login
Search
Hmm... Really,it is damn too much time to wait! It's really better idea to replace the code of this person as said before instead of waiting such 90+ years! -
exactly, however as others are pointing out, there are a lot of active developers who do not agree with some of the key points of the GPLv3 (including Linus), so until you convince them that the GPLv3 is better it really doesn't matter how hard it is to deal with the people who you can't contact. David Lang -
Last heard, Linus was quite impressed with the toned down version of the final draft of GPLv3. I think Linus, and other major developers should make their stand on this issue clear so that the kernel community can discuss the future steps. -
"future steps"? Hah. My code is going to stay GPLv2 as the v3 license is horrible for kernel code for all of the reasons I have said in the past, plus a few more (what, I can make an "industrial" product but not a commercial one? That's horrible...) thanks, greg k-h -
^^^^^^^^^^^^^ What exactly in GPLv3 forbids you from making a commercial product? -
Nothing "forbids" me, it's just the artifical distinstion of the two is, in my opinion, stupid and foolish. You are trying to define use-cases to justify their notion that you must give up the hardware keys for one type of device, yet not for another. Even the people that feel that v2 says you need to give up the keys think this is dumb. But we've been through all of that before (see previous long thread about v3 and why the kernel developers hate it, it all still applys to the final draft.) greg k-h -
You mean all the misunderstandings? ;-) -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
My impression as well is that there are many misunderstandings, even
concerning the status of Linux itself. Linus is much better at kernel
hacking than at license issues, and that's true for most other kernel
hackers, too - that's why we have Eben Moglen to hack the license.
I want to add my two cents on what I think the legal status of the
individual contributions to Linux are. The thing in question is not the
GPLv2 itself (which is pretty clear that code without explicit statements
is under "any", and if you make an explicit statement, it should read "GPL
version two, or (at your option) any later version"), it's this text on the
top of /usr/src/linux/COPYING:
"Also note that the only valid version of the GPL as far as the kernel
is concerned is _this_ particular version of the license (ie v2, not
v2.2 or v3.x or whatever), unless explicitly otherwise stated.
Linus Torvalds"
This text was added in or around 2.4.0-test9, but without asking for
permission (neither from the FSF, which has the copyright of the GPL, nor
from the other authors of the Linux kernel), and with some controversion
afterwards.
This particular comment to how the GPL is applied to the Linux kernel
therefore doesn't change the GPL as such (it can't without breaking
copyright), neither does it change the licensing conditions the original
authors put on their contribution (it can't without breaking copyright,
either), but may only provide interpretations downstream (for the user).
Linus is also entitled to make clairifications there, which the first
paragraph obviously does (i.e. the text Linus added is not a change of the
license, but a comment on it).
Again: What Linus is entitled to do is to *select* the license under which
he redistributes the code downstream. What he can't do is to *change* the
intention of the original author. So if you can choose what this somewhat
ambiguous message means, and restrict yourself to reasoning that doesn't go ...That's not exactly true. A work without explicit statements is not How about derived works? Am I free to get BSD source, incorporate it in GPL project, and release the whole under GPL? I don't think the law works like that. By default you have no rights to someone's work (file or project). The only licence I can find with Linux is GPL v2, isn't it? And even that wasn't stated explicite until that 2.4.0something (though there is a consensus that the COPYING file was indeed a licence for the whole kernel). Then you may have additional rights, such as those given in various What exactly is the "GPL regime" and how is it defined by copyright law and/or the GPL licence itself (or will of copyright holders etc.)? -- Krzysztof Halasa -
If you choose the GPL as license, the text of the GPL are the conditions.=20 Derivated work is a product of several authors, therefore each author may=20 put different conditions on his part of the work - as long as they are=20 compatible, it's ok. A derivated work originally under BSD, now with a=20 patch under GPL can only be distributed under GPL, but not under BSD=20 (because GPL requires redistribution under GPL, whereas BSD doesn't care).= =20 If you take out the patch, and revert the work to the BSD one, you are free= =20 to redistribute it under BSD. There's no point of discussing that the Linux kernel *as a whole* (as=20 compilation) currently is under GPLv2 only, since it sais so, and a few=20 files also explicitely say so. The whole combination is GPLv2 only, but=20 Copyright law defines "work", and in terms of computer programs, source=20 code "work" goes into files. Or patches, but patches end up distributed=20 over several files. The nice thing about this is that you can make automatic checks about the=20 license you actually have to fulfill. E.g. if you compile a hypothetical=20 Linux 2.8.15.3 without ZFS and dtrace in 2009, you may end up with=20 compiling only GPLv2-compatible code, and therefore can tivoize your system= =20 (unless you sell it to Germany, where the GPLv2 outlaws tivoizing by=20 Yes, and the GPLv2 sais "if the FSF releases a new version of the GPL, you= =20 may update" (section 9): " 9. The Free Software Foundation may publish revised and/or new versions of the General Public License from time to time. Such new versions will be similar in spirit to the present version, but may differ in detail to address new problems or concerns. Each version is given a distinguishing version number. If the Program specifies a version number of this License which applies to it and "any later version", you have the option of following the terms and conditions either of that version or of any later version published by the Free Software ...
You claim that any source files without a notices are 'any version of the GPL'. But I read the license and you are totally wrong about that. The GPL applies to "the Program" which in this case is the Linux kernel as a whole and it in fact does indicate a specific version. All code submitted and included in this program has has been submitted with the understanding that the work as a whole is specifically licensed as GPLv2. Some authors have granted additional rights, such as dual BSD/GPL or GPLv2 and later and explicitly added such a notice. All other code is simply copyrighted, and the only available license is the GPLv2. Take for example fs/inode.c. Notice how it doesn't have GPL boilerplate, but it is clearly indicating that it is copyrighted. So taking that file by itself out of the context of the kernel and then distributing it would clearly be a copyright violation. The only one reason you can distribute that code is because of the GPLv2 that covers The kernel is explicitly licensed as GPLv2, any contributions (source files/parts of the work) that wish to grant additional rights have to specify so explicitly, and not the other way around however much you'd Reread section 9 and consider that "the Program" is the Linux kernel, which does explicitly state a version and does not include the "and any later" option. Any source that does not explicitly specify additional rights is GPLv2. Jan -
Since the Linux kernel as a whole does not have a single author, it is impossible to license it as a whole. Nobody has the authority to do that. (The GPL is not a copyright assignment type license.) Fortunately, the GPL clears this up: "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. You may not impose any further restrictions on the recipients' exercise of the rights granted herein. You are not responsible for enforcing compliance by third parties to this License." Linus cannot impose any further restrictions on the recipients' exercise of the rights granted. When you download a copy of the Linux kernel, you do not receive one license because nobody could grant you one license. You receive a logically separate license from each original licensor. You receive from Linus only a license to his contributions. Note that you cannot take a GPLv2+ work and redistribute it as GPLv3 only. You can license your contributions as GPLv3 only of course. However, each recipient still receives a GPLv2+ license to the parts that were originally licensed that way. The people you distribute the work from receive licenses from the original licensors to those parts, and you have no right to modify that license. (See GPL section 6, quoted above.) DS -
Actually, Linus Torvalds, as maintainer, probably has a compilation copyright. See "compilations and abridgements" in If you combine dual licensed code (such as MPL + GPL) with code under only one of those licenses (MPL only), the resulting derived work cannot be distributed under the dual license, only under one license. The giant derived work knows as Linux has only been distributable under exactly one license (GPLv2, the complete text of which is included in the source tarball and it's harder to be more explicit than that about which license you mean) since version 0.12. By the way, this entire "oh no, we can use it GPLv3 no matter what you say" line of argument is rude. Linus and most of his lieutenants have explicitly said "our contributions are GPLv2 only". Linus said this explicitly seven years ago: http://www.uwsg.iu.edu/hypermail/linux/kernel/0009.1/0096.html He confirmed and elaborated his position when people first started pestering about v3: http://lwn.net/Articles/169825/ In James Bottomley's position paper last year, a number of prominent kernel developers stated their objection and that their contributions were GPLv2 only: http://lkml.org/lkml/2006/9/22/217 James E.J. Bottomley Mauro Carvalho Chehab Thomas Gleixner Christoph Hellwig Dave Jones Greg Kroah-Hartman Tony Luck Andrew Morton Trond Myklebust David Woodhouse Let me translate this into simpler terms: <lolcats> GPLv3: Does not want! </lolcats> The _reason_ it's rude to go on about it is that several people have chosen to see this entire debate as an interesting intellectual exercise, "how much code could a GPLv3 licensed project lift from the Linux kernel". Yet if you substitute "BSD Licensed" in there, it's easy to recognize how obnoxious the pestering is, despite much of the code in Linux having come from BSD sources. You don't take Linux kernel code and stick it into a BSD project, ...
