of this loophole by making each of their employees take turns violating the GPL. Or Microsoft might decide to pay non-employees for violating the GPL on their behalf.</p><p>\nAnd it\u00c3\u00a2\u00c2\u0080\u00c2\u0099s not clear what counts as an official violation. Is it being accused of violating? Is it being sued by the copyright holder? Is it losing a court case? How about settling out of court?</p><p>\nAll this paragraph does is suck up to powerful infringers and makes enforcing the GPL harder for the copyright holders. Please remove it!</p>\ngpl3.termination.p1.s1\nvladc\nlogin to agree\n9\n2466\n\n\nMake it clear that only the license from that particular copyright holder is reinstated. If there are other holders with respect to which this is not the first violation, the license from them may remain revoked.\ngpl3.termination.p1.s1\nblinken\nlogin to agree\n3\n2530\n\n\n<p style="display: inline">How does reinstating work otherwise? What if the copyright holder is dead? It makes much more sense to me for there to be a single simple test - if you are in compliance with the terms of the license then you may propagate and convey, if you aren't in compliance then you can't. \n\nAn underlying goal of the GPL is ensuring that source code can't be locked up into a modified proprietary fork - by ensuring that we can build higher by standing on each others shoulders. Illicit distribution of a modified fork of gpl code without the source should be able to be easily cured by simply following the basic contract of the GPL - convey your modifications to the source and you are in compliance. \n\nI would suggest replacing this paragraph with a much simpler one: "If you are in violation of this license for a particular covered work, your license is automatically reinstated by your compliance with the terms of this license."</p><p>\nLikewise, remove the complex 60 day notification clause with an automatic termination clause - the only thing that grants you rights to propagate or convey a covered work is compliance with the license - violate the license and your rights to propagate and convey automatically terminate. </p>\ngpl3.termination.p1.s1\nmole\nlogin to agree\n0\n3004\n\n\n", "2466:2519:2530:2857:3004": "\n\n\n\n<p style="display: inline">Please drop this whole paragraph. Don\u00c3\u00a2\u00c2\u0080\u00c2\u0099t give first-time offenders special privileges, because it is impossible to determine whether or not a legal entity is a first-time offender. If a Microsoft employee violated the GPL, does that violation count toward his employer or towards the individual? Imagine Microsoft taking advantage of this loophole by making each of their employees take turns violating the GPL. Or Microsoft might decide to pay non-employees for violating the GPL on their behalf.</p><p>\nAnd it\u00c3\u00a2\u00c2\u0080\u00c2\u0099s not clear what counts as an official violation. Is it being accused of violating? Is it being sued by the copyright holder? Is it losing a court case? How about settling out of court?</p><p>\nAll this paragraph does is suck up to powerful infringers and makes enforcing the GPL harder for the copyright holders. Please remove it!</p>\ngpl3.termination.p1.s1\nvladc\nlogin to agree\n9\n2466\n\n\nMake it the first violation of the license, period. Otherwise it could be systematically abused against one copyright holder after another.\ngpl3.termination.p1.s1\njring\nlogin to agree\n2\n2519\n\n\nMake it clear that only the license from that particular copyright holder is reinstated. If there are other holders with respect to which this is not the first violation, the license from them may remain revoked.\ngpl3.termination.p1.s1\nblinken\nlogin to agree\n3\n2530\n\n\nThere is no guidance here as to what constitutes a "cure". I think that makes the paragraph useless.\ngpl3.termination.p1.s1\njkoenig\nlogin to agree\n0\n2857\n\n\n<p style="display: inline">How does reinstating work otherwise? What if the copyright holder is dead? It makes much more sense to me for there to be a single simple test - if you are in compliance with the terms of the license then you may propagate and convey, if you aren't in compliance then you can't. \n\nAn underlying goal of the GPL is ensuring that source code can't be locked up into a modified proprietary fork - by ensuring that we can build higher by standing on each others shoulders. Illicit distribution of a modified fork of gpl code without the source should be able to be easily cured by simply following the basic contract of the GPL - convey your modifications to the source and you are in compliance. \n\nI would suggest replacing this paragraph with a much simpler one: "If you are in violation of this license for a particular covered work, your license is automatically reinstated by your compliance with the terms of this license."</p><p>\nLikewise, remove the complex 60 day notification clause with an automatic termination clause - the only thing that grants you rights to propagate or convey a covered work is compliance with the license - violate the license and your rights to propagate and convey automatically terminate. </p>\ngpl3.termination.p1.s1\nmole\nlogin to agree\n0\n3004\n\n\n", "2990": "\n\n\n\nDoes this say "In the event your rights are terminated, parties who have received copies, or rights, from you under this License will have their licenses terminated unless they themselves correct their inherited non-compliance of this license" or does it say something else?\ngpl3.termination.p2.s1\ntinjon\nlogin to agree\n0\n2990\n\n\n", "2520:2990": "\n\n\n\nWhat if the original violation makes it impossible to remain in compliance? As could happen if the full source was not disclosed? Think binary blobs in the Linux kernel etc ... \ngpl3.termination.p2.s1\njring\nlogin to agree\n0\n2520\n\n\nDoes this say "In the event your rights are terminated, parties who have received copies, or rights, from you under this License will have their licenses terminated unless they themselves correct their inherited non-compliance of this license" or does it say something else?\ngpl3.termination.p2.s1\ntinjon\nlogin to agree\n0\n2990\n\n\n", "2739:2773": "\n\n\n\nThis seems like a crucial idea which is often misunderstood. Should this be moved into the Preamble, or at least include this idea into the Preamble somehow?\ngpl3.notacontract.0.0\nmkorman\nlogin to agree\n2\n2739\n\n\nAnother section that fails to mention fair use and similar rights and possible separate licenses. Section 9 should be improved by acknowledging these possibilities.\ngpl3.notacontract.0.0\nfrx\nlogin to agree\n1\n2773\n\n\n", "3122": "\n\n\n\n<p style="display: inline">As a non-lawyer I'm totally dumbfounded by this sentence. I always thought that copyright law requires a license to run a program. Otherwise, why would it be illegal for me to run a copy of Windows that's installed on a used computer I buy (assuming that the original media and license is not included in the purchase)? And AFAIK authorities are allowed to confiscate illegal copies of software. And now you're telling me that copyright law does not require a license for possession or running of software? Or what are you trying to say? That copyright law does require a license to possess and run software but that the GPL is not that license? What then gives me the right to run GPLed software?<br><br></p><p>\nWhat I'm trying to say: Why make it so complicated? What's the purpose of disclaiming the right to govern execution of software? Why not just say: "The GPL gives you the unlimited right to possess and run this software." That's reassuring, that's simple. And if copyright law doesn't require a license for running (have you checked that this is true everywhere in the world?) it doesn't hurt.</p>\ngpl3.notacontract.p0.s1\nmux2005\nlogin to agree\n1\n3122\n\n\n", "2514:3122": "\n\n\n\nDo require that you need to accept the disclaimer of liability in order to run the program or section 9 could cause trouble for developers in some countries. \ngpl3.notacontract.p0.s1\njring\nlogin to agree\n4\n2514\n\n\n<p style="display: inline">As a non-lawyer I'm totally dumbfounded by this sentence. I always thought that copyright law requires a license to run a program. Otherwise, why would it be illegal for me to run a copy of Windows that's installed on a used computer I buy (assuming that the original media and license is not included in the purchase)? And AFAIK authorities are allowed to confiscate illegal copies of software. And now you're telling me that copyright law does not require a license for possession or running of software? Or what are you trying to say? That copyright law does require a license to possess and run software but that the GPL is not that license? What then gives me the right to run GPLed software?<br><br></p><p>\nWhat I'm trying to say: Why make it so complicated? What's the purpose of disclaiming the right to govern execution of software? Why not just say: "The GPL gives you the unlimited right to possess and run this software." That's reassuring, that's simple. And if copyright law doesn't require a license for running (have you checked that this is true everywhere in the world?) it doesn't hurt.</p>\ngpl3.notacontract.p0.s1\nmux2005\nlogin to agree\n1\n3122\n\n\n", "3092:3149": "\n\n\n\n<p style="display: inline">This clause is a really lumpy addition to the not-a-contract section. Section 4 (conveying verbatim copies) would be a better location.</p><p>\nIn fact, why do we need anything at all about peer-to-peer, since "ancillary propagation by peer-to-peer" is always going to be verbatim copying, which is already allowed by section 4?\n</p>\ngpl3.notacontract.p0.s2\nsepreece\nlogin to agree\n0\n3092\n\n\nThis sentence should be moved to the section about peer-to-peer copying in section 6. It's confusing where it is now, and distributing copies of the software, even if it's "ancillary propagation", is different from having and using a copy.\ngpl3.notacontract.p0.s2\njohnston\nlogin to agree\n0\n3149\n\n\n", "3090": "\n\n\n\nSome parts of the GPL explicitly allow separate licenses to confer permissions to propagate or modify covered works. Shouldn't this section acknowledge that? \ngpl3.notacontract.p0.s3\ntfelker\nlogin to agree\n0\n3090\n\n\n", "2859": "\n\n\n\nReplace "other than" with "except".\ngpl3.notacontract.p0.s3\njkoenig\nlogin to agree\n1\n2859\n\n\n", "2858": "\n\n\n\n<p style="display: inline">I think it would be more accurate to say:</p><p>\n"to otherwise propagate"</p>\ngpl3.notacontract.p0.s3\njkoenig\nlogin to agree\n0\n2858\n\n\n", "2549": "\n\n\n\nWhy must a person accept the GPL to *modify* the code for their own use? Doesn't a person have that right regardless of this license? I suggest removing the words "or modify" if they are unnecessary.\ngpl3.notacontract.p0.s3\nccady\nlogin to agree\n2\n2549\n\n\n", "3123": "\n\n\n\nBecause "These actions infringe copyright if you do not accept this License" this whole section is pointless and just serves to confuse people. Just like the corresponding "You are not required to accept this License, since you have not signed it." language in the GPL 2 this serves no useful legal purpose that I can see. If you want this as background information, then put it into the Preamble.\ngpl3.notacontract.p0.s4\nmux2005\nlogin to agree\n0\n3123\n\n\n", "2860:2991": "\n\n\n\nI would replace this sentence with: \n\n"By modifying or propagating a covered work, you accept this License."\ngpl3.notacontract.p0.s5\njkoenig\nlogin to agree\n0\n2860\n\n\n(This comment reflects this and the previous two paragraphs together)\nHow does this related to for example bittorrent, where peers act as distributors (making copies) at the same time they perform the act of acquiring the software? It seems to me that bittorrent and this license are not compatible unless the peer explicitly accepts the license before starting to download the software. Accepting the license implicitly (as described in section 10) as part of starting to download the software via the torrent would put the peer automatically in violation unless the software was packaged in the proper way (with the license included etc). This could open up for patent lawsuits and other nasty actions. I think this section may need another pass over.\ngpl3.notacontract.p0.s5\ntinjon\nlogin to agree\n0\n2991\n\n\n", "3123:3149": "\n\n\n\nBecause "These actions infringe copyright if you do not accept this License" this whole section is pointless and just serves to confuse people. Just like the corresponding "You are not required to accept this License, since you have not signed it." language in the GPL 2 this serves no useful legal purpose that I can see. If you want this as background information, then put it into the Preamble.\ngpl3.notacontract.p0.s4\nmux2005\nlogin to agree\n0\n3123\n\n\nThis sentence should be moved to the section about peer-to-peer copying in section 6. It's confusing where it is now, and distributing copies of the software, even if it's "ancillary propagation", is different from having and using a copy.\ngpl3.notacontract.p0.s2\njohnston\nlogin to agree\n0\n3149\n\n\n", "2521:3135": "\n\n\n\nMake this a requirement rather than a statement of fact. As it is it seems to say that this happens automatically under copyright law. Make it say "in order to propagate this work you must give the recipient a license to ...".\ngpl3.autolicense.p0.s1\njring\nlogin to agree\n1\n2521\n\n\nAre "you" appointed agent of the original licensors in order to grant the rights here? If so, there are some questions as to how this agency relationship arises in the absence of a contractual relationship. Alternatively, is it saying that each licensor is, in seeking to license under GPL3, granting permission at the outset for all subsequent recipients to use his/her contributions under GPL3? The latter makes more sense. Either way, it seems that in a work which is the subject of complex development, this creates a shower of parallel licences.\ngpl3.autolicense.p0.s1\nandrewk\nlogin to agree\n0\n3135\n\n\n", "3124": "\n\n\n\nI don't see a definition of "original licensors". Is that the same as the defined term "contributors"? That would make sense, but my naive understanding of "original licensors" is "the very very first licensors in the licensing chain", but I don't see how that would make sense here. Don't I automatically receive a license from every contributor?\ngpl3.autolicense.p0.s1\nmux2005\nlogin to agree\n1\n3124\n\n\n", "2521:3124:3135": "\n\n\n\nMake this a requirement rather than a statement of fact. As it is it seems to say that this happens automatically under copyright law. Make it say "in order to propagate this work you must give the recipient a license to ...".\ngpl3.autolicense.p0.s1\njring\nlogin to agree\n1\n2521\n\n\nI don't see a definition of "original licensors". Is that the same as the defined term "contributors"? That would make sense, but my naive understanding of "original licensors" is "the very very first licensors in the licensing chain", but I don't see how that would make sense here. Don't I automatically receive a license from every contributor?\ngpl3.autolicense.p0.s1\nmux2005\nlogin to agree\n1\n3124\n\n\nAre "you" appointed agent of the original licensors in order to grant the rights here? If so, there are some questions as to how this agency relationship arises in the absence of a contractual relationship. Alternatively, is it saying that each licensor is, in seeking to license under GPL3, granting permission at the outset for all subsequent recipients to use his/her contributions under GPL3? The latter makes more sense. Either way, it seems that in a work which is the subject of complex development, this creates a shower of parallel licences.\ngpl3.autolicense.p0.s1\nandrewk\nlogin to agree\n0\n3135\n\n\n", "2793": "\n\n\n\nIt is possible to receive a program licensed using GPLv3, including a written offer to get the source code, valid for a limited term (at least 3 years). After this term, the offer is no longer valid, and so the entity may have lost the right to possess the Corresponding Source (this happens if it has no other means to get it). In such a case, this sentence demands the entity to give away a right it doesn't have itself.\ngpl3.autolicense.p1.s2\nwijnen\nlogin to agree\n0\n2793\n\n\n", "3125": "\n\n\n\nPlease add "(except as permitted by section 7)" as a cross-reference.\ngpl3.autolicense.p2.s1\nmux2005\nlogin to agree\n0\n3125\n\n\n", "2679": "\n\n\n\n<p style="display: inline">Suppose I improve a GPLv3'ed compression program by adding one of my patented algorithms to it. I distribute it, and cannot sue my licensees. Fine!</p><p>\nBut now a licensee makes a contribution by adding one of my other patented algorithms to the program. It seems like I lose my GPLv3 license if I sue them. This case isn't so clearly okay.</p><p>\nI can see an argument for it, but is it the intended consequence of this term? </p>\ngpl3.autolicense.p2.s2\nsanjoy\nlogin to agree\n0\n2679\n\n\n", "2571:2840:3068": "\n\n\n\nThe change in this draft about the difference in applicance of sections 10 and 11 for unmodified vs modified distribution, as explained in the Rationale, is a very good idea. I like it.\ngpl3.licensingpatents.0.0\nadhemar\nlogin to agree\n0\n2571\n\n\n<p style="display: inline">[Background context:] I am a concerned bystander without legal training. I have not read any of the GPL3 drafts in full.</p><p>\nI am concerned that clause 11 of GPL3 draft3 can be circumvented as follows. A company, XYZ, uses, distributes, contributes to (etc) GPL3 code. They own no patents. They know of no patents they are violating. They have not arranged for any kind patent license related to the work they contribute. In a subscenario, XYZ owns interests in ABC. ABC, among other things, collects royalties and enforces a patent porfolio. ABC may (or may not) have acquired patents from XYZ at some point in time. Maybe some employees of XYZ do patent independent contract work for ABC. XYZ gets some sort of ongoing return on their investment (eg, dividends or simple stock appreciation for shares owned). In a second subscenario, XYZ sells patents to ABC for a lump sum, but XYZ has no ownership stake in ABC. In either scenario, ABC does not deal in or care about GPL3 software except to make patent claims on code that may very well have come from XYZ. Further, XYZ has no license covering the work they have done which falls under the claim.\n \nTwo questions:</p><p>\nQ1 -- Does the GPL3 draft3 restrict (apply to) XYZ under any of these or similar scenarios once ABC tries to enforce its rights on anyone? [I don't know if ABC can enforce its patent claims against others while not seeking anything from XYZ (eg, in US law). Assuming ABC can be selective without forfeiting any rights, XYZ can always technically be at risk though they might never know or care. Assume that there is no formal covenant between ABC and XYZ. XYZ simply takes a chance ABC will not come after it (eg, an unofficial quid pro quo).]</p><p>\nQ2 -- Should the GPL3 include protections against this? [I'm assuming the draft does not already cover this case.]\n</p>\ngpl3.licensingpatents.0.0\nhozelda\nlogin to agree\n0\n2840\n\n\n<p style="display: inline">I should be explicitly stated somewhere in the license, that grant of patent license not presume nor assert validity of licensed patents. I.e.</p><p>\n1) licensor should not be liable if licensed patent will be found invalid later.</p><p>\n2) licensee should retain the right to challenge validity of licensed patent.</p>\ngpl3.licensingpatents.0.0\nlqp\nlogin to agree\n0\n3068\n\n\n", "2433:2493:2508:2723:2804:3003": "\n\n\n\n<p style="display: inline">Suggest following: Each patent licence contribution to a Program must grant a yadda-yadda licence to all works covered by the GPL, or that licence cannot be accepted.</p><p>\nBasically, patent licence must be free enough to ensure future freedom, or patent licence cannot be accepted, and program incorporating ideas by that unacceptably licenced patent cannot be distributed AT ALL.</p><p>\nFreedom or death.</p>\ngpl3.licensingpatents.p0.s1\ngustavb\nlogin to agree\n0\n2433\n\n\nWhy not anybody else making money from, distributing or using the contribution?\nHow about a company paying a freelance programmer to write and contribute some code covered by a patent they own? They could later use or even sell the product without being the contributor. Thay could even arrange one of the later described agreements.\ngpl3.licensingpatents.p0.s1\nibor\nlogin to agree\n4\n2493\n\n\nHow can the GPL make this assurance to the licensee, now that the 2nd half of this paragraph has been deleted? The remaining patent grants requirements are conditional and potentially weaker.\ngpl3.licensingpatents.p0.s1\nblinken\nlogin to agree\n1\n2508\n\n\nThis should read distributor (or conveyor I guess). There's no reason distributors should be able to distribute without granting a patent licenses. Redistributors should not be able to use their patents against recipients, at all. It's irrelevant whether distributors know which patents the program infringes. Having distributors license patents clearly helps freedom. Termination upon a lawsuit is not enough. They shouldn't be able to sue at all.\ngpl3.licensingpatents.p0.s1\nflaschen\nlogin to agree\n3\n2723\n\n\nThis is much better than in version 2. There, a complete licence wasn't required. This is clearer.\ngpl3.licensingpatents.p0.s1\ntimi\nlogin to agree\n0\n2804\n\n\n<p style="display: inline">Missing here seems to be a clear statement of your licensing your patent claims to a covered work that you convey. How about adding a paragraph:</p><p>\nOn conveying a covered work, you grant a non-exclusive royalty-free patent license under your essential patent claims to the work, to make, use, sell, offer for sale, import and otherwise run modify and propagate the covered work.</p>\ngpl3.licensingpatents.p0.s1\nmole\nlogin to agree\n0\n3003\n\n\n", "2508:2804:3003": "\n\n\n\nHow can the GPL make this assurance to the licensee, now that the 2nd half of this paragraph has been deleted? The remaining patent grants requirements are conditional and potentially weaker.\ngpl3.licensingpatents.p0.s1\nblinken\nlogin to agree\n1\n2508\n\n\nThis is much better than in version 2. There, a complete licence wasn't required. This is clearer.\ngpl3.licensingpatents.p0.s1\ntimi\nlogin to agree\n0\n2804\n\n\n<p style="display: inline">Missing here seems to be a clear statement of your licensing your patent claims to a covered work that you convey. How about adding a paragraph:</p><p>\nOn conveying a covered work, you grant a non-exclusive royalty-free patent license under your essential patent claims to the work, to make, use, sell, offer for sale, import and otherwise run modify and propagate the covered work.</p>\ngpl3.licensingpatents.p0.s1\nmole\nlogin to agree\n0\n3003\n\n\n", "2471:2508:2804:3003": "\n\n\n\nWith only a worldwide license I could charge NASA fees to use my patent on a Mars rover. If I code and contribute my work on the moon (or a space station/ship or some other world) I could then charge fees to users on Earth. Even if this grants a license for each world I am using the work on it still does not protect items not on a world (a communications satellite or interstellar probe for example).\ngpl3.licensingpatents.p0.s1\nixphin\nlogin to agree\n7\n2471\n\n\nHow can the GPL make this assurance to the licensee, now that the 2nd half of this paragraph has been deleted? The remaining patent grants requirements are conditional and potentially weaker.\ngpl3.licensingpatents.p0.s1\nblinken\nlogin to agree\n1\n2508\n\n\nThis is much better than in version 2. There, a complete licence wasn't required. This is clearer.\ngpl3.licensingpatents.p0.s1\ntimi\nlogin to agree\n0\n2804\n\n\n<p style="display: inline">Missing here seems to be a clear statement of your licensing your patent claims to a covered work that you convey. How about adding a paragraph:</p><p>\nOn conveying a covered work, you grant a non-exclusive royalty-free patent license under your essential patent claims to the work, to make, use, sell, offer for sale, import and otherwise run modify and propagate the covered work.</p>\ngpl3.licensingpatents.p0.s1\nmole\nlogin to agree\n0\n3003\n\n\n", "3003": "\n\n\n\n<p style="display: inline">Missing here seems to be a clear statement of your licensing your patent claims to a covered work that you convey. How about adding a paragraph:</p><p>\nOn conveying a covered work, you grant a non-exclusive royalty-free patent license under your essential patent claims to the work, to make, use, sell, offer for sale, import and otherwise run modify and propagate the covered work.</p>\ngpl3.licensingpatents.p0.s1\nmole\nlogin to agree\n0\n3003\n\n\n", "2874": "\n\n\n\n<p style="display: inline">Granting the patent license only for use with the code it covers means that the code cannot at any stage be rewritten.</p><p>\nEssentially, the license holder could sue you for using their patent, but /not/ using their code.</p><p>\nWhen code is contributed for use under the GPL, it is not tied to a particular work, it can be reused in other works. Patents contributed for use under the GPL should not be tied to particular code.</p>\ngpl3.licensingpatents.p0.s1\njamesgnz\nlogin to agree\n0\n2874\n\n\n", "2433:2471:2493:2508:2723:2804:2874:3003": "\n\n\n\n<p style="display: inline">Suggest following: Each patent licence contribution to a Program must grant a yadda-yadda licence to all works covered by the GPL, or that licence cannot be accepted.</p><p>\nBasically, patent licence must be free enough to ensure future freedom, or patent licence cannot be accepted, and program incorporating ideas by that unacceptably licenced patent cannot be distributed AT ALL.</p><p>\nFreedom or death.</p>\ngpl3.licensingpatents.p0.s1\ngustavb\nlogin to agree\n0\n2433\n\n\nWith only a worldwide license I could charge NASA fees to use my patent on a Mars rover. If I code and contribute my work on the moon (or a space station/ship or some other world) I could then charge fees to users on Earth. Even if this grants a license for each world I am using the work on it still does not protect items not on a world (a communications satellite or interstellar probe for example).\ngpl3.licensingpatents.p0.s1\nixphin\nlogin to agree\n7\n2471\n\n\nWhy not anybody else making money from, distributing or using the contribution?\nHow about a company paying a freelance programmer to write and contribute some code covered by a patent they own? They could later use or even sell the product without being the contributor. Thay could even arrange one of the later described agreements.\ngpl3.licensingpatents.p0.s1\nibor\nlogin to agree\n4\n2493\n\n\nHow can the GPL make this assurance to the licensee, now that the 2nd half of this paragraph has been deleted? The remaining patent grants requirements are conditional and potentially weaker.\ngpl3.licensingpatents.p0.s1\nblinken\nlogin to agree\n1\n2508\n\n\nThis should read distributor (or conveyor I guess). There's no reason distributors should be able to distribute without granting a patent licenses. Redistributors should not be able to use their patents against recipients, at all. It's irrelevant whether distributors know which patents the program infringes. Having distributors license patents clearly helps freedom. Termination upon a lawsuit is not enough. They shouldn't be able to sue at all.\ngpl3.licensingpatents.p0.s1\nflaschen\nlogin to agree\n3\n2723\n\n\nThis is much better than in version 2. There, a complete licence wasn't required. This is clearer.\ngpl3.licensingpatents.p0.s1\ntimi\nlogin to agree\n0\n2804\n\n\n<p style="display: inline">Granting the patent license only for use with the code it covers means that the code cannot at any stage be rewritten.</p><p>\nEssentially, the license holder could sue you for using their patent, but /not/ using their code.</p><p>\nWhen code is contributed for use under the GPL, it is not tied to a particular work, it can be reused in other works. Patents contributed for use under the GPL should not be tied to particular code.</p>\ngpl3.licensingpatents.p0.s1\njamesgnz\nlogin to agree\n0\n2874\n\n\n<p style="display: inline">Missing here seems to be a clear statement of your licensing your patent claims to a covered work that you convey. How about adding a paragraph:</p><p>\nOn conveying a covered work, you grant a non-exclusive royalty-free patent license under your essential patent claims to the work, to make, use, sell, offer for sale, import and otherwise run modify and propagate the covered work.</p>\ngpl3.licensingpatents.p0.s1\nmole\nlogin to agree\n0\n3003\n\n\n", "3137": "\n\n\n\nA commitment suggests something of a binding nature, which, in practice, is likely to be contractual. It therefore excludes a statement which, although enforceable (through the principle of promissory estoppel, for example), is not contractually binding. I would favour "promise".\ngpl3.licensingpatents.p1.s1\nandrewk\nlogin to agree\n0\n3137\n\n\n", "2931": "\n\n\n\n\nPatents, section 11:\nI am puzzled by the following paragraph:\n"If you convey a covered work, knowingly relying on a patent license,\nand the Corresponding Source of the work is not available for anyone\nto copy, free of charge and under the terms of this License, through a\npublicly available network server or other readily accessible means,\nthen..."\nWhen can a covered work be conveyed without corresponding source fully\navailable?\n\n\n\n\ngpl3.licensingpatents.p2.s1\nw1\nlogin to agree\n0\n2931\n\n\n", "2459:2931": "\n\n\n\nShould we not constrain this provision to Patent license only, but any agreement that is in violation of GPL? Ditto for next paragraph.\ngpl3.licensingpatents.p2.s1\nwmshi\nlogin to agree\n0\n2459\n\n\n\nPatents, section 11:\nI am puzzled by the following paragraph:\n"If you convey a covered work, knowingly relying on a patent license,\nand the Corresponding Source of the work is not available for anyone\nto copy, free of charge and under the terms of this License, through a\npublicly available network server or other readily accessible means,\nthen..."\nWhen can a covered work be conveyed without corresponding source fully\navailable?\n\n\n\n\ngpl3.licensingpatents.p2.s1\nw1\nlogin to agree\n0\n2931\n\n\n", "2446:2931": "\n\n\n\nThe GNU GPL does not require, in general, that anything be "free of charge"; for example, source code can be offered for the reasonable charge of printing a copy. I suggest replacing this part with "under the preceding conditions of this license", or such.\ngpl3.licensingpatents.p2.s1\ncyd\nlogin to agree\n1\n2446\n\n\n\nPatents, section 11:\nI am puzzled by the following paragraph:\n"If you convey a covered work, knowingly relying on a patent license,\nand the Corresponding Source of the work is not available for anyone\nto copy, free of charge and under the terms of this License, through a\npublicly available network server or other readily accessible means,\nthen..."\nWhen can a covered work be conveyed without corresponding source fully\navailable?\n\n\n\n\ngpl3.licensingpatents.p2.s1\nw1\nlogin to agree\n0\n2931\n\n\n", "2446:2931:3126": "\n\n\n\nThe GNU GPL does not require, in general, that anything be "free of charge"; for example, source code can be offered for the reasonable charge of printing a copy. I suggest replacing this part with "under the preceding conditions of this license", or such.\ngpl3.licensingpatents.p2.s1\ncyd\nlogin to agree\n1\n2446\n\n\n\nPatents, section 11:\nI am puzzled by the following paragraph:\n"If you convey a covered work, knowingly relying on a patent license,\nand the Corresponding Source of the work is not available for anyone\nto copy, free of charge and under the terms of this License, through a\npublicly available network server or other readily accessible means,\nthen..."\nWhen can a covered work be conveyed without corresponding source fully\navailable?\n\n\n\n\ngpl3.licensingpatents.p2.s1\nw1\nlogin to agree\n0\n2931\n\n\nThis should be "everyone", right?\ngpl3.licensingpatents.p2.s1\nmux2005\nlogin to agree\n0\n3126\n\n\n", "2446": "\n\n\n\nThe GNU GPL does not require, in general, that anything be "free of charge"; for example, source code can be offered for the reasonable charge of printing a copy. I suggest replacing this part with "under the preceding conditions of this license", or such.\ngpl3.licensingpatents.p2.s1\ncyd\nlogin to agree\n1\n2446\n\n\n", "2775": "\n\n\n\n<p style="display: inline">I don't quite understand how (1) can be seen as a specific form of shielding downstream recipients. If I am a downstream recipient who does not have a patent license, what protection (against patent infringement lawsuits) would I get from the existence of a network server which makes source available to the public?</p><p>\nI'm puzzled.</p>\ngpl3.licensingpatents.p2.s1\nfrx\nlogin to agree\n2\n2775\n\n\n", "2510:2775": "\n\n\n\n<p style="display: inline">Why was public source code availability deemed sufficient to absolve a patent license? Having a source code that is unusable due to patents is of little use.</p><p>\nNeither draft 2 nor draft 3 provided a rationale for this allowance. Moreover, it was raised in several comments on draft 2 but not answered there either.</p>\ngpl3.licensingpatents.p2.s1\nblinken\nlogin to agree\n4\n2510\n\n\n<p style="display: inline">I don't quite understand how (1) can be seen as a specific form of shielding downstream recipients. If I am a downstream recipient who does not have a patent license, what protection (against patent infringement lawsuits) would I get from the existence of a network server which makes source available to the public?</p><p>\nI'm puzzled.</p>\ngpl3.licensingpatents.p2.s1\nfrx\nlogin to agree\n2\n2775\n\n\n", "2574:3069": "\n\n\n\nIANAL but shouldn't the burden of proof be put on the licensee by changing this to "do not have reason to believe are invalid"? Otherwise he could possibly claim that a patent was "probably invalid" without having to explain why or proving it.\ngpl3.licensingpatents.p2.s2\nhorsten\nlogin to agree\n1\n2574\n\n\nMany peoples believes that ALL software patents are invalid. \n\nSo, for them such clause takes a very little sense.\ngpl3.licensingpatents.p2.s2\nlqp\nlogin to agree\n2\n3069\n\n\n", "2574:2775:2931:3069:3126": "\n\n\n\nIANAL but shouldn't the burden of proof be put on the licensee by changing this to "do not have reason to believe are invalid"? Otherwise he could possibly claim that a patent was "probably invalid" without having to explain why or proving it.\ngpl3.licensingpatents.p2.s2\nhorsten\nlogin to agree\n1\n2574\n\n\n<p style="display: inline">I don't quite understand how (1) can be seen as a specific form of shielding downstream recipients. If I am a downstream recipient who does not have a patent license, what protection (against patent infringement lawsuits) would I get from the existence of a network server which makes source available to the public?</p><p>\nI'm puzzled.</p>\ngpl3.licensingpatents.p2.s1\nfrx\nlogin to agree\n2\n2775\n\n\n\nPatents, section 11:\nI am puzzled by the following paragraph:\n"If you convey a covered work, knowingly relying on a patent license,\nand the Corresponding Source of the work is not available for anyone\nto copy, free of charge and under the terms of this License, through a\npublicly available network server or other readily accessible means,\nthen..."\nWhen can a covered work be conveyed without corresponding source fully\navailable?\n\n\n\n\ngpl3.licensingpatents.p2.s1\nw1\nlogin to agree\n0\n2931\n\n\nMany peoples believes that ALL software patents are invalid. \n\nSo, for them such clause takes a very little sense.\ngpl3.licensingpatents.p2.s2\nlqp\nlogin to agree\n2\n3069\n\n\nThis should be "everyone", right?\ngpl3.licensingpatents.p2.s1\nmux2005\nlogin to agree\n0\n3126\n\n\n", "2486:2724:2776:3117": "\n\n\n\nI know this is a legal document and all, but one of the beauties of GPLv2 was that a normal person could read it and understand it. I'm a pretty smart guy and I don't have the first clue what the hell this paragraph means. It'd be nice if the lawyer-speak were toned down as much as possible so people like me still know what exactly we're proposing when we license our code under GPLv3.\ngpl3.licensingpatents.p2b.s1\njag\nlogin to agree\n7\n2486\n\n\nAgain, it should be "If you convey a covered work, you grant a patent license providing freedom to use, propogate, modify or convey the covered work to all parties receiving the covered work". Backing down on this weakens freedom.\ngpl3.licensingpatents.p2b.s1\nflaschen\nlogin to agree\n0\n2724\n\n\nThis clause seems to be intended to prevent the negative consequences of Microsoft-Novell-like deals. It seems to be well drafted, since it's based on a no discrimination principle: you cannot grant a patent license to some recipients only; if you grant a patent license to someone, you automatically extend it to every other recipient. Good.\ngpl3.licensingpatents.p2b.s1\nfrx\nlogin to agree\n0\n2776\n\n\n<p style="display: inline">If I give my friend a some money for a copy of GPL3 licenced software, to give to a friend, then I've procured conveyance of a covered work, but I've not agreed to the licence. If you could include this sort of clause in in a licence and it were enforceable, you'd see a lot of clauses requiring people to give up their firstborn children, their goods and chattels. And that's absurd.