Talk:GNU General Public License/Archive 5

Archive 1

←

Archive 3

Archive 4

Archive 5

Archive 6

Copyright on derivatives

1. Who owns the copyright to a derivative work?
2. If Work B is derived from Work A, and Work C is derived from Work B, and Work C violates the GPL, does the copyright owner of Work A, as well as the copyright owner of Work B, have the right to sue the creator of Work C?

VanishingUser 03:57, 25 April 2007 (UTC)

For any code you write under GPL you own the copyright to that. (Similarly anything you write on Wikipedia you also own the copyright to, see WP:C.) GPL'd code can have many copyright holders. If you work for a company, you normally sign a waver to assign copyright to the company: this allows them to enforce their copyright more effectively.

On the second point, they can both sue. Realistically though they will ask C to fix the problems so that C releases their code under GPL. They can use the FSF's "GPL Compliance Lab" to do this for them. Legal action is possible, but won't normally happen unless discussion fails. --h2g2bob 04:13, 25 April 2007 (UTC)

Flawed Criticism Section

I think that the major problem why many BSD camp is against GPL is actually the fact that merely linking (not even modifying the library) would force the entire application license to be GPL. This is particularly the case in the Java world. Thus one cannot distribute GPL libraries along with the commercial or open/free programs that uses a non-GPL license. There is simply no way around the problem. People have no problem with LGPL in this case, in contrast.

This distinction between LGPL and GPL has been quite important. MySQL for instance, changed the license of MySQL JDBC driver license from LGPL to GPL for v3 driver, intentionally. This change basically bars open source applications not using GPL license from using the driver. For open source applications that have to stick to a particular open source license, there is simply no way around the problem. There is no loss to MySQL anyways since they don't get paid in this case. Of course, this approach also forces commercial applications to purchase the commercial license of the driver. In fact, this strategy has been quite popular. It is ironic, that GPL has been used as a way to force people to pay up, or not to use the software as the result, as opposed to be free. 71.105.96.173 18:32, 22 May 2007 (UTC)

The difference of opinion about licenses between the pro-BSD and pro-GPL camps is largely philosophical, just as are the differences between the Open Source and Free Software communities, and between the Free Software Foundation and commercial software vendors. In the BSD v. GPL case, it goes to the core of one's personal definition of "freedom". BSD advocates often describe the GPL's "freedoms" as "restrictions", because in the BSD community code is allowed to be used in just about any way one can think of. GPL advocates often describe the BSD license as "unfree" because it doesn't ensure that the rights granted are passed intact or extended to cover code changes. The two groups generally have different primary goals, and that's expressed in their choice of licenses.

To make a long story short, I guess I disagree with your suggestion that the Criticism section is flawed in some undefined manner. It seems to sum up the actual divide pretty well. RossPatterson 22:32, 22 May 2007 (UTC)

I think that you missed the point. The criticism said "GPL terms require that all modified versions of the software must in turn be licensed under the GPL." That is not the main criticism of GPL. I am saying that for LGPL, the modification only needs to be reveal for the part that is relevant to the library, which people has little objection. Second, merely linking GPL (say a GPL MySQL java JDBC driver) would force the entire software (say a generic front end to various databases) to be GPL, but for LGPL one can still use other license.

The result of GPL's restriction is that it forces softwares (even free ones that not using GPL) to come up solutions that ultimately makes it difficult for the end user. In the above example, the generic front end to various databases would not be able to distribute with the GPL driver. The users would have to follow an instruction to download and install by themselves. This creates a burden for the end user. 71.105.110.181 12:52, 29 May 2007 (UTC)

Well, everyone's entitled to their opinion, you as well as I. But neither of us are entitled to voice our opinions in Wikipedia articles. If you can find published examples that support your opinion that the "main criticism of [the] GPL" is as you say (to be honest, I can't figure out what you say the main criticism is), go ahead and add it to the article. I lack the time at the moment to find citations that support my claim above, so I won't be adding it, at least not right now. RossPatterson 01:55, 30 May 2007 (UTC)