It doesn't matter. He can license you his compilation, but that doesn't license you the underlying elements. I can make a compilation CD of great works of Rock N' Roll. I can hold a copilation copyright in the compilation. I can license that compilation copryight. That doesn't mean you can make, copy, or sell a CD with my compilation on it, because you are also copying and distributing the That is a common simplification. The GPL is clear that it applies automatically with distribution. If you distribute a GPL'd work (or elements that are GPL'd inside a larger work), those elements are relicensed under the GPL automatically. You *cannot* prevent this from happening. If I take the Linux kernel, modify it, and then give you a copy, you get a license under GPLv2 from Linus to all of those elements that he placed under the GPL. I cannot stop or modify this. It applies even if I get separate permission from Linus to distribute his contributions under some other No, not true. I don't have the court citations handy, but it is well-settled law that a right to distribute a derivative work is useless without also having the right to distribute the original work from which the derivative Huh? I have never argued that any contribution made by Linus could be or had been licensed under GPLv3. Linus has clearly indicated, along with the works that he distributes, that the code is only offered under GPLv2. However, Linus cannot remove rights that other people grant to their code, even if he I don't know who you are talking to or what you are talking about. I haven't seen anybody doing what you claim in this thread or anywhere else and I No, not true. Please read and understand GPLv2 section 6. If a work is available under GPLv2+, and you receive that work (even if it's as part of Do you seriously not understand that a compilation right in "Great Works of Fiction 2002-2006" doesn't give you the right to actually distribute the works in it? You also receive Linus' compilation ...
I'm asking what is the _point_ of the discussion? Linux, the project, is available under GPLv2 only. It is not available under GPLv3, and its maintainers (both Linus, his lieutenants, and numerous other contributors) have expressed an explicit desire NOT to license it as such. So what are the people talking about GPLv3 trying to accomplish? Are they: A) Trying to unanimously change the mind of Linus, his lieutentants, and all the other contributors who have spoken up in favor of GPLv2 only, so that future versions of Linux grew a new license? (Doesn't matter if this new license is GPlv3, MPL, or BSD. It's a new license Linux is not currently distributed under. Bits of Linux are separately distributed under other licenses such as BSD, but Linux is not and won't be any time soon.) B) Proposing the creation of a fork of Linux which identifies and replaces all the code that can't be licensed under GPLv3? C) Moving to another codebase (Solaris? The Hurd) and trying to identify Linux code that can be ported to that other OS under another license? D) Blowing smoke to no actual purpose? Right now, it's looking like D. Is there an E that I'm not seeing? Rob -
You have a good point. It can be argued that contributions before 2.4.0-test8 were in fact GPLv2+, but anything after that point has clearly been contributed as GPLv2 only. So now we have a bunch of pre-2.4.0-test8 code that may possibly be v2+ and files that explicitly state v2+ in their boiler plate. However many of these files may have had additional contributions from other authors which (unless otherwise specified) were GPLv2-only. And because v2 and v3 are incompatible, all those files with v2-only contributions will become v2-only when version 3 is released. Of course it may be that all those copyright owners do not mind re-releasing their copyrighted code as v2+, but they will have to be contacted. Several maintainers did pay attention to such details. I once submitted a patch that among others touched reiserfs, and I promptly got a friendly email from Hans asking me to sign off any rights he needed to re-release the related code under a different license, so he made sure the combined work wouldn't end up GPLv2 only. Jan -
The licence can't redefine the copyright laws. It doesn't make it pure That would be the case if "the Program" (the whole or individual file(s)) contained something like: "you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation". This is very different (though unrelated) - patches are new work and I hope you're free to use your MS Windows under old conditions if you FSF has exactly nothing to say here (except that they've created a useful licence). The author can choose whatever conditions he/she It? What "it"? I don't get it. If you say the licence is v2 only, then how can it have First, the local and international laws apply. It's not like selling your There is no assumption of "GPL", you can only assume GPL v2 as the kernel is v2. And it's not left for assumptions anymore, see "signed-off-by" and licence tags (though the tags often specify "GPL" when the actual Sure, you can rewrite all non "GPLv2 or later" code and have v3 Linux. The problem is you think only "few" files are v2. -- Krzysztof Halasa -
By section 9. The license is v2, and basically allows to update the=20 license - and it makes this a choice of the user (who also has rights to=20 Contract law means that first and foremost the contract itself defines the= =20 rules, and only if it is not or contradicts the law, the law jumps in. The= =20 GPL is not really a contract, it's a license, but the law is not much=20 different here, especially once you accept the GPL. If you put your code=20 under GPL, the text in the GPL is the deal. The law is only the framework=20 under which the deal works. If you accept the M$ EULA, international law still applies, yet you are=20 Because only few files say so, and they must say what they mean, because GP= L=20 is rather clear that if you put a file which doesn't say which version=20 applies under GPL, it's "any GPL". Why is it so difficult to grok section 9= =20 of the current GPLv2, which people claim is well understood? A number of kernel hacker deliberately want their work under GPLv2 only=20 (like Al Viro), and they are fully entitled to do that - but they must=20 announce it in a propper place (not lkml or lwn.org), and a comment in=20 COPYING signed by Linus Torvalds doesn't seem to be propper to me,=20 especially when the GPLv2 gives a procedure how to do it (look for the=20 appendix: "How to Apply These Terms to Your New Programs"). There are good reasons to follow the advice there, and those who did follow= =20 the advice in the Linux kernel in the vast majority said "GPLv2 or later".= =20 Verbatim copy without understanding? Or is it rather that the other people= =20 who didn't follow the advice didn't read the GPL, and therefore understand= =20 it even less ;-)? =2D-=20 Bernd Paysan "If you want it done right, you have to do it yourself" http://www.jwdt.com/~paysan/
A court deposition if somebody tries to do relicensing. At that point I believe that I made myself sufficiently clear, so I really doubt that "all files without explicit license get the license defendant would like and not the one located in the tree" would fly. But you are welcome to test that, of course - will make for nice punitive damages. Just make sure to test it yourself - giving somebody else an advice that will land them in trouble is not nice... -
If I test it, it would be in Germany, and I really doubt that relicensing=20 one copyleft to another can ever cause puntative damages here. You are only= =20 entitled to collect damages here when you have actual losses (that's why=20 Harald Welte never gets a dime except for his defense expenses and=20 voluntary donations to the FSF), but you can demand compliance. That would= =20 basically mean that the hypothetical linux-something.subversion-bp with=20 GPLv3 parts in it can't be shipped further, because I can't fulfill all my= =20 obligations. It's probably completely hypothetical, but if I really liked to be nasty, I= =20 could release the blackfin sound driver I've written for our digital=20 amplifyer under GPLv3 or later. The code I've modified is explicitely under= =20 As a non-lawyer, I can't give anybody legal advice in Germany, and I'd like= =20 to extend that to the rest of the world. This is my opinion, my=20 interpretation of the GPLv2 and what's my logical reasoning what these=20 three lines on top of /usr/src/linux/COPYING really mean. And there are=20 only two possibilities: * Either it means what it says, then it's quite likely a copyright=20 infingement done by Linus to all those authors of linux-2.4.0-test8 and=20 before, and you all may need to stop distributing Linux*, since you can't=20 meet your obligations (and restart from linux-2.4.0-test8, which is the=20 last legal version), or * it does not exactly mean what it says, then you still can distribute=20 Linux, but you can't really stop anyone who's updating it to GPLv3 - except= =20 for those few files that have explicit version numbers assigned. BTW: If I grep through Linux, I find two files where you have noted your=20 copyright and the release conditions (GPL v2), and I think last time I did= =20 the same thing, I found two GPLv2-files, as well - all other files with "Al= =20 Viro" in it apparently have multiple authors. These two files may be the=20 same ones, or maybe there are two ...