</p><p>\nWhat you could do is require anyone who conveys on the procurement of another to have that other party agree to be bound by the terms of the licence too. Without that agreement, the conveyer should have no license to convey. \n\nIn this way, if my friend asks me to distribute software, I'd only be allowed to do so under the GPL3 if they had agreed to be bound by the license too. \n\nAnd in being bound, they'd be agreeing that they license patents covering software distributed under the licence. \n\nTo some extent this opens up a possible attack on distributed software:Where is the agreement of the procurer to be found? \n\nJeff \n\n</p>\ngpl3.licensingpatents.p2b.s1\njeffveit\nlogin to agree\n0\n3117\n\n\n", "2486:2724:2776:2862:3117": "\n\n\n\nI know this is a legal document and all, but one of the beauties of GPLv2 was that a normal person could read it and understand it. I'm a pretty smart guy and I don't have the first clue what the hell this paragraph means. It'd be nice if the lawyer-speak were toned down as much as possible so people like me still know what exactly we're proposing when we license our code under GPLv3.\ngpl3.licensingpatents.p2b.s1\njag\nlogin to agree\n7\n2486\n\n\nAgain, it should be "If you convey a covered work, you grant a patent license providing freedom to use, propogate, modify or convey the covered work to all parties receiving the covered work". Backing down on this weakens freedom.\ngpl3.licensingpatents.p2b.s1\nflaschen\nlogin to agree\n0\n2724\n\n\nThis clause seems to be intended to prevent the negative consequences of Microsoft-Novell-like deals. It seems to be well drafted, since it's based on a no discrimination principle: you cannot grant a patent license to some recipients only; if you grant a patent license to someone, you automatically extend it to every other recipient. Good.\ngpl3.licensingpatents.p2b.s1\nfrx\nlogin to agree\n0\n2776\n\n\nReplace "it." with "the Program."\ngpl3.licensingpatents.p2b.s1\njkoenig\nlogin to agree\n0\n2862\n\n\n<p style="display: inline">If I give my friend a some money for a copy of GPL3 licenced software, to give to a friend, then I've procured conveyance of a covered work, but I've not agreed to the licence. If you could include this sort of clause in in a licence and it were enforceable, you'd see a lot of clauses requiring people to give up their firstborn children, their goods and chattels. And that's absurd.</p><p>\nWhat you could do is require anyone who conveys on the procurement of another to have that other party agree to be bound by the terms of the licence too. Without that agreement, the conveyer should have no license to convey. \n\nIn this way, if my friend asks me to distribute software, I'd only be allowed to do so under the GPL3 if they had agreed to be bound by the license too. \n\nAnd in being bound, they'd be agreeing that they license patents covering software distributed under the licence. \n\nTo some extent this opens up a possible attack on distributed software:Where is the agreement of the procurer to be found? \n\nJeff \n\n</p>\ngpl3.licensingpatents.p2b.s1\njeffveit\nlogin to agree\n0\n3117\n\n\n", "2486:2512:2724:2776:3117": "\n\n\n\nI know this is a legal document and all, but one of the beauties of GPLv2 was that a normal person could read it and understand it. I'm a pretty smart guy and I don't have the first clue what the hell this paragraph means. It'd be nice if the lawyer-speak were toned down as much as possible so people like me still know what exactly we're proposing when we license our code under GPLv3.\ngpl3.licensingpatents.p2b.s1\njag\nlogin to agree\n7\n2486\n\n\nThis is easier to parse if phrased "any single transaction".\ngpl3.licensingpatents.p2b.s1\nblinken\nlogin to agree\n0\n2512\n\n\nAgain, it should be "If you convey a covered work, you grant a patent license providing freedom to use, propogate, modify or convey the covered work to all parties receiving the covered work". Backing down on this weakens freedom.\ngpl3.licensingpatents.p2b.s1\nflaschen\nlogin to agree\n0\n2724\n\n\nThis clause seems to be intended to prevent the negative consequences of Microsoft-Novell-like deals. It seems to be well drafted, since it's based on a no discrimination principle: you cannot grant a patent license to some recipients only; if you grant a patent license to someone, you automatically extend it to every other recipient. Good.\ngpl3.licensingpatents.p2b.s1\nfrx\nlogin to agree\n0\n2776\n\n\n<p style="display: inline">If I give my friend a some money for a copy of GPL3 licenced software, to give to a friend, then I've procured conveyance of a covered work, but I've not agreed to the licence. If you could include this sort of clause in in a licence and it were enforceable, you'd see a lot of clauses requiring people to give up their firstborn children, their goods and chattels. And that's absurd.</p><p>\nWhat you could do is require anyone who conveys on the procurement of another to have that other party agree to be bound by the terms of the licence too. Without that agreement, the conveyer should have no license to convey. \n\nIn this way, if my friend asks me to distribute software, I'd only be allowed to do so under the GPL3 if they had agreed to be bound by the license too. \n\nAnd in being bound, they'd be agreeing that they license patents covering software distributed under the licence. \n\nTo some extent this opens up a possible attack on distributed software:Where is the agreement of the procurer to be found? \n\nJeff \n\n</p>\ngpl3.licensingpatents.p2b.s1\njeffveit\nlogin to agree\n0\n3117\n\n\n", "2486:2724:2776:2841:3117": "\n\n\n\nI know this is a legal document and all, but one of the beauties of GPLv2 was that a normal person could read it and understand it. I'm a pretty smart guy and I don't have the first clue what the hell this paragraph means. It'd be nice if the lawyer-speak were toned down as much as possible so people like me still know what exactly we're proposing when we license our code under GPLv3.\ngpl3.licensingpatents.p2b.s1\njag\nlogin to agree\n7\n2486\n\n\nAgain, it should be "If you convey a covered work, you grant a patent license providing freedom to use, propogate, modify or convey the covered work to all parties receiving the covered work". Backing down on this weakens freedom.\ngpl3.licensingpatents.p2b.s1\nflaschen\nlogin to agree\n0\n2724\n\n\nThis clause seems to be intended to prevent the negative consequences of Microsoft-Novell-like deals. It seems to be well drafted, since it's based on a no discrimination principle: you cannot grant a patent license to some recipients only; if you grant a patent license to someone, you automatically extend it to every other recipient. Good.\ngpl3.licensingpatents.p2b.s1\nfrx\nlogin to agree\n0\n2776\n\n\nIn my opinion, this clause wouldn't have prevented the Novell-Microsoft deal, and it won't make the patent license apply to everyone (or stop Novell from distributing). It simply doesn't apply because only Novell is distributing (discounting Windows Terminal Services for UNIX, which is likely to continue using GPLv2 versions or just move to proprietary UNIX), but they aren't the ones granting the patents. Microsoft is granting the patents, and indeed only they can because they're the owner. Since they aren't going to be distributing GPLv3 software, this clause can not make them extend the patent license. They're beyond the scope.\ngpl3.licensingpatents.p2b.s1\nflaschen\nlogin to agree\n0\n2841\n\n\n<p style="display: inline">If I give my friend a some money for a copy of GPL3 licenced software, to give to a friend, then I've procured conveyance of a covered work, but I've not agreed to the licence. If you could include this sort of clause in in a licence and it were enforceable, you'd see a lot of clauses requiring people to give up their firstborn children, their goods and chattels. And that's absurd.</p><p>\nWhat you could do is require anyone who conveys on the procurement of another to have that other party agree to be bound by the terms of the licence too. Without that agreement, the conveyer should have no license to convey. \n\nIn this way, if my friend asks me to distribute software, I'd only be allowed to do so under the GPL3 if they had agreed to be bound by the license too. \n\nAnd in being bound, they'd be agreeing that they license patents covering software distributed under the licence. \n\nTo some extent this opens up a possible attack on distributed software:Where is the agreement of the procurer to be found? \n\nJeff \n\n</p>\ngpl3.licensingpatents.p2b.s1\njeffveit\nlogin to agree\n0\n3117\n\n\n", "2486:2776:2841:3117": "\n\n\n\nI know this is a legal document and all, but one of the beauties of GPLv2 was that a normal person could read it and understand it. I'm a pretty smart guy and I don't have the first clue what the hell this paragraph means. It'd be nice if the lawyer-speak were toned down as much as possible so people like me still know what exactly we're proposing when we license our code under GPLv3.\ngpl3.licensingpatents.p2b.s1\njag\nlogin to agree\n7\n2486\n\n\nThis clause seems to be intended to prevent the negative consequences of Microsoft-Novell-like deals. It seems to be well drafted, since it's based on a no discrimination principle: you cannot grant a patent license to some recipients only; if you grant a patent license to someone, you automatically extend it to every other recipient. Good.\ngpl3.licensingpatents.p2b.s1\nfrx\nlogin to agree\n0\n2776\n\n\nIn my opinion, this clause wouldn't have prevented the Novell-Microsoft deal, and it won't make the patent license apply to everyone (or stop Novell from distributing). It simply doesn't apply because only Novell is distributing (discounting Windows Terminal Services for UNIX, which is likely to continue using GPLv2 versions or just move to proprietary UNIX), but they aren't the ones granting the patents. Microsoft is granting the patents, and indeed only they can because they're the owner. Since they aren't going to be distributing GPLv3 software, this clause can not make them extend the patent license. They're beyond the scope.\ngpl3.licensingpatents.p2b.s1\nflaschen\nlogin to agree\n0\n2841\n\n\n<p style="display: inline">If I give my friend a some money for a copy of GPL3 licenced software, to give to a friend, then I've procured conveyance of a covered work, but I've not agreed to the licence. If you could include this sort of clause in in a licence and it were enforceable, you'd see a lot of clauses requiring people to give up their firstborn children, their goods and chattels. And that's absurd.</p><p>\nWhat you could do is require anyone who conveys on the procurement of another to have that other party agree to be bound by the terms of the licence too. Without that agreement, the conveyer should have no license to convey. \n\nIn this way, if my friend asks me to distribute software, I'd only be allowed to do so under the GPL3 if they had agreed to be bound by the license too. \n\nAnd in being bound, they'd be agreeing that they license patents covering software distributed under the licence. \n\nTo some extent this opens up a possible attack on distributed software:Where is the agreement of the procurer to be found? \n\nJeff \n\n</p>\ngpl3.licensingpatents.p2b.s1\njeffveit\nlogin to agree\n0\n3117\n\n\n", "2486:2776:2841:3117:3119": "\n\n\n\nI know this is a legal document and all, but one of the beauties of GPLv2 was that a normal person could read it and understand it. I'm a pretty smart guy and I don't have the first clue what the hell this paragraph means. It'd be nice if the lawyer-speak were toned down as much as possible so people like me still know what exactly we're proposing when we license our code under GPLv3.\ngpl3.licensingpatents.p2b.s1\njag\nlogin to agree\n7\n2486\n\n\nThis clause seems to be intended to prevent the negative consequences of Microsoft-Novell-like deals. It seems to be well drafted, since it's based on a no discrimination principle: you cannot grant a patent license to some recipients only; if you grant a patent license to someone, you automatically extend it to every other recipient. Good.\ngpl3.licensingpatents.p2b.s1\nfrx\nlogin to agree\n0\n2776\n\n\nIn my opinion, this clause wouldn't have prevented the Novell-Microsoft deal, and it won't make the patent license apply to everyone (or stop Novell from distributing). It simply doesn't apply because only Novell is distributing (discounting Windows Terminal Services for UNIX, which is likely to continue using GPLv2 versions or just move to proprietary UNIX), but they aren't the ones granting the patents. Microsoft is granting the patents, and indeed only they can because they're the owner. Since they aren't going to be distributing GPLv3 software, this clause can not make them extend the patent license. They're beyond the scope.\ngpl3.licensingpatents.p2b.s1\nflaschen\nlogin to agree\n0\n2841\n\n\n<p style="display: inline">If I give my friend a some money for a copy of GPL3 licenced software, to give to a friend, then I've procured conveyance of a covered work, but I've not agreed to the licence. If you could include this sort of clause in in a licence and it were enforceable, you'd see a lot of clauses requiring people to give up their firstborn children, their goods and chattels. And that's absurd.</p><p>\nWhat you could do is require anyone who conveys on the procurement of another to have that other party agree to be bound by the terms of the licence too. Without that agreement, the conveyer should have no license to convey. \n\nIn this way, if my friend asks me to distribute software, I'd only be allowed to do so under the GPL3 if they had agreed to be bound by the license too. \n\nAnd in being bound, they'd be agreeing that they license patents covering software distributed under the licence. \n\nTo some extent this opens up a possible attack on distributed software:Where is the agreement of the procurer to be found? \n\nJeff \n\n</p>\ngpl3.licensingpatents.p2b.s1\njeffveit\nlogin to agree\n0\n3117\n\n\nI don't see any definition of the word "procuring" or "procure" in the license (this appears to be the only place the word appears). Since this paragraph is apparently very important for the Microsoft patent issue, shouldn't the word be defined within the license?\ngpl3.licensingpatents.p2b.s1\nanjrew\nlogin to agree\n1\n3119\n\n\n", "2486:2776:2806:2841:3117": "\n\n\n\nI know this is a legal document and all, but one of the beauties of GPLv2 was that a normal person could read it and understand it. I'm a pretty smart guy and I don't have the first clue what the hell this paragraph means. It'd be nice if the lawyer-speak were toned down as much as possible so people like me still know what exactly we're proposing when we license our code under GPLv3.\ngpl3.licensingpatents.p2b.s1\njag\nlogin to agree\n7\n2486\n\n\nThis clause seems to be intended to prevent the negative consequences of Microsoft-Novell-like deals. It seems to be well drafted, since it's based on a no discrimination principle: you cannot grant a patent license to some recipients only; if you grant a patent license to someone, you automatically extend it to every other recipient. Good.\ngpl3.licensingpatents.p2b.s1\nfrx\nlogin to agree\n0\n2776\n\n\n<p style="display: inline">In the draft 2, the FSF didn't want to include a royaty free patent license. Now here it is. Well, I think this is a good idea, not only the first paragraph, but also this fourth paragraph.</p><p>\nHowever, now you want to do something against Novell. And you have now this patent license in th GPL (good) but all the rest (paragraph 5) is only ideological.</p>\ngpl3.licensingpatents.p2b.s1\ntimi\nlogin to agree\n0\n2806\n\n\nIn my opinion, this clause wouldn't have prevented the Novell-Microsoft deal, and it won't make the patent license apply to everyone (or stop Novell from distributing). It simply doesn't apply because only Novell is distributing (discounting Windows Terminal Services for UNIX, which is likely to continue using GPLv2 versions or just move to proprietary UNIX), but they aren't the ones granting the patents. Microsoft is granting the patents, and indeed only they can because they're the owner. Since they aren't going to be distributing GPLv3 software, this clause can not make them extend the patent license. They're beyond the scope.\ngpl3.licensingpatents.p2b.s1\nflaschen\nlogin to agree\n0\n2841\n\n\n<p style="display: inline">If I give my friend a some money for a copy of GPL3 licenced software, to give to a friend, then I've procured conveyance of a covered work, but I've not agreed to the licence. If you could include this sort of clause in in a licence and it were enforceable, you'd see a lot of clauses requiring people to give up their firstborn children, their goods and chattels. And that's absurd.</p><p>\nWhat you could do is require anyone who conveys on the procurement of another to have that other party agree to be bound by the terms of the licence too. Without that agreement, the conveyer should have no license to convey. \n\nIn this way, if my friend asks me to distribute software, I'd only be allowed to do so under the GPL3 if they had agreed to be bound by the license too. \n\nAnd in being bound, they'd be agreeing that they license patents covering software distributed under the licence. \n\nTo some extent this opens up a possible attack on distributed software:Where is the agreement of the procurer to be found? \n\nJeff \n\n</p>\ngpl3.licensingpatents.p2b.s1\njeffveit\nlogin to agree\n0\n3117\n\n\n", "2486:2776:3117": "\n\n\n\nI know this is a legal document and all, but one of the beauties of GPLv2 was that a normal person could read it and understand it. I'm a pretty smart guy and I don't have the first clue what the hell this paragraph means. It'd be nice if the lawyer-speak were toned down as much as possible so people like me still know what exactly we're proposing when we license our code under GPLv3.\ngpl3.licensingpatents.p2b.s1\njag\nlogin to agree\n7\n2486\n\n\nThis clause seems to be intended to prevent the negative consequences of Microsoft-Novell-like deals. It seems to be well drafted, since it's based on a no discrimination principle: you cannot grant a patent license to some recipients only; if you grant a patent license to someone, you automatically extend it to every other recipient. Good.\ngpl3.licensingpatents.p2b.s1\nfrx\nlogin to agree\n0\n2776\n\n\n<p style="display: inline">If I give my friend a some money for a copy of GPL3 licenced software, to give to a friend, then I've procured conveyance of a covered work, but I've not agreed to the licence. If you could include this sort of clause in in a licence and it were enforceable, you'd see a lot of clauses requiring people to give up their firstborn children, their goods and chattels. And that's absurd.</p><p>\nWhat you could do is require anyone who conveys on the procurement of another to have that other party agree to be bound by the terms of the licence too. Without that agreement, the conveyer should have no license to convey. \n\nIn this way, if my friend asks me to distribute software, I'd only be allowed to do so under the GPL3 if they had agreed to be bound by the license too. \n\nAnd in being bound, they'd be agreeing that they license patents covering software distributed under the licence. \n\nTo some extent this opens up a possible attack on distributed software:Where is the agreement of the procurer to be found? \n\nJeff \n\n</p>\ngpl3.licensingpatents.p2b.s1\njeffveit\nlogin to agree\n0\n3117\n\n\n", "3145": "\n\n\n\n<p style="display: inline">Consider the <b> terms </b> of such an extended license.\nIs there anything in the GPLv3 draft explicitly requiring the terms of such an extended license to be consistent with the GPLv3 draft licensing terms; in the spirit of such requirements under GPLv2 section 2b.\nIs there anything in the GPLv3 draft requiring such license to be perpetual.\nWhat is the maximum damage to business (and maximum damage should always be presumed) if development was done based on and integrated with such temporary offerings and at the end of a given term the license holder said "your license is up, I'm not renewing it for any amount of money, and btw none of the four forms of esstopple or waiver apply since the only reason I wasn't suing you was your temporary license"</p><p>\nTemporary licenses are all well and good for self sufficient closed source offerings which have fleeting lifetimes anyway; but GPL code is designed to slowly build on itself and can't afford weaknesses in its foundation, especially in an era where IP is becoming increasingly immortal and where the only use our children's children may be able to make of technology without paying prohibitive fees are uses and functionality protected by the GPL. (And yes the GPL is a religion and as such protected by the freedom of religion clauses of the United States and other countries)</p>\ngpl3.licensingpatents.p2b.s1\ndrcj\nlogin to agree\n0\n3145\n\n\n", "2499:3145": "\n\n\n\n<p style="display: inline">I cannot see how to enforce this provision. The transgresser is a third party who, assumably, had not accepted this license.</p><p>\nThis clause is specifically aimed at MS in the context of MS/Novell agreement. Say they ink the agreement after this draft license go into force. Microsoft then sue me. How can I exercise this defense as a GPL software user or contributor?</p><p>\n(Aside : Arguably, Microsoft has not convey or propagate any GPL software in the agreement. Can it can argue it only convey SUSE coupons (service agreement) and has nothing to do with software? I believe the key point here is does the definition of "propagate" covers it?)</p><p>\nIANAL. My reading is that this paragraph only apples to the transgressor if and only if (s)he is using GPL software, as this is the only way to say that (s)he had accepted the license. In my worst case scenario, it could only apply to the trangressor if he himself is using that particular software he sues another party. If (s)he does not use GPL software, (s)he falls through this net. \n\nOne particular grievence I have with MS/Novell agreement is that it is basically an agreement of MS with SUSE's customer, in disguise as an agreement of MS with Novell to side step the GPL. I am concerned that this wording falls into the same category. It can be read as a agreement between me and the person who supply software to a third party vendor of GPL software in disguise as an agreement between me and the third party vendor. This means we lose the moral highground.</p><p>\nMoreover, the ugliness of MS/Novell agreement is that it works because it is a agreement that will work for the third party (SUSE's customers) by default, i.e., MS don't sue. However, I cannot see how this paragraph helps me because the vendor signed an GPL infringing agreement.I am not even the "third party" in this agreement.\n\n</p>\ngpl3.licensingpatents.p2b.s1\nwmshi\nlogin to agree\n2\n2499\n\n\n<p style="display: inline">Consider the <b> terms </b> of such an extended license.\nIs there anything in the GPLv3 draft explicitly requiring the terms of such an extended license to be consistent with the GPLv3 draft licensing terms; in the spirit of such requirements under GPLv2 section 2b.\nIs there anything in the GPLv3 draft requiring such license to be perpetual.\nWhat is the maximum damage to business (and maximum damage should always be presumed) if development was done based on and integrated with such temporary offerings and at the end of a given term the license holder said "your license is up, I'm not renewing it for any amount of money, and btw none of the four forms of esstopple or waiver apply since the only reason I wasn't suing you was your temporary license"</p><p>\nTemporary licenses are all well and good for self sufficient closed source offerings which have fleeting lifetimes anyway; but GPL code is designed to slowly build on itself and can't afford weaknesses in its foundation, especially in an era where IP is becoming increasingly immortal and where the only use our children's children may be able to make of technology without paying prohibitive fees are uses and functionality protected by the GPL. (And yes the GPL is a religion and as such protected by the freedom of religion clauses of the United States and other countries)</p>\ngpl3.licensingpatents.p2b.s1\ndrcj\nlogin to agree\n0\n3145\n\n\n", "2499": "\n\n\n\n<p style="display: inline">I cannot see how to enforce this provision. The transgresser is a third party who, assumably, had not accepted this license.</p><p>\nThis clause is specifically aimed at MS in the context of MS/Novell agreement. Say they ink the agreement after this draft license go into force. Microsoft then sue me. How can I exercise this defense as a GPL software user or contributor?</p><p>\n(Aside : Arguably, Microsoft has not convey or propagate any GPL software in the agreement. Can it can argue it only convey SUSE coupons (service agreement) and has nothing to do with software? I believe the key point here is does the definition of "propagate" covers it?)</p><p>\nIANAL. My reading is that this paragraph only apples to the transgressor if and only if (s)he is using GPL software, as this is the only way to say that (s)he had accepted the license. In my worst case scenario, it could only apply to the trangressor if he himself is using that particular software he sues another party. If (s)he does not use GPL software, (s)he falls through this net. \n\nOne particular grievence I have with MS/Novell agreement is that it is basically an agreement of MS with SUSE's customer, in disguise as an agreement of MS with Novell to side step the GPL. I am concerned that this wording falls into the same category. It can be read as a agreement between me and the person who supply software to a third party vendor of GPL software in disguise as an agreement between me and the third party vendor. This means we lose the moral highground.</p><p>\nMoreover, the ugliness of MS/Novell agreement is that it works because it is a agreement that will work for the third party (SUSE's customers) by default, i.e., MS don't sue. However, I cannot see how this paragraph helps me because the vendor signed an GPL infringing agreement.I am not even the "third party" in this agreement.\n\n</p>\ngpl3.licensingpatents.p2b.s1\nwmshi\nlogin to agree\n2\n2499\n\n\n", "2499:2861": "\n\n\n\n<p style="display: inline">I cannot see how to enforce this provision. The transgresser is a third party who, assumably, had not accepted this license.</p><p>\nThis clause is specifically aimed at MS in the context of MS/Novell agreement. Say they ink the agreement after this draft license go into force. Microsoft then sue me. How can I exercise this defense as a GPL software user or contributor?</p><p>\n(Aside : Arguably, Microsoft has not convey or propagate any GPL software in the agreement. Can it can argue it only convey SUSE coupons (service agreement) and has nothing to do with software? I believe the key point here is does the definition of "propagate" covers it?)</p><p>\nIANAL. My reading is that this paragraph only apples to the transgressor if and only if (s)he is using GPL software, as this is the only way to say that (s)he had accepted the license. In my worst case scenario, it could only apply to the trangressor if he himself is using that particular software he sues another party. If (s)he does not use GPL software, (s)he falls through this net. \n\nOne particular grievence I have with MS/Novell agreement is that it is basically an agreement of MS with SUSE's customer, in disguise as an agreement of MS with Novell to side step the GPL. I am concerned that this wording falls into the same category. It can be read as a agreement between me and the person who supply software to a third party vendor of GPL software in disguise as an agreement between me and the third party vendor. This means we lose the moral highground.</p><p>\nMoreover, the ugliness of MS/Novell agreement is that it works because it is a agreement that will work for the third party (SUSE's customers) by default, i.e., MS don't sue. However, I cannot see how this paragraph helps me because the vendor signed an GPL infringing agreement.I am not even the "third party" in this agreement.\n\n</p>\ngpl3.licensingpatents.p2b.s1\nwmshi\nlogin to agree\n2\n2499\n\n\nI would replace "on it." with "the Program."\ngpl3.licensingpatents.p2b.s1\njkoenig\nlogin to agree\n0\n2861\n\n\n", "2486:2499:2512:2724:2776:2806:2841:2861:2862:3117:3119:3145": "\n\n\n\nI know this is a legal document and all, but one of the beauties of GPLv2 was that a normal person could read it and understand it. I'm a pretty smart guy and I don't have the first clue what the hell this paragraph means. It'd be nice if the lawyer-speak were toned down as much as possible so people like me still know what exactly we're proposing when we license our code under GPLv3.\ngpl3.licensingpatents.p2b.s1\njag\nlogin to agree\n7\n2486\n\n\n<p style="display: inline">I cannot see how to enforce this provision. The transgresser is a third party who, assumably, had not accepted this license.</p><p>\nThis clause is specifically aimed at MS in the context of MS/Novell agreement. Say they ink the agreement after this draft license go into force. Microsoft then sue me. How can I exercise this defense as a GPL software user or contributor?</p><p>\n(Aside : Arguably, Microsoft has not convey or propagate any GPL software in the agreement. Can it can argue it only convey SUSE coupons (service agreement) and has nothing to do with software? I believe the key point here is does the definition of "propagate" covers it?)</p><p>\nIANAL. My reading is that this paragraph only apples to the transgressor if and only if (s)he is using GPL software, as this is the only way to say that (s)he had accepted the license. In my worst case scenario, it could only apply to the trangressor if he himself is using that particular software he sues another party. If (s)he does not use GPL software, (s)he falls through this net. \n\nOne particular grievence I have with MS/Novell agreement is that it is basically an agreement of MS with SUSE's customer, in disguise as an agreement of MS with Novell to side step the GPL. I am concerned that this wording falls into the same category. It can be read as a agreement between me and the person who supply software to a third party vendor of GPL software in disguise as an agreement between me and the third party vendor. This means we lose the moral highground.</p><p>\nMoreover, the ugliness of MS/Novell agreement is that it works because it is a agreement that will work for the third party (SUSE's customers) by default, i.e., MS don't sue. However, I cannot see how this paragraph helps me because the vendor signed an GPL infringing agreement.I am not even the "third party" in this agreement.\n\n</p>\ngpl3.licensingpatents.p2b.s1\nwmshi\nlogin to agree\n2\n2499\n\n\nThis is easier to parse if phrased "any single transaction".\ngpl3.licensingpatents.p2b.s1\nblinken\nlogin to agree\n0\n2512\n\n\nAgain, it should be "If you convey a covered work, you grant a patent license providing freedom to use, propogate, modify or convey the covered work to all parties receiving the covered work". Backing down on this weakens freedom.\ngpl3.licensingpatents.p2b.s1\nflaschen\nlogin to agree\n0\n2724\n\n\nThis clause seems to be intended to prevent the negative consequences of Microsoft-Novell-like deals. It seems to be well drafted, since it's based on a no discrimination principle: you cannot grant a patent license to some recipients only; if you grant a patent license to someone, you automatically extend it to every other recipient. Good.\ngpl3.licensingpatents.p2b.s1\nfrx\nlogin to agree\n0\n2776\n\n\n<p style="display: inline">In the draft 2, the FSF didn't want to include a royaty free patent license. Now here it is. Well, I think this is a good idea, not only the first paragraph, but also this fourth paragraph.</p><p>\nHowever, now you want to do something against Novell. And you have now this patent license in th GPL (good) but all the rest (paragraph 5) is only ideological.</p>\ngpl3.licensingpatents.p2b.s1\ntimi\nlogin to agree\n0\n2806\n\n\nIn my opinion, this clause wouldn't have prevented the Novell-Microsoft deal, and it won't make the patent license apply to everyone (or stop Novell from distributing). It simply doesn't apply because only Novell is distributing (discounting Windows Terminal Services for UNIX, which is likely to continue using GPLv2 versions or just move to proprietary UNIX), but they aren't the ones granting the patents. Microsoft is granting the patents, and indeed only they can because they're the owner. Since they aren't going to be distributing GPLv3 software, this clause can not make them extend the patent license. They're beyond the scope.\ngpl3.licensingpatents.p2b.s1\nflaschen\nlogin to agree\n0\n2841\n\n\nI would replace "on it." with "the Program."\ngpl3.licensingpatents.p2b.s1\njkoenig\nlogin to agree\n0\n2861\n\n\nReplace "it." with "the Program."\ngpl3.licensingpatents.p2b.s1\njkoenig\nlogin to agree\n0\n2862\n\n\n<p style="display: inline">If I give my friend a some money for a copy of GPL3 licenced software, to give to a friend, then I've procured conveyance of a covered work, but I've not agreed to the licence. If you could include this sort of clause in in a licence and it were enforceable, you'd see a lot of clauses requiring people to give up their firstborn children, their goods and chattels. And that's absurd.</p><p>\nWhat you could do is require anyone who conveys on the procurement of another to have that other party agree to be bound by the terms of the licence too. Without that agreement, the conveyer should have no license to convey. \n\nIn this way, if my friend asks me to distribute software, I'd only be allowed to do so under the GPL3 if they had agreed to be bound by the license too. \n\nAnd in being bound, they'd be agreeing that they license patents covering software distributed under the licence. \n\nTo some extent this opens up a possible attack on distributed software:Where is the agreement of the procurer to be found? \n\nJeff \n\n</p>\ngpl3.licensingpatents.p2b.s1\njeffveit\nlogin to agree\n0\n3117\n\n\nI don't see any definition of the word "procuring" or "procure" in the license (this appears to be the only place the word appears). Since this paragraph is apparently very important for the Microsoft patent issue, shouldn't the word be defined within the license?\ngpl3.licensingpatents.p2b.s1\nanjrew\nlogin to agree\n1\n3119\n\n\n<p style="display: inline">Consider the <b> terms </b> of such an extended license.\nIs there anything in the GPLv3 draft explicitly requiring the terms of such an extended license to be consistent with the GPLv3 draft licensing terms; in the spirit of such requirements under GPLv2 section 2b.\nIs there anything in the GPLv3 draft requiring such license to be perpetual.\nWhat is the maximum damage to business (and maximum damage should always be presumed) if development was done based on and integrated with such temporary offerings and at the end of a given term the license holder said "your license is up, I'm not renewing it for any amount of money, and btw none of the four forms of esstopple or waiver apply since the only reason I wasn't suing you was your temporary license"</p><p>\nTemporary licenses are all well and good for self sufficient closed source offerings which have fleeting lifetimes anyway; but GPL code is designed to slowly build on itself and can't afford weaknesses in its foundation, especially in an era where IP is becoming increasingly immortal and where the only use our children's children may be able to make of technology without paying prohibitive fees are uses and functionality protected by the GPL. (And yes the GPL is a religion and as such protected by the freedom of religion clauses of the United States and other countries)</p>\ngpl3.licensingpatents.p2b.s1\ndrcj\nlogin to agree\n0\n3145\n\n\n", "2777:2790:2796:2842:2964:3054:3347": "\n\n\n\nThis clause seems to be based on the general principle that no further restrictions can be added beyond the terms of this License. In the present case, it is stated that one cannot add further restrictions through restrictive patent licenses. Looks OK.\ngpl3.licensingpatents.p3.s1\nfrx\nlogin to agree\n0\n2777\n\n\n<p style="display: inline">I don't think that this is a good idea. First it isn't good to use a license to reach political reasons.</p><p>\nBut second, when you want to use patent protected software (like some parts of the TrueType interpreter of FreeType) and you want to buy such a license, it would be impossible to buy this from a company who distributes the software. This resctriction is stupid. It makes life more complicat without bringing any benefit against patents!</p>\ngpl3.licensingpatents.p3.s1\ntimi\nlogin to agree\n0\n2790\n\n\n<p style="display: inline">There is a very huge problem with this clause. Imagine what happens if Microsoft start sueing people all over the place, while simultaneously offering a very cheap deal under which they won't sue you. \n\nSmall companies may find they have no choice but to accept the deal, at which point they are suddenly banned from redistributing GPLv3 software. Thus this clause could allow a huge company to abuse the patent system in order to effectively kill small vendors ability to use GPLv3 software.</p><p>\nI don't see any easy solution to this problem. The only alternative seems to be to allow Microsoft - Novell like deals, which is unacceptable, but the other possibility could result in proprietary software vendor's trying to push companies into signing seemingly innocent deals without realizing they are giving up their right to use the GPLv3. </p>\ngpl3.licensingpatents.p3.s1\njring\nlogin to agree\n2\n2796\n\n\nThe GPL should stand proudly on its own with out need to cater to those who violated the spirit of the version 2 license prior to its obvious intent being SPELLED OUT in version 3.