Re: Artistic and LGPL compatibility in jar files
Hi Anthony! On Sat, 19 Dec 2009 22:46:16 +, Anthony W. Youngman deb...@thewolery.demon.co.uk wrote: Ah, so you changed your position. Well, it doesn't matter whether the license grant is part of the work or not. If it is, it can be modified according to GPL. If it is not, GPL doesn't require to distribute it. Let's assume the GPL doesn't require it to be distributed. So you distribute a copy of my code, and a copy of the GPL. What right does your recipient have to distribute my code? They have two separate works, the GPL and my code, and NOTHING to link the two. So they can't distribute my code, because they have nothing that says they can. They have nothing that says the GPL that you gave them applies to the code (mine) that you gave them. Looks like they either don't have right to distribute your code or have it but don't know about it. To me, the former seems more probable if take the position that a license is a contract, the latter otherwise. Not sure I want to dig into it right now (or ever:). Both outcomes are undesirable so the explicit requirement of GPL to preserve references to GPL is important. It's not bullet-proof though. Sources could lack any references to licenses at all -- the license could be granted on a web-page, in an email or personally. (It's an interesting question how much legal force have such grants but that's a separate question.) In this case GPL seems not to require to pass a license grant. Hm, even a license grant in a separate README is suspicious... GPL itself requires to keep intact all notices stating that GPL applies to the code. So no problem here. Ah! Keep *intact* *all* notices stating that the GPL applies to the code. So you would argue that, for example, changing my statement that GPLv2+ applies to GPL v3 applies is keeping *all* notices *intact*? I'd describe that as butchering my notices, actually :-) This is very dependent on the form and the place of the notices. If the license grant is like this: This program can be used under GPLv3. Alternatively, it can be used under BSD. it can be seen as two separate notices and the section 4 of the GPLv3 permits to strip the second part. If you release your code with a notice on a web page saying this code [link] is licenced as 'GPL version 2 or later' then GPLv3 doesn't say that this license grant must be added to the Program's source code. But wait, GPLv3 actually permits to modify a license notices: 5. Conveying Modified Source Versions. You may convey a work based on the Program, or the modifications to produce it from the Program, in the form of source code under the terms of section 4, provided that you also meet all of these conditions: [...] b) The work must carry prominent notices stating that it is released under this License and any conditions added under section 7. This requirement modifies the requirement in section 4 to keep intact all notices. So GPLv3 permits to delete all previous notices about any licenses and add new prominent notices about GPLv3 only. You got me on a couple of points, I agree:-) I think I've got you on this one :-) I disagree:-) Actually I'm also convinced now (more or less) that it's NOT a grey area. Everybody preserves full license grants for upstream dual-licensed projects just because it's The Right Thing, not because it's legally required. And this is a feature, not a bug. It's quite in line with the section 7 of GPLv3: When you convey a copy of a covered work, you may at your option remove any additional permissions from that copy, or from any part of it. BTW, Mozilla tri-license boilerplate also explicitly spells it out: http://www.mozilla.org/MPL/boilerplate-1.1/mpl-tri-license-txt Alternatively, the contents of this file may be used under the terms of either the GNU General Public License Version 2 or later (the GPL), or the GNU Lesser General Public License Version 2.1 or later (the LGPL), in which case the provisions of the GPL or the LGPL are applicable instead of those above. If you wish to allow use of your version of this file only under the terms of either the GPL or the LGPL, and not to allow others to use your version of this file under the terms of the MPL, indicate your decision by deleting the provisions above and replace them with the notice and other provisions required by the GPL or the LGPL. If you do not delete the provisions above, a recipient may use your version of this file under the terms of any one of the MPL, the GPL or the LGPL. Notice that this gives you EXPLICIT PERMISSION to edit the licence grant I think this explicit permission is just an explanation of what is already there and not some additional permission. - permission which is notably lacking from the GPL. Example from the section 7 of the GPLv3 is above. Alexander Cherepanov -- To UNSUBSCRIBE, email to
Re: Artistic and LGPL compatibility in jar files
Hi Anthony! On Fri, 18 Dec 2009 10:17:48 +, Anthony W. Youngman deb...@thewolery.demon.co.uk wrote: Or if they receive an UNALTERED copy from you! Because if you change the licence (which you're not allowed to do) it's not an unaltered copy :-) Please don't not mix licenses and license grants:-) Let's consider it in more details: suppose I distribute your source code non-altered or non-creatively altered (so I don't have any copyright in this work) with GPLv3 attached and all references to other licenses (whether GPLv2 or BSD) stripped. AFAICS it's clearly permitted under clauses 4 and/or 5 of GPLv3. In which case, you HAVE altered my work. You've removed part of it, namely the licence grant. Not necessarily but for simplicity let's consider the case when yes, I have altered you work. Oh - and that probably is a very definite copyright violation :-) Please provide relevant quotes. I didn't grant you a licence to do that, Sure you did, it's called GPLv3. I granted you a licence to alter the program :-) You take the position that the program != the work? I'm suprised. And as someone else in this thread said, if they get one copy via one route that is GPL, and another via another route that is BSD, they think they can apply either licence to either copy. This is a very vague area. The copyright covers works, not copies. Accordingly GPL applies to works, not copies. So it doesn't metter which copy to use while it's the same work. But as far as I am concerned, legal niceties aside, if I dual-licence my work (such as, let's say, making it GPL v2+), if you strip off the v2 and change it to v3+ you are misrepresenting me to my users, I don't misrepresent you, I don't represent you at all. I just pass your work under GPLv3 and don't imply that you have licensed it under GPLv3 only in the first place. and you are stripping my users of the rights I granted them. Doesn't the GPL 2 itself say you mustn't impose further restrictions? Sure. And requirement to pass arbitrary license grants from original author is exactly further restriction. So you cannot require it (if you want your program to be distributable). What is removing the option to use v2, if not an unpermitted further restriction? Then let's see what GPLv2 really says about it. From section 6: You may not impose any further restrictions on the recipients' exercise of the rights granted herein. ^ So I cannot impose restrictions outside GPLv2, not outside some license grant. While this may be a legal grey area, it isn't a grey moral area - it's just unacceptable. I fully agree that it's important question. And would like to see some solid base here. Unfortunately, for now, I only see appeals to judge's common sense, morality etc. Alexander Cherepanov -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
In message 65986059fd940d55852a9fc4350fadd5.chere...@mccme.ru, Alexander Cherepanov chere...@mccme.ru writes Hi Anthony! On Fri, 18 Dec 2009 10:17:48 +, Anthony W. Youngman deb...@thewolery.demon.co.uk wrote: Or if they receive an UNALTERED copy from you! Because if you change the licence (which you're not allowed to do) it's not an unaltered copy :-) Please don't not mix licenses and license grants:-) Let's consider it in more details: suppose I distribute your source code non-altered or non-creatively altered (so I don't have any copyright in this work) with GPLv3 attached and all references to other licenses (whether GPLv2 or BSD) stripped. AFAICS it's clearly permitted under clauses 4 and/or 5 of GPLv3. In which case, you HAVE altered my work. You've removed part of it, namely the licence grant. Not necessarily but for simplicity let's consider the case when yes, I have altered you work. Oh - and that probably is a very definite copyright violation :-) Please provide relevant quotes. I didn't grant you a licence to do that, Sure you did, it's called GPLv3. In that case, seeing as you're taking the work to INCLUDE the licence grant, I assume the work also includes the licence? In which case I've just given you permission to alter the GPL :-) Which I *haven't* because I *can't*. I granted you a licence to alter the program :-) You take the position that the program != the work? I'm suprised. No. I'm taking the position that the licence *grant*, like the licence *itself*, is OUTSIDE of the work. I don't actually see how it can, legally, be part of the work itself, seeing as it's granting permissions to the work. And as someone else in this thread said, if they get one copy via one route that is GPL, and another via another route that is BSD, they think they can apply either licence to either copy. This is a very vague area. The copyright covers works, not copies. Accordingly GPL applies to works, not copies. So it doesn't metter which copy to use while it's the same work. But the GPL (or BSD or whatever) ONLY applies if, separate from BOTH the work, AND the licence, you have a document that tells you that the licence applies to the work. But as far as I am concerned, legal niceties aside, if I dual-licence my work (such as, let's say, making it GPL v2+), if you strip off the v2 and change it to v3+ you are misrepresenting me to my users, I don't misrepresent you, I don't represent you at all. I just pass your work under GPLv3 and don't imply that you have licensed it under GPLv3 only in the first place. Well, you MUST use GPLv3, or GPLv2, or GPL-whatever, to pass the work on. If your argument is correct, as I have said repeatedly, the mere act of passing on a v2+ work (*completely* *unaltered*) would, before v3 came out, have stripped the plus permission because it would have been distributed under v2. and you are stripping my users of the rights I granted them. Doesn't the GPL 2 itself say you mustn't impose further restrictions? Sure. And requirement to pass arbitrary license grants from original author is exactly further restriction. So you cannot require it (if you want your program to be distributable). But that is a requirement *I* am imposing on *you* (which, as copyright holder, I can do). Licence requirements NEVER apply to the copyright holder. So I *can* require it of you. What is removing the option to use v2, if not an unpermitted further restriction? Then let's see what GPLv2 really says about it. From section 6: You may not impose any further restrictions on the recipients' exercise of the rights granted herein. ^ So I cannot impose restrictions outside GPLv2, not outside some license grant. But without the licence grant, the GPLv2 doesn't apply. You seem to be assuming the licence grant is unimportant. No grant, no licence! While this may be a legal grey area, it isn't a grey moral area - it's just unacceptable. I fully agree that it's important question. And would like to see some solid base here. Unfortunately, for now, I only see appeals to judge's common sense, morality etc. Actually, as you can see, I'm convinced it's NOT a grey area. The licence grant tells you that you can use the licence with relation to the work. Therefore, the three (the licence, the grant and the work) are *legally* *separate*. Therefore, while the GPL gives you permission to alter the work, it does not give you permission to alter the (legally separate from the work) licence grant, which tells you that you have that permission! Cheers, Wol -- Anthony W. Youngman - anth...@thewolery.demon.co.uk -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
Hi Anthony! On Fri, 18 Dec 2009 22:33:41 +, Anthony W. Youngman deb...@thewolery.demon.co.uk wrote: Basically, you can choose which licence you want to apply to YOU. But you pass on my package as a whole (including my permission to choose which licence). So that's where your recipients get the same choices you got. I pass your code and GPLv3, there is no requirement to pass your full license grant. Just spotted something important :-) WITHOUT MY COPYING FILE your recipient has no evidence that the GPLv3 bears any relevance to my code. They have no evidences in any case. If they see some text like this program is licensed under GPL, how would they know that it was not arbitrarily added by some third-party in the way? You've just stripped all licencing from my code and In our imaginary case I've stripped only irrelevant licensing and have kept everything relevant to GPLv3. that MOST DEFINITELY IS a pretty blatant GPL violation! Can you please back at least some of your statements by anything aside from your own words? So to sum up, the GPL (whatever variant) is meaningless on its own. Passing the code on without my licencing grant is a GPL violation. And the GPL does NOT give you permission to change my grant. Here is my take on it. The requirements for preserving copyright notices etc. are in the section 4 of GPLv3: 4. Conveying Verbatim Copies. You may convey verbatim copies of the Program's source code as you receive it, in any medium, provided that you conspicuously and appropriately publish on each copy an appropriate copyright notice; Yes, I have published Copyright 2009 Anthony W. Youngman. keep intact all notices stating that this License and any Yes, all references to GPLv3 preserved. References to everything else (GPLv2, GPLv4+, BSD etc.) stripped. non-permissive terms added in accord with section 7 apply to the code; Let's suppose there are no non-permissive terms. keep intact all notices of the absence of any warranty; Yes, No warranty preserved. and give all recipients a copy of this License along with the Program. Sure, GPLv3 attached. If I modify your work I need to meet the conditions of the section 5 but this is not relevant to our discussion. As a bottom line, all conditions of GPL (either version) are met, GPL doesn't require to preserve license grants for any other licenses, and requiring to preserve them is not permitted (at least under GPLv2). My grant does give you the right to choose which licence applies to YOU. In fact, as I said elsewhere, you HAVE TO CHOOSE A SPECIFIC licence to apply to you. Yes. If you choosing a specific licence stripped your recipients' right to choose which licence applied to them, there would be no point to the or any later version wording because that would be invalid for any recipient beyond the first person to get it direct from the copyright holder. The point is that if I keep your full license grants then (and only then) my recipients can choose a license. But if I keep only the license which applies to me then they cannot choose a license. Alexander Cherepanov -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