Rot. "Multiple authors" doesn't get you out of that. If you take a code available under GPLv2 or later and combine it with code under specific version of GPL, result is under than specific version of GPL. If you want to argue against that, make sure to Cc RMS on that, I would really like to hear his opinion. Multiple authors == need permission from each author with enough contributions to that file to make the contributions in question copyrightable. And in my case (and case of gregkh, and...) that would be considerably more than a couple of files. Really. -
I would expect that if you contribute to a file that explicitely says "GPL v2 or later" and you do not change that wording then you agree GPL v2 or later for that particular contribution. So for example drivers/net/plip.c could be changed to GPL v3 even though you contributed to it. -- Dmitry -
After you exclude such cases it's still more than a couple of files... -
Undoubtedly. I was just responding to neet to contact multiple authors point. -- Dmitry -
FWIW, $ find -name "*.c" | xargs grep "any later version" | wc -l 3138 $ find -name "*.c" | wc -l 9482 Watching the output of the first grep without "wc -l" shows that, although it is not 100% accurate, it is still ok just to get a rough estimate. So yes, ~6300 files are definitely more than a couple ;) -- Paulo Marques - www.grupopie.com "God is love. Love is blind. Ray Charles is blind. Ray Charles is God." -
How many of these don't mention version 2? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
It does not matter. GPL v2 and later can be reduced to v2 by recepient. Linus did just that so unless individual source file explicitely carries "and later" it is v2. -- Dmitry -
Well, if it said "licenced under GPL" it would mean any GPL. Though it's probably uncommon. -- Krzysztof Halasa -
And expanded by the next recipient to GPLv2 or later, as long as the first= =20 recipient does not make a substantial modification ("substantial" is a=20 copyright term - there is no precise definition how much must be modified,= =20 but a line or two may not count as "substantial"). This is because you=20 receive the license from the original author, not from the man in the=20 middle. What's still open is how you can change the conditions if you do make=20 substantial changes. My position is: If you modify work (i.e. work with=20 multiple licensors), you are not in a position to change the conditions,=20 since you have to pass on the rights you have (and that included "you may=20 use any GPL" or "you may use GPLv2 or later"). If you create work, you are= =20 the only licensor, so you can choose (the created work needs to be=20 sufficiently independent, which e.g. a ZFS from OpenSolaris clearly would=20 be). If you combine work, you can ship the combined work only under a GPL=20 version that matches the common subset, but you cannot change the license=20 of the parts. By adding stuff under GPLv2 only, and then combining the work= =20 to a larger work, you may achieve the effect that the larger work is then=20 GPLv2 only. You cannot achieve that people take out the GPLv2 only work,=20 and recombine it for themselves - these people then can choose other=20 license, and combine it e.g. with GPLv3 code. If you distribute work under multiple possible license, you can also choose= =20 which conditions you want to fulfill. But that's not imposing restrictions= =20 to the next recipient, so the next recipient can choose again. It's so simple: Only the author can impose restrictions, everybody else,=20 when using the GPL, has to pass on all the rights he got. If you get a=20 court verdict depends on the law system, and in an anglo-saxonian (roman)=20 system, you might get away by exploiting loopholes, but in a Code Napoleon= =20 system, you don't, because exploiting loopholes ...
No, you do receive the license from the person or entity you received the program. You have an _option_ to go to the original author and get copy of original code with original license (or maybe other license). -- Dmitry -
You receive the licence from the original author. The GPL contains no text allowing a third party to grant new licences. You may well receive the COPYING file from a different party. The licence may well place duties on the person who supplied you but the actual right to use comes from the original author. -
GPL itself does not. But the author(s) may when they specify "any later version", "dual GPL/BSD", etc. In this case (IMHO) distributor in fact relicenses the code and may reduce license to sipmply BSD or simply GPL, or "GPL v3 from now on". To "restore" license you would need to go upstream and get the code from there. -- Dmitry -
I don't see anything in the GPL that permits a redistributor to change the licence a piece of code is distributed under. If my code is GPL v2 or later you cannot take away the "or later" unless explicitly granted powers by the author to vary the licence. What you most certainly can do is modify it and decide your modifications are GPLv3 only thus creating a derived work which is GPLv3 only. However anyone receiving your modified version and reverting the modifications is back at v2 or later. Alan -
Yes, I agree. When I am saying "distributor" it is someone like RedHat or TiVO who do modify the code, not merely use it in ints original form. -- Dmitry -
But that begs the question: How do you know what has been modified so you can revert the modifications? There won't necessarily be any indication of which files have been modified. So I think Dmitry's point is valid. Don't you need to go upstream at least far enough to verify that you have unmodified code? And how does the copyright work for kernel patches? Consider a dual licensed (i.e. anything beyond GPLv2 only) file. Someone supplies patches to Linus, he applies them, the resulting file is distributed with the kernel as GPLv2. What precisely has to happen for someone to get that same file with equivalent patches applied that can be distributed with the original dual license? Somehow it seems to me that Linus would have to take the dual licensed files from his kernel repository and copy them to a separate archive and people would have to copy from there to keep the dual license. Aren't the files you extract from a linux tarball only licensed to you under the terms of GPLv2? -- Dan -
There will
2a) You must cause the modified files to carry prominent notices
stating that you changed the files and the date of any change.