\ngpl3.licensingpatents.p3.s1\njohnrkro\nlogin to agree\n0\n2842\n\n\nThis should also include any third party which purchases Patents for the sole purpose of collecting royalities (Patent Trolls), any third party charged with licensing or collecting licensing fees and royalities for patents on behalf of the patent holder, or basically any third party seeking payments for patent rights.\ngpl3.licensingpatents.p3.s1\nghuber\nlogin to agree\n0\n2964\n\n\n<p style="display: inline">patent holders' leverage over small free software distributors.</p><p>\nThis provision was drafted to prevent a free software distributor from\nvoluntarily entering into an agreement with a patent holder in order to\nget a competitive advantage over other distributors. The unintended\neffect of Paragraphs 3 and 5 taken together, however, may be to take\noptions away from distributors who are threatened or sued by a patent\nholder. Such distributors - particularly small businesses and\nindividuals who have no intent to split the community - need to have the\nfreedom to negotiate a settlement that removes their liability and\npreserves their ability to continue to distribute free software.</p><p>\nConsider the situation in which a large software company sues a small\nfree software distributor for patent infringement. The distributor has\nonly four options that are compatible with paragraphs 3 and 5 of this draft:</p><p>\n(1) Fight the patent in court. This option is unlikely to be feasible\nfor a small distributor. Litigating a patent case to a judicial\nresolution can take years and cost millions of dollars in legal fees.\nMoreover, patents are presumed valid, and a favorable outcome is never\ncertain, so the distributor risks a significant adverse judgment by\nproceeding with the lawsuit.</p><p>\n(2) Negotiate a patent license for all downstream recipients. Putting\naside the fairness issues involved in forcing the first distributor sued\nto pay for whole community, this option is not likely to be feasible\neither. A patent holder who is willing to license the whole free\nsoftware community on affordable terms probably would not sue a small\ndistributor in the first place.</p><p>\n(3) Negotiate a patent license that does not involve a payment for the extent of the distributor's activity of conveying the work and then\nmake all of the work's source code publicly available free of charge. This option may be feasible in some instances and not in others.\nThe terms of the patent license, of course, are subject to the patent holder's agreement. A patent holder who is willing to accept a\nsmall per-unit royalty might demand a prohibitively large lump-sum payment to cover all possible future sales. Thus, not being able to\nnegotiate a license based on the extent of activity severely limits the settlement options available to the distributor.</p><p>\nAnd the safe harbor of the free, public distribution of the work's\nsource code may not always be available to a particular distributor\n(e.g., if the work was developed under NDA or other conditions of\nprivacy, if the work is used in a classified environment, or if no-cost\ndistribution of the whole work to non-customers would undermine the\ndistributor's business). Moreover, unless the distributor's license is\nsufficiently broad, the publication of source code could make the\ndistributor liable to the patent holder for indirect infringement.</p><p>\n(4) Pay the patent holder to settle the lawsuit and stop distributing\nthe free software. By default, this is the likely outcome of any patent\nlawsuit (or even threat of a patent lawsuit) against a small\ndistributor. That means that the patent holder may be able to shut down\ndistributors one by one until the work is not available at all.</p><p>\nWithout paragraphs 3 and 5, the distributor would have the option to\nnegotiate a license (regardless of its opinion of the validity of the\npatent) that covers itself and its direct customers - thereby staying in\nbusiness, continuing to distribute free software, and providing a\nbenefit to its customers in the form of protection from a patent lawsuit.</p><p>\nNothing in this draft prohibits a distributor from indemnifying its\ndirect customers against patent infringement or buying insurance for its\ncustomers. Why should negotiating a license (whether in response to a\nlawsuit or otherwise) be any different? And why shouldn't the option of\nproviding protection from patents be available to small distributors?</p><p>\nFurther, if this license makes settling a patent lawsuit infeasible for a small distributor, then indemnification and insurance will also be\nunaffordable for all but the largest distributors (because the distributor's options for negotiating a patent license for its indemnified\ncustomer who is sued for patent infringement will be limited as discussed above). And if indemnification and insurance are less practical,\nthe commercial use of free software may be reduced, as potential customers choose other solutions.</p><p>\nLike it or not, software patents exist and large software companies have\nthem. This license should not do anything to strengthen the patent\nholders' position. Rather, this license should continue to encourage\ndevelopers to create and distribute as much free software as possible -\nand give users the freedom to decide whether or not to use it based on\ntheir own assessment of the patent risk.</p><p>\nVery truly yours,\nBeth Mitchell\nCounsel\nCodeSourcery, Inc.\n650-331-3385 x703\n\n</p>\ngpl3.licensingpatents.p3.s1\nbeth1\nlogin to agree\n0\n3054\n\n\n<p style="display: inline">What here is not conveyed by the "if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you..." clause of GPLv2?</p><p>\nThis is directly addressed at a single deal, Novell/MSFT, which was ominous but did not restrict the freedoms of any user. The fact that MSFT then penned a similar deal with Xandros after March indicates that this clause is useless as the deterrent it is intended to be anyway.</p>\ngpl3.licensingpatents.p3.s1\nbk\nlogin to agree\n0\n3347\n\n\n", "2749:2777:2790:2796:2842:2964:3054:3347": "\n\n\n\n<p style="display: inline">By limiting this clause to agreements with third party software distributors the door may be opened for patent trolls (a la Rambus).</p><p>\nThis could permit a distributor to enter into a discriminatory agreement with an entity in the business of filing and licensing software patents but that refrains from distributing software. Potentially, a software distributor could transfer its patent rights to another entity that would be in the business of licensing those patents.</p><p>\nCan this be phrased so that the limitation applies to agreements covering software patents regardless of the third party's business? As technology evolves might it become difficult to distinguish software patents from other types of patents. It would be best to start off as inclusively as reasonable and loosen it up if legitimate issues arise.</p>\ngpl3.licensingpatents.p3.s1\npolymath\nlogin to agree\n1\n2749\n\n\nThis clause seems to be based on the general principle that no further restrictions can be added beyond the terms of this License. In the present case, it is stated that one cannot add further restrictions through restrictive patent licenses. Looks OK.\ngpl3.licensingpatents.p3.s1\nfrx\nlogin to agree\n0\n2777\n\n\n<p style="display: inline">I don't think that this is a good idea. First it isn't good to use a license to reach political reasons.</p><p>\nBut second, when you want to use patent protected software (like some parts of the TrueType interpreter of FreeType) and you want to buy such a license, it would be impossible to buy this from a company who distributes the software. This resctriction is stupid. It makes life more complicat without bringing any benefit against patents!</p>\ngpl3.licensingpatents.p3.s1\ntimi\nlogin to agree\n0\n2790\n\n\n<p style="display: inline">There is a very huge problem with this clause. Imagine what happens if Microsoft start sueing people all over the place, while simultaneously offering a very cheap deal under which they won't sue you. \n\nSmall companies may find they have no choice but to accept the deal, at which point they are suddenly banned from redistributing GPLv3 software. Thus this clause could allow a huge company to abuse the patent system in order to effectively kill small vendors ability to use GPLv3 software.</p><p>\nI don't see any easy solution to this problem. The only alternative seems to be to allow Microsoft - Novell like deals, which is unacceptable, but the other possibility could result in proprietary software vendor's trying to push companies into signing seemingly innocent deals without realizing they are giving up their right to use the GPLv3. </p>\ngpl3.licensingpatents.p3.s1\njring\nlogin to agree\n2\n2796\n\n\nThe GPL should stand proudly on its own with out need to cater to those who violated the spirit of the version 2 license prior to its obvious intent being SPELLED OUT in version 3.\ngpl3.licensingpatents.p3.s1\njohnrkro\nlogin to agree\n0\n2842\n\n\nThis should also include any third party which purchases Patents for the sole purpose of collecting royalities (Patent Trolls), any third party charged with licensing or collecting licensing fees and royalities for patents on behalf of the patent holder, or basically any third party seeking payments for patent rights.\ngpl3.licensingpatents.p3.s1\nghuber\nlogin to agree\n0\n2964\n\n\n<p style="display: inline">patent holders' leverage over small free software distributors.</p><p>\nThis provision was drafted to prevent a free software distributor from\nvoluntarily entering into an agreement with a patent holder in order to\nget a competitive advantage over other distributors. The unintended\neffect of Paragraphs 3 and 5 taken together, however, may be to take\noptions away from distributors who are threatened or sued by a patent\nholder. Such distributors - particularly small businesses and\nindividuals who have no intent to split the community - need to have the\nfreedom to negotiate a settlement that removes their liability and\npreserves their ability to continue to distribute free software.</p><p>\nConsider the situation in which a large software company sues a small\nfree software distributor for patent infringement. The distributor has\nonly four options that are compatible with paragraphs 3 and 5 of this draft:</p><p>\n(1) Fight the patent in court. This option is unlikely to be feasible\nfor a small distributor. Litigating a patent case to a judicial\nresolution can take years and cost millions of dollars in legal fees.\nMoreover, patents are presumed valid, and a favorable outcome is never\ncertain, so the distributor risks a significant adverse judgment by\nproceeding with the lawsuit.</p><p>\n(2) Negotiate a patent license for all downstream recipients. Putting\naside the fairness issues involved in forcing the first distributor sued\nto pay for whole community, this option is not likely to be feasible\neither. A patent holder who is willing to license the whole free\nsoftware community on affordable terms probably would not sue a small\ndistributor in the first place.</p><p>\n(3) Negotiate a patent license that does not involve a payment for the extent of the distributor's activity of conveying the work and then\nmake all of the work's source code publicly available free of charge. This option may be feasible in some instances and not in others.\nThe terms of the patent license, of course, are subject to the patent holder's agreement. A patent holder who is willing to accept a\nsmall per-unit royalty might demand a prohibitively large lump-sum payment to cover all possible future sales. Thus, not being able to\nnegotiate a license based on the extent of activity severely limits the settlement options available to the distributor.</p><p>\nAnd the safe harbor of the free, public distribution of the work's\nsource code may not always be available to a particular distributor\n(e.g., if the work was developed under NDA or other conditions of\nprivacy, if the work is used in a classified environment, or if no-cost\ndistribution of the whole work to non-customers would undermine the\ndistributor's business). Moreover, unless the distributor's license is\nsufficiently broad, the publication of source code could make the\ndistributor liable to the patent holder for indirect infringement.</p><p>\n(4) Pay the patent holder to settle the lawsuit and stop distributing\nthe free software. By default, this is the likely outcome of any patent\nlawsuit (or even threat of a patent lawsuit) against a small\ndistributor. That means that the patent holder may be able to shut down\ndistributors one by one until the work is not available at all.</p><p>\nWithout paragraphs 3 and 5, the distributor would have the option to\nnegotiate a license (regardless of its opinion of the validity of the\npatent) that covers itself and its direct customers - thereby staying in\nbusiness, continuing to distribute free software, and providing a\nbenefit to its customers in the form of protection from a patent lawsuit.</p><p>\nNothing in this draft prohibits a distributor from indemnifying its\ndirect customers against patent infringement or buying insurance for its\ncustomers. Why should negotiating a license (whether in response to a\nlawsuit or otherwise) be any different? And why shouldn't the option of\nproviding protection from patents be available to small distributors?</p><p>\nFurther, if this license makes settling a patent lawsuit infeasible for a small distributor, then indemnification and insurance will also be\nunaffordable for all but the largest distributors (because the distributor's options for negotiating a patent license for its indemnified\ncustomer who is sued for patent infringement will be limited as discussed above). And if indemnification and insurance are less practical,\nthe commercial use of free software may be reduced, as potential customers choose other solutions.</p><p>\nLike it or not, software patents exist and large software companies have\nthem. This license should not do anything to strengthen the patent\nholders' position. Rather, this license should continue to encourage\ndevelopers to create and distribute as much free software as possible -\nand give users the freedom to decide whether or not to use it based on\ntheir own assessment of the patent risk.</p><p>\nVery truly yours,\nBeth Mitchell\nCounsel\nCodeSourcery, Inc.\n650-331-3385 x703\n\n</p>\ngpl3.licensingpatents.p3.s1\nbeth1\nlogin to agree\n0\n3054\n\n\n<p style="display: inline">What here is not conveyed by the "if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you..." clause of GPLv2?</p><p>\nThis is directly addressed at a single deal, Novell/MSFT, which was ominous but did not restrict the freedoms of any user. The fact that MSFT then penned a similar deal with Xandros after March indicates that this clause is useless as the deterrent it is intended to be anyway.</p>\ngpl3.licensingpatents.p3.s1\nbk\nlogin to agree\n0\n3347\n\n\n", "2777:2796:2842:2964:3347": "\n\n\n\nThis clause seems to be based on the general principle that no further restrictions can be added beyond the terms of this License. In the present case, it is stated that one cannot add further restrictions through restrictive patent licenses. Looks OK.\ngpl3.licensingpatents.p3.s1\nfrx\nlogin to agree\n0\n2777\n\n\n<p style="display: inline">There is a very huge problem with this clause. Imagine what happens if Microsoft start sueing people all over the place, while simultaneously offering a very cheap deal under which they won't sue you. \n\nSmall companies may find they have no choice but to accept the deal, at which point they are suddenly banned from redistributing GPLv3 software. Thus this clause could allow a huge company to abuse the patent system in order to effectively kill small vendors ability to use GPLv3 software.</p><p>\nI don't see any easy solution to this problem. The only alternative seems to be to allow Microsoft - Novell like deals, which is unacceptable, but the other possibility could result in proprietary software vendor's trying to push companies into signing seemingly innocent deals without realizing they are giving up their right to use the GPLv3. </p>\ngpl3.licensingpatents.p3.s1\njring\nlogin to agree\n2\n2796\n\n\nThe GPL should stand proudly on its own with out need to cater to those who violated the spirit of the version 2 license prior to its obvious intent being SPELLED OUT in version 3.\ngpl3.licensingpatents.p3.s1\njohnrkro\nlogin to agree\n0\n2842\n\n\nThis should also include any third party which purchases Patents for the sole purpose of collecting royalities (Patent Trolls), any third party charged with licensing or collecting licensing fees and royalities for patents on behalf of the patent holder, or basically any third party seeking payments for patent rights.\ngpl3.licensingpatents.p3.s1\nghuber\nlogin to agree\n0\n2964\n\n\n<p style="display: inline">What here is not conveyed by the "if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you..." clause of GPLv2?</p><p>\nThis is directly addressed at a single deal, Novell/MSFT, which was ominous but did not restrict the freedoms of any user. The fact that MSFT then penned a similar deal with Xandros after March indicates that this clause is useless as the deterrent it is intended to be anyway.</p>\ngpl3.licensingpatents.p3.s1\nbk\nlogin to agree\n0\n3347\n\n\n", "2749:2777:2796:2842:2964:3347": "\n\n\n\n<p style="display: inline">By limiting this clause to agreements with third party software distributors the door may be opened for patent trolls (a la Rambus).</p><p>\nThis could permit a distributor to enter into a discriminatory agreement with an entity in the business of filing and licensing software patents but that refrains from distributing software. Potentially, a software distributor could transfer its patent rights to another entity that would be in the business of licensing those patents.</p><p>\nCan this be phrased so that the limitation applies to agreements covering software patents regardless of the third party's business? As technology evolves might it become difficult to distinguish software patents from other types of patents. It would be best to start off as inclusively as reasonable and loosen it up if legitimate issues arise.</p>\ngpl3.licensingpatents.p3.s1\npolymath\nlogin to agree\n1\n2749\n\n\nThis clause seems to be based on the general principle that no further restrictions can be added beyond the terms of this License. In the present case, it is stated that one cannot add further restrictions through restrictive patent licenses. Looks OK.\ngpl3.licensingpatents.p3.s1\nfrx\nlogin to agree\n0\n2777\n\n\n<p style="display: inline">There is a very huge problem with this clause. Imagine what happens if Microsoft start sueing people all over the place, while simultaneously offering a very cheap deal under which they won't sue you. \n\nSmall companies may find they have no choice but to accept the deal, at which point they are suddenly banned from redistributing GPLv3 software. Thus this clause could allow a huge company to abuse the patent system in order to effectively kill small vendors ability to use GPLv3 software.</p><p>\nI don't see any easy solution to this problem. The only alternative seems to be to allow Microsoft - Novell like deals, which is unacceptable, but the other possibility could result in proprietary software vendor's trying to push companies into signing seemingly innocent deals without realizing they are giving up their right to use the GPLv3. </p>\ngpl3.licensingpatents.p3.s1\njring\nlogin to agree\n2\n2796\n\n\nThe GPL should stand proudly on its own with out need to cater to those who violated the spirit of the version 2 license prior to its obvious intent being SPELLED OUT in version 3.\ngpl3.licensingpatents.p3.s1\njohnrkro\nlogin to agree\n0\n2842\n\n\nThis should also include any third party which purchases Patents for the sole purpose of collecting royalities (Patent Trolls), any third party charged with licensing or collecting licensing fees and royalities for patents on behalf of the patent holder, or basically any third party seeking payments for patent rights.\ngpl3.licensingpatents.p3.s1\nghuber\nlogin to agree\n0\n2964\n\n\n<p style="display: inline">What here is not conveyed by the "if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you..." clause of GPLv2?</p><p>\nThis is directly addressed at a single deal, Novell/MSFT, which was ominous but did not restrict the freedoms of any user. The fact that MSFT then penned a similar deal with Xandros after March indicates that this clause is useless as the deterrent it is intended to be anyway.</p>\ngpl3.licensingpatents.p3.s1\nbk\nlogin to agree\n0\n3347\n\n\n", "2460:2540:2555:2558:2576:2655:2725:2749:2777:2796:2842:2964:3143:3347": "\n\n\n\n<p style="display: inline">Removing these words would seem to make the license stronger with no adverse effects. The business of the third party doesn't appear relevant.</p><p>\nI'm guessing that the payment terms are there to only bring the clause into effect when the distributor is an active participant in the arrangement. However, it appears to open a loop-hole - what if the payment terms are NOT 'based on the extent of your activity of conveying the work' The payment terms seem to be so specific that they can be trivially worked around.</p>\ngpl3.licensingpatents.p3.s1\npjm\nlogin to agree\n15\n2460\n\n\nWhether the evil company that tries to make some GPL'd software non-free is in the bsuiness of distributing software is irrelevant. Imagine instead of the Microsoft-Novell deal a similar deal between a patent troll company (which could possibly be secretly funded by Microsoft) and a company which distributes GPL'd software commercially.\ngpl3.licensingpatents.p3.s1\nnb\nlogin to agree\n18\n2540\n\n\n<p style="display: inline">This section seems to very neatly cover the Novell/Microsoft deal without covering any variations to that deal, much less any deal with similar effects but different mechanics.</p><p>\nThe threat here is *any* patent license to some but not all of the users the a GPL covered work is distributed to.</p><p>\n"You may not convey a covered work if you are a party to an agreement with a third party that would give a patent license to some, but not all, recipients of that work. Further, you may not publicize a limited license nor may you use it as a sales differentiator, even if it is not the result of an explicit agreement."</p>\ngpl3.licensingpatents.p3.s1\nchandon\nlogin to agree\n6\n2555\n\n\nwhatever the business of the third party is, it should not be off the hook. What if the business is changed from distributing software to suing software vendors or supporters of free software?\nJust remove that portion as irrelevant, the entrance to a freedom-narrowing agreement should be more than enough.\ngpl3.licensingpatents.p3.s1\ngoldie\nlogin to agree\n7\n2558\n\n\nThis seems to be pointless, and makes this into a clause to catch exactly AND ONLY the thing done by MS and Novel. What if a hardware manufacturer offers a patent licence to force usage on their hardware? You cannot predict such things and the licence should be as general as possible.\ngpl3.licensingpatents.p3.s1\nrhouse\nlogin to agree\n8\n2576\n\n\n<p style="display: inline">This eliminates protection from patent trolls. Why was this sentence added? The rationale doesn't say. I cannot think of any kind of acceptable (to FSF) patent agreement that would be impossible if this sentence was removed.</p><p>\nThe only reason I can imagine is that removing this bit would make it easier for a patent troll to put a Free Software project out of business. The project would fold when it settles and makes a patent agreement or it looses the suit. Many projects do not have the money to successfully fight a troll's patent suit.</p><p>\nThen again, with this clause MS could do exactly the same to a free software project.</p>\ngpl3.licensingpatents.p3.s1\nmarechal\nlogin to agree\n4\n2655\n\n\nThis entire clause is at most useless, and at worst overly restrictive. It should be removed. GPLv3 should be as simple as possible, not as precise as possible.\ngpl3.licensingpatents.p3.s1\nflaschen\nlogin to agree\n2\n2725\n\n\n<p style="display: inline">By limiting this clause to agreements with third party software distributors the door may be opened for patent trolls (a la Rambus).</p><p>\nThis could permit a distributor to enter into a discriminatory agreement with an entity in the business of filing and licensing software patents but that refrains from distributing software. Potentially, a software distributor could transfer its patent rights to another entity that would be in the business of licensing those patents.</p><p>\nCan this be phrased so that the limitation applies to agreements covering software patents regardless of the third party's business? As technology evolves might it become difficult to distinguish software patents from other types of patents. It would be best to start off as inclusively as reasonable and loosen it up if legitimate issues arise.</p>\ngpl3.licensingpatents.p3.s1\npolymath\nlogin to agree\n1\n2749\n\n\nThis clause seems to be based on the general principle that no further restrictions can be added beyond the terms of this License. In the present case, it is stated that one cannot add further restrictions through restrictive patent licenses. Looks OK.\ngpl3.licensingpatents.p3.s1\nfrx\nlogin to agree\n0\n2777\n\n\n<p style="display: inline">There is a very huge problem with this clause. Imagine what happens if Microsoft start sueing people all over the place, while simultaneously offering a very cheap deal under which they won't sue you. \n\nSmall companies may find they have no choice but to accept the deal, at which point they are suddenly banned from redistributing GPLv3 software. Thus this clause could allow a huge company to abuse the patent system in order to effectively kill small vendors ability to use GPLv3 software.</p><p>\nI don't see any easy solution to this problem. The only alternative seems to be to allow Microsoft - Novell like deals, which is unacceptable, but the other possibility could result in proprietary software vendor's trying to push companies into signing seemingly innocent deals without realizing they are giving up their right to use the GPLv3. </p>\ngpl3.licensingpatents.p3.s1\njring\nlogin to agree\n2\n2796\n\n\nThe GPL should stand proudly on its own with out need to cater to those who violated the spirit of the version 2 license prior to its obvious intent being SPELLED OUT in version 3.\ngpl3.licensingpatents.p3.s1\njohnrkro\nlogin to agree\n0\n2842\n\n\nThis should also include any third party which purchases Patents for the sole purpose of collecting royalities (Patent Trolls), any third party charged with licensing or collecting licensing fees and royalities for patents on behalf of the patent holder, or basically any third party seeking payments for patent rights.\ngpl3.licensingpatents.p3.s1\nghuber\nlogin to agree\n0\n2964\n\n\n<p style="display: inline">\n\nAnd I don't need a reply, thanks.</p><p>\nIf you need more context, it's in comments under the Sontag article\nposted on groklaw Saturday 5/26.</p><p>\nThanks for all the work!</p><p>\nJon\n\n\n------------\n"Concern about this "Covenant Not to Sue" in relation to GPLv3\nAuthored by: map on Sunday, May 27 2007 @ 09:48 AM EDT\nWill SCO be the first to jump on the GPLv3 bandwagon.\nMy concern is that the Novell/MS loophole in the proposed GPLv3 may\nallow things not allowed in GPLv2.\nSpecifically GPLv2 section 2b:</p><p>\nb) You must cause any work that you distribute or publish, that in whole\nor in part contains or is derived from the Program or any part thereof,\nto be licensed as a whole at no charge to all third parties under the\nterms of this License.</p><p>\nThis would seem to me to imply that licenses for any infringed or\npotentially infringed IP be required for distribution of the whole, and\nsuch licenses have terms that are compliant with GPLv2.</p><p>\nBefore GPLv3 is ratified serious consideration should be given to how it\naffects various parties obligations under GPLv2 section 2b,4, and 6; and\nalso any potential impact it may have on the IBM/SCO/MS case.</p><p>\nUnnecessary language in the GPLv3 draft may be a problem such as section\n11 p5:</p><p>\n...that is in the business of distributing software...</p><p>\nI can almost guarantee you that MS is dumping IP into non-distributing\npuppet companies to get through that loophole.</p><p>\nIt never hurts to have an extra lawyer review a document, I hope FSF\nwill consult with IBM, et al. about this.</p><p>\nIANAL"\n\n\n</p>\ngpl3.licensingpatents.p3.s1\njons\nlogin to agree\n1\n3143\n\n\n<p style="display: inline">What here is not conveyed by the "if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you..." clause of GPLv2?</p><p>\nThis is directly addressed at a single deal, Novell/MSFT, which was ominous but did not restrict the freedoms of any user. The fact that MSFT then penned a similar deal with Xandros after March indicates that this clause is useless as the deterrent it is intended to be anyway.</p>\ngpl3.licensingpatents.p3.s1\nbk\nlogin to agree\n0\n3347\n\n\n", "2460:2540:2555:2558:2576:2655:2692:2725:2749:2777:2796:2842:2934:2964:3120:3143:3347": "\n\n\n\n<p style="display: inline">Removing these words would seem to make the license stronger with no adverse effects. The business of the third party doesn't appear relevant.</p><p>\nI'm guessing that the payment terms are there to only bring the clause into effect when the distributor is an active participant in the arrangement. However, it appears to open a loop-hole - what if the payment terms are NOT 'based on the extent of your activity of conveying the work' The payment terms seem to be so specific that they can be trivially worked around.</p>\ngpl3.licensingpatents.p3.s1\npjm\nlogin to agree\n15\n2460\n\n\nWhether the evil company that tries to make some GPL'd software non-free is in the bsuiness of distributing software is irrelevant. Imagine instead of the Microsoft-Novell deal a similar deal between a patent troll company (which could possibly be secretly funded by Microsoft) and a company which distributes GPL'd software commercially.\ngpl3.licensingpatents.p3.s1\nnb\nlogin to agree\n18\n2540\n\n\n<p style="display: inline">This section seems to very neatly cover the Novell/Microsoft deal without covering any variations to that deal, much less any deal with similar effects but different mechanics.</p><p>\nThe threat here is *any* patent license to some but not all of the users the a GPL covered work is distributed to.</p><p>\n"You may not convey a covered work if you are a party to an agreement with a third party that would give a patent license to some, but not all, recipients of that work. Further, you may not publicize a limited license nor may you use it as a sales differentiator, even if it is not the result of an explicit agreement."</p>\ngpl3.licensingpatents.p3.s1\nchandon\nlogin to agree\n6\n2555\n\n\nwhatever the business of the third party is, it should not be off the hook. What if the business is changed from distributing software to suing software vendors or supporters of free software?\nJust remove that portion as irrelevant, the entrance to a freedom-narrowing agreement should be more than enough.\ngpl3.licensingpatents.p3.s1\ngoldie\nlogin to agree\n7\n2558\n\n\nThis seems to be pointless, and makes this into a clause to catch exactly AND ONLY the thing done by MS and Novel. What if a hardware manufacturer offers a patent licence to force usage on their hardware? You cannot predict such things and the licence should be as general as possible.\ngpl3.licensingpatents.p3.s1\nrhouse\nlogin to agree\n8\n2576\n\n\n<p style="display: inline">This eliminates protection from patent trolls. Why was this sentence added? The rationale doesn't say. I cannot think of any kind of acceptable (to FSF) patent agreement that would be impossible if this sentence was removed.</p><p>\nThe only reason I can imagine is that removing this bit would make it easier for a patent troll to put a Free Software project out of business. The project would fold when it settles and makes a patent agreement or it looses the suit. Many projects do not have the money to successfully fight a troll's patent suit.</p><p>\nThen again, with this clause MS could do exactly the same to a free software project.</p>\ngpl3.licensingpatents.p3.s1\nmarechal\nlogin to agree\n4\n2655\n\n\n"a third party that is in the business of distributing software" ? What is to stop Microsoft or another company selling patents to a patent troll, and then using that company to do their dirty work for them ? It seems a little short-sighted to me.\n\ngpl3.licensingpatents.p3.s1\nsalsaman\nlogin to agree\n5\n2692\n\n\nThis entire clause is at most useless, and at worst overly restrictive. It should be removed. GPLv3 should be as simple as possible, not as precise as possible.\ngpl3.licensingpatents.p3.s1\nflaschen\nlogin to agree\n2\n2725\n\n\n<p style="display: inline">By limiting this clause to agreements with third party software distributors the door may be opened for patent trolls (a la Rambus).</p><p>\nThis could permit a distributor to enter into a discriminatory agreement with an entity in the business of filing and licensing software patents but that refrains from distributing software. Potentially, a software distributor could transfer its patent rights to another entity that would be in the business of licensing those patents.</p><p>\nCan this be phrased so that the limitation applies to agreements covering software patents regardless of the third party's business? As technology evolves might it become difficult to distinguish software patents from other types of patents. It would be best to start off as inclusively as reasonable and loosen it up if legitimate issues arise.</p>\ngpl3.licensingpatents.p3.s1\npolymath\nlogin to agree\n1\n2749\n\n\nThis clause seems to be based on the general principle that no further restrictions can be added beyond the terms of this License. In the present case, it is stated that one cannot add further restrictions through restrictive patent licenses. Looks OK.\ngpl3.licensingpatents.p3.s1\nfrx\nlogin to agree\n0\n2777\n\n\n<p style="display: inline">There is a very huge problem with this clause. Imagine what happens if Microsoft start sueing people all over the place, while simultaneously offering a very cheap deal under which they won't sue you. \n\nSmall companies may find they have no choice but to accept the deal, at which point they are suddenly banned from redistributing GPLv3 software. Thus this clause could allow a huge company to abuse the patent system in order to effectively kill small vendors ability to use GPLv3 software.</p><p>\nI don't see any easy solution to this problem. The only alternative seems to be to allow Microsoft - Novell like deals, which is unacceptable, but the other possibility could result in proprietary software vendor's trying to push companies into signing seemingly innocent deals without realizing they are giving up their right to use the GPLv3. </p>\ngpl3.licensingpatents.p3.s1\njring\nlogin to agree\n2\n2796\n\n\nThe GPL should stand proudly on its own with out need to cater to those who violated the spirit of the version 2 license prior to its obvious intent being SPELLED OUT in version 3.\ngpl3.licensingpatents.p3.s1\njohnrkro\nlogin to agree\n0\n2842\n\n\n<p style="display: inline">A cleaner way to resolve this is instead of distinguishing vs. the type of business, discriminate on the type of patent deal.</p><p>\nIt is much easier for the free software community to deal with the threat of a specific patent suit than the nebulous thread of a possible suit from a large portfolio.</p><p>\nA patent troll threatening litigation of a specific patent must identify the infringing code. There are a number of defenses available, such as finding prior art or re-writing the code to eliminate the patented technique.</p><p>\nHowever, no such defenses exist when the patent and the infringing code has not been identified with specificity. Thus, deals such as the Novell/Microsoft deal, where no specific litigation has been threatened, are actually more harmful than a single patent troll.</p><p>\nThus, I would propose the following rule: A "general" indemnification against an entire patent portfolio must be "generally" available to all users of the software. Against a specific patent threat where the patent and the infringing code have been identified, the distributor should be granted a bit more leeway.</p>\ngpl3.licensingpatents.p3.s1\ntalin\nlogin to agree\n0\n2934\n\n\nThis should also include any third party which purchases Patents for the sole purpose of collecting royalities (Patent Trolls), any third party charged with licensing or collecting licensing fees and royalities for patents on behalf of the patent holder, or basically any third party seeking payments for patent rights.\ngpl3.licensingpatents.p3.s1\nghuber\nlogin to agree\n0\n2964\n\n\n<p style="display: inline">distributing software"</p><p>\nThis clause seems to open up a lot of loop holes. Consider: A company that is\nin the business of selling medical equipment which happens to run software\nthat is licensed under the GPL, is not clearly in the business of\ndistributing software. (They may not even do so at all, say by pointing their\ncustomers to someone else who hosts the software specifically tailored for\ntheir system.) However they can still hold a clear interest in the software\nand could be involved in a lawsuit. This whole situation becomes a lot more\nconfusing if the work that is licensed under the GPL would not classically be\nconsidered "software".