Hi Anthony! On Sat, 19 Dec 2009 11:03:45 +, Anthony W. Youngman deb...@thewolery.demon.co.uk wrote: Or if they receive an UNALTERED copy from you! Because if you change the licence (which you're not allowed to do) it's not an unaltered copy :-) Please don't not mix licenses and license grants:-) Let's consider it in more details: suppose I distribute your source code non-altered or non-creatively altered (so I don't have any copyright in this work) with GPLv3 attached and all references to other licenses (whether GPLv2 or BSD) stripped. AFAICS it's clearly permitted under clauses 4 and/or 5 of GPLv3. In which case, you HAVE altered my work. You've removed part of it, ^^^ namely the licence grant. Not necessarily but for simplicity let's consider the case when yes, I have altered you work. Oh - and that probably is a very definite copyright violation :-) Please provide relevant quotes. I didn't grant you a licence to do that, Sure you did, it's called GPLv3. In that case, seeing as you're taking the work to INCLUDE the licence grant, No, that's you who said that the work includes the license grant -- see above, underlined. I assume the work also includes the licence? No. In which case I've just given you permission to alter the GPL :-) Which I *haven't* because I *can't*. Sure. I granted you a licence to alter the program :-) You take the position that the program != the work? I'm suprised. No. I'm taking the position that the licence *grant*, like the licence *itself*, is OUTSIDE of the work. I don't actually see how it can, legally, be part of the work itself, seeing as it's granting permissions to the work. Ah, so you changed your position. Well, it doesn't matter whether the license grant is part of the work or not. If it is, it can be modified according to GPL. If it is not, GPL doesn't require to distribute it. And as someone else in this thread said, if they get one copy via one route that is GPL, and another via another route that is BSD, they think they can apply either licence to either copy. This is a very vague area. The copyright covers works, not copies. Accordingly GPL applies to works, not copies. So it doesn't metter which copy to use while it's the same work. But the GPL (or BSD or whatever) ONLY applies if, separate from BOTH the work, AND the licence, you have a document that tells you that the licence applies to the work. Even if so, this only requires that you need two copies of such documents, not two copies of the work. But as far as I am concerned, legal niceties aside, if I dual-licence my work (such as, let's say, making it GPL v2+), if you strip off the v2 and change it to v3+ you are misrepresenting me to my users, I don't misrepresent you, I don't represent you at all. I just pass your work under GPLv3 and don't imply that you have licensed it under GPLv3 only in the first place. Well, you MUST use GPLv3, or GPLv2, or GPL-whatever, to pass the work on. Yes. If your argument is correct, as I have said repeatedly, the mere act of passing on a v2+ work (*completely* *unaltered*) would, before v3 came out, have stripped the plus permission because it would have been distributed under v2. No. That's unreleated question. Whether it's possible to modify license grant doesn't matter when it's not modified. and you are stripping my users of the rights I granted them. Doesn't the GPL 2 itself say you mustn't impose further restrictions? Sure. And requirement to pass arbitrary license grants from original author is exactly further restriction. So you cannot require it (if ^^ you want your program to be distributable). ^ But that is a requirement *I* am imposing on *you* (which, as copyright holder, I can do). Licence requirements NEVER apply to the copyright holder. So I *can* require it of you. Sure, that's why I added the part in the parentheses -- see above. What is removing the option to use v2, if not an unpermitted further restriction? Then let's see what GPLv2 really says about it. From section 6: You may not impose any further restrictions on the recipients' exercise of the rights granted herein. ^ So I cannot impose restrictions outside GPLv2, not outside some license grant. But without the licence grant, the GPLv2 doesn't apply. You seem to be assuming the licence grant is unimportant. No grant, no licence! GPL itself requires to keep intact all notices stating that GPL applies to the code. So no problem here. While this may be a legal grey area, it isn't a grey moral area - it's just unacceptable. I fully agree that it's important question. And would like to see some solid base here. Unfortunately, for now, I
Re: Artistic and LGPL compatibility in jar files
On Sat, 19 Dec 2009 17:47:29 +0300 Alexander Cherepanov wrote: [...] Actually I'm also convinced now (more or less) that it's NOT a grey area. Everybody preserves full license grants for upstream dual-licensed projects just because it's The Right Thing, not because it's legally required. And this is a feature, not a bug. For what it's worth, I am under the impression that Alexander is right here... -- New location for my website! Update your bookmarks! http://www.inventati.org/frx . Francesco Poli . GnuPG key fpr == C979 F34B 27CE 5CD8 DC12 31B5 78F4 279B DD6D FCF4 pgpP47eMKVkQB.pgp Description: PGP signature
Re: Artistic and LGPL compatibility in jar files
In message 7fdf4c21068c1acb3ed732c0cf862c1e.chere...@mccme.ru, Alexander Cherepanov chere...@mccme.ru writes Hi Anthony! On Sat, 19 Dec 2009 11:03:45 +, Anthony W. Youngman deb...@thewolery.demon.co.uk wrote: Or if they receive an UNALTERED copy from you! Because if you change the licence (which you're not allowed to do) it's not an unaltered copy :-) Please don't not mix licenses and license grants:-) Let's consider it in more details: suppose I distribute your source code non-altered or non-creatively altered (so I don't have any copyright in this work) with GPLv3 attached and all references to other licenses (whether GPLv2 or BSD) stripped. AFAICS it's clearly permitted under clauses 4 and/or 5 of GPLv3. In which case, you HAVE altered my work. You've removed part of it, ^^^ namely the licence grant. Not necessarily but for simplicity let's consider the case when yes, I have altered you work. Oh - and that probably is a very definite copyright violation :-) Please provide relevant quotes. I didn't grant you a licence to do that, Sure you did, it's called GPLv3. In that case, seeing as you're taking the work to INCLUDE the licence grant, No, that's you who said that the work includes the license grant -- see above, underlined. I assume the work also includes the licence? No. In which case I've just given you permission to alter the GPL :-) Which I *haven't* because I *can't*. Sure. I granted you a licence to alter the program :-) You take the position that the program != the work? I'm suprised. No. I'm taking the position that the licence *grant*, like the licence *itself*, is OUTSIDE of the work. I don't actually see how it can, legally, be part of the work itself, seeing as it's granting permissions to the work. Ah, so you changed your position. Well, it doesn't matter whether the license grant is part of the work or not. If it is, it can be modified according to GPL. If it is not, GPL doesn't require to distribute it. Let's assume the GPL doesn't require it to be distributed. So you distribute a copy of my code, and a copy of the GPL. What right does your recipient have to distribute my code? They have two separate works, the GPL and my code, and NOTHING to link the two. So they can't distribute my code, because they have nothing that says they can. They have nothing that says the GPL that you gave them applies to the code (mine) that you gave them. And as someone else in this thread said, if they get one copy via one route that is GPL, and another via another route that is BSD, they think they can apply either licence to either copy. This is a very vague area. The copyright covers works, not copies. Accordingly GPL applies to works, not copies. So it doesn't metter which copy to use while it's the same work. But the GPL (or BSD or whatever) ONLY applies if, separate from BOTH the work, AND the licence, you have a document that tells you that the licence applies to the work. Even if so, this only requires that you need two copies of such documents, not two copies of the work. But as far as I am concerned, legal niceties aside, if I dual-licence my work (such as, let's say, making it GPL v2+), if you strip off the v2 and change it to v3+ you are misrepresenting me to my users, I don't misrepresent you, I don't represent you at all. I just pass your work under GPLv3 and don't imply that you have licensed it under GPLv3 only in the first place. Well, you MUST use GPLv3, or GPLv2, or GPL-whatever, to pass the work on. Yes. If your argument is correct, as I have said repeatedly, the mere act of passing on a v2+ work (*completely* *unaltered*) would, before v3 came out, have stripped the plus permission because it would have been distributed under v2. No. That's unreleated question. Whether it's possible to modify license grant doesn't matter when it's not modified. and you are stripping my users of the rights I granted them. Doesn't the GPL 2 itself say you mustn't impose further restrictions? Sure. And requirement to pass arbitrary license grants from original author is exactly further restriction. So you cannot require it (if ^^ you want your program to be distributable). ^ But that is a requirement *I* am imposing on *you* (which, as copyright holder, I can do). Licence requirements NEVER apply to the copyright holder. So I *can* require it of you. Sure, that's why I added the part in the parentheses -- see above. What is removing the option to use v2, if not an unpermitted further restriction? Then let's see what GPLv2 really says about it. From section 6: You may not impose any further restrictions on the recipients' exercise of the rights granted herein.
Re: Artistic and LGPL compatibility in jar files
In message 9f4091d0c9afc9ede2ecc519bd6830bb.chere...@mccme.ru, Alexander Cherepanov chere...@mccme.ru writes Or if they receive an UNALTERED copy from you! Because if you change the licence (which you're not allowed to do) it's not an unaltered copy :-) Please don't not mix licenses and license grants:-) Let's consider it in more details: suppose I distribute your source code non-altered or non-creatively altered (so I don't have any copyright in this work) with GPLv3 attached and all references to other licenses (whether GPLv2 or BSD) stripped. AFAICS it's clearly permitted under clauses 4 and/or 5 of GPLv3. In which case, you HAVE altered my work. You've removed part of it, namely the licence grant. Oh - and that probably is a very definite copyright violation :-) I didn't grant you a licence to do that, I granted you a licence to alter the program :-) And as someone else in this thread said, if they get one copy via one route that is GPL, and another via another route that is BSD, they think they can apply either licence to either copy. This is a very vague area. But as far as I am concerned, legal niceties aside, if I dual-licence my work (such as, let's say, making it GPL v2+), if you strip off the v2 and change it to v3+ you are misrepresenting me to my users, and you are stripping my users of the rights I granted them. Doesn't the GPL 2 itself say you mustn't impose further restrictions? What is removing the option to use v2, if not an unpermitted further restriction? While this may be a legal grey area, it isn't a grey moral area - it's just unacceptable. Cheers, Wol -- Anthony W. Youngman - anth...@thewolery.demon.co.uk -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
In message 9f4091d0c9afc9ede2ecc519bd6830bb.chere...@mccme.ru, Alexander Cherepanov chere...@mccme.ru writes Basically, you can choose which licence you want to apply to YOU. But you pass on my package as a whole (including my permission to choose which licence). So that's where your recipients get the same choices you got. I pass your code and GPLv3, there is no requirement to pass your full license grant. Just spotted something important :-) WITHOUT MY COPYING FILE your recipient has no evidence that the GPLv3 bears any relevance to my code. You've just stripped all licencing from my code and that MOST DEFINITELY IS a pretty blatant GPL violation! So to sum up, the GPL (whatever variant) is meaningless on its own. Passing the code on without my licencing grant is a GPL violation. And the GPL does NOT give you permission to change my grant. My grant does give you the right to choose which licence applies to YOU. In fact, as I said elsewhere, you HAVE TO CHOOSE A SPECIFIC licence to apply to you. If you choosing a specific licence stripped your recipients' right to choose which licence applied to them, there would be no point to the or any later version wording because that would be invalid for any recipient beyond the first person to get it direct from the copyright holder. Cheers, Wol -- Anthony W. Youngman - anth...@thewolery.demon.co.uk -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
In message 20091217024135.af5a9f7...@nail.towers.org.uk, MJ Ray m...@phonecoop.coop writes Andrew Dalke wrote: On Dec 14, 2009, at 9:16 PM, Anthony W. Youngman wrote: I can't be bothered to read the book, but if it's the book I think it is, then I already have read it and came to the conclusion that the author was blind. [...] Read it for yourself, make sure you've got a copy of the GPL next you so you can *check* every reference he makes, and see if you come to the same conclusion I did, namely that the black letter of the GPL flatly contradicted the core assumption on which a large part of this book is based. You haven't read it and you made that conclusion? It sounds like you are promulgating hearsay and rumor. There's a free online copy which I linked to, and if what you are saying is right then it should be easy to point out some of the contradictions. This part followed if it's the book I think it is, then I already have read it. Maybe the contradictions aren't in the part of the book linked, but elsewhere in the book read. The link seemed to be to a PDF of part of a book and Anthony W. Youngman wrote that he couldn't be bothered to read it. Maybe a proper citation instead of a bare URL would have helped avoid this confusion. (Line wraps would help too.) Spot on. I tried to get back to find the TOC of the book, but once I stripped the page url, all I got was the home page, with no useful links I could find to get at the rest of the book. Further, Anthony W. Youngman isn't the only debian-legal contributor to think Larry Rosen's interpretations should not be taken wholesale, nor the only one who can't give full citations because those impressions were formed by interactions as much as literature. I'm another and I'm pretty sure there are others. Okay, I'll explain LONG-hand my problems with Larry. His critique of the GPL is based *entirely* on the premise that the GPL *implicitly* allows sublicensing. The GPL itself in black letter states if the GPL does not explicitly allow sublicencing, then it isn't allowed (not an exact quote I'll admit ... but it's what the GPL says). I think I pointed to that very clause ... I'm sorry, but if a lawyer can't understand BASIC legalese, then I don't trust them to be able to understand anything else! Cheers, Wol -- Anthony W. Youngman - anth...@thewolery.demon.co.uk -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