-
So how do you deduce what to revert (without going upstream for a pristine copy) if you see the following notice (I assume it is /* * ATTENTION!!! * This file was modified from it's original version on 30-FEB-2345 by XYZ */ ? -- Dmitry -
Most of them don't say anything, so they are "any GPL" by the author. When= =20 do you people accept that Linus can't change the GPL, he can only add=20 comments of what he thinks is the case! His interpretation of the GPLv2=20 might be that not saying anything about the version means "v2 only", but if= =20 he does so, he's simply wrong. He was wrong in the module case, as well,=20 and dropped this comment a while ago. He might drop this comment in future,= =20 as well. In fact, anybody can drop this comment, as it's just a comment. The kernel *as a whole* is clearly under GPLv2 only from Linus' comment,=20 which is in fact true, since the common subset of GPL versions from all=20 authors is indeed GPLv2 (by virtue of some files from Al Viro, and maybe=20 some other explicit GPL v2 files). The author must specify the version=20 himself, there simply is no other way. If you don't specify any, it's "any= =20 version", because I can license all patches straight from the authors. The= =20 way the GPLv2 allows you to explicitely specify "any version" is by not=20 saying anything about the version at all. Linus isn't in the positition to= =20 change that unless he does a substantial change to the file, and also adds= =20 a comment that this file is now GPLv2 only. =2D-=20 Bernd Paysan "If you want it done right, you have to do it yourself" http://www.jwdt.com/~paysan/
I've contributed some code for the kernel (unlike yourself, AFAICT), and believe me, I did so under GPL v2. The COPYING file is pretty much self explanatory, so I didn't need to add any explicit license statement to Linus can't and is not _changing_ the GPL. He can however use whatever license he sees fit for _his_ code just like all the other kernel developers do. People seem to forget that the kernel license in COPYING *never had* the "v2 or later" clause. Never. Period. The only change in license was from the previous hand-made one from Linus into GPL v2 only. And that is perfectly fine since the previous No, it is not "any version". It is the license specified in COPYING and Man, I sure ain't a lawyer, but people in these discussions seem to not understand the basics at all. And the basics are: "people who write the code decide the license to give it". And that's just it. And people who write kernel code are perfectly aware that the kernel license is GPL v2 only, and always has been (except for the initial linus license). So don't go around saying that because people don't put explicit license statements they don't care about the license. I care very much about the license, and would have never contributed to the kernel if it had a BSD license of some sort. Putting a license statement in _every_ file in the kernel tree would just be idiotic when there is such a clear COPYING file in the root of the kernel tree. -- Paulo Marques - www.grupopie.com "Oh dear, I think you'll find reality's on the blink again." Marvin The Paranoid Android -
It's not, it's a personal comment from a misunderstanding of the GPL text.= =20 It's as valid as the "closed source kernel modules are legal" comment that= =20 ??? Linus changed his own less permissive license (which=20 excluded "commercial use", and certainly the TiVO device is commercial) in= =20 0.0x time-frame. He added and deleted comments on top of COPYING in later=20 years, some simply wrong like the assertion on proprietary kernel modules.= =20 He added his interpretation about the version issue in 2.4.0-test9, and he= =20 COPYING says in section 9 that there may be other versions, and if you as=20 Yo, then fucking do it! Write it in the files you contribute! If you don't,= =20 you haven't! You decide, not Linus Torvalds. Make it clear you have=20 It's a personal comment from Linus, and not clear in any way. Do it the way= =20 the file COPYING itself suggests. It's not "idiotic", it's the most obvious= =20 way to do it. =2D-=20 Bernd Paysan "If you want it done right, you have to do it yourself" http://www.jwdt.com/~paysan/
Please read this sentence over and over until it sinks: "If the Program specifies a version number of this License which applies to it and "any later version", you have the option of following the terms and conditions either of that version or of any later version published by the Free Software Foundation." Pay most attention to this part "If the Program specifies xxx and "any later version"", especially word "specifies". -- Dmitry -
I believe he was talking about the sentence just after the one you quoted: If the Program does not specify a version number of this License, you may choose any version ever published by the Free Software Foundation. Linux files don't all specify version 2, but Linus, Al Viro and other authors very clearly mean their contributions to be version 2 only, while others very clearly mean their contributions to be v2+. The moment anyone makes copyrightable changes to any such files, and offers them under GPLv2 only (if that's at all possible; I used to believe so, but I've read interesting, even if surprising, arguments indicating it might not be), the result of the modification is GPLv2 only. So there's no doubt that the whole of the kernel is meant to be under GPLv2 only, even if some individual authors may choose to make their contributions available under other licenses, and be willing to make such offers when they are legally entitled to do so. I don't quite understand what this fuss is all about. Even if a majority of the Linux authors had chosen GPLv2+, or GPLvany, if any single author makes a contribution under GPLv2 only, and that contribution is integrated, that's a veto for distributing the whole under any other license. This single contributor could dictate his choice upon others, as long as his contribution was present. IANAL, but I believe that's how it works. And this means Linux is under GPLv2, no matter how much of the code in it is available under any other versions of the GPL, or even different (but compatible) licenses. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
My response to this is that by including an entire copy of specific version of GPL in the release the version number was specified. You can't say that inclusion of copy of GPL is enough to specify class of licenses (all GPL) but not specific version. -- Dmitry -
It's not that simple. Including a copy of the license is a license requirement for any redistributor, yes. But if you, a sole copyright holder, were to distribute your program, without any copy of the GPL, claiming "it's under the GPL", you're not a violator. Then, any redistributor adds a copy of any version of the GPL (because you didn't specify a version number). At this point, is the program licensed by *you* only under this specific license? Now, if you picked one of the various versions of the license, to make things easier for redistributors, does it mean you're choosing that particular version of the license, even though the license itself I can't say either of these, indeed. Or rather, I can, but I wouldn't know whether I was right ;-) -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Distributing a copy of GPL is not a requirement for me as a licensor however I did chose to include a copy of a specific version of a specific license and did not make any other statements that is the If they did not make any changes then they have to include the earliest version of GPL that applies. If they did modifications and chose GPLv4 they will have to include GPL v4 (if such requirement is in GPLv4) because License does not say otherwise. License says that if there is an _additional_ stipulation my the licensor then some other license (non existing yet license) may be used. They had to use this wording because these licensed do not exist yet. If GPL would say: "If the Program specifies a version number of this License which applies to it and "or BSD license", you have the option of following the terms and conditions either of that version or of BSD license" would you still say that BSD is allowed by default by GPL? "GPL v2 and later versions" is not different from "GPL v2 or BSD" or -- Dmitry -
And the copy you chose to include says the above. Are you absolutely sure you could terminate the license of a distributor that refrains to pass on a patent license it obtained, if you included a copy of the GPLv2 without any other indication that you're choosing GPLv2 and no other version of the GPL, in spite of the above? Would it change anything if you had released the program back when GPLv3 wasn't under discussion, and GPLv1 was long forgotten, so most No, you're referring to the portion you quoted, but I'm referring to Agreed. But this is not what this is about. This is about the license saying something like: If the program does not specify a license version number, then you're permitted to relicense the program under the BSD license. Since the license file itself is not part of the program (if it were, a program under the GPL would require the GPL itself to be under the GPL, right?), I claim the program does not specify a license version number. Now what? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Earliest is wrong I suppose. What I meant is post permissive. Otherwise it does not make any sense. And what about if there is a version of GPL that does not require passing a copy of the license along with the program? I guess it does not matter because somewhere it would still state "this program is released under GPL" (as you said there is no version number) so receient can look up what versions of GPL were ever released. Why don't you claim that actually the program is in public domain and the license file just got there by mistake? Attaching a specific license (and GPL v2 is a distinctive license, not a bumped up version of other license) places work under this (and only this) license. In my book this is different form just saying "the program is under GPL". I guess we'll have to agree to disagree. -- Dmitry -