</p><p>\nAll in all this seems like an aria that is bound to be exploited if not worded\nvery carefully. Given that, I am not sure how much value it really adds.\nPerhaps the desired results can be achieved through rewording the clause\nitself rather than adding an exception.\n\n</p>\ngpl3.licensingpatents.p3.s1\ntkaitchuck\nlogin to agree\n1\n3120\n\n\n<p style="display: inline">\n\nAnd I don't need a reply, thanks.</p><p>\nIf you need more context, it's in comments under the Sontag article\nposted on groklaw Saturday 5/26.</p><p>\nThanks for all the work!</p><p>\nJon\n\n\n------------\n"Concern about this "Covenant Not to Sue" in relation to GPLv3\nAuthored by: map on Sunday, May 27 2007 @ 09:48 AM EDT\nWill SCO be the first to jump on the GPLv3 bandwagon.\nMy concern is that the Novell/MS loophole in the proposed GPLv3 may\nallow things not allowed in GPLv2.\nSpecifically GPLv2 section 2b:</p><p>\nb) You must cause any work that you distribute or publish, that in whole\nor in part contains or is derived from the Program or any part thereof,\nto be licensed as a whole at no charge to all third parties under the\nterms of this License.</p><p>\nThis would seem to me to imply that licenses for any infringed or\npotentially infringed IP be required for distribution of the whole, and\nsuch licenses have terms that are compliant with GPLv2.</p><p>\nBefore GPLv3 is ratified serious consideration should be given to how it\naffects various parties obligations under GPLv2 section 2b,4, and 6; and\nalso any potential impact it may have on the IBM/SCO/MS case.</p><p>\nUnnecessary language in the GPLv3 draft may be a problem such as section\n11 p5:</p><p>\n...that is in the business of distributing software...</p><p>\nI can almost guarantee you that MS is dumping IP into non-distributing\npuppet companies to get through that loophole.</p><p>\nIt never hurts to have an extra lawyer review a document, I hope FSF\nwill consult with IBM, et al. about this.</p><p>\nIANAL"\n\n\n</p>\ngpl3.licensingpatents.p3.s1\njons\nlogin to agree\n1\n3143\n\n\n<p style="display: inline">What here is not conveyed by the "if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you..." clause of GPLv2?</p><p>\nThis is directly addressed at a single deal, Novell/MSFT, which was ominous but did not restrict the freedoms of any user. The fact that MSFT then penned a similar deal with Xandros after March indicates that this clause is useless as the deterrent it is intended to be anyway.</p>\ngpl3.licensingpatents.p3.s1\nbk\nlogin to agree\n0\n3347\n\n\n", "2460:2555:2576:2725:2777:2796:2842:3347": "\n\n\n\n<p style="display: inline">Removing these words would seem to make the license stronger with no adverse effects. The business of the third party doesn't appear relevant.</p><p>\nI'm guessing that the payment terms are there to only bring the clause into effect when the distributor is an active participant in the arrangement. However, it appears to open a loop-hole - what if the payment terms are NOT 'based on the extent of your activity of conveying the work' The payment terms seem to be so specific that they can be trivially worked around.</p>\ngpl3.licensingpatents.p3.s1\npjm\nlogin to agree\n15\n2460\n\n\n<p style="display: inline">This section seems to very neatly cover the Novell/Microsoft deal without covering any variations to that deal, much less any deal with similar effects but different mechanics.</p><p>\nThe threat here is *any* patent license to some but not all of the users the a GPL covered work is distributed to.</p><p>\n"You may not convey a covered work if you are a party to an agreement with a third party that would give a patent license to some, but not all, recipients of that work. Further, you may not publicize a limited license nor may you use it as a sales differentiator, even if it is not the result of an explicit agreement."</p>\ngpl3.licensingpatents.p3.s1\nchandon\nlogin to agree\n6\n2555\n\n\nThis seems to be pointless, and makes this into a clause to catch exactly AND ONLY the thing done by MS and Novel. What if a hardware manufacturer offers a patent licence to force usage on their hardware? You cannot predict such things and the licence should be as general as possible.\ngpl3.licensingpatents.p3.s1\nrhouse\nlogin to agree\n8\n2576\n\n\nThis entire clause is at most useless, and at worst overly restrictive. It should be removed. GPLv3 should be as simple as possible, not as precise as possible.\ngpl3.licensingpatents.p3.s1\nflaschen\nlogin to agree\n2\n2725\n\n\nThis clause seems to be based on the general principle that no further restrictions can be added beyond the terms of this License. In the present case, it is stated that one cannot add further restrictions through restrictive patent licenses. Looks OK.\ngpl3.licensingpatents.p3.s1\nfrx\nlogin to agree\n0\n2777\n\n\n<p style="display: inline">There is a very huge problem with this clause. Imagine what happens if Microsoft start sueing people all over the place, while simultaneously offering a very cheap deal under which they won't sue you. \n\nSmall companies may find they have no choice but to accept the deal, at which point they are suddenly banned from redistributing GPLv3 software. Thus this clause could allow a huge company to abuse the patent system in order to effectively kill small vendors ability to use GPLv3 software.</p><p>\nI don't see any easy solution to this problem. The only alternative seems to be to allow Microsoft - Novell like deals, which is unacceptable, but the other possibility could result in proprietary software vendor's trying to push companies into signing seemingly innocent deals without realizing they are giving up their right to use the GPLv3. </p>\ngpl3.licensingpatents.p3.s1\njring\nlogin to agree\n2\n2796\n\n\nThe GPL should stand proudly on its own with out need to cater to those who violated the spirit of the version 2 license prior to its obvious intent being SPELLED OUT in version 3.\ngpl3.licensingpatents.p3.s1\njohnrkro\nlogin to agree\n0\n2842\n\n\n<p style="display: inline">What here is not conveyed by the "if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you..." clause of GPLv2?</p><p>\nThis is directly addressed at a single deal, Novell/MSFT, which was ominous but did not restrict the freedoms of any user. The fact that MSFT then penned a similar deal with Xandros after March indicates that this clause is useless as the deterrent it is intended to be anyway.</p>\ngpl3.licensingpatents.p3.s1\nbk\nlogin to agree\n0\n3347\n\n\n", "2460:2555:2562:2576:2725:2777:2785:2796:2842:2863:3347": "\n\n\n\n<p style="display: inline">Removing these words would seem to make the license stronger with no adverse effects. The business of the third party doesn't appear relevant.</p><p>\nI'm guessing that the payment terms are there to only bring the clause into effect when the distributor is an active participant in the arrangement. However, it appears to open a loop-hole - what if the payment terms are NOT 'based on the extent of your activity of conveying the work' The payment terms seem to be so specific that they can be trivially worked around.</p>\ngpl3.licensingpatents.p3.s1\npjm\nlogin to agree\n15\n2460\n\n\n<p style="display: inline">This section seems to very neatly cover the Novell/Microsoft deal without covering any variations to that deal, much less any deal with similar effects but different mechanics.</p><p>\nThe threat here is *any* patent license to some but not all of the users the a GPL covered work is distributed to.</p><p>\n"You may not convey a covered work if you are a party to an agreement with a third party that would give a patent license to some, but not all, recipients of that work. Further, you may not publicize a limited license nor may you use it as a sales differentiator, even if it is not the result of an explicit agreement."</p>\ngpl3.licensingpatents.p3.s1\nchandon\nlogin to agree\n6\n2555\n\n\nWhat if the payment is a one-time lump sum?\ngpl3.licensingpatents.p3.s1\nnovalis\nlogin to agree\n1\n2562\n\n\nThis seems to be pointless, and makes this into a clause to catch exactly AND ONLY the thing done by MS and Novel. What if a hardware manufacturer offers a patent licence to force usage on their hardware? You cannot predict such things and the licence should be as general as possible.\ngpl3.licensingpatents.p3.s1\nrhouse\nlogin to agree\n8\n2576\n\n\nThis entire clause is at most useless, and at worst overly restrictive. It should be removed. GPLv3 should be as simple as possible, not as precise as possible.\ngpl3.licensingpatents.p3.s1\nflaschen\nlogin to agree\n2\n2725\n\n\nThis clause seems to be based on the general principle that no further restrictions can be added beyond the terms of this License. In the present case, it is stated that one cannot add further restrictions through restrictive patent licenses. Looks OK.\ngpl3.licensingpatents.p3.s1\nfrx\nlogin to agree\n0\n2777\n\n\nAs I read this phrase, I cannot enter into a license agreement with a third party by which recipients of the covered work from me are indemnified against patent claims, if such agreement requires payments proportionate to the number of recipients, or number of copies distributed. However, if I simply make a lump sum payment for any number of recipients or any number of copies, then my payment is arguably not "based on the extent of your activity of conveying the work", and yet arguably such an agreement should still be prohibited by this agreement.\ngpl3.licensingpatents.p3.s1\nskissane\nlogin to agree\n0\n2785\n\n\n<p style="display: inline">There is a very huge problem with this clause. Imagine what happens if Microsoft start sueing people all over the place, while simultaneously offering a very cheap deal under which they won't sue you. \n\nSmall companies may find they have no choice but to accept the deal, at which point they are suddenly banned from redistributing GPLv3 software. Thus this clause could allow a huge company to abuse the patent system in order to effectively kill small vendors ability to use GPLv3 software.</p><p>\nI don't see any easy solution to this problem. The only alternative seems to be to allow Microsoft - Novell like deals, which is unacceptable, but the other possibility could result in proprietary software vendor's trying to push companies into signing seemingly innocent deals without realizing they are giving up their right to use the GPLv3. </p>\ngpl3.licensingpatents.p3.s1\njring\nlogin to agree\n2\n2796\n\n\nThe GPL should stand proudly on its own with out need to cater to those who violated the spirit of the version 2 license prior to its obvious intent being SPELLED OUT in version 3.\ngpl3.licensingpatents.p3.s1\njohnrkro\nlogin to agree\n0\n2842\n\n\nReplace the clause: "based on the extent of your activity of" with the word "for".\ngpl3.licensingpatents.p3.s1\njkoenig\nlogin to agree\n0\n2863\n\n\n<p style="display: inline">What here is not conveyed by the "if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you..." clause of GPLv2?</p><p>\nThis is directly addressed at a single deal, Novell/MSFT, which was ominous but did not restrict the freedoms of any user. The fact that MSFT then penned a similar deal with Xandros after March indicates that this clause is useless as the deterrent it is intended to be anyway.</p>\ngpl3.licensingpatents.p3.s1\nbk\nlogin to agree\n0\n3347\n\n\n", "2460:2555:2562:2576:2725:2777:2785:2796:2842:3347": "\n\n\n\n<p style="display: inline">Removing these words would seem to make the license stronger with no adverse effects. The business of the third party doesn't appear relevant.</p><p>\nI'm guessing that the payment terms are there to only bring the clause into effect when the distributor is an active participant in the arrangement. However, it appears to open a loop-hole - what if the payment terms are NOT 'based on the extent of your activity of conveying the work' The payment terms seem to be so specific that they can be trivially worked around.</p>\ngpl3.licensingpatents.p3.s1\npjm\nlogin to agree\n15\n2460\n\n\n<p style="display: inline">This section seems to very neatly cover the Novell/Microsoft deal without covering any variations to that deal, much less any deal with similar effects but different mechanics.</p><p>\nThe threat here is *any* patent license to some but not all of the users the a GPL covered work is distributed to.</p><p>\n"You may not convey a covered work if you are a party to an agreement with a third party that would give a patent license to some, but not all, recipients of that work. Further, you may not publicize a limited license nor may you use it as a sales differentiator, even if it is not the result of an explicit agreement."</p>\ngpl3.licensingpatents.p3.s1\nchandon\nlogin to agree\n6\n2555\n\n\nWhat if the payment is a one-time lump sum?\ngpl3.licensingpatents.p3.s1\nnovalis\nlogin to agree\n1\n2562\n\n\nThis seems to be pointless, and makes this into a clause to catch exactly AND ONLY the thing done by MS and Novel. What if a hardware manufacturer offers a patent licence to force usage on their hardware? You cannot predict such things and the licence should be as general as possible.\ngpl3.licensingpatents.p3.s1\nrhouse\nlogin to agree\n8\n2576\n\n\nThis entire clause is at most useless, and at worst overly restrictive. It should be removed. GPLv3 should be as simple as possible, not as precise as possible.\ngpl3.licensingpatents.p3.s1\nflaschen\nlogin to agree\n2\n2725\n\n\nThis clause seems to be based on the general principle that no further restrictions can be added beyond the terms of this License. In the present case, it is stated that one cannot add further restrictions through restrictive patent licenses. Looks OK.\ngpl3.licensingpatents.p3.s1\nfrx\nlogin to agree\n0\n2777\n\n\nAs I read this phrase, I cannot enter into a license agreement with a third party by which recipients of the covered work from me are indemnified against patent claims, if such agreement requires payments proportionate to the number of recipients, or number of copies distributed. However, if I simply make a lump sum payment for any number of recipients or any number of copies, then my payment is arguably not "based on the extent of your activity of conveying the work", and yet arguably such an agreement should still be prohibited by this agreement.\ngpl3.licensingpatents.p3.s1\nskissane\nlogin to agree\n0\n2785\n\n\n<p style="display: inline">There is a very huge problem with this clause. Imagine what happens if Microsoft start sueing people all over the place, while simultaneously offering a very cheap deal under which they won't sue you. \n\nSmall companies may find they have no choice but to accept the deal, at which point they are suddenly banned from redistributing GPLv3 software. Thus this clause could allow a huge company to abuse the patent system in order to effectively kill small vendors ability to use GPLv3 software.</p><p>\nI don't see any easy solution to this problem. The only alternative seems to be to allow Microsoft - Novell like deals, which is unacceptable, but the other possibility could result in proprietary software vendor's trying to push companies into signing seemingly innocent deals without realizing they are giving up their right to use the GPLv3. </p>\ngpl3.licensingpatents.p3.s1\njring\nlogin to agree\n2\n2796\n\n\nThe GPL should stand proudly on its own with out need to cater to those who violated the spirit of the version 2 license prior to its obvious intent being SPELLED OUT in version 3.\ngpl3.licensingpatents.p3.s1\njohnrkro\nlogin to agree\n0\n2842\n\n\n<p style="display: inline">What here is not conveyed by the "if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you..." clause of GPLv2?</p><p>\nThis is directly addressed at a single deal, Novell/MSFT, which was ominous but did not restrict the freedoms of any user. The fact that MSFT then penned a similar deal with Xandros after March indicates that this clause is useless as the deterrent it is intended to be anyway.</p>\ngpl3.licensingpatents.p3.s1\nbk\nlogin to agree\n0\n3347\n\n\n", "2777:2796:2842:3347": "\n\n\n\nThis clause seems to be based on the general principle that no further restrictions can be added beyond the terms of this License. In the present case, it is stated that one cannot add further restrictions through restrictive patent licenses. Looks OK.\ngpl3.licensingpatents.p3.s1\nfrx\nlogin to agree\n0\n2777\n\n\n<p style="display: inline">There is a very huge problem with this clause. Imagine what happens if Microsoft start sueing people all over the place, while simultaneously offering a very cheap deal under which they won't sue you. \n\nSmall companies may find they have no choice but to accept the deal, at which point they are suddenly banned from redistributing GPLv3 software. Thus this clause could allow a huge company to abuse the patent system in order to effectively kill small vendors ability to use GPLv3 software.</p><p>\nI don't see any easy solution to this problem. The only alternative seems to be to allow Microsoft - Novell like deals, which is unacceptable, but the other possibility could result in proprietary software vendor's trying to push companies into signing seemingly innocent deals without realizing they are giving up their right to use the GPLv3. </p>\ngpl3.licensingpatents.p3.s1\njring\nlogin to agree\n2\n2796\n\n\nThe GPL should stand proudly on its own with out need to cater to those who violated the spirit of the version 2 license prior to its obvious intent being SPELLED OUT in version 3.\ngpl3.licensingpatents.p3.s1\njohnrkro\nlogin to agree\n0\n2842\n\n\n<p style="display: inline">What here is not conveyed by the "if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you..." clause of GPLv2?</p><p>\nThis is directly addressed at a single deal, Novell/MSFT, which was ominous but did not restrict the freedoms of any user. The fact that MSFT then penned a similar deal with Xandros after March indicates that this clause is useless as the deterrent it is intended to be anyway.</p>\ngpl3.licensingpatents.p3.s1\nbk\nlogin to agree\n0\n3347\n\n\n", "2416:2429:2430:2515:2539:2563:2577:2702:2707:2708:2837:2873:3067:3146": "\n\n\n\nWhat do the square brackets mean in this license? Elsewhere it refers to section numbers from version 2, but it's entirely unclear what they mean here.\ngpl3.licensingpatents.p3.s1\njag\nlogin to agree\n0\n2416\n\n\nDoes this bracketed phrase mean that Novell and their agreement with Microsoft is grandfathered in?\ngpl3.licensingpatents.p3.s1\nfreestu\nlogin to agree\n1\n2429\n\n\nIs the purpose of this bracketed clause to grandfather the agreement between Novell and Microsoft? If so, I think its a bad idea.\ngpl3.licensingpatents.p3.s1\nfreestu\nlogin to agree\n21\n2430\n\n\nThis section is not only good for Novell. It lets all enterprises supporting the FSF know that they will not be blind-sighted by the community. There are many enterprises that have contributed much and pushed the cause forward in one way or another. Non-retroactive licensing protects their interests and encourages their support and confidence. Corporate confidence is important, no matter what we idealistic geeks may prefer. \ngpl3.licensingpatents.p3.s1\nquestin\nlogin to agree\n3\n2515\n\n\n<p style="display: inline">In some jurisdictions at least, there is no requirement for this kind of agreement to be in writing. Any two parties who wish to make an evil deal like the MS-Novell deal can simply claim having orally agreed on a broad agreement of this kind before whatever date is specified, and it is impossible to prove such a claim to be false.</p><p>\nThe text in brackets must be deleted for this part to have any teeth.</p>\ngpl3.licensingpatents.p3.s1\nnb\nlogin to agree\n8\n2539\n\n\nRemove: That [bracketed clause] will make it harder to claim prior art. First, given a date - anybody can say the arrangement happened before or after to their benefit. Second, there are granted patents even though prior art does exist. The prior art predates the patents by many years. This clause is bad when the assumption has been that the patent claimant has anything to do with software under GPL. Third, a company may still distribute patented software under the GPL with patents claimed by someone not even dealt with the software without knowledge of such patents. (The linked-list patents, for example) A licensee's only hope to protect against such patent claims is to reveal prior art. Even though it is not explicit, the bracketed clause creates an assumption that prior art doesn't exist under such arrangement because knowledge of such patents are assumed - even though it is not explicitly stated that way just look at the logic down the road of how it would be played out in court. Forth, it is best to keep the patent terms uniform and remove the bracketed clause. It will make sure that cases brought up about the GPL are uniform. The bracketed clause presents a greater legal headache for many individuals then fewer that it intends grandfather. Fifth, if a clause like this must be added - remove the date and specifically state the intent, likewise "unless the arrangement upholds the GPLv2 and includes in the arrangement a certification to not adopt any license being a successor to GPLv2." Notice how I used the word "certification" to prevent wrongful retroactivity.\ngpl3.licensingpatents.p3.s1\ndzonatas\nlogin to agree\n5\n2563\n\n\nLet's face it, the only such agreement in force is the MS Novell one; it was not a naive mistake, but a deliberate end run around abundantly clear language in GPL2, and constitutes deliberate misuse (whether legally recognised or not) of others' generously donated IP. It was done eyes wide open and now the penalty should be paid: newly licenced or new software under GPL3 should NOT be available for their commercial purposes. Let Novell struggle on with legacy software or find their own way out of the mess of their deliberate and unethical creation. Please DELETE THIS sunset clause- this is the most important change you should make.\ngpl3.licensingpatents.p3.s1\nrhouse\nlogin to agree\n7\n2577\n\n\nGive this a time fuse, make it "...prior to 3/28/07, in which case breach shall be excused until 4/28/08." This will allow time for the business entities to make other arrangements, changes to contract etc. Don't perpetually excuse this as it allows ill gotten gain to remain with Novell for as long as it lasts and provides them no incentive to change their pact with Microsoft.\ngpl3.licensingpatents.p3.s1\naaronmf\nlogin to agree\n1\n2702\n\n\n<p style="display: inline">In summary: BIG MISTAKE, DON'T DO IT.</p><p>\nI really can't say in words just how bad of an idea this is. The right thing to do is hang Microsoft (and unfortunately, Novell with them... this is what happens when you make a deal with the devil) out to dry. Leave them in the GPLv2 world, we will move on to the GPLv3; they will eventually figure out this is our game, played by our rules. This exception makes as much sense as giving Tivo a free pass, that is, no sense at all. Microsoft does not respect the users' freedoms (we know this from experience). As an adversary, Microsoft is not worthy of any leniency at all. How much leniency have they given us? Have they even released MS-DOS 1.0 under the GPL, much less one of the later versions like 5.0 or 6.22, or even earlier versions of Windows? If they have, I have not heard of it and it certainly would be all over the free software-related Web sites and IRC channels I check.</p><p>\nIf Microsoft chooses to join the free software community, they may choose to. This is not the way to go about encouraging them to do it; this is a gift to Microsoft, most undeserved, and Microsoft will do what Microsoft usually does: take the gift, say (either with words or actions) what amounts to "thanks, and by the way, #$%& you" and continue to attack us as before. I'd hate to see all the hard work of the free software movement thrown away because of something this silly.</p>\ngpl3.licensingpatents.p3.s1\nskquinn\nlogin to agree\n6\n2707\n\n\nIt should be treated like other "special permissions" above and beyond the boilerplate GPL (that you can then subsequently remove on a specific copy). Incorporating a specific date into a license is going to look idiotic 15 years from now, and it should be up to the individual software team whether they want to reward Novell and similar GPL-molesters by granting this permission.\ngpl3.licensingpatents.p3.s1\nhorsten\nlogin to agree\n2\n2708\n\n\n<p style="display: inline">Take out this clause. It introduces unnecessary uncertainty as to what rights the user of a GPLV3 licensed program actually has. Who knows what agreements are out there? People who choose to use GPLV3-ed software and programmers who choose to license their code under the GPLV3 deserve to know exactly what they're getting.</p><p>\nUsers of the GPLV2 are not being "blindsided" if this clause is removed from the GPLV3. A) The FSF has always been very clear as to what freedoms the GPL is intended to secure. The GPLV3, without this clause, is entirely consistent with the stated goals of the GPLV2 B) No rights are being taken away from any user of any software that's out there right now. Nobody ever guarantees that they will continue to forever license their code under whatever license is currently used. The portion of the "corporate world", whatever that is, that is complaining about a "change in license terms" is asking the FSF and it's supporters for something that they themselves would never agree to. Especially because the intent of the license is essentially unchanged from the GPLV2. If there was some legitimacy to the claims of being "blindsided" I could see at granting a grace period during which claims could not be pressed, but never an absolute cutoff date that says claims can never be pressed.</p><p>\nFor these reasons the uncertainty this clause introduces into the license is unnecessary. Regardless of the justification, such uncertainty is unwanted.</p>\ngpl3.licensingpatents.p3.s1\nkop\nlogin to agree\n3\n2837\n\n\nCould this not be further modified to include a sunset clause terminating the\nexception at some future date?\n7 years in the future is about the limit of most business plans while 20 years would allow relevant patents to expire in many jurisdictions. \ngpl3.licensingpatents.p3.s1\ndumain\nlogin to agree\n0\n2873\n\n\n<p style="display: inline">It breaks the license!</p><p>\nIt will allow products to exist that violate the terms: it would set false precedent for people setting out to do the very same thing; it will cause confusion in courts, possibly also then allowing products getting out violating this section; it will bolster a reputation of "heavyweight" bully-ish corporations that can 'get around' the limitations seen by others.</p><p>\nIt's no good.</p>\ngpl3.licensingpatents.p3.s1\nmrcoco\nlogin to agree\n0\n3067\n\n\nIn a recent Groklaw article SCO seems to be positioning to take advantage of this.\nPossible negative effects on the IBM/SCO/MS case.\ngpl3.licensingpatents.p3.s1\ndrcj\nlogin to agree\n0\n3146\n\n\n", "2416:2429:2430:2515:2539:2547:2563:2577:2702:2707:2708:2729:2778:2802:2837:2873:2956:3067:3138:3146:3150": "\n\n\n\nWhat do the square brackets mean in this license? Elsewhere it refers to section numbers from version 2, but it's entirely unclear what they mean here.\ngpl3.licensingpatents.p3.s1\njag\nlogin to agree\n0\n2416\n\n\nDoes this bracketed phrase mean that Novell and their agreement with Microsoft is grandfathered in?\ngpl3.licensingpatents.p3.s1\nfreestu\nlogin to agree\n1\n2429\n\n\nIs the purpose of this bracketed clause to grandfather the agreement between Novell and Microsoft? If so, I think its a bad idea.\ngpl3.licensingpatents.p3.s1\nfreestu\nlogin to agree\n21\n2430\n\n\nThis section is not only good for Novell. It lets all enterprises supporting the FSF know that they will not be blind-sighted by the community. There are many enterprises that have contributed much and pushed the cause forward in one way or another. Non-retroactive licensing protects their interests and encourages their support and confidence. Corporate confidence is important, no matter what we idealistic geeks may prefer. \ngpl3.licensingpatents.p3.s1\nquestin\nlogin to agree\n3\n2515\n\n\n<p style="display: inline">In some jurisdictions at least, there is no requirement for this kind of agreement to be in writing. Any two parties who wish to make an evil deal like the MS-Novell deal can simply claim having orally agreed on a broad agreement of this kind before whatever date is specified, and it is impossible to prove such a claim to be false.</p><p>\nThe text in brackets must be deleted for this part to have any teeth.</p>\ngpl3.licensingpatents.p3.s1\nnb\nlogin to agree\n8\n2539\n\n\n<p style="display: inline">Completely unnecissary. Microsoft and Novell will always be able to use their covenant on the software available at the time.</p><p>\nThere is no reason that software written 5 years from now should still have this liability.</p><p>\nPlease don't include this.</p>\ngpl3.licensingpatents.p3.s1\njalockli\nlogin to agree\n12\n2547\n\n\nRemove: That [bracketed clause] will make it harder to claim prior art. First, given a date - anybody can say the arrangement happened before or after to their benefit. Second, there are granted patents even though prior art does exist. The prior art predates the patents by many years. This clause is bad when the assumption has been that the patent claimant has anything to do with software under GPL. Third, a company may still distribute patented software under the GPL with patents claimed by someone not even dealt with the software without knowledge of such patents. (The linked-list patents, for example) A licensee's only hope to protect against such patent claims is to reveal prior art. Even though it is not explicit, the bracketed clause creates an assumption that prior art doesn't exist under such arrangement because knowledge of such patents are assumed - even though it is not explicitly stated that way just look at the logic down the road of how it would be played out in court. Forth, it is best to keep the patent terms uniform and remove the bracketed clause. It will make sure that cases brought up about the GPL are uniform. The bracketed clause presents a greater legal headache for many individuals then fewer that it intends grandfather. Fifth, if a clause like this must be added - remove the date and specifically state the intent, likewise "unless the arrangement upholds the GPLv2 and includes in the arrangement a certification to not adopt any license being a successor to GPLv2." Notice how I used the word "certification" to prevent wrongful retroactivity.\ngpl3.licensingpatents.p3.s1\ndzonatas\nlogin to agree\n5\n2563\n\n\nLet's face it, the only such agreement in force is the MS Novell one; it was not a naive mistake, but a deliberate end run around abundantly clear language in GPL2, and constitutes deliberate misuse (whether legally recognised or not) of others' generously donated IP. It was done eyes wide open and now the penalty should be paid: newly licenced or new software under GPL3 should NOT be available for their commercial purposes. Let Novell struggle on with legacy software or find their own way out of the mess of their deliberate and unethical creation. Please DELETE THIS sunset clause- this is the most important change you should make.\ngpl3.licensingpatents.p3.s1\nrhouse\nlogin to agree\n7\n2577\n\n\nGive this a time fuse, make it "...prior to 3/28/07, in which case breach shall be excused until 4/28/08." This will allow time for the business entities to make other arrangements, changes to contract etc. Don't perpetually excuse this as it allows ill gotten gain to remain with Novell for as long as it lasts and provides them no incentive to change their pact with Microsoft.\ngpl3.licensingpatents.p3.s1\naaronmf\nlogin to agree\n1\n2702\n\n\n<p style="display: inline">In summary: BIG MISTAKE, DON'T DO IT.</p><p>\nI really can't say in words just how bad of an idea this is. The right thing to do is hang Microsoft (and unfortunately, Novell with them... this is what happens when you make a deal with the devil) out to dry. Leave them in the GPLv2 world, we will move on to the GPLv3; they will eventually figure out this is our game, played by our rules. This exception makes as much sense as giving Tivo a free pass, that is, no sense at all. Microsoft does not respect the users' freedoms (we know this from experience). As an adversary, Microsoft is not worthy of any leniency at all. How much leniency have they given us? Have they even released MS-DOS 1.0 under the GPL, much less one of the later versions like 5.0 or 6.22, or even earlier versions of Windows? If they have, I have not heard of it and it certainly would be all over the free software-related Web sites and IRC channels I check.</p><p>\nIf Microsoft chooses to join the free software community, they may choose to. This is not the way to go about encouraging them to do it; this is a gift to Microsoft, most undeserved, and Microsoft will do what Microsoft usually does: take the gift, say (either with words or actions) what amounts to "thanks, and by the way, #$%& you" and continue to attack us as before. I'd hate to see all the hard work of the free software movement thrown away because of something this silly.</p>\ngpl3.licensingpatents.p3.s1\nskquinn\nlogin to agree\n6\n2707\n\n\nIt should be treated like other "special permissions" above and beyond the boilerplate GPL (that you can then subsequently remove on a specific copy). Incorporating a specific date into a license is going to look idiotic 15 years from now, and it should be up to the individual software team whether they want to reward Novell and similar GPL-molesters by granting this permission.\ngpl3.licensingpatents.p3.s1\nhorsten\nlogin to agree\n2\n2708\n\n\n<p style="display: inline">They knew what they were doing was dodgy. \nThey didn't consult the community. \nThey kept it from the community. \nThey did not care when the community complained. \nThey didn't even tell their own developers as the negotiations were going on.</p><p>\nThey made this mess, they can clean it up themselves.</p>\ngpl3.licensingpatents.p3.s1\njring\nlogin to agree\n3\n2729\n\n\n<p style="display: inline">Why this date? What's special about it? Is it just the date of publication of the third draft? Was it chosen in order to avoid placing retroactive restrictions, by chance? If this is the motivation, I don't think this date limit is needed: since the GNU GPLv3 is not yet released, no work has been licensed under its terms yet, and hence no provision can be retroactive...</p><p>\nThe draft rationale explains that this date limit is intended to avoid forcing companies to face problems due to past agreements that cannot be changed anymore. I'm not convinced that this would be unfair: agreements like those are bad things and hence making them incompatible with conveying GPLv3'd works does not seem to be unfair.</p>\ngpl3.licensingpatents.p3.s1\nfrx\nlogin to agree\n5\n2778\n\n\n<p style="display: inline">How does such a deal, entered into on March 27, 2007 violate freedom anymore than one entered into on April 1, 2007?</p><p>\nIf this act is a threat to freedom, it should be disallowed, in all ways, at all time. If it is acceptable to allow before that, then it should be acceptable after.</p><p>\nThis line should be removed.\n</p>\ngpl3.licensingpatents.p3.s1\nkdean06\nlogin to agree\n7\n2802\n\n\n<p style="display: inline">Take out this clause. It introduces unnecessary uncertainty as to what rights the user of a GPLV3 licensed program actually has. Who knows what agreements are out there? People who choose to use GPLV3-ed software and programmers who choose to license their code under the GPLV3 deserve to know exactly what they're getting.</p><p>\nUsers of the GPLV2 are not being "blindsided" if this clause is removed from the GPLV3. A) The FSF has always been very clear as to what freedoms the GPL is intended to secure. The GPLV3, without this clause, is entirely consistent with the stated goals of the GPLV2 B) No rights are being taken away from any user of any software that's out there right now. Nobody ever guarantees that they will continue to forever license their code under whatever license is currently used. The portion of the "corporate world", whatever that is, that is complaining about a "change in license terms" is asking the FSF and it's supporters for something that they themselves would never agree to. Especially because the intent of the license is essentially unchanged from the GPLV2. If there was some legitimacy to the claims of being "blindsided" I could see at granting a grace period during which claims could not be pressed, but never an absolute cutoff date that says claims can never be pressed.</p><p>\nFor these reasons the uncertainty this clause introduces into the license is unnecessary. Regardless of the justification, such uncertainty is unwanted.</p>\ngpl3.licensingpatents.p3.s1\nkop\nlogin to agree\n3\n2837\n\n\nCould this not be further modified to include a sunset clause terminating the\nexception at some future date?\n7 years in the future is about the limit of most business plans while 20 years would allow relevant patents to expire in many jurisdictions. \ngpl3.licensingpatents.p3.s1\ndumain\nlogin to agree\n0\n2873\n\n\n<p style="display: inline">I think this date is neither necessary nor useful.</p><p>\nOne reason for the date might be that you cannot cover agreements that have been made before the publication of a license. I cannot see a problem here; the worst thing that could happen would be that a court decides that the clause is invalid. But until then the clause is better without the date and although I am no lawyer, I think the clause would be valid in most jurisdictions without the date.