Andrew Dalke wrote: On Dec 17, 2009, at 3:41 AM, MJ Ray wrote: Maybe a proper citation instead of a bare URL would have helped avoid this confusion. (Line wraps would help too.) Since my first post, of which I think you are talking about, also included the book title and author name, I figured that was sufficient. [...] It included a variation of the book title, so it's not reasonable to blame anyone for being confused about whether it was exactly the same book or not. I think you're the first person in about 12 years to mention that linewraps are a problem. I stopped carefully linewrapping [...] Don't carefully linewrap - fix the mail client. I suspect Apple Mail is missing format=flowed when you want it, but I don't know for sure. I guess this nicely spikey reply might be why people stopped mentioning the breakage about 12 years ago. Not fun. I don't bother that much any more, but this list hasn't generally succumbed to the waves of Outlook easy-to-write-but-hard-to-read emails yet. So people who were persuaded to buy the book were persuaded by the book - is that surprising for this type of book? Pardon? One isn't required to purchase an item via Amazon before one can comment on said item, at least to my understanding. I believe one could get the book from the library and also comment on Amazon. Or read parts of it online and gratis, as I did. I click Create a Review. The options are No, I am a new customer and Yes, I have a password. Nothing suggests a way for non-customers to comment. Maybe it's possible, but it seems like a minority sport. Also, remember that Amazon ... It seemed an appropriate source to try to understand if the views of Youngman were singular, rare, or widely espoused. and I explained why the click-to-buy patenter's site might not be an appropriate source. OK? [...] Again, I was not thorough. Given that the response came so quickly I would assume it's a matter of a few moments to point to something definite, and that my details responses would indicate that it's not a trivially found and widely expressed idea. Assumption is the mother of all mistakes. There's also the point that Larry Rosen does seem to be a smart lawyer who can make a convincing argument which mere developers can't perfectly refute even if we're sure the conclusions like pay-my-lawyers are wrong. [...] I still hold that Youngman is wrong in saying that relicensing takes away user rights, as a universal statement. The best counter example is the GFDL-Creative Commons relicensing, when the original GFDL's license grant is essentially identical to the GPLs. OK, if that's the counter example, please show how a user can obtain the same rights they would have had with the GFDL'd copy when they only obtain one under a CC licence? Relicensing seems to remove a possible licence for everyone downstream. [...] As far as I recall (I read it too long ago), the book was partly a sales pitch for Rosen's licences I did not notice anything in the chapters I read which mentioned any of his licenses. I did not read the entire book. Nor do I know of the 5-point definition of which you also spoke. It may have occurred after he published the book. So you haven't read it and you made these conclusions? It sounds like you are going off at half-cock. There's a free online copy, you know. [...] There's also the doctrinaire point that Debian considers the Artistic License to be free, in opposition to GNU. This isn't that black-and-white. It's far fuzzier than that. GNU actually says We cannot say that this is a free software license because it is too vague; some passages are too clever for their own good, and their meaning is not clear. http://www.gnu.org/licenses/license-list.html#ArtisticLicense So, they are not willing to rule it in, but they don't really rule it out and they do accept it as part of perl's licence: http://www.gnu.org/licenses/license-list.html#PerlLicense For whatever reason, ftpmasters may have decided it's vague but good enough. Thanks to Artistic 2, this is a disappearing problem. [...] rather than blunt statements about my need to do yet more work, or vague and not easily confirmed statements regarding the character of the people involved. Sure, but it's a bit hard if someone challenges one to justify an opinion of another person on a mutual project years ago with documentary evidence. Given how long this project has been running, sometimes debian-legal contributors are primary sources, odd as that may seem. They still benefit from verifying in secondary sources, but inability to find such secondary sources doesn't make them wrong necessarily, just unverified. Regards, -- MJR/slef My Opinion Only: see http://people.debian.org/~mjr/ Please follow http://www.uk.debian.org/MailingLists/#codeofconduct -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact
Re: Artistic and LGPL compatibility in jar files
On Thu, 17 Dec 2009 05:06:41 +0100 Andrew Dalke wrote: [...] The best counter example is the GFDL-Creative Commons relicensing, when the original GFDL's license grant is essentially identical to the GPLs. Sorry, but I strongly disagree with your statement about the presumed essential identity between the GFDL and the GPL. The Debian Project has established (through General Resolution GR-2006-001: http://www.debian.org/vote/2006/vote_001 ) that GFDL'ed works without invariant material comply with the DFSG, but recognizes that even such works are *not* free of trouble. No comparable problematic clauses are found in the GPL. Moreover, I personally think that even GFDL'ed works without invariant material *fail* to comply with the DFSG, due to the other problematic clauses. If you take a look at the GR outcome, you'll see that my opinion is shared by a significant minority of Debian developers. -- New location for my website! Update your bookmarks! http://www.inventati.org/frx . Francesco Poli . GnuPG key fpr == C979 F34B 27CE 5CD8 DC12 31B5 78F4 279B DD6D FCF4 pgpEdJS69idS6.pgp Description: PGP signature
Re: Artistic and LGPL compatibility in jar files
Hi Anthony! On Tue, 15 Dec 2009 01:34:27 +, Anthony W. Youngman deb...@thewolery.demon.co.uk wrote: In section 10 (GPLv3): 10. Automatic Licensing of Downstream Recipients. Each time you convey a covered work, the recipient automatically receives a license from the original licensors, to run, modify and propagate that work, subject to this License. [...] ^^^ [skip] Actually, that then totally destroys the whole point of v3 or later if you choosing v3 takes away your recipients rights to choose according to the original author's grant! They are always free to get the program directly from original author (put aside the case of a program combined from different sources for a moment:-). Then they have a choice of license. But the law (generally) given the choice between a sensible interpretation, and an alternative that is either ludicrous or obviously not what was intended, will *usually* choose the sensible one. Yep, but in this case we have a unequivocal statement in the license. To say that what is written is obviously not what was intended, is, well, quite a stretch IMHO. Yet another variation: suppose you licensed your program to Alice under BSD and to Bob under GPLv3. Does recipients which get your program from Bob get BSD or GPLv3 or just GPLv3? Bob's recipients get just GPLv3. That's all he got, that's all he can pass on. To make it even worse, if somebody got one copy from Alice and one from Bob, Yes, this is even more interesting. I guess technically they'd have to keep the two copies (and associated licences) separate unless they contacted me and got my permission to combine them! I don't think so. Licenses apply to works, not copies. At the end of the day, YOU need a licence to distribute my code. My grant gives you a choice of v2 or v3. Whether you choose v2 or v3, your recipient then gets the same grant as you did, Sorry, I don't see where it comes from. Basically, you can choose which licence you want to apply to YOU. But you pass on my package as a whole (including my permission to choose which licence). So that's where your recipients get the same choices you got. I pass your code and GPLv3, there is no requirement to pass your full license grant. and they can also choose v2 or v3. If your choice of v3 took away your recipients choice of v2 I would consider that a VERY retrograde step. I agree and would be happy to learn where I'm wrong. But at the end of the day, it's simple. If I say v2 or v3 then I granted EVERY recipient of my code the right to *choose*. Yes, if they receive from you directly. Or if they receive an UNALTERED copy from you! Because if you change the licence (which you're not allowed to do) it's not an unaltered copy :-) Please don't not mix licenses and license grants:-) Let's consider it in more details: suppose I distribute your source code non-altered or non-creatively altered (so I don't have any copyright in this work) with GPLv3 attached and all references to other licenses (whether GPLv2 or BSD) stripped. AFAICS it's clearly permitted under clauses 4 and/or 5 of GPLv3. Alexander Cherepanov -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
On Dec 15, 2009, at 10:20 AM, Matthew Johnson wrote: Clause c and the fact that the author may have claims to the JUMBO name under trademark law means he can certainly require a name change. I don't think he can stop you from claiming that you can read and write his format, however. A secondary thing here, however, is that you generally want to get on with your upstream. If you start doing things he doesn't like, then he will make life difficult for you (see: ion3). Yeah. Since the biggest users of the Jumbo software, and also promotors of that CML format, distribute a patched version of the software, it's something they'll have to work out amongst themselves. I think it won't stay for all that long. Either that or I'll be an annoying bastard and harp on it in emails. ;) The feedback here has helped. The CML maintainers are going to split off the CC-BY-ND into another file which can go into non-free, the rest of the JUMBO code will clarified to be Apache 2.0, the CML developers are going through all their code to check that there are no other outstanding licensing details like that. There's the minor point outstanding of it Apache 2.0's relicense clause allows LGPL, but the only time that will come into play is if as of yet non-existent downstream providers package the software and distribute the derived system with a license fee. My judgement is that that is unlikely, what CML has done is enough, that the result is free (since it can all go to GPL), and therefore these changes fit into Debian's policy. Cheers! Andrew da...@dalkescientific.com -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
On Thu Dec 17 00:06, Andrew Dalke wrote: The feedback here has helped. The CML maintainers are going to split off the CC-BY-ND into another file which can go into non-free, the rest of the JUMBO code will clarified to be Apache 2.0, the CML developers are going through all their code to check that there are no other outstanding licensing details like that. I assume, then, that it can function without that non-free file? Matt -- Matthew Johnson signature.asc Description: Digital signature
Re: Artistic and LGPL compatibility in jar files
On Dec 17, 2009, at 12:19 AM, Matthew Johnson wrote: I assume, then, that it can function without that non-free file? Yes. Either it provides validation capabilities they don't need, or they have some hand-written code to deal with the parts that were automated because of having the schema around. Andrew da...@dalkescientific.com -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
Andrew Dalke wrote: On Dec 14, 2009, at 9:16 PM, Anthony W. Youngman wrote: I can't be bothered to read the book, but if it's the book I think it is, then I already have read it and came to the conclusion that the author was blind. [...] Read it for yourself, make sure you've got a copy of the GPL next to you so you can *check* every reference he makes, and see if you come to the same conclusion I did, namely that the black letter of the GPL flatly contradicted the core assumption on which a large part of this book is based. You haven't read it and you made that conclusion? It sounds like you are promulgating hearsay and rumor. There's a free online copy which I linked to, and if what you are saying is right then it should be easy to point out some of the contradictions. This part followed if it's the book I think it is, then I already have read it. Maybe the contradictions aren't in the part of the book linked, but elsewhere in the book read. The link seemed to be to a PDF of part of a book and Anthony W. Youngman wrote that he couldn't be bothered to read it. Maybe a proper citation instead of a bare URL would have helped avoid this confusion. (Line wraps would help too.) Further, Anthony W. Youngman isn't the only debian-legal contributor to think Larry Rosen's interpretations should not be taken wholesale, nor the only one who can't give full citations because those impressions were formed by interactions as much as literature. I'm another and I'm pretty sure there are others. BTW, none of the reviewers on Amazon agree with you http://www.amazon.com/Open-Source-Licensing-Software-Intellectual/product-reviews/0131487876/ref=dp_top_cm_cr_acr_txt?ie=UTF8showViewpoints=1 and I thought that if the the book would be that poorly written then there would be some evidence. [...] So people who were persuaded to buy the book were persuaded by the book - is that surprising for this type of book? Also, remember that Amazon filed the notorious click-to-buy patent, uses DRM/TPM to erase books from their e-book reader (RMS called it the Amazon Swindle) and tries to overthrow laws they don't like (such as France's Lang Law), so some free software fans won't touch them with a bargepole. It's not a good place to go for reviews of free software related books. It scores 3.8 our of 5 on http://www.librarything.com/work/72601 (compared to 4.17 for Free Software, Free Society: Selected Essays of Richard M. Stallman http://www.librarything.com/work/179957 which I think is the highest-rated book in the cluster: read them yet?) As far as I recall (I read it too long ago), the book was partly a sales pitch for Rosen's licences and also included an attempt to correct one of the big mistakes of the Open Source Initiative and pick a 5-point definition of Open Source which could actually compete with the 4-point Free Software Definition. I think OSI still use 10 points, so that's how convincing the book is. Even without knowing the problems of the choice-of-venue and pay-my-lawyers clauses in Larry Rosen's quesionable licences, it should be immediately obvious that that book is probably going to have an inflammatory perspective. Its title is Open Source Licensing: Software Freedom and Intellectual Property Law which manages to squeeze two of http://www.gnu.org/philosophy/words-to-avoid.html into one book title. Hope that illuminates, -- MJR/slef My Opinion Only: see http://people.debian.org/~mjr/ Please follow http://www.uk.debian.org/MailingLists/#codeofconduct -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
