Re: cdrecord: weird GPL interpretation
On Thu, Sep 09, 2004 at 05:37:35AM -0700, Josh Triplett wrote: Sven Luther wrote: On Thu, Sep 02, 2004 at 02:12:54PM -0700, Adam McKenna wrote: On Thu, Sep 02, 2004 at 04:30:04PM -0400, Glenn Maynard wrote: [1] http://www.washington.edu/pine/faq/legal.html#10.2 (Accusing Free Software programmers of perverting the license by doing things they were clearly granted permission to do; that's wonderful.) Wasn't the force behind the license re-interpreation to stop people from distributing Pine with a maildir patch? Like cdrecord with a dvdrecord patch, and yaboot with anything but pristing upstream source ? Altough at lest in the yaboot case, it is not licencing that is used, but threat to remove the debian yaboot mention from the official yaboot web page of supported yaboto versions. That's interesting; could you point me towards a reference for this? I can't seem to find one via Google. (Or was it in private mail?) It was in the now archived bug against yaboot : http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=242348 Me : | BTW, i also have a patch which adds amiga partitioning support to | yaboot, would you consider including it ? Ethan : | not in 1.x. | | if debian starts patching yaboot severely i will add debian to the | unsupported dists category. The code is one which couldn't have any effect on existing plateforms using yaboot, since there is only one interaction, which is the probing for an amiga partition table, which happens only if no MBR or Mac partition tables or iso filesystems were found. As a result, i cannot get a yaboot version which works on my pegasos into debian, and need to resort to a local built package. Friendly, Sven Luther
Re: cdrecord: weird GPL interpretation
Sven Luther wrote: On Thu, Sep 02, 2004 at 02:12:54PM -0700, Adam McKenna wrote: On Thu, Sep 02, 2004 at 04:30:04PM -0400, Glenn Maynard wrote: [1] http://www.washington.edu/pine/faq/legal.html#10.2 (Accusing Free Software programmers of perverting the license by doing things they were clearly granted permission to do; that's wonderful.) Wasn't the force behind the license re-interpreation to stop people from distributing Pine with a maildir patch? Like cdrecord with a dvdrecord patch, and yaboot with anything but pristing upstream source ? Altough at lest in the yaboot case, it is not licencing that is used, but threat to remove the debian yaboot mention from the official yaboot web page of supported yaboto versions. That's interesting; could you point me towards a reference for this? I can't seem to find one via Google. (Or was it in private mail?) - Josh Triplett signature.asc Description: OpenPGP digital signature
Re: cdrecord: weird GPL interpretation
On Fri, 3 Sep 2004 09:40:49 + (UTC) Andreas Metzler wrote: [...] Hello, This was about the recent change of license in a36 that was widely covered in the news, e.g. lwn or heise.de http://weblogs.mozillazine.org/gerv/archives/006193.html We (cdrools Debian maintainers) were in indeed private contact with Joerg about this and successfully managed to resolve this, a38 undid the newly introduced non-freeness issue, the code in question (linuxcheck()) is not encumbered by a specific license anymore, it may be removed like anything else. This is resolved and completely orthogonal to this thread, so please ignore it. So, IIUC, debian bug #265546 (http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=265546) is related to linuxcheck() only. I instead thought that it was related to the whole bunch of weird restrictions that go well beyond the GPL license, and that it was filed when the non-freeness was even worse and then marked resolved when one of the issues went away. So, since I noticed that many issues were still there, I wondered if the bug should be reopened. I apologize for the misunderstanding. Well, if I *now* understand the situation correctly, we have version 2.01a38 with *another* set of non-freeness and non-distributability issues to deal with... So probably a different bug should be filed, at least if sid already includes version 2.01a38... -- | GnuPG Key ID = DD6DFCF4 | $ fortune Francesco |Key fingerprint = | Q: What is purple Poli| C979 F34B 27CE 5CD8 DC12 | and commutes? | 31B5 78F4 279B DD6D FCF4 | A: A boolean grape. pgpCNLDB19B2G.pgp Description: PGP signature
Re: cdrecord: weird GPL interpretation