• "In the above example, the generic front end to various databases would not be able to distribute with the GPL driver". "not able to distribute"!? Sorry, it would be impossible to distribute without making it GPL, which is not the same. The distributor is free to either make all her software GPL, or not include GPL code at all. Where's the problem? If all free software were licensed under the LGPL, FS could never beat commercial software, except by price (because any company could "steal" FS and include it in a proprietary bundle). Licensing under the GPL means that in the long run proprietary software will fall behind the FS technically, because they will not be able to "steal" the efforts of the FS community (and it's more obvious day by day that they can not keep up reproducing the work of the FS community). Guess which license ensures a more free future for software? To know more about it, read GPL vs LGPL at gnu.org. — isilanes (talk|contribs) 17:00, 29 May 2007 (UTC)

This page and the BSD license page seems to be written by GPL people. I feels that it is appropriate, particularly given the fact the criticism section is totally lacking in substance, to add FreeBSD's criticism on GPL. I felt funny since quoting the CEO from the company many OSS programmers despise only seem to give GPL more sympathy. That's a stealthy way of getting around true criticisms. Coconut99 99 23:13, 5 June 2007 (UTC)

That FreeBSD page should probably be mentioned alright, but it's pretty nonsensicle and those statements are neither backed up by reason or evidence. Gronky 10:21, 6 June 2007 (UTC)

The criticism section isn't a criticism section at all. It is actually an essay in defense of GPLv3. Not only is this a straw man, its also against Wikipedia's standards (no opinion essays). The people who wrote this are trying to silence criticism of the GPL. User:cksgk8 —Preceding unsigned comment added by 98.156.121.226 (talk) 21:17, 4 January 2010 (UTC)

Logo Author

I skipped through this the article and some other GNU related but I could not find a word on the artist that has designed the GNU Logo. Did I got blind or there are no info on it? On the Image page I found as Author "Aurelio A. Heckert" but I guest he is the author of the file and not the designer of the logo. Should we not have some info on him? --192.33.238.6 15:10, 4 June 2007 (UTC)

You're right, Wikipedia should have this info, but it should be on the GNU or GNU project page(s), not this one. Gronky 13:41, 5 June 2007 (UTC)

Stuff I removed from the criticisms section

The criticisms section contained a load of nonsense, which was, of course, not supported by any references. I removed three or so paragraphs, here's the last version before I did so: [1]. Gronky 10:04, 6 June 2007 (UTC)

I've deadheaded the external links section

The external links section had turned into a sprawl, so I've deleted most of them. Many of the links that I deleted were interesting though, and would make good references for parts of this article, so do take a look at the last version of the page that had all those links: [2]. Gronky 10:29, 6 June 2007 (UTC)

some rewriting

Reading this article earlier today, I saw two problems: (1) The lead was pov-ish, and dominated by boring "rah-rah" statistics about how popular the GPL is. (2) The History section (as noted in the template at the top) was a timeline of versions of the GPL, rather than a discussion of the history of the license in the sense of its importance to a social movement. The result of all this was that you could read several screenfuls of the article without ever finding out anything that would be of interest to the typical person. To try to fix this, I've rewritten the lead to be more npov, and to say more about the social significance of the license. I've also renamed the History section to Versions, and inserted a new History section above it, with a discussion about the social history, as opposed to the version history.--76.81.164.27 18:40, 11 June 2007 (UTC)

The last sentence on version 3, "in the aftermath of the Novell-Microsoft agreement, Linux kernel developer Linus Torvalds expressed interest in reconsidering GPLv3" is not or is no longer accurate. It would be more accurate to say that Linus has expressed that he is satisfied with the GPL 2 because it is a superior license to the GPL 3. See here, here, here here, and here. I don't want to get into a pissing match with the many free-bies here, but with a FSF board member trolling the LKML, I think Linus has dropped the diplomacy. Wasn't his strong suit, anyway. But the point is, Linus has made it very clear that he is not open to GPL 3. 216.175.85.40 07:22, 26 August 2007 (UTC)

owner?