Again, why? In the absence of a version number, why wouldn't the Yes. The initial recipient knows that, because he received the announcement by e-mail, where the "released under GPL" was. But how about downstream recipients? (yeah, I'm filling in blanks and making How can the downstream recipient tell this case from the case in which you attached one specific version of the license and didn't write License file there by mistake is a possibility, but this wouldn't make the program public domain, it would rather turn actions controlled by copyright law into copyright infringement, but as long as the recipient acted within the unclear intent of the licensor, the licensor probably wouldn't enforce the license anyway. And then, if he did, there'd be a number of defenses available for the We'll see if that works when someone tries to takes advantage of any of the holes in GPLv1 that GPLv2 plugged and you try to enforce GPLv2. I'm not sure whether to hope it will (such that this implied v2 gets better freedom protection than v1) or won't (such that I could I hope you've consulted a lawyer about this. If not, it might be Works for me ;-) Best regards, -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
These are not changes to the license text. These are just clarifications to help people understand the license. They don't change what the The section 9 is meant to explain how you select one version of the license in a program without having to copy the entire license text to it, i.e., in simple programs you can just put the small text, suggested by FSF at the bottom of the gpl, and have the version number there, and that should be enough to reference the entire text. But COPYING *is* the entire text and starts with: " GNU GENERAL PUBLIC LICENSE Version 2, June 1991" Why do you say "Wrong"? Have you contributed some code to the kernel thinking that the kernel was "v2 or later", only to find out later that it wasn't? In case you haven't followed previous discussions, here's a pointer: http://lkml.org/lkml/2006/9/22/176 The major kernel developers (and probably most of the total number of developers) are perfectly aware of the kernel license and chose GPL v2. I'm getting pretty tired of listening to people that just _know_ what I should do with _my_ code. And people who treat kernel developers as morons who can't read a license. We definitely need more Al Viro style comments on this thread ;) -- Paulo Marques - www.grupopie.com "The Mexicans have the Chupacabra. We have Al Viro. If you hear him roar, just _pray_ he's about to dissect somebody elses code than yours.. There is no point in running." Linus Torvalds -
The version of the COPYING file (and the licence document), not of the A fair bit of the kernel is probably v2 or later but not all of it and that shouldn't really matter as regards the kernel anyway, the GPLv2 only bits (if v2 only is a valid status) anchor it. -
So we are violently agreeing, then? This is a somewhat crude measure but it shows that only about 30% of the kernel is "v2 or later" and those pieces could be used on some other "v2 or later" project (including v3). But the kernel as a whole is v2 and my point was that the claim that there are just a few "v2 only" files was bogus. -- Paulo Marques - www.grupopie.com "As far as we know, our computer has never had an undetected error." Weisert -
Using this logic one can say that Linux kernel is BSD or even public domain and COPYING is there just for kicks. -- Dmitry -
No. Only the original author can specify the license. If no license at all is specified only the author has any rights to the work, other people don't have any right to distribute, modify or whatever. So if the COPYING file doesn't specify the license for work without a license clause directly in the file, then only the author has any rights, you can't just then move in and assign an arbitrary license. But I think you would find it very hard to argue that files contributed to the Linux kernel without an explicit license notice does not fall under the terms set forth in the COPYING document. -- Jesper Juhl <jesper.juhl@gmail.com> Don't top-post http://www.catb.org/~esr/jargon/html/T/top-post.html Plain text mails only, please http://www.expita.com/nomime.html -
On Fri, 15 Jun 2007 10:52:27 -0400 Not really because 1. The file is called COPYING which rather suggests its purpose 2. There is a note at the top of it 3. Lots of the code contains GPL headers Any sane Judge is going to come to the conclusion that this was the intended licence of the code. The fact people have said so also settles the matter pretty much. One of the big differences between law (at least UK/US law) and code is that the legal process seeks in part to figure out the intention of a licence or contract. Civil law is a dispute resolution process. Not a very good one, not a very cheap one, but as the previous system involved sending large blokes around to the opponents HQ with swords and axes it was found to have distinct benefit. Alan -
This was actually a part of a larger reply - but I decided not to sent that. Didn't want to throw away the work I did to get accurate numbers though ;) I investigated: find -name "*.c" | wc --lines 11100 find -name "*.c" | xargs egrep "(version 2.*([Ll]icense|redistribute|released|GNU|GPL|any later version|Foundation))|(([Ll]icense|redistribute|released|GNU|GPL|any later version|Foundation).*version 2)" | sed -e 's/:.*//' | sort -u | wc --lines 4042 find -name "*.c" | xargs egrep "(version 2.*([Ll]icense|redistribute|released|GNU|GPL|any later version|Foundation))|(([Ll]icense|redistribute|released|GNU|GPL|any later version|Foundation).*version 2)" | egrep -v '(either version 2(\.[0-9])*|version 2(\.[0-9])*( of the License)*,* or)' | sed -e 's/:.*//' | sort -u | wc --lines 1377 find -name "*.c" | xargs egrep "(version 2.*([Ll]icense|redistribute|released|GNU|GPL|any later version|Foundation))|(([Ll]icense|redistribute|released|GNU|GPL|any later version|Foundation).*version 2)" | egrep -v '(either version 2(\.[0-9])*|version 2(\.[0-9])*( of the License)*,* or)' | sed -e 's/:.*//' | sort -u | xargs grep 'any later version' | wc --lines 4 Hence there are 1377 - 4 = 1373 .c files that explicitely say 'version 2', and 4042 - 1373 = 2669 .c files that say 'version 2 or later'. There are 11100 - 1373 - 2669 = 7058 files that do not say anything. -
You forgot something. Namely, that file *without* any mentioning of GPL or other license is either illegal to distribute at all, under any license, *OR* inherits the default license of the project. Which is to say, what is stated in COPYING. Take your pick. For what it's worth, it would be interesting to hear the opinion of RMS - both on how much of the kernel is possible to distribute under v3 according to him and on the morality of your position. -
You got any case law for this? Seriously, I could use this for FSFLA's IRPF2007-Livre project. http://fsfla.org/svnwiki/blogs/lxo/pub/freeing-the-lion -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Umm... What other license choices are there? Seriously, if file *does* get a license from somewhere (and if it doesn't, it can't be distributed at all), where else would that license come from? I can see one arguing that it shouldn't be distributed at all (and we obviously don't want that), I can see one arguing that copyright statement floating in root of tree in file called "COPYING" and not tied to specific parts of that tree should apply, but I don't see how one would argue that some other license he happens to like should apply here. No specific case law, but I'd expect serious [eventual] trouble for somebody trying to slap some different license in such case. Not sure if anybody actually ever tried that... IIRC, the usual argument for slapping copyright into every file is along the lines of "making sure that it doesn't get lost when file is lifted into another project", not "it's free for grabbing by anyone" or "it can't be distributed at all"... -
Consider this (to make the freeing-the-lion story short): Jar file with .class files, with a copy of LGPL in the root of the tree. No other license anywhere to be seen. Is it safe to assume the whole thing is under the LGPL? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
It certainly sounds like a reasonable first assumption; unless you are aware of couterexamples, you probably would be able at least to prove that you've acted in good faith if somebody starts to complain. IANAL, obviously, so ask FSF lawyers. Really. Especially if you are doing that for a text associated with FSF-LA in any way. That's what they are for. -
I've covered my grounds and talked to lawyers in Brazil, where this all happened. But it wouldn't hurt me to have cases of law abroad, which is why I asked. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
If it wasn't clear someday, it seems to be utterly clear now, after
all this discussion (that is not the first related, AFAIK). Why is the
"spirit" so important for the GPLv{2,3} understanding, and not
important at all here? The intention that the kernel is gplv2 is very
clearly stated.
--
Glauber de Oliveira Costa.
"Free as in Freedom"
http://glommer.net
"The less confident you are, the more serious you have to act."