On Dec 17, 2009, at 3:41 AM, MJ Ray wrote: This part followed if it's the book I think it is, then I already have read it. Maybe the contradictions aren't in the part of the book linked, but elsewhere in the book read. Indeed. BTW, I should have interpreted the original phrase as read the linked document rather than read the book. I have not found those contradictions, and as I asked in my earlier response I would like an example. Maybe a proper citation instead of a bare URL would have helped avoid this confusion. (Line wraps would help too.) Since my first post, of which I think you are talking about, also included the book title and author name, I figured that was sufficient. Should I have also included publication year and publishing company? Or do I have to give the proper citation every time I repeat the same book link in a thread? I think you're the first person in about 12 years to mention that linewraps are a problem. I stopped carefully linewrapping when I started seeing all my nicely wrapped text look ugly once quoted a few times and displayed on systems which had automatic wrapping. I thought that nearly all of the email programs did that these days, including the text-based ones. Linewrapping at fixed column sizes also looks very ragged when viewed with proportional fonts. Further, Anthony W. Youngman isn't the only debian-legal contributor to think Larry Rosen's interpretations should not be taken wholesale, nor the only one who can't give full citations because those impressions were formed by interactions as much as literature. I'm another and I'm pretty sure there are others. Eternal September. I've never posted here before, and I'll be unsubscribing soon, once this thread is over. They did not come up in my searches for more information about this topic. So people who were persuaded to buy the book were persuaded by the book - is that surprising for this type of book? Pardon? One isn't required to purchase an item via Amazon before one can comment on said item, at least to my understanding. I believe one could get the book from the library and also comment on Amazon. Or read parts of it online and gratis, as I did. Also, remember that Amazon ... It seemed an appropriate source to try to understand if the views of Youngman were singular, rare, or widely espoused. It wasn't my only information source used to construct my reply, and I gave references to those other sources, including two letters by Stallman defending Rosen from more egregious statements made by reviewers of Rosen's book. In one of them Stallman does point out that Rosen's criticism did not hold up in court, but that is the only criticism I could find regarding the book that I could find from a freedom perspective. Again, I was not thorough. Given that the response came so quickly I would assume it's a matter of a few moments to point to something definite, and that my details responses would indicate that it's not a trivially found and widely expressed idea. It scores 3.8 our of 5 on http://www.librarything.com/work/72601 (compared to 4.17 for Free Software, Free Society: Selected Essays of Richard M. Stallman http://www.librarything.com/work/179957 which I think is the highest-rated book in the cluster: read them yet?) I had never heard of librarything before this. I will have to look at it some more. I have not read that collection of essays by Stallman. The point I was researching was in regards to Youngman's comment I'm always wary of explicitly relicencing. The GPL doesn't permit it, and by doing so you are taking away user rights. Searching Stallman's book now I see that relicense is not mentioned and sublicense is only mentioned as parts of the quoted GNU licenses. It provides no extra information to this topic. I still hold that Youngman is wrong in saying that relicensing takes away user rights, as a universal statement. The best counter example is the GFDL-Creative Commons relicensing, when the original GFDL's license grant is essentially identical to the GPLs. He urged me to Read what the GPL says, CAREFULLY, but I see nothing in GPLv2 which prevents the addition of a relicensing clause of the kind which occurred with GFDL. Rosen's book, on the other hand, did specifically discuss the need for sublicensing and relicensing, and helped me understand some of the changes that went into GPLv3. As well, it helped me understand some of the nuances between the BSD and MIT licenses. As far as I recall (I read it too long ago), the book was partly a sales pitch for Rosen's licences I did not notice anything in the chapters I read which mentioned any of his licenses. I did not read the entire book. Nor do I know of the 5-point definition of which you also spoke. It may have occurred after he published the book. it should be immediately obvious that that book is probably going to have an inflammatory perspective. Its title is Open Source Licensing: Software Freedom
Re: Artistic and LGPL compatibility in jar files
On Tue Dec 15 00:42, Andrew Dalke wrote: How do I interpret this LICENSE.txt? The Artistic License 2.0 allows relicensing to the GPL. I'm well and clear about that (though there's still a subtle question of if it allows relicensing to the LGPL). However, if I use clause 4(c)(ii) to switch the GPL, am I and my downstream users still prohibited from: - distributing the software under the name JUMBO (or a derivative) (Jumbo, Jr, Dumbo, Elephant and Timothy all seem derivative) - calling a modified version a compliant CML system - asserting that a modified version can read and write CML? That is, are these clauses additions to the Artistic License 2.0 which must be preserved even after 4(c)(ii) relicensing to the GPL? My suspicion is that derivatives must still be prohibited from those activities. It's interesting, the GPLv3 says explicitly: ... c) Prohibiting misrepresentation of the origin of that material, or requiring that modified versions of such material be marked in reasonable ways as different from the original version; or All other non-permissive additional terms are considered further restrictions within the meaning of section 10. If the Program as you received it, or any part of it, contains a notice stating that it is governed by this License along with a term that is a further restriction, you may remove that term. If a license document contains a further restriction but permits relicensing or conveying under this License, you may add to a covered work material governed by the terms of that license document, provided that the further restriction does not survive such relicensing or conveying. Clause c and the fact that the author may have claims to the JUMBO name under trademark law means he can certainly require a name change. I don't think he can stop you from claiming that you can read and write his format, however. A secondary thing here, however, is that you generally want to get on with your upstream. If you start doing things he doesn't like, then he will make life difficult for you (see: ion3). Is the resulting software (with these extra limitations) free software enough for Debian? Yes, there's ample example of rename clauses. Iceweasel is a high-profile example. DFSG4 says: ... The license may require derived works to carry a different name or version number from the original software. Matt -- Matthew Johnson signature.asc Description: Digital signature
Re: Artistic and LGPL compatibility in jar files
In message 76e62a33-41da-414c-a485-7819eb35f...@dalkescientific.com, Andrew Dalke da...@dalkescientific.com writes On Dec 13, 2009, at 2:24 AM, Anthony W. Youngman wrote: In message f4ccec28-fe42-4af3-b0c0-c832a6b0d...@dalkescientific.com, Andrew Dalke da...@dalkescientific.com writes I'm always wary of explicitly relicencing. The GPL doesn't permit it, and by doing so you are taking away user rights. Well, the GPL does allow relicensing to newer versions of the GPL... IT DOESN'T, ACTUALLY !!! Read what the GPL says, CAREFULLY. I didn't realize this was such a hot point to need the use of capital letters. Sorry. But this does press one of my hot buttons. There's an awful lot of people around who don't properly grok the GPL. (And you might guess I read groklaw avidly, where there's a lot of emphasis on getting things right.) I've made a lot of mistakes understanding the GPL, and been set right both on groklaw and here, and it's made me a bit fanatical about making sure other people get it right too... Pretend I said LGPL instead of GPL. In that case I can talk about relicensing, yes, since the LGPL explicitly allows relicensing to the GPL: http://www.gnu.org/licenses/gpl-faq.html#compat-matrix-footnote-7 7: LGPLv2.1 gives you permission to relicense the code under any version of the GPL since GPLv2. If you can switch the LGPLed code in this case to using an appropriate version of the GPL instead (as noted in the table), you can make this combination. I didn't realise the LGPL allowed that! Still, I've never really gone near it - all my experience with licencing has been to do with the GPL. LGPL is, after all, the Lesser GPL. In v3 the LGPL is specifically designed to give additional permissions than those of the GPL. You talked about how relicensing takes away user rights but in that case relicensing from LGPL to GPL is more taking away user permissions, yes? Yes, I'd agree with you there. But I think there it's only LGPL 2.1. Talking out of my hat here, the GPL v2 and LGPL v2.1 are different licences, so allowing the LGPL to be converted to the GPL makes sense. After all, rms isn't keen on the LGPL - it's just a useful stepping stone on the way to full GPL as far as he's concerned. And having seen that, I'd be rather wary of the LGPL 2.1! As you say, LGPL 3 just gives extra permissions over and above GPL 3, so if you mix the two licences the combination is just GPL 3 - no relicencing required :-) Still, the LGPL is designed to be relicensed to the GPL. What about something which doesn't have a built-in relicensing? Pretend I had said GFDL instead of GPL, in which case this quote from Stallman is highly relevant: http://www.fsf.org/blogs/licensing/2008-12-fdl-open-letter The relicensing option in GFDL 1.3 is fully consistent with the spirit and purpose of the GFDL. Stallman used the term 'relicense' several times in that open letter, and as a highly-visible response to the accusations of misdeeds during the GFDL/CC-BY-SA change, where 1.3 has an explicit section titled RELICENSING while 1.2 did not. He cannot have used it by mistake or as a poor word choice. Does that relicensing take away any user rights which are part of the spirit and purpose of the GFDL? (It does obviously take away the right to revert the license to 1.2, but is that an important right?) Let's go back to what I originally wrote - I'm wary of relicencing. While I don't think rms has done anything wrong (as far as I can see he has just enabled switching from one strong-copyleft licence to another), it still throws up the spectre of relicencing! That's why, actually, given the choice of LGPL 2.1 or 3, much as I haven't investigated 3 very much, I'll almost certainly prefer 3 to 2.1 because it means other people CAN'T relicence my code :-) Let's say I write a load of code, and release it with a notice saying this code is licenced as 'GPL version 2 or later' . The FSF suggests that you should write it thusly: 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 2 of the License, or (at your option) any later version. Okay, I'd use the FSF-recommended wording, fine. (Actually, personal choice, I'd probably take a leaf out of Linus' book and use the wording either version 2 or version 3.) But note, the GPL *itself* says that the recipient gets their licence from *me*. And the licence I would grant is 2+ or 2 or 3. Compare to the suggested text for the GFDL Copyright (c) YEAR YOUR NAME. Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation; with no Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts. A copy of the license is included in the section entitled GNU
Re: Artistic and LGPL compatibility in jar files
In message ded71701-a16b-4597-ac48-b1e0f7916...@dalkescientific.com, Andrew Dalke da...@dalkescientific.com writes On Dec 13, 2009, at 2:24 AM, Anthony W. Youngman wrote: In message f4ccec28-fe42-4af3-b0c0-c832a6b0d...@dalkescientific.com, Andrew Dalke da...@dalkescientific.com writes Well, the GPL does allow relicensing to newer versions of the GPL... IT DOESN'T, ACTUALLY !!! Read what the GPL says, CAREFULLY. Here is relevant commentary in Rosen's book Open Source Licensing book at http://rosenlaw.com/Rosen%5FCh06.pdf OMG! Larry Rosen! I can't be bothered to read the book, but if it's the book I think it is, then I already have read it and came to the conclusion that the author was blind. Read it for yourself, make sure you've got a copy of the GPL next to you so you can *check* every reference he makes, and see if you come to the same conclusion I did, namely that the black letter of the GPL flatly contradicted the core assumption on which a large part of this book is based. Oh - and I've more than enough experience of lawyers who's grasp of the law appears tenuous, I don't kow-tow to them until they've earnt my respect. (I respect them as a *person* until they *earn* my respect as a lawyer. If this is who I think he is, he lost that ... :-( Cheers, Wol -- Anthony W. Youngman - anth...@thewolery.demon.co.uk -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
In message e2301abde0ad8a5dfe21279f96b8fda6.chere...@mccme.ru, Alexander Cherepanov chere...@mccme.ru writes Hi Anthony! On Sun, 13 Dec 2009 01:24:36 +, Anthony W. Youngman deb...@thewolery.demon.co.uk wrote: Well, the GPL does allow relicensing to newer versions of the GPL... IT DOESN'T, ACTUALLY !!! Read what the GPL says, CAREFULLY. Let's say I write a load of code, and release it with a notice saying this code is licenced as 'GPL version 2 or later' . Typical dual-license scenario, good. Could you please elaborate some more how both licenses propagates in this case? This seems to be a very common notion but it's not clear to me. What this give YOU is the right to redistribute the code according to the terms of the GPL v3. BUT - READ THE GPL - the people to whom you give the code get their licence from ME, NOT YOU. Right, this is section 6 of GPLv2 of section 10 of GPLv3. Let's quote the latter: 10. Automatic Licensing of Downstream Recipients. Each time you convey a covered work, the recipient automatically receives a license from the original licensors, to run, modify and propagate that work, subject to this License. [...] And I granted the licence as v2 or later. The text of the GPL says subject to this License, i.e. GPLv3, not subject to all Licenses. So I don't see how your conclusion follows. As I said in another post, you're confusing the licence *grant* with the licence *itself*. Let's say I write some software and - as I would - I stick a notice that says this software is licenced v2 or v3. That is my grant. You now look at the code. You like v3, so you say v3 is my licence and distribute it as v3. Your recipients also get *my* grant, so any one of them can say actually, I like v *2* so I'll take that as my licence. Now let's say you write some code, add it to my work to make a derivative work and, being a trusting bloke your grant says your code is v3 or later. You can choose to distribute the code under v2 or v3, because you need to comply with my grant for my code. You can do what you like with your own code. Your recipients, now, can ONLY distribute under V3 ONLY. They can choose v2 for my code, but that won't let them distribute yours, so they can't distribute the derivative work under v2. They can choose v4 (when it comes out) for your code, but that won't let them distribute mine, so they can't use that for the derivative work. They CAN choose v3 which is valid for both your code, and mine, so the project COPYING file should say the only licence valid for the work as a whole is v3, but individual parts have their own licence and may be copied under a different licence, if appropriate. Cheers, Wol -- Anthony W. Youngman - anth...@thewolery.demon.co.uk -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