If the resulting software is kept only for use by the modifier, 
no disclosure of source code is required.
The GPL is automatically revoked upon any violation of its terms,
but copyright owners of works licensed with the GPL are free to 
negotiate alternate terms with authors of derived works

But who's the owner?. Because some GNU projects are the works of a community, or then "first to create the projects" can negotiate the works of others?.—Preceding unsigned comment added by 200.73.30.108 (talk • contribs) 15:45, 5 July 2007

The "owner" is a matter of law, not of license (or logic, for that matter). Under copyright law, nothing is "the work[] of a community", but rather the individual works of the members of the communities. This question is exactly the reason that official projects of the Free Software Foundation (e.g., GCC, Emacs) require that contributors sign an agreement assigning their copyright interests to the FSF. That way there is no question under law of who the owner is, or of who has the right to compel licensees to comply with the GPL terms. The undisputed owner in such projects is the FSF. As a counter example, Linus Torvalds has never required copyright assignment for code contributed to the Linux kernel, and the "ownership" of code within the kernel is an oft-discussed question. Indeed, when GPL3 was first proposed, several prominent Linux community members noted that it was probably impossible to change the kernel license simply because it would be so difficult to identify the various contributors who own the copyrights. RossPatterson 23:34, 5 July 2007 (UTC)

GPL in closed source projects.

I would like to use #ziplib licensed under GPL for my closed source project. It says some legal terms which led me to think I'm not allowed to use it, and then the bottom line said otherwise:

Bottom line - In plain English this means you can use this library in commercial closed-source applications.

I went here to verify if this was the case, and I'm still not sure. Can I use GPL software in non-GPL, closed source projects or can't I?

Ripper234 12:41, 14 July 2007 (UTC)

No, you can't use GPL software in closed source applications. But SharpZipLib is a special case — its authors offer you an exception to that rule, similar in concept and in text to the exception for the GNU Classpath implementation of the standard Java library. If you don't trust their "Bottom line" statement, then you really need to had the entire license off to your lawyer for a reliable opinion on the use of SharpZipLib. But regardless of how that turns out, the exception doesn't apply to software that is licensed solely under the GPL. RossPatterson 15:19, 14 July 2007 (UTC)

GPLv3 clarification needed

The article claims that the GPLv3 offers better compatibility to other licenses but GPLv3 text looks rather more restrictive than the GPLv2. If no lawyer explains why there should be better compatibility, this claim should be removed from wikipedia

I don't know if it is a relevant source, since it's not third-party (i.e., it's published by the authors of GPLv3), but the rationale documents from the GPLv3 legislative history talk at great length about how the additional compatibility works, and those rationale documents were written primarily by SFLC attorney, Richard Fontana. Frankly, I don't think there are any attorneys who are particularly qualified to make the assessment anyway, even if SFLC is a biased party since they are FSF's attorneys. -- bkuhn 21:54, 16 August 2007 (UTC)

On dynamic linking

Without really discussing the legal implications, as someone well-versed in the possible technicalities but not the legalities, I would like to just note here that it is possible for software to dynamically link to libraries that weren't even written at the moment the software was released, even if the original developers didn't intend for this to happen. What libraries a program dynamically links to is beyond the control of the developer. Shinobu (talk) 04:14, 6 December 2007 (UTC)

Absolutely agree. Also, an application can be written to use a Java interface without being aware of any particular implementation of that interface. It's possible that the user who causes that application to run with a GPL'd implementation of that interface is creating a "derived work" (a term which GPLv3 doesn't actually use, by the way), but since what they have created is transient and unlikely to be distributed, it hardly matters.