-
Woah! Stop right there. Since when a work without a license spelled Yes, you can. In this case you get _different_ software, maybe even under different license. I bet if you go straight to the authors you Let me quote GPLv2 for you: "If the Program specifies a version number of this License which applies to it and "any later version", you have the option of following the terms and conditions either of that version or of any later version published by the Free Software Foundation." As you can see for "ant later version" to apply the program must explicitely say "This program is distributed under GPLv2 and any later version". Supplying text of GPL v2 with the program does not automatically add that clause. IOW, if copying file would litrally read "Distributed under GPL v2. For the text of GPL v2 go to www.fsf.org" woudl you still argue that it is "GPLv2 and later"? -- Dmitry -
Which would only have effect on future additions, not the current content of the file - of course. Now - what if someone would write a patch for such a file (that was 'any version' before and then had a header added saying 'just version 2') under the license 'version 2 or later'? This patch could be transformed to 'just 2', and then applied to said source file - but, it could also be added to the previous version of that file (without the new header) as 'version 2 or later'. Hence, the file can still be constructed at any moment (provided the header is removed) as license 'version 2 or later', UNLESS someone adds a *crucial* patch (that cannot be removed as well, along with the header) that is explicitely made version 2 ONLY by its author. Bottom line - adding a header to those files with "version 2 only" by Linus is pointless. -- Carlo Wood <carlo@alinoe.com> -
My personal guess as to the reasoning behind this decision is that consumer devices are the ones that require most attention, mainly because the home users are the ones with least (individual) power to demand respect for their freedoms. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
I see the smiley, but I hate it how the FSF thinks others are morons and cannot read or think for themselves. Any time you disagree with the FSF, you "misunderstand" (insert condescending voice) the issue. _Please_ don't continue that idiocy. Disagreement and thinking that the FSF is controlling and putting its fingers where they don't belong is _not_ misunderstanding. It's just not "blind and unquestioning obedience". Linus -
Look, there was room for misunderstandings in earlier drafts of the license. Based on the public comments, the wording was improved. I'd like to think the issues that arose from misunderstandings of the earlier drafts are no longer an issue. Is it not so? Keeping on making false claims about the license drafts can be one of two things: misunderstandings, out of ambiguity in the text or preconceptions, or ill intentions. I'd rather believe it's the former. Now, of course you can look at the licenses and decide that you never agreed with the spirit of the GPL in the first place, and that GPLv2 models better your intentions than GPLv3. Your assessment about sharing of code between Linux and OpenSolaris very much makes it seem like that the spirit of sharing, of letting others run, study, modify and share the code as long as they respect others' freedoms, has never been what moved you. Rather, you seem to perceive the GPL as demanding some form of payback, of contribution, rather than the respect for others' freedoms that it requires. In fact, you said something along these lines yourself many months ago. With this different frame of mind, it is not surprising at all that you don't find GPLv3 a better license. With different goals in mind, reasonable people can reach different conclusions. But claiming that GPLv3 is changing the spirit of the license, or that it prohibits certain kinds of software, is plain false. In fact, the spirit has always been described in its preamble, and it didn't change at all: it's all about respecting others' freedoms. Sure, this evokes a number of other nice behaviors in various players, and it's clear to me that it's in these other nice behaviors that you seek when you choose GPLv2. There's nothing inherently wrong in that. However, it seems to me that GPLv3 would do an even better job at serving these goals than GPLv2, even if the holes v3 plugs that enabled players to disrespect others' freedoms might steer away ...
I believe a number of people don't think the GPL v3 is in the same spirit as the GPL v2. I guess it comes down to what people thought the spirit of the GPL v2 was. There certainly seems to be a variety of opinions on that, and I am not sure the FSF's opinion on it agrees with what most others believe, but that would be rather difficult to determine. -- Len Sorensen -
So let's go back to the preamble, that provides motivations and some guidance as to the interpretation of the legal text (i.e., the spirit of the license): [...] the GNU General Public License is intended to guarantee your freedom to share and change free software--to make sure the software is free for all its users. [...] [...] Our General Public Licenses are designed to make sure that you have the freedom to distribute copies of free software (and charge for this service if you wish), that you receive source code or can get it if you want it, that you can change the software or use pieces of it in new free programs; and that you know you can do these things. To protect your rights, we need to make restrictions that forbid anyone to deny you these rights or to ask you to surrender the rights. These restrictions translate to certain responsibilities for you if you distribute copies of the software, or if you modify it. [...] if you distribute copies of such a program, whether gratis or for a fee, you must give the recipients all the rights that you have Can anyone show me how any of the provisions of GPLv3 fails to meet this spirit? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
What kind of logic is that? It sounds like "Can you prove that God doesn't exist?" The fact is, Tivo didn't take those rights away from you, yet the FSF says that what Tivo did was "against the spirit". That's *bullshit*. So the whole "to protect these rights, we take away other rigths" argument hinges on the false premise that the new language in GPLv3 is somehow needed. It's not. You still had the right to distribute the software (and modify it), even if the *hardware* is limited to only one version. In other words, GPLv3 restricts rights that do not need to be restricted, and yes, I think that violates the spirit of the GPLv2 preamble! Think of it this way: what if the GPLv3 had an addition saying "You can not use this software to make a weapon". Do you see the problem? It restricts peoples rights, would you agree? Would you _also_ agree that it doesn't actually follow that "To protect your rights" logic AT ALL? And this is exactly where the GPLv3 *diverges* from the above logic. If I build hardware, and sell it with software installed, you can still copy and modify the software. You may not do so within the confines of the hardware I built, but the hardware was never under the license in the first place. In other words, GPLv3 *restricts* peoples freedoms more than it protects them. It does *not* cause any additional stated freedoms - quite the reverse. It tries to free up stuff that was never mentioned in the first place. And then the FSF has the gall to call themselves the "protector of freedoms", and claim that everybody else is evil. What a crock. In other words, if you want to argue for the changes in GPLv3, you need to CHANGE THE PREAMBLE TOO! You should change: When we speak of free software, we are referring to freedom, not price. Our General Public Licenses are designed to make sure that you have the freedom to distribute copies of free software (and charge for this service if you wish), that you receive source code or ...
By this reasoning, it sounds like you've been claiming that "God does exist", even though you can't prove it. It shouldn't be anywhere that difficult to show that the GPLv3 fails to meet the spirit of the GPLs. You just have to show a single counter-example. Since there are so many objections to the changes, Oh, good, let's take this one. if you distribute copies of such a program, [...] you must give the recipients all the rights that you have So, TiVo includes a copy of Linux in its DVR. TiVo retains the right to modify that copy of Linux as it sees fit. It doesn't give the recipients the same right. Oops. Sounds like a violation of the spirit to me. That's correct. They don't need to be restricted. The whole idea of copyleft, implemented through the GPL, is not based on needs, but rather on the wish to defend the freedoms established in the preamble from those who would rather not respect them. Do you deny that TiVo prevents you (or at least a random customer) from modifying the copy of Linux that they ship in their DVR? This would make GPLv3 a non-Free Software license. But the GPLv3 last call draft doesn't say anything along these lines. You can use the software as much as you like, even in DVRs, and even to implement DRM in them, as long as you respect the users' freedoms to change and share the software. Per the GPLv3 (paraphrased), if it is possible to install modified versions of the covered program in the device, you must tell the recipient how to do it. Otherwise, the freedom to modify the program is being too severely limited. And, in the particular case of TiVo, it's a case of distributing incomplete source code, of refraining from including functional While you look at it from the point of view of TiVo, who wants to be free to prohibit people from modifying the workings of the device it sells while it can still modify it itself, and it does that in order to prohibit people from removing locks that stop them from ...
On Wednesday 13 June 2007 19:49:23 Alexandre Oliva wrote: Are you an idiot, or do you just choose to ignore all proof that doesn't fit your preconceived beliefs? TiVO gives you every right to the Linux kernel that they recieved. What they don't give you the right to do is use modified versions on their *HARDWARE* - which they have *NEVER* given you any rights to, except for "normal use". (And no, it isn't legal to put those 200G hard Exactly. They don't. What TiVO prevents is using that modified version on their hardware. And they have that right, because the Hardware *ISN'T* covered by the GPL. And this unnaturally restricts the freedom of hardware manufacturers. If they add a custom, internal connector so a repair shop can restore the hardware to its *FACTORY* state then it is "possible to install modified versions", provided the person doing it has the specialized hardware needed. And this is what the FSF, RMS and yes, *YOU*, Alexandre, fail to realize - the GPL covers *ONLY* the software. It has *ZERO* legal standing when applied to hardware. Not even the most draconian of MS EULA's tries to apply itself to the hardware. In the case of 99% of the hardware targeted by the clause of the GPLv3 you elucidate on, the "ability to install modified versions of the software" was *NOT* intended for that use, nor was it intended for *ANYONE* *EXCEPT* trained service personell to have *ACCESS* to that functionality. Arguing otherwise is just idiotic - I have never found a piece of "high tech" hardware (like a TiVO) that was designed for the end-user to modify. (yes, And? They distribute the kernel source - as they recieved it - in compliance with the GPL. Their additions - whether they be "modules" or just the UI - do not, necessarily, fall under the GPL. (Yes, there have been discussions about whether a kernel module is a derived work, but most of the time those What "Legally Entitled" things? And... You do realize that almost every difference ...