Bernhard R. Link wrote: * M?ns Rullg?rd [EMAIL PROTECTED] [040902 17:11]: In particular, he seems to be relying on German Authors' Rights, and claims to be in discussion with Debian people. That's nearly a month ago. More specifically, he claims to be in discussion with Debian how to stop SuSE from doing what they have every right to do. I know nothing about German law, so I can't comment on that bit. Only thing in German law I could imagine is that he could withdraw some of his work. Which would require him to pay everyone what it is currently worth before the ban can take effect and make it impossible for him to allow others to use it. Anyone knows some reading if there is any other possibility and if this possibilty is even applicaple to software? I think he's trying to say that his moral rights were violated by SuSE when they made a broken version of cdrecord. You are indeed not allowed under German law to modify a work in such a way as to damage the original author's reputation or good name. However that doesn't cancel the license. It merely is a ground for a civil lawsuit to obtain damages. Arnoud -- Arnoud Engelfriet, Dutch patent attorney - Speaking only for myself Patents, copyright and IPR explained for techies: http://www.iusmentis.com/
Re: cdrecord: weird GPL interpretation
On Thu, Sep 02, 2004 at 02:12:54PM -0700, Adam McKenna wrote: On Thu, Sep 02, 2004 at 04:30:04PM -0400, Glenn Maynard wrote: [1] http://www.washington.edu/pine/faq/legal.html#10.2 (Accusing Free Software programmers of perverting the license by doing things they were clearly granted permission to do; that's wonderful.) Wasn't the force behind the license re-interpreation to stop people from distributing Pine with a maildir patch? Like cdrecord with a dvdrecord patch, and yaboot with anything but pristing upstream source ? Altough at lest in the yaboot case, it is not licencing that is used, but threat to remove the debian yaboot mention from the official yaboot web page of supported yaboto versions. Friendly, Sven Luther
Re: cdrecord: weird GPL interpretation
Brian Thomas Sniffen bts at alum.mit.edu writes: [...] On the other hand, I find this message interesting: http://lkml.org/lkml/2004/8/19/111 In particular, he seems to be relying on German Authors' Rights, and claims to be in discussion with Debian people. That's nearly a month ago. Hello, This was about the recent change of license in a36 that was widely covered in the news, e.g. lwn or heise.de http://weblogs.mozillazine.org/gerv/archives/006193.html We (cdrools Debian maintainers) were in indeed private contact with Joerg about this and successfully managed to resolve this, a38 undid the newly introduced non-freeness issue, the code in question (linuxcheck()) is not encumbered by a specific license anymore, it may be removed like anything else. This is resolved and completely orthogonal to this thread, so please ignore it. cu andreas
Re: cdrecord: weird GPL interpretation
Raul Miller moth at debian.org writes: [...] I've taken a look at a copy from January, and it has the same problem. I don't know how far back we'd have to go to find a legally distributable copy. Probably February or January 2002. cu andreas
Re: cdrecord: weird GPL interpretation
Andreas Metzler [EMAIL PROTECTED] writes: Brian Thomas Sniffen bts at alum.mit.edu writes: Raul Miller moth at debian.org writes: [...] There's an additional problem: cdrtools, at least as Debian distributes it, uses some code for which Schilling is not the copyright holder. The HFS support, for example, is copyright Robert Leslie, and licensed under the normal, sanely interpreted GPL. cdrecord is not distributable by anybody, including Schilling, in this state. [...] cdrtools consists of a bunch of largely independent applications and libraries (e.g cdrecord, readcd, mkisofs, cdda2wav), debian/copyright lists the licenses and copyright holders in detail. The two issues mentioned in this thread influence different parts of cdrtools: * defaults.c /* * WARNING you are only allowed to change this filename if you also snip This one is used and linked against all applications of cdrtools since 2.01a26 (previously only in cdrecord). If it is GPL incompatible it indeed breaks the e.g. mkisofs' and cdda2wav's original copyrights. That's certainly GPL-incompatible. It's an extra restriction. Since it affects functional behavior, I'd call it non-free. You must change this filename requirements are generally considered non-free, so I'd expect You may not change this filename without paying this fee requirements to be non-free. The second issue * If you modify cdrecord you need to include additional version * printing code that [...] in cdrecord/cdrecord.c only applies to cdrecord which is completely copyrighted by JS. Therefore he is able to license it as GPL+restrictions and if the restrictions are still DFSG free we are able to ship it as part of Debian/main. - If cdrtools stopped being distributed as whole and would be split into separate tarballs for the different applications, because otherwise this part of GPL ... I think if that could easily be done, and the packages didn't Depend on each other, then you could say that they're separate works and merely aggregated into one package. -- But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it. -- ... could give us a headache. cu andreas -- Brian Sniffen [EMAIL PROTECTED]