</p><p>\nThe other argument for the date would be the fact that we should not take away software freedoms retroactively. But I do not think that this is a problem either. Novell (and any other party that made such agreements) can use all GPLv2-covered software forever. As far as new software or software that will be newly released under GPLv3 is concerned, Novell will simply have to deal with that.\nWhen I write a piece of software and release it under GPLv2, then there is never a guaranty that I will stick to that license. I only grant certain freedoms to users of specific versions of my software. I may well decide to release future versions under different licenses and I may even make it proprietary software. The only guaranty is that versions released under GPLv2 will stay free. There is no difference here, Novell can still distribute software covered by GPLv2 and if they want to use GPLv3-covered software, they need to agree to the new license.</p>\ngpl3.licensingpatents.p3.s1\nSoong\nlogin to agree\n1\n2956\n\n\n<p style="display: inline">It breaks the license!</p><p>\nIt will allow products to exist that violate the terms: it would set false precedent for people setting out to do the very same thing; it will cause confusion in courts, possibly also then allowing products getting out violating this section; it will bolster a reputation of "heavyweight" bully-ish corporations that can 'get around' the limitations seen by others.</p><p>\nIt's no good.</p>\ngpl3.licensingpatents.p3.s1\nmrcoco\nlogin to agree\n0\n3067\n\n\n<p style="display: inline">There should be no exception made for Novell. Just as there should be no exception to the Samsung & Microsoft patent cross-licensing agreement that has been signed on patents Microsoft claims have been illegally borrowed by the Linux operating system. I wonder how many other such agreements are out there like that, which we have yet to hear about.</p><p>\nThese corporations have attorneys and should have known the possible consequences of what they were agreeing to. I will be very disappointed if this exception remains as is.\n</p>\ngpl3.licensingpatents.p3.s1\ncodeblue\nlogin to agree\n1\n3138\n\n\nIn a recent Groklaw article SCO seems to be positioning to take advantage of this.\nPossible negative effects on the IBM/SCO/MS case.\ngpl3.licensingpatents.p3.s1\ndrcj\nlogin to agree\n0\n3146\n\n\nThis exclusion is such a BAD idea. There should be NO exclusions, especially for Novell and Microsoft, otherwise why bother?\ngpl3.licensingpatents.p3.s1\ntonythed\nlogin to agree\n0\n3150\n\n\n", "2416:2429:2430:2447:2495:2515:2532:2537:2539:2547:2563:2577:2647:2693:2702:2707:2708:2726:2729:2778:2802:2837:2838:2873:2956:3067:3078:3138:3146:3150": "\n\n\n\nWhat do the square brackets mean in this license? Elsewhere it refers to section numbers from version 2, but it's entirely unclear what they mean here.\ngpl3.licensingpatents.p3.s1\njag\nlogin to agree\n0\n2416\n\n\nDoes this bracketed phrase mean that Novell and their agreement with Microsoft is grandfathered in?\ngpl3.licensingpatents.p3.s1\nfreestu\nlogin to agree\n1\n2429\n\n\nIs the purpose of this bracketed clause to grandfather the agreement between Novell and Microsoft? If so, I think its a bad idea.\ngpl3.licensingpatents.p3.s1\nfreestu\nlogin to agree\n21\n2430\n\n\nLet's see how many people will come along showing agreements allegedly signed prior to that date.\ngpl3.licensingpatents.p3.s1\nibor\nlogin to agree\n3\n2447\n\n\nLook... it's not even about Novell. It's the fact that Novell did something with Microsoft. However, there could already exist business dealing involving patents that do not involve the high profile Novell-Microsoft pact. With that said, I have no idea if this statement belongs in the document though. After all, we're talking about GPLv3... no deals have happened under GPLv3.\ngpl3.licensingpatents.p3.s1\ncjcox\nlogin to agree\n0\n2495\n\n\nThis section is not only good for Novell. It lets all enterprises supporting the FSF know that they will not be blind-sighted by the community. There are many enterprises that have contributed much and pushed the cause forward in one way or another. Non-retroactive licensing protects their interests and encourages their support and confidence. Corporate confidence is important, no matter what we idealistic geeks may prefer. \ngpl3.licensingpatents.p3.s1\nquestin\nlogin to agree\n3\n2515\n\n\n<p style="display: inline">I do not think we should give Novell a get-out-of-jail free card.</p><p>\nThey clearly and deliberatly violated the principles of free software, even though the GPLv2 didn't explicitly forbid it.</p>\ngpl3.licensingpatents.p3.s1\nnocturn\nlogin to agree\n9\n2532\n\n\nno date-based exceptions please.\ngpl3.licensingpatents.p3.s1\nhkbst\nlogin to agree\n10\n2537\n\n\n<p style="display: inline">In some jurisdictions at least, there is no requirement for this kind of agreement to be in writing. Any two parties who wish to make an evil deal like the MS-Novell deal can simply claim having orally agreed on a broad agreement of this kind before whatever date is specified, and it is impossible to prove such a claim to be false.</p><p>\nThe text in brackets must be deleted for this part to have any teeth.</p>\ngpl3.licensingpatents.p3.s1\nnb\nlogin to agree\n8\n2539\n\n\n<p style="display: inline">Completely unnecissary. Microsoft and Novell will always be able to use their covenant on the software available at the time.</p><p>\nThere is no reason that software written 5 years from now should still have this liability.</p><p>\nPlease don't include this.</p>\ngpl3.licensingpatents.p3.s1\njalockli\nlogin to agree\n12\n2547\n\n\nRemove: That [bracketed clause] will make it harder to claim prior art. First, given a date - anybody can say the arrangement happened before or after to their benefit. Second, there are granted patents even though prior art does exist. The prior art predates the patents by many years. This clause is bad when the assumption has been that the patent claimant has anything to do with software under GPL. Third, a company may still distribute patented software under the GPL with patents claimed by someone not even dealt with the software without knowledge of such patents. (The linked-list patents, for example) A licensee's only hope to protect against such patent claims is to reveal prior art. Even though it is not explicit, the bracketed clause creates an assumption that prior art doesn't exist under such arrangement because knowledge of such patents are assumed - even though it is not explicitly stated that way just look at the logic down the road of how it would be played out in court. Forth, it is best to keep the patent terms uniform and remove the bracketed clause. It will make sure that cases brought up about the GPL are uniform. The bracketed clause presents a greater legal headache for many individuals then fewer that it intends grandfather. Fifth, if a clause like this must be added - remove the date and specifically state the intent, likewise "unless the arrangement upholds the GPLv2 and includes in the arrangement a certification to not adopt any license being a successor to GPLv2." Notice how I used the word "certification" to prevent wrongful retroactivity.\ngpl3.licensingpatents.p3.s1\ndzonatas\nlogin to agree\n5\n2563\n\n\nLet's face it, the only such agreement in force is the MS Novell one; it was not a naive mistake, but a deliberate end run around abundantly clear language in GPL2, and constitutes deliberate misuse (whether legally recognised or not) of others' generously donated IP. It was done eyes wide open and now the penalty should be paid: newly licenced or new software under GPL3 should NOT be available for their commercial purposes. Let Novell struggle on with legacy software or find their own way out of the mess of their deliberate and unethical creation. Please DELETE THIS sunset clause- this is the most important change you should make.\ngpl3.licensingpatents.p3.s1\nrhouse\nlogin to agree\n7\n2577\n\n\n<p style="display: inline">The license should protect software, it is failing to protect users freedoms if you leave this in. Worse still an added clause protecting such freedoms would be "an additional restriction"</p><p>\nInstead let users add such a prior to ... as an optional extra if they wish, then they will be giving extra permissions and the balance is correct\n</p>\ngpl3.licensingpatents.p3.s1\nAlanCox\nlogin to agree\n3\n2647\n\n\nDid Novell's lawyers suggest this wording ?\ngpl3.licensingpatents.p3.s1\nsalsaman\nlogin to agree\n3\n2693\n\n\nGive this a time fuse, make it "...prior to 3/28/07, in which case breach shall be excused until 4/28/08." This will allow time for the business entities to make other arrangements, changes to contract etc. Don't perpetually excuse this as it allows ill gotten gain to remain with Novell for as long as it lasts and provides them no incentive to change their pact with Microsoft.\ngpl3.licensingpatents.p3.s1\naaronmf\nlogin to agree\n1\n2702\n\n\n<p style="display: inline">In summary: BIG MISTAKE, DON'T DO IT.</p><p>\nI really can't say in words just how bad of an idea this is. The right thing to do is hang Microsoft (and unfortunately, Novell with them... this is what happens when you make a deal with the devil) out to dry. Leave them in the GPLv2 world, we will move on to the GPLv3; they will eventually figure out this is our game, played by our rules. This exception makes as much sense as giving Tivo a free pass, that is, no sense at all. Microsoft does not respect the users' freedoms (we know this from experience). As an adversary, Microsoft is not worthy of any leniency at all. How much leniency have they given us? Have they even released MS-DOS 1.0 under the GPL, much less one of the later versions like 5.0 or 6.22, or even earlier versions of Windows? If they have, I have not heard of it and it certainly would be all over the free software-related Web sites and IRC channels I check.</p><p>\nIf Microsoft chooses to join the free software community, they may choose to. This is not the way to go about encouraging them to do it; this is a gift to Microsoft, most undeserved, and Microsoft will do what Microsoft usually does: take the gift, say (either with words or actions) what amounts to "thanks, and by the way, #$%& you" and continue to attack us as before. I'd hate to see all the hard work of the free software movement thrown away because of something this silly.</p>\ngpl3.licensingpatents.p3.s1\nskquinn\nlogin to agree\n6\n2707\n\n\nIt should be treated like other "special permissions" above and beyond the boilerplate GPL (that you can then subsequently remove on a specific copy). Incorporating a specific date into a license is going to look idiotic 15 years from now, and it should be up to the individual software team whether they want to reward Novell and similar GPL-molesters by granting this permission.\ngpl3.licensingpatents.p3.s1\nhorsten\nlogin to agree\n2\n2708\n\n\n<p style="display: inline">I basically agree with all the previous criticisms. I want to specifically note:</p><p>\na. Novell already has permanent GPLv2 rights, and they don't "deserve" anything special for GPLv3.</p><p>\nb. Any date makes the license even more over-specific and awkward than it already is.</p><p>\nc. It implies Novell was innocent and didn't know they were being unethical. In fact, they did, and even thought the legal case was uncertain enough to consult with SFLC.</p><p>\nd. It is unfair to others who want to abuse the GPL. Novell shouldn't get preference because they did it first.</p>\ngpl3.licensingpatents.p3.s1\nflaschen\nlogin to agree\n4\n2726\n\n\n<p style="display: inline">They knew what they were doing was dodgy. \nThey didn't consult the community. \nThey kept it from the community. \nThey did not care when the community complained. \nThey didn't even tell their own developers as the negotiations were going on.</p><p>\nThey made this mess, they can clean it up themselves.</p>\ngpl3.licensingpatents.p3.s1\njring\nlogin to agree\n3\n2729\n\n\n<p style="display: inline">Why this date? What's special about it? Is it just the date of publication of the third draft? Was it chosen in order to avoid placing retroactive restrictions, by chance? If this is the motivation, I don't think this date limit is needed: since the GNU GPLv3 is not yet released, no work has been licensed under its terms yet, and hence no provision can be retroactive...</p><p>\nThe draft rationale explains that this date limit is intended to avoid forcing companies to face problems due to past agreements that cannot be changed anymore. I'm not convinced that this would be unfair: agreements like those are bad things and hence making them incompatible with conveying GPLv3'd works does not seem to be unfair.</p>\ngpl3.licensingpatents.p3.s1\nfrx\nlogin to agree\n5\n2778\n\n\n<p style="display: inline">How does such a deal, entered into on March 27, 2007 violate freedom anymore than one entered into on April 1, 2007?</p><p>\nIf this act is a threat to freedom, it should be disallowed, in all ways, at all time. If it is acceptable to allow before that, then it should be acceptable after.</p><p>\nThis line should be removed.\n</p>\ngpl3.licensingpatents.p3.s1\nkdean06\nlogin to agree\n7\n2802\n\n\n<p style="display: inline">Take out this clause. It introduces unnecessary uncertainty as to what rights the user of a GPLV3 licensed program actually has. Who knows what agreements are out there? People who choose to use GPLV3-ed software and programmers who choose to license their code under the GPLV3 deserve to know exactly what they're getting.</p><p>\nUsers of the GPLV2 are not being "blindsided" if this clause is removed from the GPLV3. A) The FSF has always been very clear as to what freedoms the GPL is intended to secure. The GPLV3, without this clause, is entirely consistent with the stated goals of the GPLV2 B) No rights are being taken away from any user of any software that's out there right now. Nobody ever guarantees that they will continue to forever license their code under whatever license is currently used. The portion of the "corporate world", whatever that is, that is complaining about a "change in license terms" is asking the FSF and it's supporters for something that they themselves would never agree to. Especially because the intent of the license is essentially unchanged from the GPLV2. If there was some legitimacy to the claims of being "blindsided" I could see at granting a grace period during which claims could not be pressed, but never an absolute cutoff date that says claims can never be pressed.</p><p>\nFor these reasons the uncertainty this clause introduces into the license is unnecessary. Regardless of the justification, such uncertainty is unwanted.</p>\ngpl3.licensingpatents.p3.s1\nkop\nlogin to agree\n3\n2837\n\n\n<p style="display: inline">Why not change this to January 1, 2005? Or even earlier.</p><p>\nI'm not sure why this should be allowed at all, but I REALLY don't think that the GPL v3 should sanction the MS/Novell "deal".\n</p>\ngpl3.licensingpatents.p3.s1\ncharlesh\nlogin to agree\n0\n2838\n\n\nCould this not be further modified to include a sunset clause terminating the\nexception at some future date?\n7 years in the future is about the limit of most business plans while 20 years would allow relevant patents to expire in many jurisdictions. \ngpl3.licensingpatents.p3.s1\ndumain\nlogin to agree\n0\n2873\n\n\n<p style="display: inline">I think this date is neither necessary nor useful.</p><p>\nOne reason for the date might be that you cannot cover agreements that have been made before the publication of a license. I cannot see a problem here; the worst thing that could happen would be that a court decides that the clause is invalid. But until then the clause is better without the date and although I am no lawyer, I think the clause would be valid in most jurisdictions without the date.</p><p>\nThe other argument for the date would be the fact that we should not take away software freedoms retroactively. But I do not think that this is a problem either. Novell (and any other party that made such agreements) can use all GPLv2-covered software forever. As far as new software or software that will be newly released under GPLv3 is concerned, Novell will simply have to deal with that.\nWhen I write a piece of software and release it under GPLv2, then there is never a guaranty that I will stick to that license. I only grant certain freedoms to users of specific versions of my software. I may well decide to release future versions under different licenses and I may even make it proprietary software. The only guaranty is that versions released under GPLv2 will stay free. There is no difference here, Novell can still distribute software covered by GPLv2 and if they want to use GPLv3-covered software, they need to agree to the new license.</p>\ngpl3.licensingpatents.p3.s1\nSoong\nlogin to agree\n1\n2956\n\n\n<p style="display: inline">It breaks the license!</p><p>\nIt will allow products to exist that violate the terms: it would set false precedent for people setting out to do the very same thing; it will cause confusion in courts, possibly also then allowing products getting out violating this section; it will bolster a reputation of "heavyweight" bully-ish corporations that can 'get around' the limitations seen by others.</p><p>\nIt's no good.</p>\ngpl3.licensingpatents.p3.s1\nmrcoco\nlogin to agree\n0\n3067\n\n\n<p style="display: inline">It makes GPL 3 uncertain.</p><p>\nIBM SUN... All big companies have agreements between each other before that date. Remove it. If its not patent clear it should not be in the source code other wise we could have people turning around when using parts in another gpl3 project that there patents don't cover that project so we have to pay up. Loop holes are a bad thing when they will all ready be exploited and could cause trouble.</p>\ngpl3.licensingpatents.p3.s1\noiaohm\nlogin to agree\n0\n3078\n\n\n<p style="display: inline">There should be no exception made for Novell. Just as there should be no exception to the Samsung & Microsoft patent cross-licensing agreement that has been signed on patents Microsoft claims have been illegally borrowed by the Linux operating system. I wonder how many other such agreements are out there like that, which we have yet to hear about.</p><p>\nThese corporations have attorneys and should have known the possible consequences of what they were agreeing to. I will be very disappointed if this exception remains as is.\n</p>\ngpl3.licensingpatents.p3.s1\ncodeblue\nlogin to agree\n1\n3138\n\n\nIn a recent Groklaw article SCO seems to be positioning to take advantage of this.\nPossible negative effects on the IBM/SCO/MS case.\ngpl3.licensingpatents.p3.s1\ndrcj\nlogin to agree\n0\n3146\n\n\nThis exclusion is such a BAD idea. There should be NO exclusions, especially for Novell and Microsoft, otherwise why bother?\ngpl3.licensingpatents.p3.s1\ntonythed\nlogin to agree\n0\n3150\n\n\n", "2416:2429:2430:2447:2495:2515:2532:2537:2539:2547:2563:2577:2647:2693:2702:2707:2708:2726:2729:2778:2802:2837:2838:2873:2914:2956:3067:3078:3138:3146:3150": "\n\n\n\nWhat do the square brackets mean in this license? Elsewhere it refers to section numbers from version 2, but it's entirely unclear what they mean here.\ngpl3.licensingpatents.p3.s1\njag\nlogin to agree\n0\n2416\n\n\nDoes this bracketed phrase mean that Novell and their agreement with Microsoft is grandfathered in?\ngpl3.licensingpatents.p3.s1\nfreestu\nlogin to agree\n1\n2429\n\n\nIs the purpose of this bracketed clause to grandfather the agreement between Novell and Microsoft? If so, I think its a bad idea.\ngpl3.licensingpatents.p3.s1\nfreestu\nlogin to agree\n21\n2430\n\n\nLet's see how many people will come along showing agreements allegedly signed prior to that date.\ngpl3.licensingpatents.p3.s1\nibor\nlogin to agree\n3\n2447\n\n\nLook... it's not even about Novell. It's the fact that Novell did something with Microsoft. However, there could already exist business dealing involving patents that do not involve the high profile Novell-Microsoft pact. With that said, I have no idea if this statement belongs in the document though. After all, we're talking about GPLv3... no deals have happened under GPLv3.\ngpl3.licensingpatents.p3.s1\ncjcox\nlogin to agree\n0\n2495\n\n\nThis section is not only good for Novell. It lets all enterprises supporting the FSF know that they will not be blind-sighted by the community. There are many enterprises that have contributed much and pushed the cause forward in one way or another. Non-retroactive licensing protects their interests and encourages their support and confidence. Corporate confidence is important, no matter what we idealistic geeks may prefer. \ngpl3.licensingpatents.p3.s1\nquestin\nlogin to agree\n3\n2515\n\n\n<p style="display: inline">I do not think we should give Novell a get-out-of-jail free card.</p><p>\nThey clearly and deliberatly violated the principles of free software, even though the GPLv2 didn't explicitly forbid it.</p>\ngpl3.licensingpatents.p3.s1\nnocturn\nlogin to agree\n9\n2532\n\n\nno date-based exceptions please.\ngpl3.licensingpatents.p3.s1\nhkbst\nlogin to agree\n10\n2537\n\n\n<p style="display: inline">In some jurisdictions at least, there is no requirement for this kind of agreement to be in writing. Any two parties who wish to make an evil deal like the MS-Novell deal can simply claim having orally agreed on a broad agreement of this kind before whatever date is specified, and it is impossible to prove such a claim to be false.</p><p>\nThe text in brackets must be deleted for this part to have any teeth.</p>\ngpl3.licensingpatents.p3.s1\nnb\nlogin to agree\n8\n2539\n\n\n<p style="display: inline">Completely unnecissary. Microsoft and Novell will always be able to use their covenant on the software available at the time.</p><p>\nThere is no reason that software written 5 years from now should still have this liability.</p><p>\nPlease don't include this.</p>\ngpl3.licensingpatents.p3.s1\njalockli\nlogin to agree\n12\n2547\n\n\nRemove: That [bracketed clause] will make it harder to claim prior art. First, given a date - anybody can say the arrangement happened before or after to their benefit. Second, there are granted patents even though prior art does exist. The prior art predates the patents by many years. This clause is bad when the assumption has been that the patent claimant has anything to do with software under GPL. Third, a company may still distribute patented software under the GPL with patents claimed by someone not even dealt with the software without knowledge of such patents. (The linked-list patents, for example) A licensee's only hope to protect against such patent claims is to reveal prior art. Even though it is not explicit, the bracketed clause creates an assumption that prior art doesn't exist under such arrangement because knowledge of such patents are assumed - even though it is not explicitly stated that way just look at the logic down the road of how it would be played out in court. Forth, it is best to keep the patent terms uniform and remove the bracketed clause. It will make sure that cases brought up about the GPL are uniform. The bracketed clause presents a greater legal headache for many individuals then fewer that it intends grandfather. Fifth, if a clause like this must be added - remove the date and specifically state the intent, likewise "unless the arrangement upholds the GPLv2 and includes in the arrangement a certification to not adopt any license being a successor to GPLv2." Notice how I used the word "certification" to prevent wrongful retroactivity.\ngpl3.licensingpatents.p3.s1\ndzonatas\nlogin to agree\n5\n2563\n\n\nLet's face it, the only such agreement in force is the MS Novell one; it was not a naive mistake, but a deliberate end run around abundantly clear language in GPL2, and constitutes deliberate misuse (whether legally recognised or not) of others' generously donated IP. It was done eyes wide open and now the penalty should be paid: newly licenced or new software under GPL3 should NOT be available for their commercial purposes. Let Novell struggle on with legacy software or find their own way out of the mess of their deliberate and unethical creation. Please DELETE THIS sunset clause- this is the most important change you should make.\ngpl3.licensingpatents.p3.s1\nrhouse\nlogin to agree\n7\n2577\n\n\n<p style="display: inline">The license should protect software, it is failing to protect users freedoms if you leave this in. Worse still an added clause protecting such freedoms would be "an additional restriction"</p><p>\nInstead let users add such a prior to ... as an optional extra if they wish, then they will be giving extra permissions and the balance is correct\n</p>\ngpl3.licensingpatents.p3.s1\nAlanCox\nlogin to agree\n3\n2647\n\n\nDid Novell's lawyers suggest this wording ?\ngpl3.licensingpatents.p3.s1\nsalsaman\nlogin to agree\n3\n2693\n\n\nGive this a time fuse, make it "...prior to 3/28/07, in which case breach shall be excused until 4/28/08." This will allow time for the business entities to make other arrangements, changes to contract etc. Don't perpetually excuse this as it allows ill gotten gain to remain with Novell for as long as it lasts and provides them no incentive to change their pact with Microsoft.\ngpl3.licensingpatents.p3.s1\naaronmf\nlogin to agree\n1\n2702\n\n\n<p style="display: inline">In summary: BIG MISTAKE, DON'T DO IT.</p><p>\nI really can't say in words just how bad of an idea this is. The right thing to do is hang Microsoft (and unfortunately, Novell with them... this is what happens when you make a deal with the devil) out to dry. Leave them in the GPLv2 world, we will move on to the GPLv3; they will eventually figure out this is our game, played by our rules. This exception makes as much sense as giving Tivo a free pass, that is, no sense at all. Microsoft does not respect the users' freedoms (we know this from experience). As an adversary, Microsoft is not worthy of any leniency at all. How much leniency have they given us? Have they even released MS-DOS 1.0 under the GPL, much less one of the later versions like 5.0 or 6.22, or even earlier versions of Windows? If they have, I have not heard of it and it certainly would be all over the free software-related Web sites and IRC channels I check.</p><p>\nIf Microsoft chooses to join the free software community, they may choose to. This is not the way to go about encouraging them to do it; this is a gift to Microsoft, most undeserved, and Microsoft will do what Microsoft usually does: take the gift, say (either with words or actions) what amounts to "thanks, and by the way, #$%& you" and continue to attack us as before. I'd hate to see all the hard work of the free software movement thrown away because of something this silly.</p>\ngpl3.licensingpatents.p3.s1\nskquinn\nlogin to agree\n6\n2707\n\n\nIt should be treated like other "special permissions" above and beyond the boilerplate GPL (that you can then subsequently remove on a specific copy). Incorporating a specific date into a license is going to look idiotic 15 years from now, and it should be up to the individual software team whether they want to reward Novell and similar GPL-molesters by granting this permission.\ngpl3.licensingpatents.p3.s1\nhorsten\nlogin to agree\n2\n2708\n\n\n<p style="display: inline">I basically agree with all the previous criticisms. I want to specifically note:</p><p>\na. Novell already has permanent GPLv2 rights, and they don't "deserve" anything special for GPLv3.</p><p>\nb. Any date makes the license even more over-specific and awkward than it already is.</p><p>\nc. It implies Novell was innocent and didn't know they were being unethical. In fact, they did, and even thought the legal case was uncertain enough to consult with SFLC.</p><p>\nd. It is unfair to others who want to abuse the GPL. Novell shouldn't get preference because they did it first.</p>\ngpl3.licensingpatents.p3.s1\nflaschen\nlogin to agree\n4\n2726\n\n\n<p style="display: inline">They knew what they were doing was dodgy. \nThey didn't consult the community. \nThey kept it from the community. \nThey did not care when the community complained. \nThey didn't even tell their own developers as the negotiations were going on.</p><p>\nThey made this mess, they can clean it up themselves.</p>\ngpl3.licensingpatents.p3.s1\njring\nlogin to agree\n3\n2729\n\n\n<p style="display: inline">Why this date? What's special about it? Is it just the date of publication of the third draft? Was it chosen in order to avoid placing retroactive restrictions, by chance? If this is the motivation, I don't think this date limit is needed: since the GNU GPLv3 is not yet released, no work has been licensed under its terms yet, and hence no provision can be retroactive...</p><p>\nThe draft rationale explains that this date limit is intended to avoid forcing companies to face problems due to past agreements that cannot be changed anymore. I'm not convinced that this would be unfair: agreements like those are bad things and hence making them incompatible with conveying GPLv3'd works does not seem to be unfair.</p>\ngpl3.licensingpatents.p3.s1\nfrx\nlogin to agree\n5\n2778\n\n\n<p style="display: inline">How does such a deal, entered into on March 27, 2007 violate freedom anymore than one entered into on April 1, 2007?</p><p>\nIf this act is a threat to freedom, it should be disallowed, in all ways, at all time. If it is acceptable to allow before that, then it should be acceptable after.</p><p>\nThis line should be removed.\n</p>\ngpl3.licensingpatents.p3.s1\nkdean06\nlogin to agree\n7\n2802\n\n\n<p style="display: inline">Take out this clause. It introduces unnecessary uncertainty as to what rights the user of a GPLV3 licensed program actually has. Who knows what agreements are out there? People who choose to use GPLV3-ed software and programmers who choose to license their code under the GPLV3 deserve to know exactly what they're getting.</p><p>\nUsers of the GPLV2 are not being "blindsided" if this clause is removed from the GPLV3. A) The FSF has always been very clear as to what freedoms the GPL is intended to secure. The GPLV3, without this clause, is entirely consistent with the stated goals of the GPLV2 B) No rights are being taken away from any user of any software that's out there right now. Nobody ever guarantees that they will continue to forever license their code under whatever license is currently used. The portion of the "corporate world", whatever that is, that is complaining about a "change in license terms" is asking the FSF and it's supporters for something that they themselves would never agree to. Especially because the intent of the license is essentially unchanged from the GPLV2. If there was some legitimacy to the claims of being "blindsided" I could see at granting a grace period during which claims could not be pressed, but never an absolute cutoff date that says claims can never be pressed.</p><p>\nFor these reasons the uncertainty this clause introduces into the license is unnecessary. Regardless of the justification, such uncertainty is unwanted.</p>\ngpl3.licensingpatents.p3.s1\nkop\nlogin to agree\n3\n2837\n\n\n<p style="display: inline">Why not change this to January 1, 2005? Or even earlier.</p><p>\nI'm not sure why this should be allowed at all, but I REALLY don't think that the GPL v3 should sanction the MS/Novell "deal".\n</p>\ngpl3.licensingpatents.p3.s1\ncharlesh\nlogin to agree\n0\n2838\n\n\nCould this not be further modified to include a sunset clause terminating the\nexception at some future date?\n7 years in the future is about the limit of most business plans while 20 years would allow relevant patents to expire in many jurisdictions. \ngpl3.licensingpatents.p3.s1\ndumain\nlogin to agree\n0\n2873\n\n\nMake it April 1, 2007. It's the start of a month, the start of a quarter, and the start of the fiscal year, so it's much neater, and it's only 4 days difference.\ngpl3.licensingpatents.p3.s1\njamesgnz\nlogin to agree\n0\n2914\n\n\n<p style="display: inline">I think this date is neither necessary nor useful.</p><p>\nOne reason for the date might be that you cannot cover agreements that have been made before the publication of a license. I cannot see a problem here; the worst thing that could happen would be that a court decides that the clause is invalid. But until then the clause is better without the date and although I am no lawyer, I think the clause would be valid in most jurisdictions without the date.</p><p>\nThe other argument for the date would be the fact that we should not take away software freedoms retroactively. But I do not think that this is a problem either. Novell (and any other party that made such agreements) can use all GPLv2-covered software forever. As far as new software or software that will be newly released under GPLv3 is concerned, Novell will simply have to deal with that.\nWhen I write a piece of software and release it under GPLv2, then there is never a guaranty that I will stick to that license. I only grant certain freedoms to users of specific versions of my software. I may well decide to release future versions under different licenses and I may even make it proprietary software. The only guaranty is that versions released under GPLv2 will stay free. There is no difference here, Novell can still distribute software covered by GPLv2 and if they want to use GPLv3-covered software, they need to agree to the new license.</p>\ngpl3.licensingpatents.p3.s1\nSoong\nlogin to agree\n1\n2956\n\n\n<p style="display: inline">It breaks the license!</p><p>\nIt will allow products to exist that violate the terms: it would set false precedent for people setting out to do the very same thing; it will cause confusion in courts, possibly also then allowing products getting out violating this section; it will bolster a reputation of "heavyweight" bully-ish corporations that can 'get around' the limitations seen by others.</p><p>\nIt's no good.</p>\ngpl3.licensingpatents.p3.s1\nmrcoco\nlogin to agree\n0\n3067\n\n\n<p style="display: inline">It makes GPL 3 uncertain.</p><p>\nIBM SUN... All big companies have agreements between each other before that date. Remove it. If its not patent clear it should not be in the source code other wise we could have people turning around when using parts in another gpl3 project that there patents don't cover that project so we have to pay up. Loop holes are a bad thing when they will all ready be exploited and could cause trouble.</p>\ngpl3.licensingpatents.p3.s1\noiaohm\nlogin to agree\n0\n3078\n\n\n<p style="display: inline">There should be no exception made for Novell. Just as there should be no exception to the Samsung & Microsoft patent cross-licensing agreement that has been signed on patents Microsoft claims have been illegally borrowed by the Linux operating system. I wonder how many other such agreements are out there like that, which we have yet to hear about.</p><p>\nThese corporations have attorneys and should have known the possible consequences of what they were agreeing to. I will be very disappointed if this exception remains as is.\n</p>\ngpl3.licensingpatents.p3.s1\ncodeblue\nlogin to agree\n1\n3138\n\n\nIn a recent Groklaw article SCO seems to be positioning to take advantage of this.\nPossible negative effects on the IBM/SCO/MS case.\ngpl3.licensingpatents.p3.s1\ndrcj\nlogin to agree\n0\n3146\n\n\nThis exclusion is such a BAD idea. There should be NO exclusions, especially for Novell and Microsoft, otherwise why bother?\ngpl3.licensingpatents.p3.s1\ntonythed\nlogin to agree\n0\n3150\n\n\n", "2416:2429:2430:2447:2460:2495:2515:2532:2537:2539:2540:2547:2555:2558:2562:2563:2576:2577:2647:2655:2692:2693:2702:2707:2708:2725:2726:2729:2749:2777:2778:2785:2790:2796:2802:2837:2838:2842:2863:2873:2914:2934:2956:2964:3054:3067:3078:3120:3138:3143:3146:3150:3347": "\n\n\n\nWhat do the square brackets mean in this license? Elsewhere it refers to section numbers from version 2, but it's entirely unclear what they mean here.\ngpl3.licensingpatents.p3.s1\njag\nlogin to agree\n0\n2416\n\n\nDoes this bracketed phrase mean that Novell and their agreement with Microsoft is grandfathered in?\ngpl3.licensingpatents.p3.s1\nfreestu\nlogin to agree\n1\n2429\n\n\nIs the purpose of this bracketed clause to grandfather the agreement between Novell and Microsoft? If so, I think its a bad idea.\ngpl3.licensingpatents.p3.s1\nfreestu\nlogin to agree\n21\n2430\n\n\nLet's see how many people will come along showing agreements allegedly signed prior to that date.\ngpl3.licensingpatents.p3.s1\nibor\nlogin to agree\n3\n2447\n\n\n<p style="display: inline">Removing these words would seem to make the license stronger with no adverse effects. The business of the third party doesn't appear relevant.</p><p>\nI'm guessing that the payment terms are there to only bring the clause into effect when the distributor is an active participant in the arrangement. However, it appears to open a loop-hole - what if the payment terms are NOT 'based on the extent of your activity of conveying the work' The payment terms seem to be so specific that they can be trivially worked around.</p>\ngpl3.licensingpatents.p3.s1\npjm\nlogin to agree\n15\n2460\n\n\nLook... it's not even about Novell. It's the fact that Novell did something with Microsoft. However, there could already exist business dealing involving patents that do not involve the high profile Novell-Microsoft pact. With that said, I have no idea if this statement belongs in the document though. After all, we're talking about GPLv3... no deals have happened under GPLv3.