On Dec 14, 2009, at 8:36 PM, Anthony W. Youngman wrote: (And you might guess I read groklaw avidly, where there's a lot of emphasis on getting things right.) Sorry, but I don't know what groklaw is, at least, not enough to guess about your interests in it. I'm contacting debian-legal because I don't know enough about what the details are concerning a package where the developers want it to be distributed as part of Debian. After all, rms isn't keen on the LGPL - it's just a useful stepping stone on the way to full GPL as far as he's concerned. And having seen that, I'd be rather wary of the LGPL 2.1! For what it's worth, the authors of these packages I'm talking about want LGPL and are removing all traces of GPL-licensed code from their package. While I'm more of an BSD/MIT kinda person. The subject line of this post is also about the LGPL, so I'm really diverting things by going into a GPL discussion. Let's go back to what I originally wrote - I'm wary of relicencing. While I don't think rms has done anything wrong (as far as I can see he has just enabled switching from one strong-copyleft licence to another), it still throws up the spectre of relicencing! Or the more complete quote I'm always wary of explicitly relicencing. The GPL doesn't permit it, and by doing so you are taking away user rights. I still would like to know what user rights I'm taking away by relicensing. Stallman seems to think that relicensing is acceptable under some circumstances so long as the essential rights are preserved, which include the rights supported by GNU and the FSF. (I say essential rights because that is what Stallman used. There are obviously differences between the licenses.) Okay, I'd use the FSF-recommended wording, fine. (Actually, personal choice, I'd probably take a leaf out of Linus' book and use the wording either version 2 or version 3.) One of the projects I work with uses source code which was explicitly GPL version 2 only. Now they are starting to have problems integrating with GPLv3 software and they are considering if a massive rewrite is in order. But note, the GPL *itself* says that the recipient gets their licence from *me*. And the licence I would grant is 2+ or 2 or 3. I pointed out the quote from a copyright lawyer with a special interest in free software who said that the GPL was ambiguous about sublicensing and if a chain of licenses was required or not. Oh - and the GFDL 1.2 does *not* allow relicensing to CC-BY-SA. Your legal logic has slipped up. You've made the elementary error of confusing the *grant* of licence with the licence *itself*. If I use the recommended wording from GNU, which is what I quoted and was using as a reference, then the phrase is Version 1.2 or any later version published by the Free Software Foundation; Obviously if the license says 1.2 and leaves out that provision for sublicensing/ relicening/ whatever you want to call it, then it's not possible to change it. The GPL even has a section where it describes the impact or later has on being able to re/sub/license. Just like if something says GPL 2 and leaves out the provision for or later then it cannot be changed to GPLv3. It's just, well, I didn't say that. I don't understand why you think I'm confused about this matter. If I licence my work as GFDL 1.2 and you relicence it as CC-BY-SA, I'll be after you for infringement like a shot! You *need* the or later wording in the grant, and that is nothing to do with the licence itself. I see. It's because when I said Well, the GPL does allow relicensing to newer versions of the GPL it's because I should have written Well, the GPL does allow relicensing to newer versions of the GPL so long as you use the recommended phrase from GNU which allows 'or later' licenses to be applied. If you get a work as GFDL 1.2 or later, *then* and *then*only* can you say to yourself stuff 1.2, I'll use 1.3 which allows you to pass the stuff on as CC-BY-SA instead of GFDL. I really do think you're reacting to what was a minor omission on my part, by my using the phrase the GPL when it was the grant that most people make to use the GPL. Pointing out that omission would have been more clarifying than saying IT DOESN'T, ACTUALLY !!! My omission came up because you outright declared that relicensing takes away rights, when it's clear from the history of free licenses that relicensing does occur and that it's possible to relicense without taking away (essential) rights. In this case, I think *YOU* are now the licensor. My legal-fu isn't up to this, but if I originally granted GFDL, I don't think a CC-BY-SA recipient can get their CC-BY-SA licence from me (they are *restrospectively* changing my grant of licence, which I don't think is possible). Likewise with LGPL 2.1, I think the person who changes the licence to GPL becomes the new licensor. Except that that wasn't at all
Re: Artistic and LGPL compatibility in jar files
Hi Anthony! On Mon, 14 Dec 2009 20:25:22 +, Anthony W. Youngman deb...@thewolery.demon.co.uk wrote: Right, this is section 6 of GPLv2 of section 10 of GPLv3. Let's quote the latter: 10. Automatic Licensing of Downstream Recipients. Each time you convey a covered work, the recipient automatically receives a license from the original licensors, to run, modify and propagate that work, subject to this License. [...] And I granted the licence as v2 or later. The text of the GPL says subject to this License, i.e. GPLv3, not subject to all Licenses. So I don't see how your conclusion follows. As I said in another post, you're confusing the licence *grant* with the licence *itself*. Let's say I write some software and - as I would - I stick a notice that says this software is licenced v2 or v3. That is my grant. Ok. You now look at the code. You like v3, so you say v3 is my licence and distribute it as v3. Ok. Your recipients also get *my* grant, so any one of them can say actually, I like v *2* so I'll take that as my licence. Why do you think that my recipients will get your entire grant? GPLv3 only says that they will get your grant for _this_ License, i.e. GPLv3. Now let's say you write some code, add it to my work to make a derivative work and, being a trusting bloke your grant says your code is v3 or later. You can choose to distribute the code under v2 or v3, because you need to comply with my grant for my code. You can do what you like with your own code. Your recipients, now, can ONLY distribute under V3 ONLY. They can choose v2 for my code, but that won't let them distribute yours, so they can't distribute the derivative work under v2. They can choose v4 (when it comes out) for your code, but that won't let them distribute mine, so they can't use that for the derivative work. They CAN choose v3 which is valid for both your code, and mine, so the project COPYING file should say the only licence valid for the work as a whole is v3, but individual parts have their own licence and may be copied under a different licence, if appropriate. I agree here. Alexander Cherepanov -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
On Mon, 14 Dec 2009 19:36:58 + Anthony W. Youngman wrote: [...] That's why, actually, given the choice of LGPL 2.1 or 3, much as I haven't investigated 3 very much, I'll almost certainly prefer 3 to 2.1 because it means other people CAN'T relicence my code :-) Please note that adopting the LGPLv3 makes the work not linkable with GPLv2 (only) works. If you choose the LGPL in order to obtain a weak copyleft that allows linking with proprietary code, forbidding linking with GPLv2 code looks a bit awkward... -- New location for my website! Update your bookmarks! http://www.inventati.org/frx . Francesco Poli . GnuPG key fpr == C979 F34B 27CE 5CD8 DC12 31B5 78F4 279B DD6D FCF4 pgpP6Acj7vvvd.pgp Description: PGP signature
Re: Artistic and LGPL compatibility in jar files
On Dec 14, 2009, at 9:16 PM, Anthony W. Youngman wrote: I can't be bothered to read the book, but if it's the book I think it is, then I already have read it and came to the conclusion that the author was blind. Still, I have given references to Stallman, to the GNU pages, to the XEmacs project, and to Rosen, in order to back up my views and understanding. I would like to see some external references to back up your statements. Read it for yourself, make sure you've got a copy of the GPL next to you so you can *check* every reference he makes, and see if you come to the same conclusion I did, namely that the black letter of the GPL flatly contradicted the core assumption on which a large part of this book is based. You haven't read it and you made that conclusion? It sounds like you are promulgating hearsay and rumor. There's a free online copy which I linked to, and if what you are saying is right then it should be easy to point out some of the contradictions. Are you sure you haven't confused Don Rosenberg with Larry Rosen? Stallman wrote http://richardstallman.sys-con.com/node/128143/mobile Don Rosenberg's review in LWM (Vol. 3, issue 4) of Larry Rosen's book, Open Source Licensing, did double-duty as a platform for FUD about the GNU GPL. and while Stallman expresses disagreements with Rosen on requirements on combined works, he is much more directed towards Rosenberg. Stallman also writes in: http://richardstallman.sys-con.com/node/48833/mobile Richard Stallman writes: Maureen O'Gara's review in Linux Business Week of Larry Rosen's book misrepresents the Free Software Foundation's views, when it says we criticized Rosen for recognizing...licenses other than the GPL. I find nothing by Stallman which expresses similar levels of dismissiveness towards Rosen as you have, which should have arisen if the book was as against the GPL as you say. BTW, none of the reviewers on Amazon agree with you http://www.amazon.com/Open-Source-Licensing-Software-Intellectual/product-reviews/0131487876/ref=dp_top_cm_cr_acr_txt?ie=UTF8showViewpoints=1 and I thought that if the the book would be that poorly written then there would be some evidence. I have now read a few chapters of Rosen and found them to be instructive, interesting, and clear. Oh - and I've more than enough experience of lawyers who's grasp of the law appears tenuous, I don't kow-tow to them until they've earnt my respect. (I respect them as a *person* until they *earn* my respect as a lawyer. If this is who I think he is, he lost that ... :-( Congratulations. As for me, citations needed. Andrew da...@dalkescientific.com -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
In message e15688105c664a6bde658d2ac687ad82.chere...@mccme.ru, Alexander Cherepanov chere...@mccme.ru writes Your recipients also get *my* grant, so any one of them can say actually, I like v *2* so I'll take that as my licence. Why do you think that my recipients will get your entire grant? GPLv3 only says that they will get your grant for _this_ License, i.e. GPLv3. WHERE does it say that? But in that case, as soon as you distribute my code using GPL2 as your licence, YOU have STOPPED them distributing under version 3! That argument cuts both ways! Actually, that then totally destroys the whole point of v3 or later if you choosing v3 takes away your recipients rights to choose according to the original author's grant! I've just checked v3, and it contains the same gets your licence from the original licensor wording as v2, so they get their grant from me, and you don't have the right (or ability) to change what I grant. At the end of the day, YOU need a licence to distribute my code. My grant gives you a choice of v2 or v3. Whether you choose v2 or v3, your recipient then gets the same grant as you did, and they can also choose v2 or v3. If your choice of v3 took away your recipients choice of v2 I would consider that a VERY retrograde step. But at the end of the day, it's simple. If I say v2 or v3 then I granted EVERY recipient of my code the right to *choose*. Both v2 and v3 are explicit that your recipients get their rights from ME not you, so your choice of v3 does not constrain their right to choose. Cheers, Wol -- Anthony W. Youngman - anth...@thewolery.demon.co.uk -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
In message 06db76b9-3d28-44ab-82c8-e23917bf3...@dalkescientific.com, Andrew Dalke da...@dalkescientific.com writes On Dec 14, 2009, at 8:36 PM, Anthony W. Youngman wrote: (And you might guess I read groklaw avidly, where there's a lot of emphasis on getting things right.) Sorry, but I don't know what groklaw is, at least, not enough to guess about your interests in it. I'm contacting debian-legal because I don't know enough about what the details are concerning a package where the developers want it to be distributed as part of Debian. It's a law site, where SCO Group's lawsuit against IBM, Novell and Linux in general is getting thoroughly dissected. If you're not interested then fair enough, but copyright and the GPL in particular are very important there. After all, rms isn't keen on the LGPL - it's just a useful stepping stone on the way to full GPL as far as he's concerned. And having seen that, I'd be rather wary of the LGPL 2.1! For what it's worth, the authors of these packages I'm talking about want LGPL and are removing all traces of GPL-licensed code from their package. While I'm more of an BSD/MIT kinda person. The subject line of this post is also about the LGPL, so I'm really diverting things by going into a GPL discussion. Sounds weird to me you're deferring to rms then :-) While he'd defend your *right* to choose BSD/MIT or LGPL, he'd be very sorry about your choice - you should be choosing GPL :-) Let's go back to what I originally wrote - I'm wary of relicencing. While I don't think rms has done anything wrong (as far as I can see he has just enabled switching from one strong-copyleft licence to another), it still throws up the spectre of relicencing! Or the more complete quote I'm always wary of explicitly relicencing. The GPL doesn't permit it, and by doing so you are taking away user rights. I still would like to know what user rights I'm taking away by relicensing. Stallman seems to think that relicensing is acceptable under some circumstances so long as the essential rights are preserved, which include the rights supported by GNU and the FSF. And those are the rights your choice of licences do NOT preserve. Bearing in mind that, above my quote, I also wrote that I thought the artistic licence was close to BSD (ie not strong copyleft). You can relicence BSD as closed source - where are your essential rights now? I obviously thought something similar could happen with artistic. (looking at it - especially artistic 2 - in more detail, I see that it's far more strong copyleft than I thought.) (I say essential rights because that is what Stallman used. There are obviously differences between the licenses.) Okay, I'd use the FSF-recommended wording, fine. (Actually, personal choice, I'd probably take a leaf out of Linus' book and use the wording either version 2 or version 3.) One of the projects I work with uses source code which was explicitly GPL version 2 only. Now they are starting to have problems integrating with GPLv3 software and they are considering if a massive rewrite is in order. Well, linux itself is explicitly v2 only :-) And I'd follow Linus lead and be very wary of the or later wording. I'd actually (with hindsight :-) do it differently and say that any version of the GPL extant * as of the copyright date * could be used. But note, the GPL *itself* says that the recipient gets their licence from *me*. And the licence I would grant is 2+ or 2 or 3. I pointed out the quote from a copyright lawyer with a special interest in free software who said that the GPL was ambiguous about sublicensing and if a chain of licenses was required or not. I see the GPL explicitly agrees with me, not Larry Rosen :-) !!! This is the GPL v3 - read the last section of 2. Basic Permissions : Conveying under any other circumstances is permitted solely under the conditions stated below. Sublicensing is not allowed; section 10 makes it unnecessary. Now let's look at section 10 ... 10. Automatic Licensing of Downstream Recipients. Each time you convey a covered work, the recipient automatically receives a license from the original licensors, to run, modify and propagate that work, subject to this License. You are not responsible for enforcing compliance by third parties with this License. This is exactly the section (maybe worded, certainly numbered, differently) that I have repeatedly been referring to from the GPL v2. Oh - and the GFDL 1.2 does *not* allow relicensing to CC-BY-SA. Your legal logic has slipped up. You've made the elementary error of confusing the *grant* of licence with the licence *itself*. If I use the recommended wording from GNU, which is what I quoted and was using as a reference, then the phrase is Version 1.2 or any later version published by the Free Software Foundation; Obviously if the license says 1.2 and leaves out that provision for sublicensing/ relicening/