Re: cdrecord: weird GPL interpretation
On Fri, Sep 03, 2004 at 09:24:00AM -0400, Brian Thomas Sniffen wrote: The two issues mentioned in this thread influence different parts of cdrtools: * defaults.c /* * WARNING you are only allowed to change this filename if you also snip This one is used and linked against all applications of cdrtools since 2.01a26 (previously only in cdrecord). If it is GPL incompatible it indeed breaks the e.g. mkisofs' and cdda2wav's original copyrights. That's certainly GPL-incompatible. It's an extra restriction. Since it affects functional behavior, I'd call it non-free. You must change this filename requirements are generally considered non-free, so I'd expect You may not change this filename without paying this fee requirements to be non-free. Why do we have to use the word fee to describe this sort of requirement? It's unnatural, and seems contrived to fit DFSG#1, which is unnecessary, since fees are not the only sort of restrictions prohibited by DFSG#1. I agree that this is non-free, as is any requirement that I explain myself before distributing a modification. -- Glenn Maynard
Re: cdrecord: weird GPL interpretation
On Fri, Sep 03, 2004 at 10:09:31AM +, Andreas Metzler wrote: The second issue * If you modify cdrecord you need to include additional version * printing code that [...] in cdrecord/cdrecord.c only applies to cdrecord which is completely copyrighted by JS. Therefore he is able to license it as GPL+restrictions and if the restrictions are still DFSG free we are able to ship it as part of Debian/main. Only if the implementation of that license is clear and consistent. I don't believe a work under the GPL with clarifications that don't follow from the GPL in any way is either. If he wants something like the GPL with extra restrictions, he should follow the procedure for modifying the GPL: rename it (the CDRPL), remove the preamble, and actually modify the text of the license. -- Glenn Maynard
Re: cdrecord: weird GPL interpretation
Because fee is an English word meaning a payment for a good or service. It really doesn't mean money only, in any context where precise language is used. If I have to perform in some way to obtain a license, then that's a fee. Do you have a better word, taking brevity and clarity into account? -Brian -- Brian Sniffen [EMAIL PROTECTED]
Re: cdrecord: weird GPL interpretation
On Fri, Sep 03, 2004 at 08:16:27PM -0400, Brian Thomas Sniffen wrote: Because fee is an English word meaning a payment for a good or service. It really doesn't mean money only, in any context where precise language is used. If I have to perform in some way to obtain a license, then that's a fee. Do you have a better word, taking brevity and clarity into account? Requirement. -- Raul
Re: cdrecord: weird GPL interpretation
On Fri, Sep 03, 2004 at 08:16:27PM -0400, Brian Thomas Sniffen wrote: Because fee is an English word meaning a payment for a good or service. It really doesn't mean money only, in any context where precise language is used. If I have to perform in some way to obtain a license, then that's a fee. Do you have a better word, taking brevity and clarity into account? Restriction. It includes all fees, and also includes things which are obviously not fees (such as, again, only on Tuesday), and is also directly tied to DFSG#1. I prefer it because contrived-feeling (whether legitimate or not) use of fee may spread the notion that only fees are covered by DFSG#1, and not other restrictions; and lead to more pointless dictionary- lawyering over whether something is a fee or not. -- Glenn Maynard
Re: cdrecord: weird GPL interpretation
Raul Miller [EMAIL PROTECTED] writes: On Fri, Sep 03, 2004 at 08:16:27PM -0400, Brian Thomas Sniffen wrote: Because fee is an English word meaning a payment for a good or service. It really doesn't mean money only, in any context where precise language is used. If I have to perform in some way to obtain a license, then that's a fee. Do you have a better word, taking brevity and clarity into account? Requirement. That's a much broader word. For example, a license which says I may only make modifications in French has a requirement, but that is not a fee. -Brian -- Brian Sniffen [EMAIL PROTECTED]
Re: cdrecord: weird GPL interpretation