\ngpl3.licensingpatents.p3.s1\ncjcox\nlogin to agree\n0\n2495\n\n\nThis section is not only good for Novell. It lets all enterprises supporting the FSF know that they will not be blind-sighted by the community. There are many enterprises that have contributed much and pushed the cause forward in one way or another. Non-retroactive licensing protects their interests and encourages their support and confidence. Corporate confidence is important, no matter what we idealistic geeks may prefer. \ngpl3.licensingpatents.p3.s1\nquestin\nlogin to agree\n3\n2515\n\n\n<p style="display: inline">I do not think we should give Novell a get-out-of-jail free card.</p><p>\nThey clearly and deliberatly violated the principles of free software, even though the GPLv2 didn't explicitly forbid it.</p>\ngpl3.licensingpatents.p3.s1\nnocturn\nlogin to agree\n9\n2532\n\n\nno date-based exceptions please.\ngpl3.licensingpatents.p3.s1\nhkbst\nlogin to agree\n10\n2537\n\n\n<p style="display: inline">In some jurisdictions at least, there is no requirement for this kind of agreement to be in writing. Any two parties who wish to make an evil deal like the MS-Novell deal can simply claim having orally agreed on a broad agreement of this kind before whatever date is specified, and it is impossible to prove such a claim to be false.</p><p>\nThe text in brackets must be deleted for this part to have any teeth.</p>\ngpl3.licensingpatents.p3.s1\nnb\nlogin to agree\n8\n2539\n\n\nWhether the evil company that tries to make some GPL'd software non-free is in the bsuiness of distributing software is irrelevant. Imagine instead of the Microsoft-Novell deal a similar deal between a patent troll company (which could possibly be secretly funded by Microsoft) and a company which distributes GPL'd software commercially.\ngpl3.licensingpatents.p3.s1\nnb\nlogin to agree\n18\n2540\n\n\n<p style="display: inline">Completely unnecissary. Microsoft and Novell will always be able to use their covenant on the software available at the time.</p><p>\nThere is no reason that software written 5 years from now should still have this liability.</p><p>\nPlease don't include this.</p>\ngpl3.licensingpatents.p3.s1\njalockli\nlogin to agree\n12\n2547\n\n\n<p style="display: inline">This section seems to very neatly cover the Novell/Microsoft deal without covering any variations to that deal, much less any deal with similar effects but different mechanics.</p><p>\nThe threat here is *any* patent license to some but not all of the users the a GPL covered work is distributed to.</p><p>\n"You may not convey a covered work if you are a party to an agreement with a third party that would give a patent license to some, but not all, recipients of that work. Further, you may not publicize a limited license nor may you use it as a sales differentiator, even if it is not the result of an explicit agreement."</p>\ngpl3.licensingpatents.p3.s1\nchandon\nlogin to agree\n6\n2555\n\n\nwhatever the business of the third party is, it should not be off the hook. What if the business is changed from distributing software to suing software vendors or supporters of free software?\nJust remove that portion as irrelevant, the entrance to a freedom-narrowing agreement should be more than enough.\ngpl3.licensingpatents.p3.s1\ngoldie\nlogin to agree\n7\n2558\n\n\nWhat if the payment is a one-time lump sum?\ngpl3.licensingpatents.p3.s1\nnovalis\nlogin to agree\n1\n2562\n\n\nRemove: That [bracketed clause] will make it harder to claim prior art. First, given a date - anybody can say the arrangement happened before or after to their benefit. Second, there are granted patents even though prior art does exist. The prior art predates the patents by many years. This clause is bad when the assumption has been that the patent claimant has anything to do with software under GPL. Third, a company may still distribute patented software under the GPL with patents claimed by someone not even dealt with the software without knowledge of such patents. (The linked-list patents, for example) A licensee's only hope to protect against such patent claims is to reveal prior art. Even though it is not explicit, the bracketed clause creates an assumption that prior art doesn't exist under such arrangement because knowledge of such patents are assumed - even though it is not explicitly stated that way just look at the logic down the road of how it would be played out in court. Forth, it is best to keep the patent terms uniform and remove the bracketed clause. It will make sure that cases brought up about the GPL are uniform. The bracketed clause presents a greater legal headache for many individuals then fewer that it intends grandfather. Fifth, if a clause like this must be added - remove the date and specifically state the intent, likewise "unless the arrangement upholds the GPLv2 and includes in the arrangement a certification to not adopt any license being a successor to GPLv2." Notice how I used the word "certification" to prevent wrongful retroactivity.\ngpl3.licensingpatents.p3.s1\ndzonatas\nlogin to agree\n5\n2563\n\n\nThis seems to be pointless, and makes this into a clause to catch exactly AND ONLY the thing done by MS and Novel. What if a hardware manufacturer offers a patent licence to force usage on their hardware? You cannot predict such things and the licence should be as general as possible.\ngpl3.licensingpatents.p3.s1\nrhouse\nlogin to agree\n8\n2576\n\n\nLet's face it, the only such agreement in force is the MS Novell one; it was not a naive mistake, but a deliberate end run around abundantly clear language in GPL2, and constitutes deliberate misuse (whether legally recognised or not) of others' generously donated IP. It was done eyes wide open and now the penalty should be paid: newly licenced or new software under GPL3 should NOT be available for their commercial purposes. Let Novell struggle on with legacy software or find their own way out of the mess of their deliberate and unethical creation. Please DELETE THIS sunset clause- this is the most important change you should make.\ngpl3.licensingpatents.p3.s1\nrhouse\nlogin to agree\n7\n2577\n\n\n<p style="display: inline">The license should protect software, it is failing to protect users freedoms if you leave this in. Worse still an added clause protecting such freedoms would be "an additional restriction"</p><p>\nInstead let users add such a prior to ... as an optional extra if they wish, then they will be giving extra permissions and the balance is correct\n</p>\ngpl3.licensingpatents.p3.s1\nAlanCox\nlogin to agree\n3\n2647\n\n\n<p style="display: inline">This eliminates protection from patent trolls. Why was this sentence added? The rationale doesn't say. I cannot think of any kind of acceptable (to FSF) patent agreement that would be impossible if this sentence was removed.</p><p>\nThe only reason I can imagine is that removing this bit would make it easier for a patent troll to put a Free Software project out of business. The project would fold when it settles and makes a patent agreement or it looses the suit. Many projects do not have the money to successfully fight a troll's patent suit.</p><p>\nThen again, with this clause MS could do exactly the same to a free software project.</p>\ngpl3.licensingpatents.p3.s1\nmarechal\nlogin to agree\n4\n2655\n\n\n"a third party that is in the business of distributing software" ? What is to stop Microsoft or another company selling patents to a patent troll, and then using that company to do their dirty work for them ? It seems a little short-sighted to me.\n\ngpl3.licensingpatents.p3.s1\nsalsaman\nlogin to agree\n5\n2692\n\n\nDid Novell's lawyers suggest this wording ?\ngpl3.licensingpatents.p3.s1\nsalsaman\nlogin to agree\n3\n2693\n\n\nGive this a time fuse, make it "...prior to 3/28/07, in which case breach shall be excused until 4/28/08." This will allow time for the business entities to make other arrangements, changes to contract etc. Don't perpetually excuse this as it allows ill gotten gain to remain with Novell for as long as it lasts and provides them no incentive to change their pact with Microsoft.\ngpl3.licensingpatents.p3.s1\naaronmf\nlogin to agree\n1\n2702\n\n\n<p style="display: inline">In summary: BIG MISTAKE, DON'T DO IT.</p><p>\nI really can't say in words just how bad of an idea this is. The right thing to do is hang Microsoft (and unfortunately, Novell with them... this is what happens when you make a deal with the devil) out to dry. Leave them in the GPLv2 world, we will move on to the GPLv3; they will eventually figure out this is our game, played by our rules. This exception makes as much sense as giving Tivo a free pass, that is, no sense at all. Microsoft does not respect the users' freedoms (we know this from experience). As an adversary, Microsoft is not worthy of any leniency at all. How much leniency have they given us? Have they even released MS-DOS 1.0 under the GPL, much less one of the later versions like 5.0 or 6.22, or even earlier versions of Windows? If they have, I have not heard of it and it certainly would be all over the free software-related Web sites and IRC channels I check.</p><p>\nIf Microsoft chooses to join the free software community, they may choose to. This is not the way to go about encouraging them to do it; this is a gift to Microsoft, most undeserved, and Microsoft will do what Microsoft usually does: take the gift, say (either with words or actions) what amounts to "thanks, and by the way, #$%& you" and continue to attack us as before. I'd hate to see all the hard work of the free software movement thrown away because of something this silly.</p>\ngpl3.licensingpatents.p3.s1\nskquinn\nlogin to agree\n6\n2707\n\n\nIt should be treated like other "special permissions" above and beyond the boilerplate GPL (that you can then subsequently remove on a specific copy). Incorporating a specific date into a license is going to look idiotic 15 years from now, and it should be up to the individual software team whether they want to reward Novell and similar GPL-molesters by granting this permission.\ngpl3.licensingpatents.p3.s1\nhorsten\nlogin to agree\n2\n2708\n\n\nThis entire clause is at most useless, and at worst overly restrictive. It should be removed. GPLv3 should be as simple as possible, not as precise as possible.\ngpl3.licensingpatents.p3.s1\nflaschen\nlogin to agree\n2\n2725\n\n\n<p style="display: inline">I basically agree with all the previous criticisms. I want to specifically note:</p><p>\na. Novell already has permanent GPLv2 rights, and they don't "deserve" anything special for GPLv3.</p><p>\nb. Any date makes the license even more over-specific and awkward than it already is.</p><p>\nc. It implies Novell was innocent and didn't know they were being unethical. In fact, they did, and even thought the legal case was uncertain enough to consult with SFLC.</p><p>\nd. It is unfair to others who want to abuse the GPL. Novell shouldn't get preference because they did it first.</p>\ngpl3.licensingpatents.p3.s1\nflaschen\nlogin to agree\n4\n2726\n\n\n<p style="display: inline">They knew what they were doing was dodgy. \nThey didn't consult the community. \nThey kept it from the community. \nThey did not care when the community complained. \nThey didn't even tell their own developers as the negotiations were going on.</p><p>\nThey made this mess, they can clean it up themselves.</p>\ngpl3.licensingpatents.p3.s1\njring\nlogin to agree\n3\n2729\n\n\n<p style="display: inline">By limiting this clause to agreements with third party software distributors the door may be opened for patent trolls (a la Rambus).</p><p>\nThis could permit a distributor to enter into a discriminatory agreement with an entity in the business of filing and licensing software patents but that refrains from distributing software. Potentially, a software distributor could transfer its patent rights to another entity that would be in the business of licensing those patents.</p><p>\nCan this be phrased so that the limitation applies to agreements covering software patents regardless of the third party's business? As technology evolves might it become difficult to distinguish software patents from other types of patents. It would be best to start off as inclusively as reasonable and loosen it up if legitimate issues arise.</p>\ngpl3.licensingpatents.p3.s1\npolymath\nlogin to agree\n1\n2749\n\n\nThis clause seems to be based on the general principle that no further restrictions can be added beyond the terms of this License. In the present case, it is stated that one cannot add further restrictions through restrictive patent licenses. Looks OK.\ngpl3.licensingpatents.p3.s1\nfrx\nlogin to agree\n0\n2777\n\n\n<p style="display: inline">Why this date? What's special about it? Is it just the date of publication of the third draft? Was it chosen in order to avoid placing retroactive restrictions, by chance? If this is the motivation, I don't think this date limit is needed: since the GNU GPLv3 is not yet released, no work has been licensed under its terms yet, and hence no provision can be retroactive...</p><p>\nThe draft rationale explains that this date limit is intended to avoid forcing companies to face problems due to past agreements that cannot be changed anymore. I'm not convinced that this would be unfair: agreements like those are bad things and hence making them incompatible with conveying GPLv3'd works does not seem to be unfair.</p>\ngpl3.licensingpatents.p3.s1\nfrx\nlogin to agree\n5\n2778\n\n\nAs I read this phrase, I cannot enter into a license agreement with a third party by which recipients of the covered work from me are indemnified against patent claims, if such agreement requires payments proportionate to the number of recipients, or number of copies distributed. However, if I simply make a lump sum payment for any number of recipients or any number of copies, then my payment is arguably not "based on the extent of your activity of conveying the work", and yet arguably such an agreement should still be prohibited by this agreement.\ngpl3.licensingpatents.p3.s1\nskissane\nlogin to agree\n0\n2785\n\n\n<p style="display: inline">I don't think that this is a good idea. First it isn't good to use a license to reach political reasons.</p><p>\nBut second, when you want to use patent protected software (like some parts of the TrueType interpreter of FreeType) and you want to buy such a license, it would be impossible to buy this from a company who distributes the software. This resctriction is stupid. It makes life more complicat without bringing any benefit against patents!</p>\ngpl3.licensingpatents.p3.s1\ntimi\nlogin to agree\n0\n2790\n\n\n<p style="display: inline">There is a very huge problem with this clause. Imagine what happens if Microsoft start sueing people all over the place, while simultaneously offering a very cheap deal under which they won't sue you. \n\nSmall companies may find they have no choice but to accept the deal, at which point they are suddenly banned from redistributing GPLv3 software. Thus this clause could allow a huge company to abuse the patent system in order to effectively kill small vendors ability to use GPLv3 software.</p><p>\nI don't see any easy solution to this problem. The only alternative seems to be to allow Microsoft - Novell like deals, which is unacceptable, but the other possibility could result in proprietary software vendor's trying to push companies into signing seemingly innocent deals without realizing they are giving up their right to use the GPLv3. </p>\ngpl3.licensingpatents.p3.s1\njring\nlogin to agree\n2\n2796\n\n\n<p style="display: inline">How does such a deal, entered into on March 27, 2007 violate freedom anymore than one entered into on April 1, 2007?</p><p>\nIf this act is a threat to freedom, it should be disallowed, in all ways, at all time. If it is acceptable to allow before that, then it should be acceptable after.</p><p>\nThis line should be removed.\n</p>\ngpl3.licensingpatents.p3.s1\nkdean06\nlogin to agree\n7\n2802\n\n\n<p style="display: inline">Take out this clause. It introduces unnecessary uncertainty as to what rights the user of a GPLV3 licensed program actually has. Who knows what agreements are out there? People who choose to use GPLV3-ed software and programmers who choose to license their code under the GPLV3 deserve to know exactly what they're getting.</p><p>\nUsers of the GPLV2 are not being "blindsided" if this clause is removed from the GPLV3. A) The FSF has always been very clear as to what freedoms the GPL is intended to secure. The GPLV3, without this clause, is entirely consistent with the stated goals of the GPLV2 B) No rights are being taken away from any user of any software that's out there right now. Nobody ever guarantees that they will continue to forever license their code under whatever license is currently used. The portion of the "corporate world", whatever that is, that is complaining about a "change in license terms" is asking the FSF and it's supporters for something that they themselves would never agree to. Especially because the intent of the license is essentially unchanged from the GPLV2. If there was some legitimacy to the claims of being "blindsided" I could see at granting a grace period during which claims could not be pressed, but never an absolute cutoff date that says claims can never be pressed.</p><p>\nFor these reasons the uncertainty this clause introduces into the license is unnecessary. Regardless of the justification, such uncertainty is unwanted.</p>\ngpl3.licensingpatents.p3.s1\nkop\nlogin to agree\n3\n2837\n\n\n<p style="display: inline">Why not change this to January 1, 2005? Or even earlier.</p><p>\nI'm not sure why this should be allowed at all, but I REALLY don't think that the GPL v3 should sanction the MS/Novell "deal".\n</p>\ngpl3.licensingpatents.p3.s1\ncharlesh\nlogin to agree\n0\n2838\n\n\nThe GPL should stand proudly on its own with out need to cater to those who violated the spirit of the version 2 license prior to its obvious intent being SPELLED OUT in version 3.\ngpl3.licensingpatents.p3.s1\njohnrkro\nlogin to agree\n0\n2842\n\n\nReplace the clause: "based on the extent of your activity of" with the word "for".\ngpl3.licensingpatents.p3.s1\njkoenig\nlogin to agree\n0\n2863\n\n\nCould this not be further modified to include a sunset clause terminating the\nexception at some future date?\n7 years in the future is about the limit of most business plans while 20 years would allow relevant patents to expire in many jurisdictions. \ngpl3.licensingpatents.p3.s1\ndumain\nlogin to agree\n0\n2873\n\n\nMake it April 1, 2007. It's the start of a month, the start of a quarter, and the start of the fiscal year, so it's much neater, and it's only 4 days difference.\ngpl3.licensingpatents.p3.s1\njamesgnz\nlogin to agree\n0\n2914\n\n\n<p style="display: inline">A cleaner way to resolve this is instead of distinguishing vs. the type of business, discriminate on the type of patent deal.</p><p>\nIt is much easier for the free software community to deal with the threat of a specific patent suit than the nebulous thread of a possible suit from a large portfolio.</p><p>\nA patent troll threatening litigation of a specific patent must identify the infringing code. There are a number of defenses available, such as finding prior art or re-writing the code to eliminate the patented technique.</p><p>\nHowever, no such defenses exist when the patent and the infringing code has not been identified with specificity. Thus, deals such as the Novell/Microsoft deal, where no specific litigation has been threatened, are actually more harmful than a single patent troll.</p><p>\nThus, I would propose the following rule: A "general" indemnification against an entire patent portfolio must be "generally" available to all users of the software. Against a specific patent threat where the patent and the infringing code have been identified, the distributor should be granted a bit more leeway.</p>\ngpl3.licensingpatents.p3.s1\ntalin\nlogin to agree\n0\n2934\n\n\n<p style="display: inline">I think this date is neither necessary nor useful.</p><p>\nOne reason for the date might be that you cannot cover agreements that have been made before the publication of a license. I cannot see a problem here; the worst thing that could happen would be that a court decides that the clause is invalid. But until then the clause is better without the date and although I am no lawyer, I think the clause would be valid in most jurisdictions without the date.</p><p>\nThe other argument for the date would be the fact that we should not take away software freedoms retroactively. But I do not think that this is a problem either. Novell (and any other party that made such agreements) can use all GPLv2-covered software forever. As far as new software or software that will be newly released under GPLv3 is concerned, Novell will simply have to deal with that.\nWhen I write a piece of software and release it under GPLv2, then there is never a guaranty that I will stick to that license. I only grant certain freedoms to users of specific versions of my software. I may well decide to release future versions under different licenses and I may even make it proprietary software. The only guaranty is that versions released under GPLv2 will stay free. There is no difference here, Novell can still distribute software covered by GPLv2 and if they want to use GPLv3-covered software, they need to agree to the new license.</p>\ngpl3.licensingpatents.p3.s1\nSoong\nlogin to agree\n1\n2956\n\n\nThis should also include any third party which purchases Patents for the sole purpose of collecting royalities (Patent Trolls), any third party charged with licensing or collecting licensing fees and royalities for patents on behalf of the patent holder, or basically any third party seeking payments for patent rights.\ngpl3.licensingpatents.p3.s1\nghuber\nlogin to agree\n0\n2964\n\n\n<p style="display: inline">patent holders' leverage over small free software distributors.</p><p>\nThis provision was drafted to prevent a free software distributor from\nvoluntarily entering into an agreement with a patent holder in order to\nget a competitive advantage over other distributors. The unintended\neffect of Paragraphs 3 and 5 taken together, however, may be to take\noptions away from distributors who are threatened or sued by a patent\nholder. Such distributors - particularly small businesses and\nindividuals who have no intent to split the community - need to have the\nfreedom to negotiate a settlement that removes their liability and\npreserves their ability to continue to distribute free software.</p><p>\nConsider the situation in which a large software company sues a small\nfree software distributor for patent infringement. The distributor has\nonly four options that are compatible with paragraphs 3 and 5 of this draft:</p><p>\n(1) Fight the patent in court. This option is unlikely to be feasible\nfor a small distributor. Litigating a patent case to a judicial\nresolution can take years and cost millions of dollars in legal fees.\nMoreover, patents are presumed valid, and a favorable outcome is never\ncertain, so the distributor risks a significant adverse judgment by\nproceeding with the lawsuit.</p><p>\n(2) Negotiate a patent license for all downstream recipients. Putting\naside the fairness issues involved in forcing the first distributor sued\nto pay for whole community, this option is not likely to be feasible\neither. A patent holder who is willing to license the whole free\nsoftware community on affordable terms probably would not sue a small\ndistributor in the first place.</p><p>\n(3) Negotiate a patent license that does not involve a payment for the extent of the distributor's activity of conveying the work and then\nmake all of the work's source code publicly available free of charge. This option may be feasible in some instances and not in others.\nThe terms of the patent license, of course, are subject to the patent holder's agreement. A patent holder who is willing to accept a\nsmall per-unit royalty might demand a prohibitively large lump-sum payment to cover all possible future sales. Thus, not being able to\nnegotiate a license based on the extent of activity severely limits the settlement options available to the distributor.</p><p>\nAnd the safe harbor of the free, public distribution of the work's\nsource code may not always be available to a particular distributor\n(e.g., if the work was developed under NDA or other conditions of\nprivacy, if the work is used in a classified environment, or if no-cost\ndistribution of the whole work to non-customers would undermine the\ndistributor's business). Moreover, unless the distributor's license is\nsufficiently broad, the publication of source code could make the\ndistributor liable to the patent holder for indirect infringement.</p><p>\n(4) Pay the patent holder to settle the lawsuit and stop distributing\nthe free software. By default, this is the likely outcome of any patent\nlawsuit (or even threat of a patent lawsuit) against a small\ndistributor. That means that the patent holder may be able to shut down\ndistributors one by one until the work is not available at all.</p><p>\nWithout paragraphs 3 and 5, the distributor would have the option to\nnegotiate a license (regardless of its opinion of the validity of the\npatent) that covers itself and its direct customers - thereby staying in\nbusiness, continuing to distribute free software, and providing a\nbenefit to its customers in the form of protection from a patent lawsuit.</p><p>\nNothing in this draft prohibits a distributor from indemnifying its\ndirect customers against patent infringement or buying insurance for its\ncustomers. Why should negotiating a license (whether in response to a\nlawsuit or otherwise) be any different? And why shouldn't the option of\nproviding protection from patents be available to small distributors?</p><p>\nFurther, if this license makes settling a patent lawsuit infeasible for a small distributor, then indemnification and insurance will also be\nunaffordable for all but the largest distributors (because the distributor's options for negotiating a patent license for its indemnified\ncustomer who is sued for patent infringement will be limited as discussed above). And if indemnification and insurance are less practical,\nthe commercial use of free software may be reduced, as potential customers choose other solutions.</p><p>\nLike it or not, software patents exist and large software companies have\nthem. This license should not do anything to strengthen the patent\nholders' position. Rather, this license should continue to encourage\ndevelopers to create and distribute as much free software as possible -\nand give users the freedom to decide whether or not to use it based on\ntheir own assessment of the patent risk.</p><p>\nVery truly yours,\nBeth Mitchell\nCounsel\nCodeSourcery, Inc.\n650-331-3385 x703\n\n</p>\ngpl3.licensingpatents.p3.s1\nbeth1\nlogin to agree\n0\n3054\n\n\n<p style="display: inline">It breaks the license!</p><p>\nIt will allow products to exist that violate the terms: it would set false precedent for people setting out to do the very same thing; it will cause confusion in courts, possibly also then allowing products getting out violating this section; it will bolster a reputation of "heavyweight" bully-ish corporations that can 'get around' the limitations seen by others.</p><p>\nIt's no good.</p>\ngpl3.licensingpatents.p3.s1\nmrcoco\nlogin to agree\n0\n3067\n\n\n<p style="display: inline">It makes GPL 3 uncertain.</p><p>\nIBM SUN... All big companies have agreements between each other before that date. Remove it. If its not patent clear it should not be in the source code other wise we could have people turning around when using parts in another gpl3 project that there patents don't cover that project so we have to pay up. Loop holes are a bad thing when they will all ready be exploited and could cause trouble.</p>\ngpl3.licensingpatents.p3.s1\noiaohm\nlogin to agree\n0\n3078\n\n\n<p style="display: inline">distributing software"</p><p>\nThis clause seems to open up a lot of loop holes. Consider: A company that is\nin the business of selling medical equipment which happens to run software\nthat is licensed under the GPL, is not clearly in the business of\ndistributing software. (They may not even do so at all, say by pointing their\ncustomers to someone else who hosts the software specifically tailored for\ntheir system.) However they can still hold a clear interest in the software\nand could be involved in a lawsuit. This whole situation becomes a lot more\nconfusing if the work that is licensed under the GPL would not classically be\nconsidered "software".</p><p>\nAll in all this seems like an aria that is bound to be exploited if not worded\nvery carefully. Given that, I am not sure how much value it really adds.\nPerhaps the desired results can be achieved through rewording the clause\nitself rather than adding an exception.\n\n</p>\ngpl3.licensingpatents.p3.s1\ntkaitchuck\nlogin to agree\n1\n3120\n\n\n<p style="display: inline">There should be no exception made for Novell. Just as there should be no exception to the Samsung & Microsoft patent cross-licensing agreement that has been signed on patents Microsoft claims have been illegally borrowed by the Linux operating system. I wonder how many other such agreements are out there like that, which we have yet to hear about.</p><p>\nThese corporations have attorneys and should have known the possible consequences of what they were agreeing to. I will be very disappointed if this exception remains as is.\n</p>\ngpl3.licensingpatents.p3.s1\ncodeblue\nlogin to agree\n1\n3138\n\n\n<p style="display: inline">\n\nAnd I don't need a reply, thanks.</p><p>\nIf you need more context, it's in comments under the Sontag article\nposted on groklaw Saturday 5/26.</p><p>\nThanks for all the work!</p><p>\nJon\n\n\n------------\n"Concern about this "Covenant Not to Sue" in relation to GPLv3\nAuthored by: map on Sunday, May 27 2007 @ 09:48 AM EDT\nWill SCO be the first to jump on the GPLv3 bandwagon.\nMy concern is that the Novell/MS loophole in the proposed GPLv3 may\nallow things not allowed in GPLv2.\nSpecifically GPLv2 section 2b:</p><p>\nb) You must cause any work that you distribute or publish, that in whole\nor in part contains or is derived from the Program or any part thereof,\nto be licensed as a whole at no charge to all third parties under the\nterms of this License.</p><p>\nThis would seem to me to imply that licenses for any infringed or\npotentially infringed IP be required for distribution of the whole, and\nsuch licenses have terms that are compliant with GPLv2.</p><p>\nBefore GPLv3 is ratified serious consideration should be given to how it\naffects various parties obligations under GPLv2 section 2b,4, and 6; and\nalso any potential impact it may have on the IBM/SCO/MS case.</p><p>\nUnnecessary language in the GPLv3 draft may be a problem such as section\n11 p5:</p><p>\n...that is in the business of distributing software...</p><p>\nI can almost guarantee you that MS is dumping IP into non-distributing\npuppet companies to get through that loophole.</p><p>\nIt never hurts to have an extra lawyer review a document, I hope FSF\nwill consult with IBM, et al. about this.</p><p>\nIANAL"\n\n\n</p>\ngpl3.licensingpatents.p3.s1\njons\nlogin to agree\n1\n3143\n\n\nIn a recent Groklaw article SCO seems to be positioning to take advantage of this.\nPossible negative effects on the IBM/SCO/MS case.\ngpl3.licensingpatents.p3.s1\ndrcj\nlogin to agree\n0\n3146\n\n\nThis exclusion is such a BAD idea. There should be NO exclusions, especially for Novell and Microsoft, otherwise why bother?\ngpl3.licensingpatents.p3.s1\ntonythed\nlogin to agree\n0\n3150\n\n\n<p style="display: inline">What here is not conveyed by the "if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you..." clause of GPLv2?</p><p>\nThis is directly addressed at a single deal, Novell/MSFT, which was ominous but did not restrict the freedoms of any user. The fact that MSFT then penned a similar deal with Xandros after March indicates that this clause is useless as the deterrent it is intended to be anyway.</p>\ngpl3.licensingpatents.p3.s1\nbk\nlogin to agree\n0\n3347\n\n\n", "2794": "\n\n\n\n<p style="display: inline">I would think "any implied license or other defenses to infringement" includes every possible thing you can do against someone infringing on your patents. If this license does not exclude you from those that would otherwise be available to you, it clearly doesn't exclude you from any defenses. Which means, AFAICS, that anything about patents can just as well be omitted.</p><p>\nI suspect this may not be true for granting patent licenses, but this is not at all clear. Please clarify this.</p>\ngpl3.licensingpatents.p4.s1\nwijnen\nlogin to agree\n0\n2794\n\n\n", "2779": "\n\n\n\nThis section seems substantially the same as section 7 of GPLv2 (and of section 12 of previous drafts), apart from minor details. I think it's OK.\ngpl3.libertyordeath.0.0\nfrx\nlogin to agree\n1\n2779\n\n\n", "2440:2445:2569": "\n\n\n\n<p style="display: inline">Given a compulsory-reciprocation license is distinct from a liberty-restoration license, it's indeed best that the licenses remain separate.</p><p>\nCompatibility at the linkage level certainly helps preserve a clear demarcation between source subject to each license.</p>\ngpl3.affero.0.0\ncrosbie\nlogin to agree\n3\n2440\n\n\nThis section should be removed; if the copyright holder wishes to cross-license with the Affero GPL, he or she should explicitly cross-license. Forcing this cross-licensing upon all users of the GNU GPLv3 would mean that any loopholes in the AGPLv2 would "contaminate" the GNU GPLv3.\ngpl3.affero.0.0\ncyd\nlogin to agree\n9\n2445\n\n\nOne major difference between Draft2.\u00c3\u0082\u00c2\u00a77.b.4 and Draft3.\u00c3\u0082\u00c2\u00a713 is that Draft2.\u00c3\u0082\u00c2\u00a77.b.4 applied to all versions of all licenses with this Affero-like clause that are compatible with the GPL, both existing and future; while Draft3.\u00c3\u0082\u00c2\u00a713 only applies to one single specific version (the as yet unpublished! version 2) of one single specific license (the Affero General Public License). This does not strike me as a good idea. [In software analogy, one would say this solution lacks robustness, is not extensible or upward-compatible by design, targets a specific implementation rather than an interface or open specification]\ngpl3.affero.0.0\nadhemar\nlogin to agree\n6\n2569\n\n\n", "2440:2445:2534:2569:2704:2706:2780:2783:2784:2803:2813:2958:3127": "\n\n\n\n<p style="display: inline">Given a compulsory-reciprocation license is distinct from a liberty-restoration license, it's indeed best that the licenses remain separate.</p><p>\nCompatibility at the linkage level certainly helps preserve a clear demarcation between source subject to each license.</p>\ngpl3.affero.0.0\ncrosbie\nlogin to agree\n3\n2440\n\n\nThis section should be removed; if the copyright holder wishes to cross-license with the Affero GPL, he or she should explicitly cross-license. Forcing this cross-licensing upon all users of the GNU GPLv3 would mean that any loopholes in the AGPLv2 would "contaminate" the GNU GPLv3.\ngpl3.affero.0.0\ncyd\nlogin to agree\n9\n2445\n\n\nYou shouldn't bloat the already wordy GPL3\nwith special casing for another licensense like this. Especially one like this which has next to no use.\ngpl3.affero.0.0\nuraeus\nlogin to agree\n7\n2534\n\n\nOne major difference between Draft2.\u00c3\u0082\u00c2\u00a77.b.4 and Draft3.\u00c3\u0082\u00c2\u00a713 is that Draft2.\u00c3\u0082\u00c2\u00a77.b.4 applied to all versions of all licenses with this Affero-like clause that are compatible with the GPL, both existing and future; while Draft3.\u00c3\u0082\u00c2\u00a713 only applies to one single specific version (the as yet unpublished! version 2) of one single specific license (the Affero General Public License). This does not strike me as a good idea. [In software analogy, one would say this solution lacks robustness, is not extensible or upward-compatible by design, targets a specific implementation rather than an interface or open specification]\ngpl3.affero.0.0\nadhemar\nlogin to agree\n6\n2569\n\n\n\ngpl3.affero.0.0\nrek2\nlogin to agree\n0\n2704\n\n\nhttp://www.linux-mag.com/id/3017/\nhave we chicken out? \nI strongly believe we should add a affero alike part under the gplv3! protect us web programmers please...\ngpl3.affero.0.0\nrek2\nlogin to agree\n3\n2706\n\n\n<p style="display: inline">This section introduces a form of compatibility with a license that is yet unreleased and thus possibly non-free: the Affero General Public License, version 2. The AfferoGPL v1 is, in my opinion, a non-free license, due to its clause 2(d). I won't restate all the reasons for my conclusions (more details in http://gplv3.fsf.org/comments/rt/readsay.html?filename=gplv3-draft-3&id=1663). As a consequence, I have few hopes that the forthcoming version 2 of the AfferoGPL will be a free license.</p><p>\nBeing compatible with an unknown (and thus possibly non-free) license destroys the copyleft mechanism of the GPLv3. The draft rationale states that most other free software licenses also feature such form of compatibility with the AfferoGPL: this is true for non-copyleft licenses, though! When I choose a long and complicated copyleft license such as the GNU GPL, I want a copyleft mechanism that actually works; or otherwise I can choose short and simple non-copyleft licenses, such as the Expat or the 2-clause BSD ones...</p><p>\nI strongly recommend dropping section 13 entirely.