Re: Artistic and LGPL compatibility in jar files
In message 20091214220044.1cc797d6@firenze.linux.it, Francesco Poli f...@firenze.linux.it writes On Mon, 14 Dec 2009 19:36:58 + Anthony W. Youngman wrote: [...] That's why, actually, given the choice of LGPL 2.1 or 3, much as I haven't investigated 3 very much, I'll almost certainly prefer 3 to 2.1 because it means other people CAN'T relicence my code :-) Please note that adopting the LGPLv3 makes the work not linkable with GPLv2 (only) works. If you choose the LGPL in order to obtain a weak copyleft that allows linking with proprietary code, forbidding linking with GPLv2 code looks a bit awkward... I'd add an exception to LGPL 3 :-) Actually, I'd probably choose pure GPL 2 or 3 :-) Cheers, Wol -- Anthony W. Youngman - anth...@thewolery.demon.co.uk -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
Hi Anthony! On Mon, 14 Dec 2009 21:44:35 +, Anthony W. Youngman deb...@thewolery.demon.co.uk wrote: Your recipients also get *my* grant, so any one of them can say actually, I like v *2* so I'll take that as my licence. Why do you think that my recipients will get your entire grant? GPLv3 only says that they will get your grant for _this_ License, i.e. GPLv3. WHERE does it say that? In section 10 (GPLv3): 10. Automatic Licensing of Downstream Recipients. Each time you convey a covered work, the recipient automatically receives a license from the original licensors, to run, modify and propagate that work, subject to this License. [...] ^^^ GPLv2 says effectively the same: 6. Each time you redistribute the Program (or any work based on the Program), the recipient automatically receives a license from the original licensor to copy, distribute or modify the Program subject to ^^ these terms and conditions. [...] ^^ But in that case, as soon as you distribute my code using GPL2 as your licence, YOU have STOPPED them distributing under version 3! That argument cuts both ways! Sure. Actually, that then totally destroys the whole point of v3 or later if you choosing v3 takes away your recipients rights to choose according to the original author's grant! They are always free to get the program directly from original author (put aside the case of a program combined from different sources for a moment:-). Then they have a choice of license. Some variation of the scenario: suppose your grant is this software is licensed under BSD or GPLv3 and I choose GPLv3. Does this mean that my recipients still get BSD or GPLv3? Yet another variation: suppose you licensed your program to Alice under BSD and to Bob under GPLv3. Does recipients which get your program from Bob get BSD or GPLv3 or just GPLv3? I've just checked v3, and it contains the same gets your licence from the original licensor wording as v2, so they get their grant from me, and you don't have the right (or ability) to change what I grant. I hope quotes above explain what I mean. At the end of the day, YOU need a licence to distribute my code. My grant gives you a choice of v2 or v3. Whether you choose v2 or v3, your recipient then gets the same grant as you did, Sorry, I don't see where it comes from. and they can also choose v2 or v3. If your choice of v3 took away your recipients choice of v2 I would consider that a VERY retrograde step. I agree and would be happy to learn where I'm wrong. But at the end of the day, it's simple. If I say v2 or v3 then I granted EVERY recipient of my code the right to *choose*. Yes, if they receive from you directly. Both v2 and v3 are explicit that your recipients get their rights from ME not you, so your choice of v3 does not constrain their right to choose. Alexander Cherepanov -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
On Dec 14, 2009, at 11:24 PM, Anthony W. Youngman wrote: It's a law site, where SCO Group's lawsuit against IBM, Novell and Linux in general is getting thoroughly dissected. If you're not interested then fair enough, but copyright and the GPL in particular are very important there. I have and enjoy using my Mac. I remember that case but did not follow the details. Sounds weird to me you're deferring to rms then :-) While he'd defend your *right* to choose BSD/MIT or LGPL, he'd be very sorry about your choice - you should be choosing GPL :-) This is an issue of Artistic License and LGPL compatibility which I'm bringing up because people I associate want to include their software with Debian. Debian's policies and those of Stallman's are well aligned, excepting that Stallman and the FSF does not consider the Artistic License to be a free license while Debian does. My deference, as you incorrectly put it, is because that's the closest proxy I have to something which is definite for this argument. My own views on BSD are largely irrelevant. I also wrote that I thought the artistic licence was close to BSD (ie not strong copyleft). And that would not be correct. The Artistic License has many clauses which are not BSD-like, such has prohibitions against license fees which the BSD does allow. You can relicence BSD as closed source - where are your essential rights now? I quoted and used the phrase essential rights because it derives from Stallman's use in the GFDL-Creative Commons relicensing, which he specifically called a relicensing which does not lose essential rights. I obviously thought something similar could happen with artistic. (looking at it - especially artistic 2 - in more detail, I see that it's far more strong copyleft than I thought.) Why did you not look at it the first time, instead of making quite incorrect assumptions about it? Well, linux itself is explicitly v2 only :-) I was unaware of that. I rarely am on a Linux box. My main server machine is a FreeBSD box. I pointed out the quote from a copyright lawyer with a special interest in free software who said that the GPL was ambiguous about sublicensing and if a chain of licenses was required or not. I see the GPL explicitly agrees with me, not Larry Rosen :-) !!! This is the GPL v3 - read the last section of 2. Basic Permissions : Which means you didn't look at the top of the first page of the link I sent you, where you would see the book was written in 2004 and therefore pre-GPLv3. It also means you didn't recall my original text where I wrote: As you can tell, a professional lawyer in this field is not clear about if the GPLv2 allows sublicensing, so I hope it's understandable how someone could view a change from GPLv2 to GPLv3 without keeping the chain of titles (which is the common practice) could be considered a relicense. I believe I have careful to only used references from that book with respect to GPLv2, and not use it as a way to interpret reading the book has helped me understand some of the improvements made in GPLv3. The above was one of the few cases where I was not. The proper behavior should be to point out that I likely was imprecise and should have written GPLv2 instead of simply GPL. 10. Automatic Licensing of Downstream Recipients. Each time you convey a covered work, the recipient automatically receives a license from the original licensors, to run, modify and propagate that work, subject to this License. You are not responsible for enforcing compliance by third parties with this License. This is exactly the section (maybe worded, certainly numbered, differently) that I have repeatedly been referring to from the GPL v2. This is the specific improvement to text which Rosen says is ambiguous in GPLv2. As you have not bothered to read the text and yet still comment on what you believe he has written, I shall copy it here: http://rosenlaw.com/Rosen%5FCh06.pdf This GPL section 4, with its negative wording, is also the only place that references the right to sublicense. One might assume from the way GPL section 4 is worded that the right to sublicense was intended in sections 1 (right to copy), 2 (right to modify) and 3 (right to distribute) as well. However, section 6 implies that there are no sublicenses but instead a direct license from each up-stream contributor: ... As to sublicensing, then, the GPL is ambiguous. I refer you to the discussion in Chapter 5 of sublicensing in the MIT license. Sublicensing rights can be very important to open source distributors for dealing properly with the chain of title to contributions. In practice, most software projects ignore the issue completely and assume that, for GPL software, only the most recent license in the chain of title matters. They assume that GPL licensed software is sublicenseable, but the GPL isn’t clear about that. Sorry, I know I'm
Re: Artistic and LGPL compatibility in jar files
Anthony W. Youngman deb...@thewolery.demon.co.uk writes: As I said in another post, you're confusing the licence *grant* with the licence *itself*. It might be clearer to say that the issue is a confusion between the license grant versus the license terms. Let's say I write some software and - as I would - I stick a notice that says this software is licenced v2 or v3. That is my grant. More precisely, the grant would need to say (words to the effect of) either: You may do X, Y, Z to this work under the following terms: foo, bar, baz. or: You may do X, Y, Z to this work under the terms of foobar license; see $EXPLICIT_REFERENCE_TO_SPECIFIC_VERSION_OF_FOOBAR_TERMS. The former is a license grant immediately followed by license terms. (That's what you get often with very short license terms, such as Expat.) The latter is a license grant only, referring the reader to another document to discover what the license terms are. (That's what you get often with longer license terms, such as the GPL.) When people say “the GPL” in the context of licensing, they're referring to *the set of terms* that make up the GPL. Those terms, floating in a document, don't grant anything to anyone. The grant of license must, as Anthony points out, be distinct and explicit, only applies to some specific work, and can only be granted by the copyright holder for that work. -- \ “He who wonders discovers that this in itself is wonder.” | `\ —Maurits Cornelis Escher | _o__) | Ben Finney -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
On Dec 15, 2009, at 12:20 AM, Ben Finney wrote: More precisely, the grant would need to say (words to the effect of) either: You may do X, Y, Z to this work under the following terms: foo, bar, baz. or: You may do X, Y, Z to this work under the terms of foobar license; see $EXPLICIT_REFERENCE_TO_SPECIFIC_VERSION_OF_FOOBAR_TERMS. which brings my back to my original question. The LICENSE.txt file from JUMBO/CML says All JUMBO code is distributed under the Open Source Artistic License (http://www.opensource.org). You are free to modify the code but if you do it may no longer be distributed under the name JUMBO (or a derivative) without permission of Peter Murray-Rust. Any distribution must acknowledge the origins and also include copies of the JUMBO source (see Artistic License for details). You may not claim that a modified version is a compliant CML system and may not assert that it reads or writes CML. In private mail the copyright owners have clarified that this is 2.0. How do I interpret this LICENSE.txt? The Artistic License 2.0 allows relicensing to the GPL. I'm well and clear about that (though there's still a subtle question of if it allows relicensing to the LGPL). However, if I use clause 4(c)(ii) to switch the GPL, am I and my downstream users still prohibited from: - distributing the software under the name JUMBO (or a derivative) (Jumbo, Jr, Dumbo, Elephant and Timothy all seem derivative) - calling a modified version a compliant CML system - asserting that a modified version can read and write CML? That is, are these clauses additions to the Artistic License 2.0 which must be preserved even after 4(c)(ii) relicensing to the GPL? My suspicion is that derivatives must still be prohibited from those activities. Is the resulting software (with these extra limitations) free software enough for Debian? Best regards, Andrew da...@dalkescientific.com -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