On Fri, Sep 03, 2004 at 11:00:28PM -0400, Brian Thomas Sniffen wrote: On Fri, Sep 03, 2004 at 08:16:27PM -0400, Brian Thomas Sniffen wrote: Because fee is an English word meaning a payment for a good or service. It really doesn't mean money only, in any context where precise language is used. If I have to perform in some way to obtain a license, then that's a fee. Do you have a better word, taking brevity and clarity into account? Requirement. That's a much broader word. For example, a license which says I may only make modifications in French has a requirement, but that is not a fee. The point was that fee is a narrower word, and its use in this context (explaining rationale) is awkward, and only invites dictionary debates. I believe both requirement and restriction are better choices here (personally preferring restriction for its easy relationship to may not restrict in the DFSG). I don't think having to explain to the world at large in the readme why you did something is a fee or payment. We don't need to agree on this point, though; it's clearly a restriction. What matters is whether the restriction is considered onerous or not; whether it's a fee is irrelevant. -- Glenn Maynard
Re: cdrecord: weird GPL interpretation
MJ Ray [EMAIL PROTECTED] writes: On 2004-09-01 23:40:43 +0100 Måns Rullgård [EMAIL PROTECTED] wrote: Francesco Poli [EMAIL PROTECTED] writes: in cdrtools-2.01a38 I found the following weird GPL interpretation. [...] - You may not modify certain copyright messages in cdrecord.c See cdrecord.c for further information. Looks like an invariant section of GFDL infamy. [...] Well, this is like a misapplied bad invariant section, one that relates to the main subject and may actually be untrue. As far as I can tell, it's not an interpretation or clarification, but a direct contradiction of the GPL. The same goes for the restrictions on the config file default location. I guess that means we don't have a viable licence :-/ I take it someone on this list followed the recent flame war on lkml. Heh. Any flame war in particular? Jörg Schilling vs. rest of list. -- Måns Rullgård [EMAIL PROTECTED]
Re: cdrecord: weird GPL interpretation
Glenn Maynard writes: On Thu, Sep 02, 2004 at 12:19:26AM +0200, Francesco Poli wrote: Joerg Schilling Please note that this is just the way I interpret the GPL and as this is my software, users should follow my interpretation of the GPL and not use their own different interpretations. This came up previously: http://lists.debian.org/debian-legal/2003/10/msg00398.html I concur with Mathieu Roy: this is not an interpretation that follows from the GPL. The GPL clearly says you can do certain things and can not do certain other things; this interpretation contradicts the text of the GPL. I don't think it's safe to distribute a work where the author says one thing in his license text and a different thing in his interpretation. This person does not want the GPL; he wants something else. I do not believe that granting a license and then applying a bizarre interpretation or clarification to those terms is an acceptable approach to setting license terms. I also don't think Debian should distribute works that claim to be under the GPL when they're being modified by weird interpretations. (This is only my own opinion, of course, and not necessarily Debian policy.) Joerg's changes are clearly non-free; I've not seen anybody arguing otherwise. We basically need to route around him at this point, and fork from a previous free version. His ridiculous statement that his new statements also apply to older (GPL) versions of cdrtools should just be ignored as the puffery that it is IMHO... -- Steve McIntyre, Cambridge, UK.[EMAIL PROTECTED] Getting a SCSI chain working is perfectly simple if you remember that there must be exactly three terminations: one on one end of the cable, one on the far end, and the goat, terminated over the SCSI chain with a silver-handled knife whilst burning *black* candles. --- Anthony DeBoer
Re: cdrecord: weird GPL interpretation
On Thu, Sep 02, 2004 at 09:19:58AM -0400, Brian Thomas Sniffen wrote: Steve McIntyre [EMAIL PROTECTED] writes: Joerg's changes are clearly non-free; I've not seen anybody arguing otherwise. We basically need to route around him at this point, and fork from a previous free version. His ridiculous statement that his new statements also apply to older (GPL) versions of cdrtools should just be ignored as the puffery that it is IMHO... While legally you're right, I think from a point of view of politeness you're wrong. Maybe somebody who isn't Debian will fork cdrtools, but in the meantime it should just be moved to non-free. We need a valid license to ship it in non-free, and this doesn't really seem to be one. -- .''`. ** Debian GNU/Linux ** | Andrew Suffield : :' : http://www.debian.org/ | `. `' | `- -- | signature.asc Description: Digital signature