</p>\ngpl3.affero.0.0\nfrx\nlogin to agree\n4\n2780\n\n\n\ngpl3.affero.0.0\nhawk\nlogin to agree\n0\n2783\n\n\n<p style="display: inline">First: allowing linking with a specific license is a great compromise.</p><p>\nHowever, why not reuse some of language from GPLv3DD2 7b instead of naming the AGPL explicitly?</p><p>\nAdvantages include:\n(1) the GPL would stand on its own,\n(2) it gives a path for graceful upgrades of the GPL without including the "or later"-clause,\n(3) it allows GPL-code to be used by those bent on a patent retaliation clause, and\n(4) if done properly, those who want more restrictive licenses could be LGPL-like and we just might avoid a lot of license proliferation.\n</p>\ngpl3.affero.0.0\nhawk\nlogin to agree\n3\n2784\n\n\n<p style="display: inline">The GPL, in all versions, should very clearly spell out rights that it protects. Referring to another license doesn't do this, I think.</p><p>\nAs stated with the reference to laws, referencing this license by NAME also irrevocably ties it to the fate of that license. Should, for some reason, the Affero GPL be rendered invalid, it also renders the GPLv3 invalid; a VERY dangerous thing.</p><p>\nThe GPL should spell out exactly what the Affero GPL protects, or this section should be removed.</p>\ngpl3.affero.0.0\nkdean06\nlogin to agree\n4\n2803\n\n\n<p style="display: inline">How about this:</p><p>\n"Notwithstanding any other provision of this License, you have permission to link any covered work with a work to which you have a license that is certified as 'GPL linkage compatible' by the FSF, and to convey the resulting combination. The terms of this License will continue to apply to your covered work but will not apply to the work with which it is linked, which will remain governed by the terms of its own license."</p><p>\nThe FSF can then list all the umpteen GPL compatible licenses and GPL affinity licenses for which linkage is permitted.</p><p>\nNo doubt the list will include AGPL.\n</p>\ngpl3.affero.0.0\ncrosbie\nlogin to agree\n2\n2813\n\n\n<p style="display: inline">We all know that copyleft is not necessary in order to consider a software as free software. But it was adopted by FSF on GPL as a strategic mechanism intended to avoid the removal of any of the 4 fundamental freedoms.</p><p>\nThe way that copyleft is implemented on GPLv2 does not make sense for web applications because the criterium for being obligated to provide the source code is the binaries distribution. As the web platform makes it possible to provide access to the application without transfering the binary, this configures as a mechanism of restricting the user's freedoms.</p><p>\nFree web applications (such as mediawiki, for example) may have modified versions for which the users are denied access to the source code. This mere possibility (that we all know that is exploited by lots of websites such as youtube, among others) shows to us a non-copyleft aspect of the license.</p><p>\nThis exemplifies how bad it is to link copyleft requirements to the binaries distribution. This way of implementing copyleft is only effective for desktop applications which are a particular case of software where access to the binary is synonym of access to the application.</p><p>\nFor the web application scenario, the most adequate solution would be to define some criteria such as, for example, access to the web application. That is: if a site is accessible from within a local network, then there would be the obligation of distributing its sources for all of the users of this local network (if they required access to it).</p><p>\nGPLv2 only manifests its copyleft aspect on non-web softwares. For web applications, GPLv2 behaves in a similar way as BSD tolerating the creation of proprietary derived web applications. That's why I believe that the inclusion of a strong "web copyleft" criterium would be important to fit GPLv3 to the FSF political attitude regarding preferencial methods of free software licensing, not being sufficient to solely create Affero compatibility clauses but, instead, making "web copyleft" an obligatory requirement explicit in GPLv3 itself (not on a attached Affero license).</p>\ngpl3.affero.0.0\nfsanches\nlogin to agree\n0\n2958\n\n\n<p style="display: inline">This section means that if someone wants to understand the GPL completely he not only has to read the GPL itself but the Affero GPL, too. This is a HUGE turn-off for hobbyists who typically don't like reading legalese. So this may cause people to refrain from relicensing their GPL2 works under GPL3 for the simple reason that they "don't have the time" (read "don't want to") to read 2 licenses.<br><br></p><p>\nThen of course there's the possibility that someone disagrees with the Affero GPL and therefore does not want to relicense under the GPL 3. I think it's better to leave it to the licensor to add this permission. If it means a lot to you I suggest you move it to the end of the GPL into the part "How to Apply these Terms to Your New Programs". Since we know from the past that people copy this boilerplate blindly, we can assume that the Affero linkage exception will become widespread.</p>\ngpl3.affero.0.0\nmux2005\nlogin to agree\n0\n3127\n\n\n", "3128": "\n\n\n\n<p style="display: inline">You need to define what "to link" means. This is a very overloaded term that can have a lot of different meanings in different contexts. The biggest problem is that it has non-technical meanings (which are the only meanings that most judges and lawyers will be acquainted with).<br><br></p><p>\nSince the GPL has been phrased to apply to "works" in general, encompassing more than just programs, it is not clear that here "to link" is to refer to the linking of programs only. Or maybe it isn't? In that case defining "to link" is even more necessary, because in the context of non-program works (e.g. icon sets, which under Windows for instance, are confusingly often stored in DLLs, i.e. dynamic LINK libraries) "to link" can mean a lot of things.<br><br></p><p>\nThen of course there's the problem that even in the technical world "to link" is far from being well-defined. There are a lot of corner-cases, arising for instance in the context of virtual machines and compilation/linking at run-time.</p>\ngpl3.affero.p0.s1\nmux2005\nlogin to agree\n0\n3128\n\n\n", "2453:2456:2489:2711": "\n\n\n\nLike 14., this should identify the legal entity that publishes Affero General Public License.\n\ngpl3.affero.p0.s1\nkaol\nlogin to agree\n5\n2453\n\n\nWhere can this AGPL version 2 be read? Is it still being drafted? Are you asking us to approve compatibility with another license blindly?\ngpl3.affero.p0.s1\nkaol\nlogin to agree\n10\n2456\n\n\n<p style="display: inline">Section 13 is an improvement over previous drafts because it enabled simplification of Section 7 by removing the provision there for an 'optional' Affero-like clause as an 'additional restriction'. Such 'additional restrictions' made the GPLv3 fundamentally more complex than the GPLv2: potentially many different licenses in effect, each different in subtle ways. \n\nI approve of the Section 13 approach, because I approve of the intent to allow an Affero-like restriction compatible with the license. Nothing else closes the 'Google loophole' in which operators of huge commercial web sites and server farms can effectively evade copyleft completely, using free software to build derived works the source to which they keep secret.</p><p>\nHowever, leaving the Affero-like feature to be specified by another organization seems unwise. It was an advantage of the old section 7 approach that FSF, not Affero, provided the exact language of the Affero-like restriction. Also, license versioning (re: future versions) is complex enough with only the FSF licenses involved.</p><p>\nI would like to see FSF publish a separate license, in the same sense that the LGPL is a separate license than the GPL, which differs from the GPLv3 only in its copyleft clause: only by including an Affero-like restriction (perhaps using language like that removed from section 7). This new FSF license and GPLv3 would refer to each other in their respective Section 13's.</p><p>\nThere is no getting around the fact that unless an Affero-like clause is (a) prohibited or (b) required by GPLv3, two licenses are needed. That is what an 'optional restriction' implies. Previous discussion seems to have decided neither (a) nor (b) is acceptable to the whole community. The only question remaining is whether FSF will take responsibility for the Affero-like alternative license. I think doing it right is important enough that FSF should not punt on it any longer.\n\n</p>\ngpl3.affero.p0.s1\nsamizdat\nlogin to agree\n7\n2489\n\n\nIf there is going to be another license running in parallel with the GPL, please try to get the numbering in sync. (Skip AGPLv2, and make it AGPLv3) It's not going to make much difference right now, but after another couple of releases it will just save everyone headaches. (Take a look at Mozilla version numbering--it's not pretty.)\ngpl3.affero.p0.s1\njamesgnz\nlogin to agree\n4\n2711\n\n\n", "2453:2489": "\n\n\n\nLike 14., this should identify the legal entity that publishes Affero General Public License.\n\ngpl3.affero.p0.s1\nkaol\nlogin to agree\n5\n2453\n\n\n<p style="display: inline">Section 13 is an improvement over previous drafts because it enabled simplification of Section 7 by removing the provision there for an 'optional' Affero-like clause as an 'additional restriction'. Such 'additional restrictions' made the GPLv3 fundamentally more complex than the GPLv2: potentially many different licenses in effect, each different in subtle ways. \n\nI approve of the Section 13 approach, because I approve of the intent to allow an Affero-like restriction compatible with the license. Nothing else closes the 'Google loophole' in which operators of huge commercial web sites and server farms can effectively evade copyleft completely, using free software to build derived works the source to which they keep secret.</p><p>\nHowever, leaving the Affero-like feature to be specified by another organization seems unwise. It was an advantage of the old section 7 approach that FSF, not Affero, provided the exact language of the Affero-like restriction. Also, license versioning (re: future versions) is complex enough with only the FSF licenses involved.</p><p>\nI would like to see FSF publish a separate license, in the same sense that the LGPL is a separate license than the GPL, which differs from the GPLv3 only in its copyleft clause: only by including an Affero-like restriction (perhaps using language like that removed from section 7). This new FSF license and GPLv3 would refer to each other in their respective Section 13's.</p><p>\nThere is no getting around the fact that unless an Affero-like clause is (a) prohibited or (b) required by GPLv3, two licenses are needed. That is what an 'optional restriction' implies. Previous discussion seems to have decided neither (a) nor (b) is acceptable to the whole community. The only question remaining is whether FSF will take responsibility for the Affero-like alternative license. I think doing it right is important enough that FSF should not punt on it any longer.\n\n</p>\ngpl3.affero.p0.s1\nsamizdat\nlogin to agree\n7\n2489\n\n\n", "2453:2489:2533:2828": "\n\n\n\nLike 14., this should identify the legal entity that publishes Affero General Public License.\n\ngpl3.affero.p0.s1\nkaol\nlogin to agree\n5\n2453\n\n\n<p style="display: inline">Section 13 is an improvement over previous drafts because it enabled simplification of Section 7 by removing the provision there for an 'optional' Affero-like clause as an 'additional restriction'. Such 'additional restrictions' made the GPLv3 fundamentally more complex than the GPLv2: potentially many different licenses in effect, each different in subtle ways. \n\nI approve of the Section 13 approach, because I approve of the intent to allow an Affero-like restriction compatible with the license. Nothing else closes the 'Google loophole' in which operators of huge commercial web sites and server farms can effectively evade copyleft completely, using free software to build derived works the source to which they keep secret.</p><p>\nHowever, leaving the Affero-like feature to be specified by another organization seems unwise. It was an advantage of the old section 7 approach that FSF, not Affero, provided the exact language of the Affero-like restriction. Also, license versioning (re: future versions) is complex enough with only the FSF licenses involved.</p><p>\nI would like to see FSF publish a separate license, in the same sense that the LGPL is a separate license than the GPL, which differs from the GPLv3 only in its copyleft clause: only by including an Affero-like restriction (perhaps using language like that removed from section 7). This new FSF license and GPLv3 would refer to each other in their respective Section 13's.</p><p>\nThere is no getting around the fact that unless an Affero-like clause is (a) prohibited or (b) required by GPLv3, two licenses are needed. That is what an 'optional restriction' implies. Previous discussion seems to have decided neither (a) nor (b) is acceptable to the whole community. The only question remaining is whether FSF will take responsibility for the Affero-like alternative license. I think doing it right is important enough that FSF should not punt on it any longer.\n\n</p>\ngpl3.affero.p0.s1\nsamizdat\nlogin to agree\n7\n2489\n\n\nI don't know what version 2 of the Affero GPL looks like, but if it is just GPLv3 with the Affero section 2(d) added to it, then it should not require that the transmission be by http, as that is not necessarily the protocol that is used on the network and the original application. \ngpl3.affero.p0.s1\njring\nlogin to agree\n1\n2533\n\n\n<p style="display: inline">The current AGPL is not adequate to protect developers from the "software as a service loophole." According to section 2d the work must be a program that is designed to be accessed over a network and which is able to download its source code to the user.</p><p>\nMy software would not apply since it is a C++ class library and not a program. Using my software would make it easy to develop a web application, and that is one of its goals, but it itself is just a collection of classes. In version 2 of the AGPL, I urge you to require the service provider to offer the source code, not the "program" itself.</p><p>\nMike\n</p>\ngpl3.affero.p0.s1\nmiked\nlogin to agree\n0\n2828\n\n\n", "3129": "\n\n\n\n<p style="display: inline">I don't understand how this is supposed to work. The GPL says that if I create a work based on a GPL-covered work, the result must be GPL-licensed. The Affero GPL says the same thing about the Affero GPL, doesn't it? So which of the 2 licenses applies to my combined work? <br><br></p><p>\nThis section 13 seems to say that the GPL applies to the combined work. And it has to be like this, because otherwise the GPL3 would be unacceptable to people who don't like the Affero GPL, because these people will not accept a GPL3 that allows relicensing a combined work under Affero. And if you alienate those people and cause them to not use the GPL3, leaving only those who are sympathetic to the Affero terms, then you can just as well add the additional Affero terms to the GPL3 itself and simplify things.<br><br></p><p>\nOf course the same argument can be made from the Affero side. Why would someone who likes the Affero restrictions accept an Affero GPL 2 that permits relicensing of a combined work under GPL3? So the Affero GPL 2 can't do that.<br><br></p><p>\nThe end result is that if either license allows the relicensing of a combined work (even if the combination is only by linking) under the other license, it becomes unacceptable to its followers. So I'd like to reiterate my question: How is this supposed to work?<br><br></p><p>\nI'm puzzled, and that's a bad thing. As I've said elsewhere, I will not relicense my GPL2 works under the GPL3 if I do not have confidence that I understand the GPL3.<br><br></p><p>\nI trust the GPL2. I believe I understand it. I especially trust the copyleft provision of the GPL2. I know that if someone combines my contribution with something else, the result MUST be licensed under the GPL2, too. With this new section 13 I'm just not sure. And if you do not manage to create a GPL3 that gives me the same feeling of confidence I have with the GPL2, I won't use it. And I'm sure I'm speaking for many hobbyists here.</p>\ngpl3.affero.p0.s2\nmux2005\nlogin to agree\n1\n3129\n\n\n", "2550:3129": "\n\n\n\n<p style="display: inline">This part of Section 13 seems to contradict the known historical teaching\nfrom FSF regarding GPL compatibility, and the "additional restrictions"\nlanguage in GPLv3DD2 Section 7. Generally speaking, under GPLv2, it was\nnot possible to create a new derivative work that combined GPLv2 works\nwith those under GPLv2-incompatible licenses.</p><p>\nA GPLv2 incompatible license was historically defined as "a licenses which\nplaces additional restrictions not permitted by GPL". GPLv3 S. 7 was\ndesigned to increase the scope of that compatibility, but it still clearly\nstates "All other non-permissive additional terms are considered 'further\nrestrictions'".</p><p>\nAs I see it, the core AGPL requirement (that source code be made available\nvia to all network users of the software) will constitute an "additional\nrestriction" as described in GPLv3 Section 10. I realize that AGPLv2\nhasn't been written yet (and I've volunteer to serve on the drafting\ncommittee for it), but I simply don't see a way that you can reach the\nclear goal of the AGPL while not imposing a "further restriction" as\ndefined in this GPLv3DD2.</p><p>\nIt appears that GPLv3DD2's S. 13 is designed to act somewhat like its own\nSection 7: allowing a certain type of additional restriction. However, it\ndoes not go far enough. It seems to imply that you can keep some files in\nlarger derivative work under GPLv3 and others under AGPLv2. In that case,\nwhat is the license of the whole work? It can't be GPLv2, because that\nwill contradict the Affero network-source-distribution requirement. It\ncan't be AGPLv2, because it will contradict GPLv3's Sections 7 and 10.</p><p>\nIn other words, this text contradicts this historical thinking on what a\nFree Software copyleft license must do. Generally, in the Free Software\nworld, we've believed that even while a particular license notice lives\nwith each file in the larger work, the entire copyrighted work has a\nsingle known license. This is precisely why GPL-compatibility always\nmeant "the other license allows the conveyor to distribute the non-GPL'd\nwork under full terms of the GPL".</p><p>\nMost will agree that even in a GPLv3/AGPLv2 compatibility situation that\neach file in the work should carry their own license notices referring to\nthe individual licenses; that's standard Free Software\nmodification/redistribution practice anyway. However, the work as a whole\nmust be under a single license. It seems that GPLv3 S. 13 simply doesn't\ngo far enough to make the core AGPL provision compatible with GPLv3, such\nthat the entire license of the work can be stated clearly without\ncontradictions.\n</p>\ngpl3.affero.p0.s2\nbkuhn\nlogin to agree\n0\n2550\n\n\n<p style="display: inline">I don't understand how this is supposed to work. The GPL says that if I create a work based on a GPL-covered work, the result must be GPL-licensed. The Affero GPL says the same thing about the Affero GPL, doesn't it? So which of the 2 licenses applies to my combined work? <br><br></p><p>\nThis section 13 seems to say that the GPL applies to the combined work. And it has to be like this, because otherwise the GPL3 would be unacceptable to people who don't like the Affero GPL, because these people will not accept a GPL3 that allows relicensing a combined work under Affero. And if you alienate those people and cause them to not use the GPL3, leaving only those who are sympathetic to the Affero terms, then you can just as well add the additional Affero terms to the GPL3 itself and simplify things.<br><br></p><p>\nOf course the same argument can be made from the Affero side. Why would someone who likes the Affero restrictions accept an Affero GPL 2 that permits relicensing of a combined work under GPL3? So the Affero GPL 2 can't do that.<br><br></p><p>\nThe end result is that if either license allows the relicensing of a combined work (even if the combination is only by linking) under the other license, it becomes unacceptable to its followers. So I'd like to reiterate my question: How is this supposed to work?<br><br></p><p>\nI'm puzzled, and that's a bad thing. As I've said elsewhere, I will not relicense my GPL2 works under the GPL3 if I do not have confidence that I understand the GPL3.<br><br></p><p>\nI trust the GPL2. I believe I understand it. I especially trust the copyleft provision of the GPL2. I know that if someone combines my contribution with something else, the result MUST be licensed under the GPL2, too. With this new section 13 I'm just not sure. And if you do not manage to create a GPL3 that gives me the same feeling of confidence I have with the GPL2, I won't use it. And I'm sure I'm speaking for many hobbyists here.</p>\ngpl3.affero.p0.s2\nmux2005\nlogin to agree\n1\n3129\n\n\n", "2830:3129": "\n\n\n\n<p style="display: inline">Web hole is open with this clause if Affero v2 includes a reciprocal clause that allow linking the code with GPLv3, with each part under its respective license. \n\nPeople could extend the Affero program using subroutines under GPLv3 license and don't show/distribute them. This is the same problem with web hole than MPL with proprietary code: MPL allow linking with proprietary code and for this reason its effectiveness copyleft is void.</p><p>\nWeb hole will be limited if AfferoV2 dose not allow to link with GPLv3 and this requires a explicit permission from copyright folder. In this case, copyrigth holder of a AfferoV2 could use a component under GPLv3 including a explicit clause that allows linking with this library if not alter the interface and the GPLv3 component is not altered to modify internal structures of the AfferoV2 program.</p>\ngpl3.affero.p0.s2\nchemaper\nlogin to agree\n0\n2830\n\n\n<p style="display: inline">I don't understand how this is supposed to work. The GPL says that if I create a work based on a GPL-covered work, the result must be GPL-licensed. The Affero GPL says the same thing about the Affero GPL, doesn't it? So which of the 2 licenses applies to my combined work? <br><br></p><p>\nThis section 13 seems to say that the GPL applies to the combined work. And it has to be like this, because otherwise the GPL3 would be unacceptable to people who don't like the Affero GPL, because these people will not accept a GPL3 that allows relicensing a combined work under Affero. And if you alienate those people and cause them to not use the GPL3, leaving only those who are sympathetic to the Affero terms, then you can just as well add the additional Affero terms to the GPL3 itself and simplify things.<br><br></p><p>\nOf course the same argument can be made from the Affero side. Why would someone who likes the Affero restrictions accept an Affero GPL 2 that permits relicensing of a combined work under GPL3? So the Affero GPL 2 can't do that.<br><br></p><p>\nThe end result is that if either license allows the relicensing of a combined work (even if the combination is only by linking) under the other license, it becomes unacceptable to its followers. So I'd like to reiterate my question: How is this supposed to work?<br><br></p><p>\nI'm puzzled, and that's a bad thing. As I've said elsewhere, I will not relicense my GPL2 works under the GPL3 if I do not have confidence that I understand the GPL3.<br><br></p><p>\nI trust the GPL2. I believe I understand it. I especially trust the copyleft provision of the GPL2. I know that if someone combines my contribution with something else, the result MUST be licensed under the GPL2, too. With this new section 13 I'm just not sure. And if you do not manage to create a GPL3 that gives me the same feeling of confidence I have with the GPL2, I won't use it. And I'm sure I'm speaking for many hobbyists here.</p>\ngpl3.affero.p0.s2\nmux2005\nlogin to agree\n1\n3129\n\n\n", "2550:2711:2828:2830:3128:3129": "\n\n\n\n<p style="display: inline">This part of Section 13 seems to contradict the known historical teaching\nfrom FSF regarding GPL compatibility, and the "additional restrictions"\nlanguage in GPLv3DD2 Section 7. Generally speaking, under GPLv2, it was\nnot possible to create a new derivative work that combined GPLv2 works\nwith those under GPLv2-incompatible licenses.</p><p>\nA GPLv2 incompatible license was historically defined as "a licenses which\nplaces additional restrictions not permitted by GPL". GPLv3 S. 7 was\ndesigned to increase the scope of that compatibility, but it still clearly\nstates "All other non-permissive additional terms are considered 'further\nrestrictions'".</p><p>\nAs I see it, the core AGPL requirement (that source code be made available\nvia to all network users of the software) will constitute an "additional\nrestriction" as described in GPLv3 Section 10. I realize that AGPLv2\nhasn't been written yet (and I've volunteer to serve on the drafting\ncommittee for it), but I simply don't see a way that you can reach the\nclear goal of the AGPL while not imposing a "further restriction" as\ndefined in this GPLv3DD2.</p><p>\nIt appears that GPLv3DD2's S. 13 is designed to act somewhat like its own\nSection 7: allowing a certain type of additional restriction. However, it\ndoes not go far enough. It seems to imply that you can keep some files in\nlarger derivative work under GPLv3 and others under AGPLv2. In that case,\nwhat is the license of the whole work? It can't be GPLv2, because that\nwill contradict the Affero network-source-distribution requirement. It\ncan't be AGPLv2, because it will contradict GPLv3's Sections 7 and 10.</p><p>\nIn other words, this text contradicts this historical thinking on what a\nFree Software copyleft license must do. Generally, in the Free Software\nworld, we've believed that even while a particular license notice lives\nwith each file in the larger work, the entire copyrighted work has a\nsingle known license. This is precisely why GPL-compatibility always\nmeant "the other license allows the conveyor to distribute the non-GPL'd\nwork under full terms of the GPL".</p><p>\nMost will agree that even in a GPLv3/AGPLv2 compatibility situation that\neach file in the work should carry their own license notices referring to\nthe individual licenses; that's standard Free Software\nmodification/redistribution practice anyway. However, the work as a whole\nmust be under a single license. It seems that GPLv3 S. 13 simply doesn't\ngo far enough to make the core AGPL provision compatible with GPLv3, such\nthat the entire license of the work can be stated clearly without\ncontradictions.\n</p>\ngpl3.affero.p0.s2\nbkuhn\nlogin to agree\n0\n2550\n\n\nIf there is going to be another license running in parallel with the GPL, please try to get the numbering in sync. (Skip AGPLv2, and make it AGPLv3) It's not going to make much difference right now, but after another couple of releases it will just save everyone headaches. (Take a look at Mozilla version numbering--it's not pretty.)\ngpl3.affero.p0.s1\njamesgnz\nlogin to agree\n4\n2711\n\n\n<p style="display: inline">The current AGPL is not adequate to protect developers from the "software as a service loophole." According to section 2d the work must be a program that is designed to be accessed over a network and which is able to download its source code to the user.</p><p>\nMy software would not apply since it is a C++ class library and not a program. Using my software would make it easy to develop a web application, and that is one of its goals, but it itself is just a collection of classes. In version 2 of the AGPL, I urge you to require the service provider to offer the source code, not the "program" itself.</p><p>\nMike\n</p>\ngpl3.affero.p0.s1\nmiked\nlogin to agree\n0\n2828\n\n\n<p style="display: inline">Web hole is open with this clause if Affero v2 includes a reciprocal clause that allow linking the code with GPLv3, with each part under its respective license. \n\nPeople could extend the Affero program using subroutines under GPLv3 license and don't show/distribute them. This is the same problem with web hole than MPL with proprietary code: MPL allow linking with proprietary code and for this reason its effectiveness copyleft is void.</p><p>\nWeb hole will be limited if AfferoV2 dose not allow to link with GPLv3 and this requires a explicit permission from copyright folder. In this case, copyrigth holder of a AfferoV2 could use a component under GPLv3 including a explicit clause that allows linking with this library if not alter the interface and the GPLv3 component is not altered to modify internal structures of the AfferoV2 program.</p>\ngpl3.affero.p0.s2\nchemaper\nlogin to agree\n0\n2830\n\n\n<p style="display: inline">You need to define what "to link" means. This is a very overloaded term that can have a lot of different meanings in different contexts. The biggest problem is that it has non-technical meanings (which are the only meanings that most judges and lawyers will be acquainted with).<br><br></p><p>\nSince the GPL has been phrased to apply to "works" in general, encompassing more than just programs, it is not clear that here "to link" is to refer to the linking of programs only. Or maybe it isn't? In that case defining "to link" is even more necessary, because in the context of non-program works (e.g. icon sets, which under Windows for instance, are confusingly often stored in DLLs, i.e. dynamic LINK libraries) "to link" can mean a lot of things.<br><br></p><p>\nThen of course there's the problem that even in the technical world "to link" is far from being well-defined. There are a lot of corner-cases, arising for instance in the context of virtual machines and compilation/linking at run-time.</p>\ngpl3.affero.p0.s1\nmux2005\nlogin to agree\n0\n3128\n\n\n<p style="display: inline">I don't understand how this is supposed to work. The GPL says that if I create a work based on a GPL-covered work, the result must be GPL-licensed. The Affero GPL says the same thing about the Affero GPL, doesn't it? So which of the 2 licenses applies to my combined work? <br><br></p><p>\nThis section 13 seems to say that the GPL applies to the combined work. And it has to be like this, because otherwise the GPL3 would be unacceptable to people who don't like the Affero GPL, because these people will not accept a GPL3 that allows relicensing a combined work under Affero. And if you alienate those people and cause them to not use the GPL3, leaving only those who are sympathetic to the Affero terms, then you can just as well add the additional Affero terms to the GPL3 itself and simplify things.<br><br></p><p>\nOf course the same argument can be made from the Affero side. Why would someone who likes the Affero restrictions accept an Affero GPL 2 that permits relicensing of a combined work under GPL3? So the Affero GPL 2 can't do that.<br><br></p><p>\nThe end result is that if either license allows the relicensing of a combined work (even if the combination is only by linking) under the other license, it becomes unacceptable to its followers. So I'd like to reiterate my question: How is this supposed to work?<br><br></p><p>\nI'm puzzled, and that's a bad thing. As I've said elsewhere, I will not relicense my GPL2 works under the GPL3 if I do not have confidence that I understand the GPL3.<br><br></p><p>\nI trust the GPL2. I believe I understand it. I especially trust the copyleft provision of the GPL2. I know that if someone combines my contribution with something else, the result MUST be licensed under the GPL2, too. With this new section 13 I'm just not sure. And if you do not manage to create a GPL3 that gives me the same feeling of confidence I have with the GPL2, I won't use it. And I'm sure I'm speaking for many hobbyists here.</p>\ngpl3.affero.p0.s2\nmux2005\nlogin to agree\n1\n3129\n\n\n", "2713": "\n\n\n\nSurely this section should be placed as a note after the end of terms and conditions?\ngpl3.revisedversions.0.0\njamesgnz\nlogin to agree\n0\n2713\n\n\n", "2713:2781": "\n\n\n\nSurely this section should be placed as a note after the end of terms and conditions?\ngpl3.revisedversions.0.0\njamesgnz\nlogin to agree\n0\n2713\n\n\nThis section is unchanged from previous drafts, and similar to section 9 of GPLv2 (apart from minor rephrasing and the addition of the final paragraph). It's good that licensors have the option of specifying a certain version of the GNU GPL "or any later version", but are not forced to do so. In other words, the upgradeability of the license is a permission that is granted outside the license text and that is a good feature to keep.\ngpl3.revisedversions.0.0\nfrx\nlogin to agree\n0\n2781\n\n\n", "2727:2805:2807:3001": "\n\n\n\nThis kind of bloat should not be in the license. The copyright holder already knows they do this.\ngpl3.revisedversions.p2.s1\nflaschen\nlogin to agree\n2\n2727\n\n\n<p style="display: inline">I believe this opens a potential danger. Not towards freedom, but the fact that proxys are specifically mentioned here, but not elsewhere, is slightly alarming.</p><p>\nA proxy should have any and/or all of the rights granted in this license, if so granted by the copyright holder, it should be applied to the entire license, or removed from this section as it can potentially limit a copyright holder's right to enforce protections through a proxy.</p><p>\nFor instance, a proxy is very clearly given the right to accept the next version of the GPL, but only the proxy agreement ensures that a proxy has the right to offer patent protection as required previously in this license. A proxy could license a work under v3 without having the legal authority to allow covered patents to be used.</p>\ngpl3.revisedversions.p2.s1\nkdean06\nlogin to agree\n1\n2805\n\n\n<p style="display: inline">Who can be a proxy? How is it specified? How does this interact with ownership of the copyright? Is the proxy designation revocable? Can a proxy be specified after-the-fact?</p><p>\nI think this idea is potentially interesting, but needs significant grounding before including it.\n</p>\ngpl3.revisedversions.p2.s1\nsepreece\nlogin to agree\n1\n2807\n\n\n<p style="display: inline">This appears to be poorly thought out bloat license and should be removed. The How to apply this license section doesn't describe the text needed to specify a proxy. What constitutes a proxy, how a proxy might be designated, when a proxy might be designated, what a proxy might be, all are undefined. Can code be released under GPL 3 and any later version and a proxy specfied who can decide that "any later version" doesn't apply to GPL 5 - can a proxy limit the license, or can a proxy only be specified if the code is licensed under the GPL 3 only... What happens if a public statement accepting a version is made by someone who appears to be the proxy but the proxy later makes a statement denying this acceptance?</p><p>\nThe choice by the copyright holders of GPL3 only or GPL3+any later is simple and clear. Stick with simple. There is already too much complex added language in this version. </p>\ngpl3.revisedversions.p2.s1\nmole\nlogin to agree\n0\n3001\n\n\n", "2807:3001": "\n\n\n\n<p style="display: inline">Who can be a proxy? How is it specified? How does this interact with ownership of the copyright? Is the proxy designation revocable? Can a proxy be specified after-the-fact?</p><p>\nI think this idea is potentially interesting, but needs significant grounding before including it.\n</p>\ngpl3.revisedversions.p2.s1\nsepreece\nlogin to agree\n1\n2807\n\n\n<p style="display: inline">This appears to be poorly thought out bloat license and should be removed. The How to apply this license section doesn't describe the text needed to specify a proxy. What constitutes a proxy, how a proxy might be designated, when a proxy might be designated, what a proxy might be, all are undefined. Can code be released under GPL 3 and any later version and a proxy specfied who can decide that "any later version" doesn't apply to GPL 5 - can a proxy limit the license, or can a proxy only be specified if the code is licensed under the GPL 3 only... What happens if a public statement accepting a version is made by someone who appears to be the proxy but the proxy later makes a statement denying this acceptance?</p><p>\nThe choice by the copyright holders of GPL3 only or GPL3+any later is simple and clear. Stick with simple. There is already too much complex added language in this version. </p>\ngpl3.revisedversions.p2.s1\nmole\nlogin to agree\n0\n3001\n\n\n", "3001": "\n\n\n\n<p style="display: inline">This appears to be poorly thought out bloat license and should be removed. The How to apply this license section doesn't describe the text needed to specify a proxy. What constitutes a proxy, how a proxy might be designated, when a proxy might be designated, what a proxy might be, all are undefined. Can code be released under GPL 3 and any later version and a proxy specfied who can decide that "any later version" doesn't apply to GPL 5 - can a proxy limit the license, or can a proxy only be specified if the code is licensed under the GPL 3 only... What happens if a public statement accepting a version is made by someone who appears to be the proxy but the proxy later makes a statement denying this acceptance?