The GPL FAQ makes a distinction between dynamic linking and fork/exec - they say that one creates a derivative work and the other doesn't. Can anyone tell me what words in the actual license could possibly lead one to this conclusion? Mhkay (talk) 18:20, 14 January 2008 (UTC)

FWIW: from a legal standpoint, I suspect that this distinction is largely, if not wholly, mythical. GPL seems to be trying to create a distinction absolutely unknown in actual copyright law here-- and failing. The definition of a derivative work is simply not a matter that the GPL can define for itself, as it wishes. KenThomas (talk) 19:31, 17 October 2009 (UTC)

"In court" section needs a reorg

At the moment it's a not-terribly-useful timeline covering people both using and attacking the GPL in a variety of jurisdictions. Is there some sensible and useful way to sort this out? I've also added a note on BusyBox, the refs for which are in that article - David Gerard (talk) 23:34, 22 December 2007 (UTC)

WP is not a list-of-lists, but this section is odd and would be improved by simply following the cases from a relatively neutral source (Slashdot's coverage is not neutral, but the list it would provide, more-or-less is). Selection here is odd; coverage is odd; statements are odd, or questionable. —Preceding unsigned comment added by KenThomas (talk • contribs) 19:34, 17 October 2009 (UTC)

Does the license encourage supplying a visible notice to the original authors?

For example, if you install a CMS that uses this license, does the license encourage you to put a footer link to the makers of the CMS? -79.182.0.62 (talk) 11:40, 10 January 2008 (UTC)

No. See SugarCRM for an example of an outcry caused by a license which did do this. Chris Cunningham (talk) 11:57, 10 January 2008 (UTC)

This article indeed mentions there was an outcry but doesn't attribute it to having to put a link in sites' footers. -79.182.0.62 (talk) 18:25, 12 January 2008 (UTC)

The license in CMS

How does the license limit clients if they install a CMS that use it? We've already established above that they won't have to put a link in the footer for the original authors. So modifications would have to be GPL too, but is there anything else they can't they do, that they could if the CMS was completely public domain? -79.182.0.62 (talk) 18:25, 12 January 2008 (UTC)

It doesn't limit them at all, beyond a prominent copyright / license notice, so long as they don't actually allow clients to download the server-side code they're using. They can modify at as they wish. Chris Cunningham (talk) 18:43, 12 January 2008 (UTC)

But we've established above they don't have to put a copyright in the site itself. -79.182.0.62 (talk) 20:31, 12 January 2008 (UTC)

No, we haven't. There's no requirement to stick the URLs of the original authors in some "footer" of a web app; there is a requirement to post a prominent copyright notice, but this need not be some banner ad.

I'd seriously encourage you to just Google any further questions regarding the licence's terms. Talk pages are not general discussion forums for the subject. Chris Cunningham (talk) 00:39, 13 January 2008 (UTC)

Well, just last question - what is "prominent" if not a link? -79.182.0.62 (talk) 12:00, 13 January 2008 (UTC)

There's no given definition. I don't believe anyone has yet been legally challenged on the issue. Chris Cunningham (talk) 12:38, 13 January 2008 (UTC)

And they won't ever be, because they are not distributing the software. The license requires a copyright notice whenever someone distributes the software, not when they use it. —Preceding unsigned comment added by 66.102.196.44 (talk) 06:21, 8 March 2008 (UTC)

If I understood you correctly, you are claiming that using a GPL'ed CMS to run a public website does NOT mean that you are "releasing" the CMS software in terms of the GPL (even so the "results" - the website(s) - of the CMS are public). Do you have any source for that? If yes, I think that should be added to the article. 138.246.51.239 (talk) 18:48, 30 November 2010 (UTC)

The question is confused. If one is "using a GPL'ed (sic) CMS to run a public website" then the CMS software is already available under the terms of the GPL. One does not need to do anything in order to follow the license as one is not distributing the software itself but simply running it. The output is never subject to the GPL as it's simply data being run through the program; if such data were subject to the GPL then practically everything ever put on the Web would be subject to the GPL as multiple pieces of GPL software are usually involved in that. Chris Cunningham (user:thumperward: not at work) - talk 21:53, 30 November 2010 (UTC)