On Wed, 2007-06-13 at 20:42 -0400, Daniel Hazelton wrote: The hardware isn't directly covered by the GPL, correct. But, if they want to use the software on the hardware, they have to comply with the GPL. The software license can then influence hardware IF they want to use it badly enough. For example, the hardware is perfectly capable of being used to break the terms of the GPL by being used to distribute a modified binary without releasing the source. But the hardware's behavior is restricted by the software for the betterment of all. This whole argument is about the spirit of the GPL. Linus and others think the spirit is one thing, the FSF guys think its something else. Since the license is clearly owned by the FSF, I think they get the final vote on what they "intended" it to be when they wrote it, no? If they say they intended it to not allow Tivoization then believe them, because they are the only ones that know what they were thinking when they wrote it! The GPLv2 seems to allow it though. If Linus and friends want to allow it, then they can stay with the GPLv2. For those who want Guys, we are all friends here. No reason to be so insulting. Its just a difference of opinion. People seem to be talking past each other instead of to one another. This usually happens when people are basing their underlying assumptions on different things and not listening to the other. Please take a step back and think about it. <SNIP> -
Only with the GPLv3. Again, don't confuse the *new* requirements in the GPLv3 with any "GPL requirements". They didn't exist before. The kernel never signed up to them. They are irrelevant for the discussion. So hardware details have *nothing* to do with compying with the GPLv2. Could you write *another* license that puts limitations on the hardware or environment that you have to comply with? Sure can. And the GPLv3 does that. But the GPLv2 does not, and that's a fundmanetal *improvement* over the GPLv3 in my opinion. Do you like licenses that force the licensee to give money back? So why do you like licenses that force the licensee to give access to hardware back? It's a form of "extra compensation" that the GPLv2 never had. The GPLv2 talks about giving access to the *source* code. The GPLv3 talks about giving access to the *hardware*. Can people really not see the difference, and why I might think it's a fundamental difference, and why I might choose to say that the GPLv3 is a worse license? And *why* would I ever downgrade to a worse license? There had better be some really pressing reason to choose the worse version of the GPL. And I just don't find that reason in the GPLv3 itself - although, as mentioned, the reason could become *external* (ie I might accept a worse license it it comes with external code attached to it that I think makes up for the license deficiency). Linus -
This is not true. The terms of the GPLv2 that say you can't impose further restrictions on the exercise of the freedoms apply to the I don't know where the 'back' in the second question amounts to, but it definitely isn't about GPLv3. In fact, the GPL isn't about giving anything back. It's about passing on. So both requirements, as you phrased them, would be equally wrong. So let's change the question to turn them into forms of passing on: Do you like licenses that force the licensee to pass money on? Do you like licenses that force the licensee to pass on access to hardware? This is still bad. This is still not what the GPLv3 is about. There's no requirement to let the user go wild and do whatever she likes on the hardware. The only requirement is the one that was always there: to respect the freedoms of the users of the software, i.e., let them modify and share the software, not imposing any further restrictions, by whatever means. So the second question would be correctly phrased as Do you like licenses that force the licensee to pass on the right to modify the software in the hardware containing it? Or, reframing it: Do you like licenses that permit the licensee to deny others the No, sir, it's still respect for the freedoms. The same "in kind" Since someone brought liberal (Original BSD, Modified BSD, MIT, etc) licenses into the picture, and you expressed dislike for them, let me Sounds a lot like the very "everybody for himself" attitude you dislike. So can you please explain to me how enabling TiVO to deny others the freedom that it received "in kind", failing to keep with the "in kind" spirit of the GPL, encourage people to work together, and to merge? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
They're not denying others the freedom that they themselves received. Tivo took GPL'd software, modified it, and distributed it with their own custom hardware. You have the right to take their changes, possibly modify them further, and distribute them (possibly with your own hardware). The fact that you can't modify the software and load it back onto the tivo is irrelevent. They are not restricting your distribution of the software in any way. Rather, they're restricting the *running* of the software on their proprietary hardware platform. Chris -
The GPLv2 talks *only* about the software. You're making everything else up, and when I point out that your reading of the GPLv2 is insane, you Because Tivo *IS NOT DENYING* those freedoms. Tivo *respected* the freedoms, and gave source back, and gave you all the same rights you had to Linux originally, and to their modifications. How stupid are you to not acknowledge that? Tivo limited their *hardware*, not the software. Linus -
Have you ever wondered *why* it limited the hardware? Is it per chance such that I cannot modify the software that runs on the hardware? How is that respecting the freedoms? How is this not imposing further restrictions? And, more importantly, how is it that permitting this makes for *better* compliance with your tit-for-tat conceptions about the GPL? I.e., if Tivoization is the only issue that you think makes GPLv3 a worse license than GPLv2, and you like GPLv2 because of this tit-for-tat, surely you should be able to explain why Tivoization promotes this tit-for-tat notion better than GPLv3, right? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
if you cannot modify the software that runs on your Tivo hardware you haven't tried very hard. true, they don't go out of their way to make it easy, but even if they didn't do the integrity checking of the system it still wouldn't be easy to load your own software on the tivo, there's no path to load the I think the software is all available at www.tivo.com/linux that provides you all the freedom that they got. -
Yes, but the GPLv2 clearly says that you don't have to try very hard. The=20 preferred form of modification has to be distributed. I can run a=20 decompiler or disassembler on a program, and I can even modify it in place= =20 with a hex editor (I have even modified programs in embedded ROMs by using= =20 focussed ion beam, so I know you can modify every program if you try hard=20 enough). It's certainly possible to crack Tivo's firmware to accept my own= =20 signature, but it's *not* the preferred form of modification, the source=20 code and Tivo's key for the signature. Since Tivo's firmware only accepts a signed kernel, the combination of=20 kernel+signature is the binary they ship. The kernel itself is useless, the= =20 signature as well. Therefore, you can imply that Tivo's key is part of=20 the "other stuff" the GPLv2 mentions, because you need it to recreate the=20 same code as Tivo did and shipped (compilers insert timestamps and such),=20 and to modify that code. The source code is just a mean, the thing they=20 shipped is the end (the binary), and they have to comply with the GPL for=20 that binary - which by all means of practical understanding includes the=20 signature. "You can imply" means: It depends on court and legal system. I'm quite=20 confident that in Germany, the legal system might favor the "GPLv2 does not= =20 allow tivoization" point of view, and in the USA, the legal sysem might do= =20 the opposite. =2D-=20 Bernd Paysan "If you want it done right, you have to do it yourself" http://www.jwdt.com/~paysan/
How is a signing key part of the "preferred form for modification"? It isn't a requirement to *modify* anything, just to *replace* something. (And I am I can find no such requirement in the GPLv2. In fact, it actually says that you don't even have to be able to *USE* the program. See section 12 of the In light of the d-link case, I'm pretty certain that the German Courts interpretation of the GPLv2 makes "Tivoization" a violation. In the US I can say that the result would be "GPLv2 does not disallow tivoization". As I've pointed out in other posts, the GPLv2 actually *limits* itself to three specific "activities". Whether it was intended to "incidentally" cover other things or not, it does *clearly* state what it's scope is. If that scope *IS* *NOT* the intent of the person and/or person who authored the license, that text *SHOULD* *NOT* exist. DRH -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
The signing key determines a critical portion of the binary form that was distributed. You cannot produce that portion of the binary form without the signing key. Without that portion, the binary form does not perform the function for which it is distributed. If you think such an input is not part of "the preferred form for modification", I have a bridge to sell you. The work that the GPL protects a recipient's right to modify and redistribute is not the source code -- it is each form the user Section 12 of the GPL(v2) is a warranty and liability disclaimer. It is not an absolution of license obligations. It limits the liability of a distributor to the end user, not to copyright owners. Michael Poole -