In message e77d3607-af7f-43f4-a7e0-11e7c541c...@dalkescientific.com, Andrew Dalke da...@dalkescientific.com writes I pointed out the quote from a copyright lawyer with a special interest in free software who said that the GPL was ambiguous about sublicensing and if a chain of licenses was required or not. I see the GPL explicitly agrees with me, not Larry Rosen :-) !!! This is the GPL v3 - read the last section of 2. Basic Permissions : Which means you didn't look at the top of the first page of the link I sent you, where you would see the book was written in 2004 and therefore pre-GPLv3. I didn't need to. I knew it was pre-GPL3. And you've confirmed this is the book which confirmed my dim view of Larry's competence ... It also means you didn't recall my original text where I wrote: As you can tell, a professional lawyer in this field is not clear about if the GPLv2 allows sublicensing, so I hope it's understandable how someone could view a change from GPLv2 to GPLv3 without keeping the chain of titles (which is the common practice) could be considered a relicense. My views on him are a direct consequence of discussing things - WITH HIM - on the lsb mailing lists. He referred me to this book and I read it in its entirety back then - quite some years ago (probably circa 2005, maybe even 2004). Basically I formed the impression he was a very capable lawyer, determined to twist everything to suit his interpretation of things regardless of fact or clear intent. And as I say, reading this book only confirmed that impression that I had already formed. I believe I have careful to only used references from that book with respect to GPLv2, and not use it as a way to interpret reading the book has helped me understand some of the improvements made in GPLv3. The above was one of the few cases where I was not. The proper behavior should be to point out that I likely was imprecise and should have written GPLv2 instead of simply GPL. 10. Automatic Licensing of Downstream Recipients. Each time you convey a covered work, the recipient automatically receives a license from the original licensors, to run, modify and propagate that work, subject to this License. You are not responsible for enforcing compliance by third parties with this License. This is exactly the section (maybe worded, certainly numbered, differently) that I have repeatedly been referring to from the GPL v2. This is the specific improvement to text which Rosen says is ambiguous in GPLv2. As you have not bothered to read the text and yet still comment on what you believe he has written, I shall copy it here: http://rosenlaw.com/Rosen%5FCh06.pdf This GPL section 4, with its negative wording, is also the only place that references the right to sublicense. One might assume from the way GPL section 4 is worded that the right to sublicense was intended in sections 1 (right to copy), 2 (right to modify) and 3 (right to distribute) as well. However, section 6 implies that there are no sublicenses but instead a direct license from each up-stream contributor: ... As to sublicensing, then, the GPL is ambiguous. I refer you to the discussion in Chapter 5 of sublicensing in the MIT license. Sublicensing rights can be very important to open source distributors for dealing properly with the chain of title to contributions. In practice, most software projects ignore the issue completely and assume that, for GPL software, only the most recent license in the chain of title matters. They assume that GPL licensed software is sublicenseable, but the GPL isn’t clear about that. I think that ALL Larry's negative comments about the GPL stem from that assumption that you can sublicence GPL code. As I understand the GPL, (a) sublicencing is unnecessary, and (b) this very section 4 clearly says you can only sublicence GPL code if the GPL explicitly says you can (which it then *doesn't*). That section in GPLv3 part 2 makes it clear that v2 did *not* intend to permit sublicencing. Obviously, rms seems to respect Larry's view a bit more than I do, in that he explicitly addressed it in v3, but equally it's clear that v2 did not (in his view) permit sublicensing. Sorry, I know I'm being nit-picky about things, but lawyers do nit-pick. If you don't, it can cost you EVERYTHING. Then nit-pick over things that actually exist. Lawyers at least get paid to nit-pick over whatever they get paid for. They also get paid to work on multiple iterations of their text, where obviously what I am writing now is a first draft. As I understand it, Larry is *willfully* *misunderstanding* the GPL. imho that's a fact :-) See above. It's the *grant* which allows YOU to choose which version of the GPL applies to YOU. As I said above, I know I'm being nit-picky. But if you don't understand what you're doing, then you're going to get burnt at some point. Point made. It could have been done without as many exclamation
Re: Artistic and LGPL compatibility in jar files
In message ab9c74f8aa8f5d509d7617b49633a35b.chere...@mccme.ru, Alexander Cherepanov chere...@mccme.ru writes Hi Anthony! On Mon, 14 Dec 2009 21:44:35 +, Anthony W. Youngman deb...@thewolery.demon.co.uk wrote: Your recipients also get *my* grant, so any one of them can say actually, I like v *2* so I'll take that as my licence. Why do you think that my recipients will get your entire grant? GPLv3 only says that they will get your grant for _this_ License, i.e. GPLv3. WHERE does it say that? In section 10 (GPLv3): 10. Automatic Licensing of Downstream Recipients. Each time you convey a covered work, the recipient automatically receives a license from the original licensors, to run, modify and propagate that work, subject to this License. [...] ^^^ GPLv2 says effectively the same: 6. Each time you redistribute the Program (or any work based on the Program), the recipient automatically receives a license from the original licensor to copy, distribute or modify the Program subject to ^^ these terms and conditions. [...] ^^ But in that case, as soon as you distribute my code using GPL2 as your licence, YOU have STOPPED them distributing under version 3! That argument cuts both ways! Sure. Actually, that then totally destroys the whole point of v3 or later if you choosing v3 takes away your recipients rights to choose according to the original author's grant! They are always free to get the program directly from original author (put aside the case of a program combined from different sources for a moment:-). Then they have a choice of license. But the law (generally) given the choice between a sensible interpretation, and an alternative that is either ludicrous or obviously not what was intended, will *usually* choose the sensible one. Some variation of the scenario: suppose your grant is this software is licensed under BSD or GPLv3 and I choose GPLv3. Does this mean that my recipients still get BSD or GPLv3? Given that, imho, BSD doesn't permit relicencing either, yes your recipients do get that choice. As soon as you modify my code, they then only get GPLv3 (unless they strip your modifications out). Yet another variation: suppose you licensed your program to Alice under BSD and to Bob under GPLv3. Does recipients which get your program from Bob get BSD or GPLv3 or just GPLv3? Bob's recipients get just GPLv3. That's all he got, that's all he can pass on. To make it even worse, if somebody got one copy from Alice and one from Bob, I guess technically they'd have to keep the two copies (and associated licences) separate unless they contacted me and got my permission to combine them! I've just checked v3, and it contains the same gets your licence from the original licensor wording as v2, so they get their grant from me, and you don't have the right (or ability) to change what I grant. I hope quotes above explain what I mean. At the end of the day, YOU need a licence to distribute my code. My grant gives you a choice of v2 or v3. Whether you choose v2 or v3, your recipient then gets the same grant as you did, Sorry, I don't see where it comes from. Basically, you can choose which licence you want to apply to YOU. But you pass on my package as a whole (including my permission to choose which licence). So that's where your recipients get the same choices you got. and they can also choose v2 or v3. If your choice of v3 took away your recipients choice of v2 I would consider that a VERY retrograde step. I agree and would be happy to learn where I'm wrong. But at the end of the day, it's simple. If I say v2 or v3 then I granted EVERY recipient of my code the right to *choose*. Yes, if they receive from you directly. Or if they receive an UNALTERED copy from you! Because if you change the licence (which you're not allowed to do) it's not an unaltered copy :-) Both v2 and v3 are explicit that your recipients get their rights from ME not you, so your choice of v3 does not constrain their right to choose. Alexander Cherepanov Cheers, Wol -- Anthony W. Youngman - anth...@thewolery.demon.co.uk -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
On Dec 13, 2009, at 2:24 AM, Anthony W. Youngman wrote: In message f4ccec28-fe42-4af3-b0c0-c832a6b0d...@dalkescientific.com, Andrew Dalke da...@dalkescientific.com writes Well, the GPL does allow relicensing to newer versions of the GPL... IT DOESN'T, ACTUALLY !!! Read what the GPL says, CAREFULLY. Here is relevant commentary in Rosen's book Open Source Licensing book at http://rosenlaw.com/Rosen%5FCh06.pdf This GPL section 4, with its negative wording, is also the only place that references the right to sublicense. One might assume from the way GPL section 4 is worded that the right to sublicense was intended in sections 1 (right to copy), 2 (right to modify) and 3 (right to distribute) as well. However, section 6 implies that there are no sublicenses but instead a direct license from each up-stream contributor: ... As to sublicensing, then, the GPL is ambiguous. I refer you to the discussion in Chapter 5 of sublicensing in the MIT license. Sublicensing rights can be very important to open source distributors for dealing properly with the chain of title to contributions. In practice, most software projects ignore the issue completely and assume that, for GPL software, only the most recent license in the chain of title matters. They assume that GPL licensed software is sublicenseable, but the GPL isn’t clear about that. I will try to use the word sublicense in the future as that seems more precise. As you can tell, a professional lawyer in this field is not clear about if the GPLv2 allows sublicensing, so I hope it's understandable how someone could view a change from GPLv2 to GPLv3 without keeping the chain of titles (which is the common practice) could be considered a relicense. Best regards, Andrew da...@dalkescientific.com -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
Hi Anthony! On Sun, 13 Dec 2009 01:24:36 +, Anthony W. Youngman deb...@thewolery.demon.co.uk wrote: Well, the GPL does allow relicensing to newer versions of the GPL... IT DOESN'T, ACTUALLY !!! Read what the GPL says, CAREFULLY. Let's say I write a load of code, and release it with a notice saying this code is licenced as 'GPL version 2 or later' . Typical dual-license scenario, good. Could you please elaborate some more how both licenses propagates in this case? This seems to be a very common notion but it's not clear to me. What this give YOU is the right to redistribute the code according to the terms of the GPL v3. BUT - READ THE GPL - the people to whom you give the code get their licence from ME, NOT YOU. Right, this is section 6 of GPLv2 of section 10 of GPLv3. Let's quote the latter: 10. Automatic Licensing of Downstream Recipients. Each time you convey a covered work, the recipient automatically receives a license from the original licensors, to run, modify and propagate that work, subject to this License. [...] And I granted the licence as v2 or later. The text of the GPL says subject to this License, i.e. GPLv3, not subject to all Licenses. So I don't see how your conclusion follows. Alexander Cherepanov -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
On Fri Dec 11 22:42, Andrew Dalke wrote: There seems to be a licensing problem with some of the chemistry software packages, at least one of which is included in Debian. I'm working with a few of the package developers to see if there really is a problem. We need some better advice than I can find. Short version: - Can an LGPL 2.1 JAR library include an Artistic License library and still be distributed under the LGPL 2.1? - What about an LGPL 2.1 JAR library including a package under the Artistic License 2.0 license? Or would the entire package need to be moved to the GPL as a relicensing which is compatible with both underlying licenses? I believe that neither of these licences specify the licence of the code they are linked with, so this will be alright. The resulting licence of the package will be _both_, applying to different parts, AIUI. - One of the XML schema files is released (most likely) under the Creative Commons - No Derivation license. This is used by the Artistic License package in order to do schema validation. Can the LGPL 2.1 library include functionality which requires this unalterable schema definition? Well, the biggest problem is that the CC-ND licence is not DFSG free, so inclusion of this at all would require putting the package in non-free. I'm going to simplify the story a bit and give a minimal example. CDK is the Chemistry Development Kit, available in Debian as http://packages.qa.debian.org/c/cdk.html It is distributed under the LGPL 2.1. It is one of a number of LGPL'ed chemistry tools which use the JUMBO and CML toolkits available from https://sourceforge.net/projects/cml/ This software library and program are available under the Artistic License, which you can read in the Sourceforge details and in the distributed pom.xml file. The distribution also includes the file LICENSE.txt, saying: This doesn't sound like the two things being included in the same package? I'd expect it to depend on them at build and runtime. Matt I have contacted one of the authors and gotten different responses. Originally he said the package was Artistic License 2.0 and then said the pom.xml file says it is 'Artistic License', and in discussions where I pointed out the existence of the LICENSE.txt file he said that he wants those additional restrictions in place, so I am going on the principle that the LICENSE.txt file is correct, and that the license is Artistic License and not Artistic License 2.0. Notably, the license text is not included in the distribution. CDK is one of the downstream chemistry toolkits which make Java jar distributions which use and repackage the jar file released by the JUMBO/CML project. One of them also includes a patched JAR file. These are not simple aggregates in a single jar file; the downstream packages make use of functionality from the JUMBO/CML package. ugh, that's generally really bad form... My understanding is that mixing the Artistic License and LGPL 2.1 is not possible. I base this primarily on the FSF statement that they consider the Artistic License to be incompatible with the GPL. I have not found a statement about compatibility between the Artistic License the LGPL. GPL is definitely != LGPL in this area, but I'd appreciate other comments on the issue I tried to read and understand the Artistic License but I got confused. The simplest conflict seems to be that the Artistic License says You may not charge a fee for this Package itself. where Package refers to the collection of files distributed by the Copyright Holder, and derivatives of that collection of files created through textual modification. This is in conflict with the LGPL 2.1 clause You may charge a fee for the physical act of transferring a copy. This may well be a problem for combining things into a single package, but I would not have thought it was an issue for things in different packages. I have talked with one of the authors of JUMBO/CML and they may be willing to relicense under the Artistic License 2.0. In doing the research for that I read that the FSF considers the 2.0 license compatible with the GPL because of the relicensing clause 4(c)(ii), which allows the GPL. In this case the whole work would be distributed under the full GPL, not the LGPL This is relevant because it would prevent CDK and other downstream packages from including libraries which are compatible with the LGPL but not compatible with the GPL. Or they would remove or reimplement the JUMBO/CML component. correct. If it is possible to relicense and be compatible with the LGPL 2.1, the main CDK developer wants to know how to relicense the software. Does he need to make a specific source release of JUMBO/CML under the LGPL 2.1 then turn around and use it inside of his code? Or can CDK include the JUMBO/CML code and just state somewhere inside the CDK documentation Originally
Re: Artistic and LGPL compatibility in jar files