Ok, to clarify it: 1. I meant if you modify the CMS and then you use it to run a public website. 2. I didn't ask if the content of the website becomes GPL'ed (*), but if you have to make the source code of your modified CMS available. (*) this expression is also used in the article - if you think that it is wrong you should correct it there first... 134.155.248.57 (talk) 10:49, 3 December 2010 (UTC)

There you have it: you are "using" the software and not "distributing" it. Sending the output of a program to someone (which is what happens when one runs a Web service) does not distribute the source or object code to that program, and so the user is not subject to any of the GPL's provisions regarding distribution of modified source. The GPL3 has a provision to modify the license such that running such software does count as distribution, and this is already covered in the article in the paragraph starting "Other notable changes include allowing authors to add certain conditions or requirements to their contributions". If you'd like to dive in and help improve the article, you could try a Web search for "GPL3 Web services" and see if you can find good references to add to that paragraph. Chris Cunningham (user:thumperward: not at work) - talk 14:46, 7 December 2010 (UTC)

Software as a service and GPL

Should this be mentioned or explained? (http://en.wikipedia.org/wiki/Software_as_a_Service) or show a diagram in detail?

GPL can be more "business friendly" than BSD

The case of Quake3 (that may improve the article if incorporated in some form or another): Quake3 is now GPL'ed. However, it can also be bought (as source) to be closed for commercial purposes. At the same time, there is no BSD (or similar) version. One has to either use GPL or pay to have a new game. i.e. BSD in that case would be less "business friendly". In general at least in that case one can clearly say Quake3 is more profitable for ID software being released in GPL (+closed dual licenced) rather than in BSD (or even alongside BSD). --Leladax (talk) 22:34, 25 February 2008 (UTC)

Every license allows for some set of business models to monetize it. It just takes an intelligent developer to use the right business and license model to benefit them (personally or their cause) the most. There's a large tradeoff between control and audience... the permissiveness of your selected license allows more and more people to use your work... but gives you less and less control over how they can use your work (or in copyleft... their audience using their work... and their audience's audience...) That may be desired... as you might be making money indirectly from the work. You may also be developing fairly useless work in isolation but really want the resultant... and want whoever makes the resultant derivative work to make it open under various conditions. It always... *always* depends on what you are attempting to achieve. There is nothing wrong with any given license... even completely proprietary... the goods and evils are in how they're implemented. Phopojijo (talk) 08:34, 23 October 2010 (UTC)

I would note Quake III isn't open source. Only the engine is. There may be advantages to id Software to using the GPL (in ensuring other proprietary software developers can't use their code) but most likely the primary reason why people will buy Quake III is because they have to. You can't get a GPL version of Quake III since it doesn't exist. Now the Quake III engine was extensively licenced at a time (primarily when it was still fairly new ) but in terms of the more general point, I don't think you need to use the Quake III example. There are many examples where companies have released their code under GPL primarily for similar reasons a lot of independent developers have released their code under GPL, i.e. not wanting people to be able to use their work unless they use the GPL (generally unless they release the sources for any binary versions) Nil Einne (talk) 12:35, 8 July 2011 (UTC)

aggregation vs combination

It's not really made clear enough in the article, but the GPL allows for linkage with closed-source applications under certain circumstances. To put it briefly, GPL allows aggregation but not combination. As an example, Say I have a closed-source application that as part of its duties needs to read in data of different file-types. To accomodate this, I have provided a generic plugin interface to allow any file-type that has a plugin available to be used. As a separate component, I supply a plugin that makes use of GNU readline for some reason or another. readline is GPL'd, which means that the plugin needs to be released with a GPL-compatible license. The program itself can remain closed source, however. This is because the plugin architecture makes no assumptions about what the plugin has to do internally. In fact, since plugins can be supplied by anyone, as the author of the original program I am unable to make any guarantees that no one will do this. This execution model is essentially the same as if I call an external executable from a shell script; Utilizing gpl'd programs inside of a shell-script doesn't mean you have to gpl your shell-script. —Preceding unsigned comment added by 66.102.196.44 (talk) 06:11, 8 March 2008 (UTC)

"to permit the Microsoft-Novell agreement"?