Re: cdrecord: weird GPL interpretation
Brian Thomas Sniffen [EMAIL PROTECTED] writes: Steve McIntyre [EMAIL PROTECTED] writes: Joerg's changes are clearly non-free; I've not seen anybody arguing otherwise. We basically need to route around him at this point, and fork from a previous free version. His ridiculous statement that his new statements also apply to older (GPL) versions of cdrtools should just be ignored as the puffery that it is IMHO... While legally you're right, I think from a point of view of politeness you're wrong. Go read some postings by JS and you won't feel any need for politeness. -- Måns Rullgård [EMAIL PROTECTED]
Re: cdrecord: weird GPL interpretation
Måns Rullgård [EMAIL PROTECTED] writes: Brian Thomas Sniffen [EMAIL PROTECTED] writes: Steve McIntyre [EMAIL PROTECTED] writes: Joerg's changes are clearly non-free; I've not seen anybody arguing otherwise. We basically need to route around him at this point, and fork from a previous free version. His ridiculous statement that his new statements also apply to older (GPL) versions of cdrtools should just be ignored as the puffery that it is IMHO... While legally you're right, I think from a point of view of politeness you're wrong. Go read some postings by JS and you won't feel any need for politeness. I've read them. It doesn't seem any worse than the drivel which shows up here regularly. Joerg tells Alan Cox he doesn't know anything about Linux systems or security. People here say things about that ridiculous once a week -- you've seen them too. On the other hand, I find this message interesting: http://lkml.org/lkml/2004/8/19/111 In particular, he seems to be relying on German Authors' Rights, and claims to be in discussion with Debian people. That's nearly a month ago. -Brian -- Brian Sniffen [EMAIL PROTECTED]
Re: cdrecord: weird GPL interpretation
On Thu, Sep 02, 2004 at 09:19:58AM -0400, Brian Thomas Sniffen wrote: While legally you're right, I think from a point of view of politeness you're wrong. Maybe somebody who isn't Debian will fork cdrtools, but in the meantime it should just be moved to non-free. Distributing a forked copy is just as polite (or impolite) as distributing a forkable copy. -- Raul
Re: cdrecord: weird GPL interpretation
On Thu, Sep 02, 2004 at 10:24:44AM -0400, Brian Thomas Sniffen wrote: http://lkml.org/lkml/2004/8/19/111 Is there any chance that someone has hacked his account? Alternatively, is there any chance that he's writing in german and relying on a program to translate what he says? Or, maybe, that he has some really significant problems understanding what a number of english words mean? -- Raul
Re: cdrecord: weird GPL interpretation
The copyright file for cdrtools is excellently done -- I wish all maintainers kept the separate threads of ownership so clear. It does make it pretty clear that cdrecord is not distributable. Followup-For: Bug #265546 Joerg Schilling's license is essentially the GNU GPL plus some extra restrictions. These restrictions are probably non-free. Call this the JS-GPL But cdrtools uses code copyrighted by others and licensed only under the real GNU GPL. The HFS code, for example, is copyright Robert Leslie. Some of the Mac isofs code is copyright James Pearson. The eltorito.c file in mkisofs is copyright Red Hat, and much of the rest is Yggdrasil's. Taken altogether, it looks like this package is not distributable by anybody with parts under the JS-GPL. -Brian -- Brian Sniffen [EMAIL PROTECTED]
Re: cdrecord: weird GPL interpretation
On Thu, Sep 02, 2004 at 01:11:42PM -0400, Brian Thomas Sniffen wrote: Taken altogether, it looks like this package is not distributable by anybody with parts under the JS-GPL. I've taken a look at a copy from January, and it has the same problem. I don't know how far back we'd have to go to find a legally distributable copy. -- Raul
Re: cdrecord: weird GPL interpretation
* M?ns Rullg?rd [EMAIL PROTECTED] [040902 17:11]: In particular, he seems to be relying on German Authors' Rights, and claims to be in discussion with Debian people. That's nearly a month ago. More specifically, he claims to be in discussion with Debian how to stop SuSE from doing what they have every right to do. I know nothing about German law, so I can't comment on that bit. Only thing in German law I could imagine is that he could withdraw some of his work. Which would require him to pay everyone what it is currently worth before the ban can take effect and make it impossible for him to allow others to use it. Anyone knows some reading if there is any other possibility and if this possibilty is even applicaple to software? Hochachtungsvoll, Bernhard R. Link -- Sendmail is like emacs: A nice operating system, but missing an editor and a MTA.