In message 20091212142624.gl10...@matthew.ath.cx, Matthew Johnson mj...@debian.org writes My understanding is that mixing the Artistic License and LGPL 2.1 is not possible. I base this primarily on the FSF statement that they consider the Artistic License to be incompatible with the GPL. I have not found a statement about compatibility between the Artistic License the LGPL. GPL is definitely != LGPL in this area, but I'd appreciate other comments on the issue I tried to read and understand the Artistic License but I got confused. The simplest conflict seems to be that the Artistic License says You may not charge a fee for this Package itself. where Package refers to the collection of files distributed by the Copyright Holder, and derivatives of that collection of files created through textual modification. This is in conflict with the LGPL 2.1 clause You may charge a fee for the physical act of transferring a copy. This may well be a problem for combining things into a single package, but I would not have thought it was an issue for things in different packages. I have talked with one of the authors of JUMBO/CML and they may be willing to relicense under the Artistic License 2.0. In doing the research for that I read that the FSF considers the 2.0 license compatible with the GPL because of the relicensing clause 4(c)(ii), which allows the GPL. In this case the whole work would be distributed under the full GPL, not the LGPL I may (well) be wrong, but I've always understood the INTENT of the artistic licence to be BSD plus a trademark licence. But, if the JUMBO/CML people are happy, why not ask them to add an extra permission, or dual-licence. If they are the copyright holders (and therefore able to change the licence from Artistic to Artistic 2), they could always change the licence to Artistic 1 or LGPL2.1 if they wanted. If, however, they do change to Artistic 2 (probably a good idea, the original Artistic licence has been slated as being very poor legalese) ... If it is possible to relicense and be compatible with the LGPL 2.1, the main CDK developer wants to know how to relicense the software. Does he need to make a specific source release of JUMBO/CML under the LGPL 2.1 then turn around and use it inside of his code? Or can CDK include the JUMBO/CML code and just state somewhere inside the CDK documentation Originally under the Artistic License 2.0 and relicensed under clause 4(c)(ii) to the LGPL 2.1? I'm always wary of explicitly relicencing. The GPL doesn't permit it, and by doing so you are taking away user rights. If you're distributing JUMBO/CML code *unchanged*, what I'd do is to keep it separate inside the package (in its own directory or something), and in the CDK documentation state that you are distributing JUMBO/CML under the LGPL as permitted by 4(c)(ii) of the Artistic licence. That way, you're leaving (the licence of) JUMBO/CML unchanged, but distributing CDK (including JUMBO/CML) under the LGPL. And the recipient of CDK can strip JUMBO/CML out of it and use it under the Artistic licence as the author intended. Cheers, Wol -- Anthony W. Youngman - anth...@thewolery.demon.co.uk -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
[ on combining LGPL and Artistic Licenses in a single JAR file as part of a Java library distribution.] On Dec 12, 2009, at 3:26 PM, Matthew Johnson wrote: I believe that neither of these licences specify the licence of the code they are linked with, so this will be alright. The resulting licence of the package will be _both_, applying to different parts, AIUI. Both licenses mention linking, although in the Artistic License it's an irrelevant reference. In 2.0 there's a direct statement about linking - section (8). I wasn't sure if the jar file counts as a derived work in its totality, or if it counts as an aggregate. If I read the Artistic License correctly then it does seem to limit itself to textual changes, which I didn't catch the first few times I read the license. There is a clause about object code. If a distribution includes object code then it must do at least one of four possibilities. Option (c) is c) accompany any non-standard executables with their corresponding Standard Version executables, giving the non-standard executables non-standard names, and clearly documenting the differences in manual pages (or equivalent), together with instructions on where to get the Standard Version. and is the easiest to do, since the JAR distributes no executable. So with the Artistic License it's clear (now) that there's no imposition on the license terms for the other packages in the jar. Interestingly, the license does mention linking in a strange way, given that this is a Java package (but obvious given the Perl source): 7. C or perl subroutines supplied by you and linked into this Package shall not be considered part of this Package. It's a moot point since CDK links to the Package and not vice versa. The above was for the Artistic License. For Artistic License 2.0 there is a section on linking. Aggregating or Linking the Package (7) You may aggregate the Package (either the Standard Version or Modified Version) with other packages and Distribute the resulting aggregation provided that you do not charge a licensing fee for the Package. Distributor Fees are permitted, and licensing fees for other components in the aggregation are permitted. The terms of this license apply to the use and Distribution of the Standard or Modified Versions as included in the aggregation. (8) You are permitted to link Modified and Standard Versions with other works, to embed the Package in a larger work of your own, or to build stand-alone binary or bytecode versions of applications that include the Package, and Distribute the result without restriction, provided the result does not expose a direct interface to the Package. If the JAR is an aggregate then (7) applies, but as CDK uses and exposes the API, I strongly suspect (8) is the one which applies. That is why the CDK will have to take the relicense clause, which is why I want to know if the Artistic License 2.0 relicense requirements are compatible with LGPL 2.1. Can the LGPL 2.1 library include functionality which requires this unalterable schema definition? Well, the biggest problem is that the CC-ND licence is not DFSG free, so inclusion of this at all would require putting the package in non-free. I've forwarded this to the CDK maintainer. He's going to look into pulling out that one file so it's not needed in the core. Thanks for the clarification! This software library and program are available under the Artistic License, which you can read in the Sourceforge details and in the distributed pom.xml file. The distribution also includes the file LICENSE.txt, saying: This doesn't sound like the two things being included in the same package? I'd expect it to depend on them at build and runtime. Matt I'm sorry, I didn't understand your comment. The source and jar distributions are named JUMBO and CML is a library distributed as part of JUMBO. I tried to read and understand the Artistic License but I got confused. The simplest conflict seems to be that the Artistic License says You may not charge a fee for this Package itself. where Package refers to the collection of files distributed by the Copyright Holder, and derivatives of that collection of files created through textual modification. This is in conflict with the LGPL 2.1 clause You may charge a fee for the physical act of transferring a copy. This may well be a problem for combining things into a single package, but I would not have thought it was an issue for things in different packages. Currently the CDK distributes everything in a single jar. One solution might be to distribute two jars. This does have some downstream difficulties. The JChemPaint Java applet uses CDK and it distributes a single applet jar. It would have to rearchitect a bit so there are two jar files, one with the JUMBO/CML code. This is possible, and I think would be an acceptable interpretation. I read that the FSF
Re: Artistic and LGPL compatibility in jar files
On Dec 12, 2009, at 11:12 PM, Anthony W. Youngman wrote: I may (well) be wrong, but I've always understood the INTENT of the artistic licence to be BSD plus a trademark licence. It has some clauses which are decidedly non-BSD-ish. See for example section (8) of the Artistic License 2.0. It is more meant to keep control over what it means to, say, be perl, which is not simply a trademark issue. But, if the JUMBO/CML people are happy, why not ask them to add an extra permission, or dual-licence. If they are the copyright holders (and therefore able to change the licence from Artistic to Artistic 2), they could always change the licence to Artistic 1 or LGPL2.1 if they wanted. *nods head* That's a couple of the possibilities I mentioned to them, and I assume they are considering the options. They do want to use the license to include trademark restrictions. I've explained how that doesn't work in practice. I don't think that made me new friends. I'm always wary of explicitly relicencing. The GPL doesn't permit it, and by doing so you are taking away user rights. Well, the GPL does allow relicensing to newer versions of the GPL... If you're distributing JUMBO/CML code *unchanged*, what I'd do is to keep it separate inside the package (in its own directory or something), and in the CDK documentation state that you are distributing JUMBO/CML under the LGPL as permitted by 4(c)(ii) of the Artistic licence. It a patched version. It also turns out that CDK wasn't distributing the patched source code as per the Artistic License requirements, but that's being addressed as part of this license cleanup. (There were other problems we've found, like some of the LGPL packages not listing in the documentation all of the third-party LGPL'ed components they were using and including. This has triggered a long-needed review of what's going into the distributions.) Thank you for your time, and best regards! Andrew da...@dalkescientific.com -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
In message f4ccec28-fe42-4af3-b0c0-c832a6b0d...@dalkescientific.com, Andrew Dalke da...@dalkescientific.com writes I'm always wary of explicitly relicencing. The GPL doesn't permit it, and by doing so you are taking away user rights. Well, the GPL does allow relicensing to newer versions of the GPL... IT DOESN'T, ACTUALLY !!! Read what the GPL says, CAREFULLY. Let's say I write a load of code, and release it with a notice saying this code is licenced as 'GPL version 2 or later' . What this give YOU is the right to redistribute the code according to the terms of the GPL v3. BUT - READ THE GPL - the people to whom you give the code get their licence from ME, NOT YOU. And I granted the licence as v2 or later. So, AT NO POINT WHATSOEVER, does my code become v3, whatever you say or do. If you modify my code and licence your stuff as v3, the resulting work then becomes v3-only because the licence of the work as a whole is the subset of the individual licences - here v3 - but my code still remains v2+. Cheers, Wol -- Anthony W. Youngman - anth...@thewolery.demon.co.uk -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Re: Artistic and LGPL compatibility in jar files
On Dec 13, 2009, at 2:24 AM, Anthony W. Youngman wrote: In message f4ccec28-fe42-4af3-b0c0-c832a6b0d...@dalkescientific.com, Andrew Dalke da...@dalkescientific.com writes I'm always wary of explicitly relicencing. The GPL doesn't permit it, and by doing so you are taking away user rights. Well, the GPL does allow relicensing to newer versions of the GPL... IT DOESN'T, ACTUALLY !!! Read what the GPL says, CAREFULLY. I didn't realize this was such a hot point to need the use of capital letters. Pretend I said LGPL instead of GPL. In that case I can talk about relicensing, yes, since the LGPL explicitly allows relicensing to the GPL: http://www.gnu.org/licenses/gpl-faq.html#compat-matrix-footnote-7 7: LGPLv2.1 gives you permission to relicense the code under any version of the GPL since GPLv2. If you can switch the LGPLed code in this case to using an appropriate version of the GPL instead (as noted in the table), you can make this combination. LGPL is, after all, the Lesser GPL. In v3 the LGPL is specifically designed to give additional permissions than those of the GPL. You talked about how relicensing takes away user rights but in that case relicensing from LGPL to GPL is more taking away user permissions, yes? Still, the LGPL is designed to be relicensed to the GPL. What about something which doesn't have a built-in relicensing? Pretend I had said GFDL instead of GPL, in which case this quote from Stallman is highly relevant: http://www.fsf.org/blogs/licensing/2008-12-fdl-open-letter The relicensing option in GFDL 1.3 is fully consistent with the spirit and purpose of the GFDL. Stallman used the term 'relicense' several times in that open letter, and as a highly-visible response to the accusations of misdeeds during the GFDL/CC-BY-SA change, where 1.3 has an explicit section titled RELICENSING while 1.2 did not. He cannot have used it by mistake or as a poor word choice. Does that relicensing take away any user rights which are part of the spirit and purpose of the GFDL? (It does obviously take away the right to revert the license to 1.2, but is that an important right?) Let's say I write a load of code, and release it with a notice saying this code is licenced as 'GPL version 2 or later' . The FSF suggests that you should write it thusly: 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 2 of the License, or (at your option) any later version. Compare to the suggested text for the GFDL Copyright (c) YEAR YOUR NAME. Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation; with no Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts. A copy of the license is included in the section entitled GNU Free Documentation License. If changing from GFDL 1.2 to GFDL 1.3 to CC-BY-SA is called a relicense by the FSF and by Stallman, and allowed under the GFDL 1.2, then I hope you can see why I did not spot the subtle difference in these texts which means I should not call switching from GPLv2 to GPLv3 a relicense. I looked again. I still don't see the difference. Could you point it out? What this give YOU is the right to redistribute the code according to the terms of the GPL v3. BUT - READ THE GPL - the people to whom you give the code get their licence from ME, NOT YOU. And I granted the licence as v2 or later. So, AT NO POINT WHATSOEVER, does my code become v3, whatever you say or do. If you modify my code and licence your stuff as v3, the resulting work then becomes v3-only because the licence of the work as a whole is the subset of the individual licences - here v3 - but my code still remains v2+. That interpretation is not consistent with the license upgrade path from GFDL 1.2-GFDL 1.3-CC-BY-SA. The 1.3 license clearly says a Massive Multiauthor Collaboration Site can relicense simply by republishing the site during the given time period. It does not require any changes to the content of the site in order to allow the relicense. I understand that some people feel Stallman and the FSF were wrong in doing this, but I again think I can excused for considering the change from v2 to v3 as a relicense. Even if you say that the GPL is different in this matter than the GFDL, can I not simply change the COPYING file because that alone is a modification to your code? Perhaps also changing the README to say Now under GPLv3 - w00t!, in order to add creative input? That would be consistent with statements like: http://calypso.tux.org/pipermail/xemacs-beta/2009-February/016039.html To relicense to GPLv3 in principle is just a global substitution, and exchanging the license text in COPYING. We should also make sure we have the