In the "Version 3" section it is said: "The fourth discussion draft[...] made an exception to permit the Microsoft-Novell agreement, saying in section 11 paragraph 6 that[...]" The intended meaning is the opposite, isn't it? —Preceding unsigned comment added by 189.95.103.132 (talk) 20:13, 15 June 2008 (UTC)

Compatibility and multi-licensing

As I wrote in Image talk:Quick-guide-gplv3-compatibility.svg, should not that image have an arrow from LGPLv3 down to GPLv3? --193.11.177.69 (talk) 04:59, 21 July 2008 (UTC)

(No header)

I don't believe a laywer looked over this wikipedia entry at all. This statement: "That is, their code can be combined with a program under the GPL without conflict (the new combination would have the GPL applied to the whole)." is patently false. You cannot simply "steal" BSD licensed code, copy and paste it into a GPL licensed program, and cover the whole thing with the GPL. BSD licensed code remains BSD licensed code, as do any modifications to it, unless you work out a deal with the author to release it under another license. —Preceding unsigned comment added by 71.233.86.139 (talk) 23:16, 27 July 2008 (UTC)

That is not the accepted wisdom. BSD does not as much differ from GPL as much as it is a subset of it. That is, BSD requires some things, and GPL requires more things, but the BSD requirements does not conflict with the GPL ones. Therefore, a copied BSD code section is technically still licensed under BSD, but since BSD is compatible with GPL, recipients of the combined work can treat the whole as if it were under the GPL with no worries. --193.11.177.69 (talk) 14:41, 31 August 2008 (UTC)

And, furthermore, if you extract the BSD section from the GPL work-as-a-whole, it remains BSD-licensed (so there's nothing "technically" about it; this distinction can indeed matter in practice). The licence of the BSD-licensed code is not affected by treating it as part of a larger work as a whole; there's nothing wrong with this as long as the terms of the BSD licence continue to be respected. Colin Watson (talk) 18:14, 31 August 2008 (UTC)

Write more clearly

can some explain what GPL actualy does ? —Preceding unsigned comment added by 203.199.118.12 (talk) 06:44, 22 August 2008 (UTC)

Under the same heading, I would suggest that language about the GPL being "restrictive" (weasel words!) should be changed to describing licenses such as the 2-clause BSD license as "more permissive". After all the starting position is that the recipient of software has no rights to do anything at all with it. From that perspective, the GPL is conditionally permissive rather than "restrictive". (The 2-clause BSD license is also conditionally permissive; its conditions merely have a different aim.) 194.187.213.95 (talk) 12:53, 23 September 2008 (UTC)

English spelling styles

Wikipedia, as you should all know, says that it doesn't matter which English spelling system is used in an article as long as it is consistent. However, making spellings consistent does not apply to proper names and Web page titles: please use the actual values of those, as they should not be altered to a spelling that does not exactly match the name or title. Todd Vierling (talk) 13:31, 23 September 2008 (UTC)

authorship

I don't think Stallman wrote the GPL by himself. There's a lawyer in Newton MA named Jerry Cohen who is an old friend of my dad's, and has done a lot of software industry work (he was Mitch Kapor's lawyer for a while). He's mentioned working with RMS creating the GPL. I'll try to remember to ask Stallman next time I see him. -- user akb4, not logged in —Preceding unsigned comment added by 68.160.144.138 (talk) 06:11, 27 September 2008 (UTC)