</p><p>\nThe choice by the copyright holders of GPL3 only or GPL3+any later is simple and clear. Stick with simple. There is already too much complex added language in this version. </p>\ngpl3.revisedversions.p2.s1\nmole\nlogin to agree\n0\n3001\n\n\n", "2808:3001": "\n\n\n\n<p style="display: inline">This provision works only in the case that only one party has copyright for a work, i think.</p><p>\n(case a)\nIf someone add this proxy clause to their modified version, and previous licensors do not agree, what will happen?</p><p>\n(case b)\nIf original version has this proxy clause, and someone develop a modified version without such proxy clause, is it violation?</p>\ngpl3.revisedversions.p2.s1\nyusuke\nlogin to agree\n0\n2808\n\n\n<p style="display: inline">This appears to be poorly thought out bloat license and should be removed. The How to apply this license section doesn't describe the text needed to specify a proxy. What constitutes a proxy, how a proxy might be designated, when a proxy might be designated, what a proxy might be, all are undefined. Can code be released under GPL 3 and any later version and a proxy specfied who can decide that "any later version" doesn't apply to GPL 5 - can a proxy limit the license, or can a proxy only be specified if the code is licensed under the GPL 3 only... What happens if a public statement accepting a version is made by someone who appears to be the proxy but the proxy later makes a statement denying this acceptance?</p><p>\nThe choice by the copyright holders of GPL3 only or GPL3+any later is simple and clear. Stick with simple. There is already too much complex added language in this version. </p>\ngpl3.revisedversions.p2.s1\nmole\nlogin to agree\n0\n3001\n\n\n", "2524:2808:3001": "\n\n\n\nClarify this to say "That proxy's public statement of acceptance for you or anybody else to use a particular version is permanent authorization for you to choose that version for the work in question." \n\nThis is necessary because currently it seems to require that the proxy gives you permission to use ANY version, and also, it doesn't make it clear that this is allowed if he expresses acceptance for the program to be put under that version rather than just expressing acceptance of teh license in the "yea that is a good license" kind of way. \ngpl3.revisedversions.p2.s1\njring\nlogin to agree\n2\n2524\n\n\n<p style="display: inline">This provision works only in the case that only one party has copyright for a work, i think.</p><p>\n(case a)\nIf someone add this proxy clause to their modified version, and previous licensors do not agree, what will happen?</p><p>\n(case b)\nIf original version has this proxy clause, and someone develop a modified version without such proxy clause, is it violation?</p>\ngpl3.revisedversions.p2.s1\nyusuke\nlogin to agree\n0\n2808\n\n\n<p style="display: inline">This appears to be poorly thought out bloat license and should be removed. The How to apply this license section doesn't describe the text needed to specify a proxy. What constitutes a proxy, how a proxy might be designated, when a proxy might be designated, what a proxy might be, all are undefined. Can code be released under GPL 3 and any later version and a proxy specfied who can decide that "any later version" doesn't apply to GPL 5 - can a proxy limit the license, or can a proxy only be specified if the code is licensed under the GPL 3 only... What happens if a public statement accepting a version is made by someone who appears to be the proxy but the proxy later makes a statement denying this acceptance?</p><p>\nThe choice by the copyright holders of GPL3 only or GPL3+any later is simple and clear. Stick with simple. There is already too much complex added language in this version. </p>\ngpl3.revisedversions.p2.s1\nmole\nlogin to agree\n0\n3001\n\n\n", "2524:2727:2805:2807:2808:3001": "\n\n\n\nClarify this to say "That proxy's public statement of acceptance for you or anybody else to use a particular version is permanent authorization for you to choose that version for the work in question." \n\nThis is necessary because currently it seems to require that the proxy gives you permission to use ANY version, and also, it doesn't make it clear that this is allowed if he expresses acceptance for the program to be put under that version rather than just expressing acceptance of teh license in the "yea that is a good license" kind of way. \ngpl3.revisedversions.p2.s1\njring\nlogin to agree\n2\n2524\n\n\nThis kind of bloat should not be in the license. The copyright holder already knows they do this.\ngpl3.revisedversions.p2.s1\nflaschen\nlogin to agree\n2\n2727\n\n\n<p style="display: inline">I believe this opens a potential danger. Not towards freedom, but the fact that proxys are specifically mentioned here, but not elsewhere, is slightly alarming.</p><p>\nA proxy should have any and/or all of the rights granted in this license, if so granted by the copyright holder, it should be applied to the entire license, or removed from this section as it can potentially limit a copyright holder's right to enforce protections through a proxy.</p><p>\nFor instance, a proxy is very clearly given the right to accept the next version of the GPL, but only the proxy agreement ensures that a proxy has the right to offer patent protection as required previously in this license. A proxy could license a work under v3 without having the legal authority to allow covered patents to be used.</p>\ngpl3.revisedversions.p2.s1\nkdean06\nlogin to agree\n1\n2805\n\n\n<p style="display: inline">Who can be a proxy? How is it specified? How does this interact with ownership of the copyright? Is the proxy designation revocable? Can a proxy be specified after-the-fact?</p><p>\nI think this idea is potentially interesting, but needs significant grounding before including it.\n</p>\ngpl3.revisedversions.p2.s1\nsepreece\nlogin to agree\n1\n2807\n\n\n<p style="display: inline">This provision works only in the case that only one party has copyright for a work, i think.</p><p>\n(case a)\nIf someone add this proxy clause to their modified version, and previous licensors do not agree, what will happen?</p><p>\n(case b)\nIf original version has this proxy clause, and someone develop a modified version without such proxy clause, is it violation?</p>\ngpl3.revisedversions.p2.s1\nyusuke\nlogin to agree\n0\n2808\n\n\n<p style="display: inline">This appears to be poorly thought out bloat license and should be removed. The How to apply this license section doesn't describe the text needed to specify a proxy. What constitutes a proxy, how a proxy might be designated, when a proxy might be designated, what a proxy might be, all are undefined. Can code be released under GPL 3 and any later version and a proxy specfied who can decide that "any later version" doesn't apply to GPL 5 - can a proxy limit the license, or can a proxy only be specified if the code is licensed under the GPL 3 only... What happens if a public statement accepting a version is made by someone who appears to be the proxy but the proxy later makes a statement denying this acceptance?</p><p>\nThe choice by the copyright holders of GPL3 only or GPL3+any later is simple and clear. Stick with simple. There is already too much complex added language in this version. </p>\ngpl3.revisedversions.p2.s1\nmole\nlogin to agree\n0\n3001\n\n\n", "2498:2525:2681:2782": "\n\n\n\nIt's upper-case again, why is this needed?\ngpl3.nowarranty.0.0\nthomas\nlogin to agree\n5\n2498\n\n\nMake it clear that this disclaimer applies even if you don't redistribute the program. Currently the license states you don't have to accept it in order to run the program, which invalidates the disclaimer. \ngpl3.nowarranty.0.0\njring\nlogin to agree\n7\n2525\n\n\nSince another warranty (disclaimer) statement can accompany the works (and potentially all derivative works under supplemental terms allowed by section 7a), why not explicitely state: "Unless another warranty (disclaimer) statement accompanies the work, and unless you have a warranty agreement, this default Disclaimer of Warranty applies:".\nIdem for the liability (limitation) statement.\ngpl3.nowarranty.0.0\nadhemar\nlogin to agree\n1\n2681\n\n\nThis section seems to be virtually the same as the joining of sections 11 and 12 of GPLv2, except for the last paragraph, which is a good explicit clarification. It looks fine to me.\ngpl3.nowarranty.0.0\nfrx\nlogin to agree\n0\n2782\n\n\n", "2740": "\n\n\n\nThis may be silly, but what if you don't want to fix it? What if someone else fixes it? This sentence seems to suggest that if the software is broken, you _must_ pay to have it fixed, if it ever gets fixed.\ngpl3.nowarranty.p0.s4\nmkorman\nlogin to agree\n1\n2740\n\n\n", "3130": "\n\n\n\nI've written a similar comment elsewhere. In our globalized world and in the context of multi-national corporations the word "local" does not have a well-defined meaning. If a liability dispute exists between entities in different jurisdictions, things get especially messy, because the GPL does not (and can not) have a choice of law provision. The GPL should not add to this mess. I think in this context here it's easiest to specify that "local" refers to the jurisdiction of the reviewing court, irrespective of where the parties are located and where the dispute arises. Or maybe the word "local" can simply be removed altoghether. If I read this paragraph without "local" in it, I think it still works (and better than with "local").\ngpl3.nowarranty.p2.s1\nmux2005\nlogin to agree\n0\n3130\n\n\n", "2526": "\n\n\n\nRemove the word "civil" as it could have implication in some legislatures that distinguishes between different types of legislation. \ngpl3.nowarranty.p2.s1\njring\nlogin to agree\n4\n2526\n\n\n", "3002": "\n\n\n\nDoes a copy mean any copy anywhere, or a copy of the program conveyed to the person who suffered damages?\ngpl3.nowarranty.p2.s1\nmole\nlogin to agree\n0\n3002\n\n\n", "2470": "\n\n\n\nWhat if some distributor is foolish enough to offer a warranty at no cost? It would be suicidal for small businesses, but larger ones could be bold enough to assume liability in certain (limited) cases without requiring any extra fee.\ngpl3.nowarranty.p2.s1\nraphael\nlogin to agree\n6\n2470\n\n\n", "2470:2526:3002:3130": "\n\n\n\nWhat if some distributor is foolish enough to offer a warranty at no cost? It would be suicidal for small businesses, but larger ones could be bold enough to assume liability in certain (limited) cases without requiring any extra fee.\ngpl3.nowarranty.p2.s1\nraphael\nlogin to agree\n6\n2470\n\n\nRemove the word "civil" as it could have implication in some legislatures that distinguishes between different types of legislation. \ngpl3.nowarranty.p2.s1\njring\nlogin to agree\n4\n2526\n\n\nDoes a copy mean any copy anywhere, or a copy of the program conveyed to the person who suffered damages?\ngpl3.nowarranty.p2.s1\nmole\nlogin to agree\n0\n3002\n\n\nI've written a similar comment elsewhere. In our globalized world and in the context of multi-national corporations the word "local" does not have a well-defined meaning. If a liability dispute exists between entities in different jurisdictions, things get especially messy, because the GPL does not (and can not) have a choice of law provision. The GPL should not add to this mess. I think in this context here it's easiest to specify that "local" refers to the jurisdiction of the reviewing court, irrespective of where the parties are located and where the dispute arises. Or maybe the word "local" can simply be removed altoghether. If I read this paragraph without "local" in it, I think it still works (and better than with "local").\ngpl3.nowarranty.p2.s1\nmux2005\nlogin to agree\n0\n3130\n\n\n", "2648": "\n\n\n\n<p style="display: inline">In some countries you cannot disclaim liability for death, it is therefore normal to add something about safety critical software. The GPL could sanely and sensibly add this with</p><p>\n"Unless specifically stated the recipient should not assume the work has been tested and proved suitable for use in a safety critical environment"</p>\ngpl3.nowarranty.p3.s1\nAlanCox\nlogin to agree\n3\n2648\n\n\n", "2554:2869": "\n\n\n\n<p style="display: inline">As noted in the other remarks about the various names used in this section it's arguably better to either use bland but descriptive names (eg. MyApplication, FirstName, FamilyName) than to expect people, particularly non-native English speakers to pick up subtle references. It might be best to just use bracketed terms like <application> throughout in order to more easily guide people around the parts that need parameterising.</p><p>\nGuidance is needed about whether this section should be included in the licence file or not. My impression is that this is optional "metadata", but I've seen many people include this text in their licence file, even though its relevance is diminished: the licence is already applied, and advice about using it - whilst desirable - is a peripheral matter. The text should say that this section is optional unless otherwise stipulated somewhere: in which case, such stipulations should appear in this section.</p>\nhowtoapply.0.0\npboddie\nlogin to agree\n5\n2554\n\n\nThis advice brings no relevant information to the License. Should it stay only in its own page at FSF site and here to have just a pointer to it, the License would be more simple.\nStaying in its page, also has the advantage to be actualized independent of the License\nhowtoapply.0.0\nlcchueri\nlogin to agree\n1\n2869\n\n\n", "2929": "\n\n\n\n<p style="display: inline">\nThanks,\nWard.</p><p>\n--\nWard Vandewege <ward@fsf.org>\nFree Software Foundation - Senior System Administrator\n\n</p>\ngpl3.howtoapply.p0.s1\nward1\nlogin to agree\n0\n2929\n\n\n", "2622": "\n\n\n\n<program> <summary>\ngpl3.howtoapply.p2.s1\nsamj\nlogin to agree\n0\n2622\n\n\n", "2623": "\n\n\n\nLower case (c)?\ngpl3.howtoapply.p2.s1\nsamj\nlogin to agree\n1\n2623\n\n\n", "2622:2623": "\n\n\n\n<program> <summary>\ngpl3.howtoapply.p2.s1\nsamj\nlogin to agree\n0\n2622\n\n\nLower case (c)?\ngpl3.howtoapply.p2.s1\nsamj\nlogin to agree\n1\n2623\n\n\n", "3047": "\n\n\n\nThis package is free; you can use, copy, distribute, study, change and\nimprove it under the terms of the GNU General Public License as published by\nthe Free Software Foundation; either version 2 of the License, or (at your\noption) any later version.\n\ngpl3.howtoapply.p3.s1\nnslater\nlogin to agree\n0\n3047\n\n\n", "2442:2624:3000:3023:3024:3046:3047": "\n\n\n\nMany people don't know that they can omit this part to make their program GPL v3 only. Please consider adding a sentence or two explaining this point.\ngpl3.howtoapply.p3.s1\ntcort\nlogin to agree\n1\n2442\n\n\nRemove '(at your option)'?\ngpl3.howtoapply.p3.s1\nsamj\nlogin to agree\n0\n2624\n\n\n<p style="display: inline">s/under the terms of the GNU General Public License/under the terms of [ version 3 of | version three or any later version of] the GNU General Public License/</p><p>\ns/; either version 3 of the License, or (at your option) any later version./. <BR> At your option specify whether only version 3 of the license applies, or whether version 3 or any subsequent version applies./</p>\ngpl3.howtoapply.p3.s1\nmole\nlogin to agree\n0\n3000\n\n\n<p style="display: inline">I think this section would read better if it emphasis the benefits of the GPL. Something like this:</p><p>\nThis package is free; you can use, copy, distribute, study, change and improve it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.</p>\ngpl3.howtoapply.p3.s1\nnslater\nlogin to agree\n0\n3023\n\n\n<p style="display: inline">This section should read like the following:</p><p>\nThis package is free; you can use, copy, distribute, study, change and\nimprove it under the terms of the GNU General Public License as published by\nthe Free Software Foundation; either version 2 of the License, or (at your\noption) any later version.</p>\ngpl3.howtoapply.p3.s1\nnslater\nlogin to agree\n0\n3024\n\n\nThis package is free; you can use, copy, distribute, study, change and\nimprove it under the terms of the GNU General Public License as published by\nthe Free Software Foundation; either version 2 of the License, or (at your\noption) any later version.\ngpl3.howtoapply.p3.s1\nnslater\nlogin to agree\n0\n3046\n\n\nThis package is free; you can use, copy, distribute, study, change and\nimprove it under the terms of the GNU General Public License as published by\nthe Free Software Foundation; either version 2 of the License, or (at your\noption) any later version.\n\ngpl3.howtoapply.p3.s1\nnslater\nlogin to agree\n0\n3047\n\n\n", "2527": "\n\n\n\nThis is not necessary. Keep it as short as possible. \ngpl3.howtoapply.p4.s1\njring\nlogin to agree\n3\n2527\n\n\n", "3136": "\n\n\n\nIf this is intended to be a bare licence, this wording is confusing. In English law, Warranties can only be implied into agreements, not bare licences, and the very act of asserting that no warranty is implied could have precisely the opposite effect to that intended: i.e. the document tries to exclude warranties, which means it must be an agreement, which means that the exclusions may not be effective under (1) common law; (2) the Unfair Contract Terms Act 1977; or (3) the Unfair Terms in Consumer Contracts Regulations 1999.\ngpl3.howtoapply.p4.s1\nandrewk\nlogin to agree\n0\n3136\n\n\n", "2528": "\n\n\n\n"GNU General Public License version 3"\ngpl3.howtoapply.p4.s2\njring\nlogin to agree\n4\n2528\n\n\n", "2528:3136": "\n\n\n\n"GNU General Public License version 3"\ngpl3.howtoapply.p4.s2\njring\nlogin to agree\n4\n2528\n\n\nIf this is intended to be a bare licence, this wording is confusing. In English law, Warranties can only be implied into agreements, not bare licences, and the very act of asserting that no warranty is implied could have precisely the opposite effect to that intended: i.e. the document tries to exclude warranties, which means it must be an agreement, which means that the exclusions may not be effective under (1) common law; (2) the Unfair Contract Terms Act 1977; or (3) the Unfair Terms in Consumer Contracts Regulations 1999.\ngpl3.howtoapply.p4.s1\nandrewk\nlogin to agree\n0\n3136\n\n\n", "2436:2617:2939": "\n\n\n\nThis URL is also missing a trailing slash and should read "http://www.gnu.org/licenses/".\ngpl3.howtoapply.p5.s1\nfraggle\nlogin to agree\n7\n2436\n\n\nPlease drop the superfluous 'www.'\ngpl3.howtoapply.p5.s1\nsamj\nlogin to agree\n1\n2617\n\n\n<p style="display: inline">The proper format of this URL is "http://www.gnu.org/licenses/". The "www" is usually superfluous, but it seems that GNU's website is always redirected to a URL that includes it. That would seem improper to me, but that's a whole other discussion.</p><p>\nThe trailing slash, however, should still be there. This link points to a directory; when following it, the GNU server will redirect you to the URL that includes the slash.</p>\ngpl3.howtoapply.p5.s1\nmsikma\nlogin to agree\n1\n2939\n\n\n", "2728:3131": "\n\n\n\nThis should be expanded so it isn't specific to one kind of interface.\ngpl3.howtoapply.p7.s1\nflaschen\nlogin to agree\n0\n2728\n\n\n<p style="display: inline">Terminal interaction, especially of the kind suggested by "show w" and "show c" is a bit antiquated. Furthermore, AFAICT aside from some console programs by the FSF (not even all of them), few programs under the GPL seem to have features like this. So this is dead weight (see also my comments that the part requiring preservation of such notices should be struck). It should either be removed or replaced with something more up-to-date such as:<br><br></p><p>\nIf the program has a graphical user interface with an About box, make it include a short notice like this:\n....\n</p>\ngpl3.howtoapply.p7.s1\nmux2005\nlogin to agree\n0\n3131\n\n\n", "2705": "\n\n\n\nAdding these notices to a program is optional, and the wording here should reflect that. As it is currently worded, it reads as though adding a notice like this is necessary for the program being under the GPL.\ngpl3.howtoapply.p7.s1\nn0dalus\nlogin to agree\n1\n2705\n\n\n", "2940": "\n\n\n\nI suggest "like the following paragraph".\ngpl3.howtoapply.p7.s1\nmsikma\nlogin to agree\n1\n2940\n\n\n", "2705:2728:2940:3131": "\n\n\n\nAdding these notices to a program is optional, and the wording here should reflect that. As it is currently worded, it reads as though adding a notice like this is necessary for the program being under the GPL.\ngpl3.howtoapply.p7.s1\nn0dalus\nlogin to agree\n1\n2705\n\n\nThis should be expanded so it isn't specific to one kind of interface.\ngpl3.howtoapply.p7.s1\nflaschen\nlogin to agree\n0\n2728\n\n\nI suggest "like the following paragraph".\ngpl3.howtoapply.p7.s1\nmsikma\nlogin to agree\n1\n2940\n\n\n<p style="display: inline">Terminal interaction, especially of the kind suggested by "show w" and "show c" is a bit antiquated. Furthermore, AFAICT aside from some console programs by the FSF (not even all of them), few programs under the GPL seem to have features like this. So this is dead weight (see also my comments that the part requiring preservation of such notices should be struck). It should either be removed or replaced with something more up-to-date such as:<br><br></p><p>\nIf the program has a graphical user interface with an About box, make it include a short notice like this:\n....\n</p>\ngpl3.howtoapply.p7.s1\nmux2005\nlogin to agree\n0\n3131\n\n\n", "2620": "\n\n\n\nUse 'Hello World'\ngpl3.howtoapply.p8.s1\nsamj\nlogin to agree\n0\n2620\n\n\n", "2417": "\n\n\n\nThat's a bit puerile, don't you think? How old are we again? If we're going to seem childish, perhaps we could use version 3.1337 instead; at least it wouldn't be perceived as an inappropriate sexual reference.\ngpl3.howtoapply.p8.s1\njag\nlogin to agree\n14\n2417\n\n\n", "2618": "\n\n\n\nUse <> to designate boilerplate text. eg Copyright (c) <year> <author>\ngpl3.howtoapply.p8.s1\nsamj\nlogin to agree\n2\n2618\n\n\n", "2618:2619": "\n\n\n\nUse <> to designate boilerplate text. eg Copyright (c) <year> <author>\ngpl3.howtoapply.p8.s1\nsamj\nlogin to agree\n2\n2618\n\n\nEvery word should earn its keep. 'name of' does not necessarily make sense in the context of a company and should be dropped.\ngpl3.howtoapply.p8.s1\nsamj\nlogin to agree\n1\n2619\n\n\n", "2613": "\n\n\n\nCould warranty and copyright not be combined? The shorter this document the better.\ngpl3.howtoapply.p8.s1\nsamj\nlogin to agree\n0\n2613\n\n\n", "2611:2613": "\n\n\n\nSuggest 'for warranty details type 'show w''\ngpl3.howtoapply.p8.s1\nsamj\nlogin to agree\n1\n2611\n\n\nCould warranty and copyright not be combined? The shorter this document the better.\ngpl3.howtoapply.p8.s1\nsamj\nlogin to agree\n0\n2613\n\n\n", "2612": "\n\n\n\nSuggest 'for copyright details type 'show c''. This follows the previous format and makes it more clear why 'c' and 'w' were chosen.\ngpl3.howtoapply.p8.s2\nsamj\nlogin to agree\n0\n2612\n\n\n", "2612:2613:2618:2619:2620": "\n\n\n\nSuggest 'for copyright details type 'show c''. This follows the previous format and makes it more clear why 'c' and 'w' were chosen.\ngpl3.howtoapply.p8.s2\nsamj\nlogin to agree\n0\n2612\n\n\nCould warranty and copyright not be combined? The shorter this document the better.\ngpl3.howtoapply.p8.s1\nsamj\nlogin to agree\n0\n2613\n\n\nUse <> to designate boilerplate text. eg Copyright (c) <year> <author>\ngpl3.howtoapply.p8.s1\nsamj\nlogin to agree\n2\n2618\n\n\nEvery word should earn its keep. 'name of' does not necessarily make sense in the context of a company and should be dropped.\ngpl3.howtoapply.p8.s1\nsamj\nlogin to agree\n1\n2619\n\n\nUse 'Hello World'\ngpl3.howtoapply.p8.s1\nsamj\nlogin to agree\n0\n2620\n\n\n", "2621": "\n\n\n\nEvery word should earn its keep. Suggest a list of examples, eg command, about box, usage instructions, installation procedure, documentation.\ngpl3.howtoapply.p9.s2\nsamj\nlogin to agree\n1\n2621\n\n\n", "2608:3079": "\n\n\n\nUsers must be authorised to assign copyright so if their work is assigned to others by contract or law (eg work for hire) then they MUST obtain this permission.\ngpl3.howtoapply.p10.s1\nsamj\nlogin to agree\n2\n2608\n\n\nRewrite this as: "If you have are under contract to a third party who may assert ownership of your work,..." to avoid limiting the situations covered.\ngpl3.howtoapply.p10.s1\nsepreece\nlogin to agree\n0\n3079\n\n\n", "2608:3076:3079": "\n\n\n\nUsers must be authorised to assign copyright so if their work is assigned to others by contract or law (eg work for hire) then they MUST obtain this permission.\ngpl3.howtoapply.p10.s1\nsamj\nlogin to agree\n2\n2608\n\n\nThis is superfluous and isn't required. Does it matter if you work as a programmer or not? \ngpl3.howtoapply.p10.s1\njamesj\nlogin to agree\n0\n3076\n\n\nRewrite this as: "If you have are under contract to a third party who may assert ownership of your work,..." to avoid limiting the situations covered.\ngpl3.howtoapply.p10.s1\nsepreece\nlogin to agree\n0\n3079\n\n\n", "2608": "\n\n\n\nUsers must be authorised to assign copyright so if their work is assigned to others by contract or law (eg work for hire) then they MUST obtain this permission.\ngpl3.howtoapply.p10.s1\nsamj\nlogin to agree\n2\n2608\n\n\n", "2941": "\n\n\n\nI suggest "The following paragraph".\ngpl3.howtoapply.p10.s2\nmsikma\nlogin to agree\n1\n2941\n\n\n", "2614": "\n\n\n\nEvery word in this document (which will be endlessly examined by expensive lawyers) should earn its keep. These do not.\ngpl3.howtoapply.p10.s2\nsamj\nlogin to agree\n0\n2614\n\n\n", "2614:2941:3076:3079": "\n\n\n\nEvery word in this document (which will be endlessly examined by expensive lawyers) should earn its keep. These do not.\ngpl3.howtoapply.p10.s2\nsamj\nlogin to agree\n0\n2614\n\n\nI suggest "The following paragraph".\ngpl3.howtoapply.p10.s2\nmsikma\nlogin to agree\n1\n2941\n\n\nThis is superfluous and isn't required. Does it matter if you work as a programmer or not? \ngpl3.howtoapply.p10.s1\njamesj\nlogin to agree\n0\n3076\n\n\nRewrite this as: "If you have are under contract to a third party who may assert ownership of your work,..." to avoid limiting the situations covered.\ngpl3.howtoapply.p10.s1\nsepreece\nlogin to agree\n0\n3079\n\n\n", "2610": "\n\n\n\nEither use a boilerplace (eg 'Company') or 'Acme, Inc.'\ngpl3.howtoapply.p11.s1\nsamj\nlogin to agree\n2\n2610\n\n\n", "2606": "\n\n\n\nSuggest 'Hello World!' as a well known alternative.\ngpl3.howtoapply.p11.s1\nsamj\nlogin to agree\n0\n2606\n\n\n", "2607": "\n\n\n\nThis is a serious, widely deployed license. Jokes are certainly unnecessary and potentially damaging. The GPL is 'growing up' and should behave like it.\ngpl3.howtoapply.p11.s1\nsamj\nlogin to agree\n2\n2607\n\n\n", "3015": "\n\n\n\nMaybe the word "Hacker" can confuse some people.\ngpl3.howtoapply.p11.s1\nmayeco\nlogin to agree\n1\n3015\n\n\n", "2606:2607:2610:3015": "\n\n\n\nSuggest 'Hello World!' as a well known alternative.\ngpl3.howtoapply.p11.s1\nsamj\nlogin to agree\n0\n2606\n\n\nThis is a serious, widely deployed license. Jokes are certainly unnecessary and potentially damaging. The GPL is 'growing up' and should behave like it.\ngpl3.howtoapply.p11.s1\nsamj\nlogin to agree\n2\n2607\n\n\nEither use a boilerplace (eg 'Company') or 'Acme, Inc.'\ngpl3.howtoapply.p11.s1\nsamj\nlogin to agree\n2\n2610\n\n\nMaybe the word "Hacker" can confuse some people.\ngpl3.howtoapply.p11.s1\nmayeco\nlogin to agree\n1\n3015\n\n\n", "2915": "\n\n\n\nI think square brackets would be better here. This is a typographical convention I am used to when there is supposed to be a graphic, table, etc., but it hasn't been inserted yet.\ngpl3.howtoapply.p12.s1\njamesgnz\nlogin to agree\n0\n2915\n\n\n", "2915:3059": "\n\n\n\nI think square brackets would be better here. This is a typographical convention I am used to when there is supposed to be a graphic, table, etc., but it hasn't been inserted yet.\ngpl3.howtoapply.p12.s1\njamesgnz\nlogin to agree\n0\n2915\n\n\nthe word "Coon" should be changed\ngpl3.howtoapply.p12.s1\nmayeco\nlogin to agree\n2\n3059\n\n\n", "2615": "\n\n\n\nA serious document should have a serious date. Suggest '1 January'.\ngpl3.howtoapply.p12.s1\nsamj\nlogin to agree\n0\n2615\n\n\n", "2616": "\n\n\n\nWhile the GPL was originally released in 1989 this version will be released in 2007/2008. Suggest using the year of release.\ngpl3.howtoapply.p12.s1\nsamj\nlogin to agree\n2\n2616\n\n\n", "2418": "\n\n\n\nWhile surely an unintentional consequence of trying to make an anti-corporate joke, given the controversy at the use of this name and its historical connotations in America as a derogatory term for black people, we should consider using a different name here.\ngpl3.howtoapply.p12.s1\njag\nlogin to agree\n11\n2418\n\n\n", "2418:2615:2616:2915:3059": "\n\n\n\nWhile surely an unintentional consequence of trying to make an anti-corporate joke, given the controversy at the use of this name and its historical connotations in America as a derogatory term for black people, we should consider using a different name here.\ngpl3.howtoapply.p12.s1\njag\nlogin to agree\n11\n2418\n\n\nA serious document should have a serious date. Suggest '1 January'.\ngpl3.howtoapply.p12.s1\nsamj\nlogin to agree\n0\n2615\n\n\nWhile the GPL was originally released in 1989 this version will be released in 2007/2008. Suggest using the year of release.\ngpl3.howtoapply.p12.s1\nsamj\nlogin to agree\n2\n2616\n\n\nI think square brackets would be better here. This is a typographical convention I am used to when there is supposed to be a graphic, table, etc., but it hasn't been inserted yet.\ngpl3.howtoapply.p12.s1\njamesgnz\nlogin to agree\n0\n2915\n\n\nthe word "Coon" should be changed\ngpl3.howtoapply.p12.s1\nmayeco\nlogin to agree\n2\n3059\n\n\n", "2434:2609:2789:2942:3132": "\n\n\n\nThis URL is missing the ending '/'.\ngpl3.howtoapply.p13.s1\nfraggle\nlogin to agree\n9\n2434\n\n\nPlease drop the superfluous 'www.' which only really makes sense when 'http://' is missing and you're dealing with clueless users (which may indeed be the case with licenses).\ngpl3.howtoapply.p13.s1\nsamj\nlogin to agree\n1\n2609\n\n\n<p style="display: inline">The URL <http://www.gnu.org/licenses.> doesn't work for me. Mostly because it's wrong, but that's the easiest thing to copy from the file.</p><p>\nHow about "... the GNU GPL, see the GNU Licenses Page (http://www.gnu.org/licenses)."</p>\ngpl3.howtoapply.p13.s1\nchandon\nlogin to agree\n0\n2789\n\n\nAs stated above for the same URL, the "www" part is actually supposed to be there. GNU's site redirects all requests to a URL that contains the "www". However, you are linking to a directory; please include the trailing slash.\ngpl3.howtoapply.p13.s1\nmsikma\nlogin to agree\n1\n2942\n\n\nYes, proper grammar requires a period at the end of the sentence, but please make an exception in this case. The final period is annoying when you do copy'n'paste (especially when you're viewing in a proportional width font where the "." sticks closely to the "s"). If you insist on proper grammar, maybe you can put the URL on its own paragraph and end the sentence with a ":".\ngpl3.howtoapply.p13.s1\nmux2005\nlogin to agree\n0\n3132\n\n\n", "2973": "\n\n\n\n<p style="display: inline"><p>Paragraph</P></p><p>\n<ul>\n<li>List 1</li>\n<li>List 2</li>\n<li>List 3</li>\n</ul></p><p>\n<p>another paragraph\nwith a carriage return\nand another.\n</p></p><p>\nText</p><p>\nMore text\n\n\nYet more text</p><p>\nMore text 2</p>\ngpl3.howtoapply.p14.s1\ngerv\nlogin to agree\n0\n2973\n\n\n", "2696:2992": "\n\n\n\n<p style="display: inline">I'm not sure where to put my comment, so I'll put it here... what about Web-based applications? A lot of free software software is available for use on many servers, but without a source code, and modifications to this code are not given back - this is really anti-community behavior and maybe it should be disallowed in this license?</p><p>\nThis is really an important issue, because the FUTURE is in web-based solutions, yet I haven't heard of any license, similar to GPL that will force the authors of server-based software to give back the changes.</p>\ngpl3.howtoapply.p14.s1\nmichals\nlogin to agree\n0\n2696\n\n\n<p style="display: inline">Please correct me if I am wrong, but it looks to me that I could use GPL libraries in closed source programs by:\nA) Defining a plugin interface (API) in my proprietary program and later on create a mediator component which mediates calls between the an, in the mediator encapsulated library (e. GPL or ASL libraries) and whatever program is on the other side of the API (e.g. my proprietary program) [1]. I would then distribute the mediator under GPL along with any third party GPL libraries and my proprietary program under an ordinary draconian license form.\nB) Create a thin program wrapper around the GPL library, exposing functions through an IPC interface. The IPC interface would not be as efficient as direct library linking into the proprietary program, but would be good enough if the interface is coarse grained.</p><p>\nBoth scenarios are common in the industry and people create copy-left plugins frequently to proprietary applications with a good end-result. The vendor gains a vibrant community and sell more products and the community can share knowledge embedded in the plugins. However, this ecology can be abused as well, as indicated.\n\n\n1. To fully cover my back, I could elect not to distribute the mediator and GPL libraries with my program but instead provide a feature which automatically downloads both components to a plugins folder if the user chooses to activate that feature. If the components are not downloaded, the required features would be disabled in my program, thus avoiding the apparent dependency (and legal argument) on the GPL code for the program to function.</p>\ngpl3.howtoapply.p14.s1\ntinjon\nlogin to agree\n0\n2992\n\n\n", "2795": "\n\n\n\nThere are only few occasions where using the LGPL instead of the GPL is a good idea, even for libraries (and the fsf agrees with me on that, AFAIK). Therefore, even though it may be good to mention it (if the LGPL continues to exist), it should be discouraged while doing so IMO.\ngpl3.howtoapply.p14.s2\nwijnen\nlogin to agree\n0\n2795\n\n\n", "2658:2916": "\n\n\n\nI thought this line would be removed as the LGPL is turning into a set of additional permissions for the GPL?\nor has this been changed?\ngpl3.howtoapply.p14.s3\nilly\nlogin to agree\n2\n2658\n\n\nThe fonts exception is another additional permission that it would be useful to mention:\nhttp://www.gnu.org/licenses/gpl-faq.html#FontException\ngpl3.howtoapply.p14.s3\njamesgnz\nlogin to agree\n0\n2916\n\n\n", "2795:2916:2973:2992": "\n\n\n\nThere are only few occasions where using the LGPL instead of the GPL is a good idea, even for libraries (and the fsf agrees with me on that, AFAIK). Therefore, even though it may be good to mention it (if the LGPL continues to exist), it should be discouraged while doing so IMO.\ngpl3.howtoapply.p14.s2\nwijnen\nlogin to agree\n0\n2795\n\n\nThe fonts exception is another additional permission that it would be useful to mention:\nhttp://www.gnu.org/licenses/gpl-faq.html#FontException\ngpl3.howtoapply.p14.s3\njamesgnz\nlogin to agree\n0\n2916\n\n\n<p style="display: inline"><p>Paragraph</P></p><p>\n<ul>\n<li>List 1</li>\n<li>List 2</li>\n<li>List 3</li>\n</ul></p><p>\n<p>another paragraph\nwith a carriage return\nand another.\n</p></p><p>\nText</p><p>\nMore text\n\n\nYet more text</p><p>\nMore text 2</p>\ngpl3.howtoapply.p14.s1\ngerv\nlogin to agree\n0\n2973\n\n\n<p style="display: inline">Please correct me if I am wrong, but it looks to me that I could use GPL libraries in closed source programs by:\nA) Defining a plugin interface (API) in my proprietary program and later on create a mediator component which mediates calls between the an, in the mediator encapsulated library (e. GPL or ASL libraries) and whatever program is on the other side of the API (e.g. my proprietary program) [1]. I would then distribute the mediator under GPL along with any third party GPL libraries and my proprietary program under an ordinary draconian license form.\nB) Create a thin program wrapper around the GPL library, exposing functions through an IPC interface. The IPC interface would not be as efficient as direct library linking into the proprietary program, but would be good enough if the interface is coarse grained.</p><p>\nBoth scenarios are common in the industry and people create copy-left plugins frequently to proprietary applications with a good end-result. The vendor gains a vibrant community and sell more products and the community can share knowledge embedded in the plugins. However, this ecology can be abused as well, as indicated.\n\n\n1. To fully cover my back, I could elect not to distribute the mediator and GPL libraries with my program but instead provide a feature which automatically downloads both components to a plugins folder if the user chooses to activate that feature. If the components are not downloaded, the required features would be disabled in my program, thus avoiding the apparent dependency (and legal argument) on the GPL code for the program to function.</p>\ngpl3.howtoapply.p14.s1\ntinjon\nlogin to agree\n0\n2992\n\n\n", "3048": "\n\n\n\n\nI'm glad this section ("How to Apply These Terms...") is back. If the point of the GPL is to encourage the development of Free Software, under the same consistent license (so that people can combine works), then this section really NEEDS to be in the license. A lot of developers won't know how to apply the GPL to their own works without this section; with it, applying this license is easy. I realize that technically this section isn't part of the "license", but if your goal is USE of the license (not just noting that it exists), this is important. Please keep this section in the final license text.\n\n\n\n\nhowtoapply\ndwheeler2\nlogin to agree\n0\n3048\n\n\n"}