Re: cdrecord: weird GPL interpretation
Bernhard R. Link [EMAIL PROTECTED] writes: * M?ns Rullg?rd [EMAIL PROTECTED] [040902 17:11]: In particular, he seems to be relying on German Authors' Rights, and claims to be in discussion with Debian people. That's nearly a month ago. More specifically, he claims to be in discussion with Debian how to stop SuSE from doing what they have every right to do. I know nothing about German law, so I can't comment on that bit. Only thing in German law I could imagine is that he could withdraw some of his work. Which would require him to pay everyone what it is currently worth before the ban can take effect and make it impossible for him to allow others to use it. Anyone knows some reading if there is any other possibility and if this possibilty is even applicaple to software? Wow. does some reading on withdrawal of works. Some people make the kookiest laws. And hundreds of them have been elected to run Europe. It's like letting the crew of the B-ark take over... these are a perfectly valid system of laws for a society that isn't anything like this one peopled by people who aren't human. It's not that free software is impossible under a system with Authors' Rights and Withdrawal; you can't do business at all! Thanks for mentioning this law, Bernhard, but I have to hope that Joerg Schilling isn't relying on it. -Brian -- Brian Sniffen [EMAIL PROTECTED]
Re: cdrecord: weird GPL interpretation
On Thu, Sep 02, 2004 at 04:30:04PM -0400, Glenn Maynard wrote: [1] http://www.washington.edu/pine/faq/legal.html#10.2 (Accusing Free Software programmers of perverting the license by doing things they were clearly granted permission to do; that's wonderful.) Wasn't the force behind the license re-interpreation to stop people from distributing Pine with a maildir patch? --Adam -- Adam McKenna [EMAIL PROTECTED] [EMAIL PROTECTED]
Re: cdrecord: weird GPL interpretation
Francesco Poli [EMAIL PROTECTED] writes: Hi all, in cdrtools-2.01a38 I found the following weird GPL interpretation. I wonder if this is considered acceptable for main (I would say that this is non-free). I don't know whether cdrecord links with (or is otherwise a derivative work of) other GPL'd software (whose copyright is held by other people): in that case I would say that this is even undistributable... :( What do you think about this? There already is a Debian BTS bug report (http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=265546) about this issue (it was filed when it was even worse, it seems...), but it says it's resoved with version 2.01a38. I wonder if you agree... NOTE: I am Cc:ing the the package maintainer (Joerg Jaspert) and the bug-report filer (Andreas Metzler). I don't know if they would like to be Mail-Followup:ed... Issue description follows: -=-=-=-= cdrecord/LICENSE =-=-=-=- This software is under GPL but you should read the following clarifications: - You may not modify certain copyright messages in cdrecord.c See cdrecord.c for further information. Looks like an invariant section of GFDL infamy. - You may (with a few exceptions) not modify the location of the configuration file /etc/default/cdrecord. See defaults.c for further information. Looks like lunacy. I don't recall ever reading anything about that in the GPL. Please note that this is just the way I interpret the GPL and as this is my software, users should follow my interpretation of the GPL and not use their own different interpretations. -=-=-=-= cdrecord/cdrecord.c (sorry for linewrapping) =-=-=-=- I take it someone on this list followed the recent flame war on lkml. -- Måns Rullgård [EMAIL PROTECTED]
Re: cdrecord: weird GPL interpretation
On Thu, Sep 02, 2004 at 12:19:26AM +0200, Francesco Poli wrote: -=-=-=-= cdrecord/LICENSE =-=-=-=- This software is under GPL but you should read the following clarifications: - You may not modify certain copyright messages in cdrecord.c See cdrecord.c for further information. - You may (with a few exceptions) not modify the location of the configuration file /etc/default/cdrecord. See defaults.c for further information. Please note that this is just the way I interpret the GPL and as this is my software, users should follow my interpretation of the GPL and not use their own different interpretations. This came up previously: http://lists.debian.org/debian-legal/2003/10/msg00398.html I concur with Mathieu Roy: this is not an interpretation that follows from the GPL. The GPL clearly says you can do certain things and can not do certain other things; this interpretation contradicts the text of the GPL. I don't think it's safe to distribute a work where the author says one thing in his license text and a different thing in his interpretation. This person does not want the GPL; he wants something else. I do not believe that granting a license and then applying a bizarre interpretation or clarification to those terms is an acceptable approach to setting license terms. I also don't think Debian should distribute works that claim to be under the GPL when they're being modified by weird interpretations. (This is only my own opinion, of course, and not necessarily Debian policy.) -=-=-=-= cdrecord/cdrecord.c (sorry for linewrapping) =-=-=-=- [...] /* * Begin restricted code for quality assurance. * * Warning: you are not allowed to modify or to remove the * Copyright and version printing code below! This is still clearly non-free. Requiring an appropriate copyright notice is acceptable; prohibiting me from modifying the code that does so is not. * See also GPL § 2 subclause c) * * If you modify cdrecord you need to include additional version * printing code that: * * - Clearly states that the current version is an * inofficial (modified) version and thus may have bugs * that are not present in the original. * * - Print your support e-mail address and tell people that * you will do complete support for this version of * cdrecord. * * Or clearly state that there is absolutely no support * for the modified version you did create. This has been improved since the last time this came up, at least. * * - Tell the users not to ask the original author for * help. I don't know if it's free to require all of this cruft. I sure don't like having to say this software is buggy, the original is probably better!. It's not clear whether this needs to be printed every time, or just in eg. --version output. I don't know if the answer to that affects freeness. * This limitation definitely also applies when you use any other * cdrecord release together with libscg-0.6 or later, or when you * use any amount of code from cdrecord-1.11a17 or later. * In fact, it applies to any version of cdrecord, see also * GPL Preamble, subsection 6. I wasn't aware that the GPL's preamble had subsections. /* * WARNING you are only allowed to change this filename if you also * change the documentation and add a statement that makes clear * where the official location of the file is why you did choose a * nonstandard location and that the nonstandard location only refers * to inofficial cdrecord versions. * * I was forced to add this because some people change cdrecord without * rational reason and then publish the result. As those people * don't contribute work and don't give support, they are causing extra * work for me and this way slow down the cdrecord development. */ return (defltopen(/etc/default/cdrecord)); Requiring that I explain (or even have) rationale for changes can't possibly be free. -- Glenn Maynard
Re: cdrecord: weird GPL interpretation
On 2004-09-01 23:40:43 +0100 Måns Rullgård [EMAIL PROTECTED] wrote: Francesco Poli [EMAIL PROTECTED] writes: in cdrtools-2.01a38 I found the following weird GPL interpretation. [...] - You may not modify certain copyright messages in cdrecord.c See cdrecord.c for further information. Looks like an invariant section of GFDL infamy. [...] Well, this is like a misapplied bad invariant section, one that relates to the main subject and may actually be untrue. As far as I can tell, it's not an interpretation or clarification, but a direct contradiction of the GPL. The same goes for the restrictions on the config file default location. I guess that means we don't have a viable licence :-/ I take it someone on this list followed the recent flame war on lkml. Heh. Any flame war in particular? -- MJR/slefMy Opinion Only and not of any group I know http://www.ttllp.co.uk/ for creative copyleft computing Please email about: BT alternative for line rental+DSL; Education on SMEs+EU FP6; office filing that works fast