What the fsck it is, linux-kernel or bleeding Council of Nikea? -
Until GPLv3 there was no requirement that the modified code be able to operate on any given device - even the one its designed for. Claiming otherwise seems But this can be done *NOW* - has been done by at least one company, IIRC. (and, IIRC again, they didn't so much as "not release the modified version" I noticed that ten or twenty messages after I made that comment. In truth, the reason I made it was, and is, because I am tired of explaining the fact that there is no "one" interpretation of the GPLv2 - or any license - *UNLESS* it has been ruled on by a court. And even then, the courts ruling only applies to the parts of the license that were in contention before it. -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
Indeed, TiVO has this legal right. But then they must not use software under the GPLv3 in it. And, arguably, they must not use It's about time for a change for better, wouldn't you think? In 95% of the desktop computers, you can't make changes to the OS that This makes it seem like you think that passing on the source code is Time shifting of any shows, creating copies of shows for personal use, letting others do so. Think fair use, and how TiVO software and DRM For those who are not willing to abide by the spirit of the license, yes. Does it look like I'm concerned about them? If they're willing to look for and maybe even find holes in the license to disrespect users' freedoms, why should I worry about the problems that plugging these holes is going to cause them? If they'd taken the spirit of the GPL for what it is, instead of looking for loopholes, this improved This is a pretty sad accusation. 2/3s of the Free Software packages use the GPL with its existing spirit, and you still haven't shown that any changes proposed in GPLv3 fail to abide by the same spirit. That some (many?) people misunderstood or disregarded the spirit is an unfortunate fact, but trying to pose the patching that's going into GPLv3 as if it was a matter of personal taste, rather than improved What are you doing lurking and spreading confusion in a list about a Oh, sorry. I missed when the meaning of the word computer was narrowed from "machine with a general-purpose microprocessor, memory and other peripherals" to whatever you decide it is. And then, the GPL doesn't talk about computers at all. It's not about Err.. The hardware licensor who includes software under the GPL be supposed to be a licensee of the software in order to have legal permission to distribute it, at which point the following provision kicks in: 6. Each time you redistribute the Program (or any work based on the Program) [...] You may not impose any further restrictions on the recipients' ...
^^^^^^^^^^^^^^ BTW as soon as I bought that thing, it is *my* hardware and no longer Do they? At least in .at, it is usually impossible to (legally) limit the rights of the *owner* a (tangible) thing (and if I bought it, I *am* the owner and no one else) - even if you put it in the sales contract since this is discussion about/within sales law. One usual example is "you buy a car and neither the car producer nor the (re)seller can restrict the brands of the tires you may use or the brand of the fuel etc.". And the same holds for pretty much everything. No one can forbid you to open a TV set and fix it (or let it fix by whoever I choose to). Yes, there are exceptions in several laws for specific things (e.g. for really dangerous ones like airbags in cars) but in general, you are allowed to do almost anything (including the simple destruction of it). And yes, if you *rent* the thing, you are not the owner and this is a Bernd -- Firmix Software GmbH http://www.firmix.at/ mobil: +43 664 4416156 fax: +43 1 7890849-55 Embedded Linux Development and Services -
eh. Perhaps I should have said that differently. And TiVO could handle it No argument there. However, that is not to say that "you bought it, now you're free to do with it whatever you please" is always what the law says (at least in the US) In the TiVO case there may be restrictions placed on the manufacturer for legal reasons or contractual reasons. Seeing as I'm not privy to the contracts between TiVO and the various production and broadcasting companies I can't comment on what contracts they have. As to the legal side there are I know of at least one company that will sell you the parts to repair your TV if its out of warranty. -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
Of course not (and I neither stated nor implied it) - there are lots of
laws which forbid killing other people etc.
But the seller of the car is not in the position to forbid anything
(which is not forbidden by the law), e.g. ha cannot forbid to replace
the motor or similar thing. I may loose guarantee or have to cope with
other consequences (if it is done badly), but that is my problem and
Frankly, I really don't care that much about legal and contractual
reasons of the *manufacturer* (starting from waste disposal regulations
up to tax regulations, etc.) and they are irrelevant to me anyways.
At most I can have
*) legal restrictions (obviously coming from the law) on the *usage* of
the device or
*) from a contract (obviously with the seller of the device since there
is no other involved - and this contract may contain inapplicable
clauses - e.g. sth. like "you are not allowed hear German music with
this device").
And I don't have a contract with the manufacturer so there can't be any
ACK. But copyright law (at least the equivalent in .at and very
probably .de - and IMHO it is probably everywhere else similar simply
because copyright/authors rights laws was actually designed and written
to deal with music, literature, etc. which are intangible by nature)
simply doesn't apply to hardware as such (pun intended;-).
Bernd
--
Firmix Software GmbH http://www.firmix.at/
mobil: +43 664 4416156 fax: +43 1 7890849-55
Embedded Linux Development and Services
-
You bought *their* design. It was your choice. And yes, you own the hardware, and you can hack it any which way you like (modulo laws and any other contracts you signed when you bought it). But they had the right to design it certain ways, and part of that design may be making it _harder_ for you to hack. For example, they may have used glue to put the thing together rather than standard phillips screws. Or poured resin over some of the chips. All of which has been done (not necessarily with Linux, but this really is an issue that has nothing to do with Linux per se). Making the firmware or hardware harder to access or modify is their choice. Your choice is whether you buy it, despite the fact that you know it's not The "when I buy it, I own it" argument is a favourite of the GPLv3 shills, but it's irrelevant. The *design* was done long before you bought it, and yes, Tivo had the right to design and build it, any which way they wanted You are missing the picture. Sure, you can do whatever you want to (within any applicable laws) _after_ you bought it. But that doesn't take away the right from the manufacturer to design it his way. And you're also *wrong*. Tivo doesn't limit the brands of electricity it uses or anything idiotic like that. You can put after-market rubber bumps on the thing to make it look sleeker, and I seriously doubt that Tivo will do aythign at all. It's about going into the innards, and different car manufacturers make that harder too, for various reasons. If the car manufacturer makes things harder to hack, it's your choice. For example, car hackers *do* actually prefer certain brands. Apparently the Subaru's are popular, and German cars are a pain to try to change. I'm told that even somethign as simple as upgrading the sound system is just _harder_ in a German car, apparently because they make things fit together so tightly, that doing after-market cabling is just much more of a problem. Same goes for things like ...
Yes. It can be argued. But I cannot find *ANYTHING* in the GPLv2 that stops anyone from doing that, unless you add extra meaning to one specific clause. I've never had a reason to want to change the way any device like a TiVO Faulty logic. I have yet to find a computer that I couldn't change the OS on. I have run Linux on 3 different Mac's, every x86 machine I've ever owned and even had it running on my Palm. Whats more is that I have *never* heard of a person that knows what they are doing not being able to change the OS on a Hrm. Strange, but thats what most companies think. Hell, it even says that you have to do just that in the GPL. If you're talking about the fact that it can be argued that they are "distributing" Linux by selling their boxes and its a I thought that time shifting and creating personal copies was what the TiVO did already. Or do you mean "transferring the recorded copies off the TiVO and on to a different medium"? If that is what you mean by "Creating Copies" Okay. So you're not concerned that you're potentially pushing companies that would otherwise be major consumers of GPL'd software away? That doesn't make Why should I repeat Linus' explanation of the ways that GPLv3 violates the spirit of GPLv2? And why shouldn't I pose it as a matter of "Personal Taste"? The biggest and most powerful voice in the FSF says "I don't like Tivoization" and "I don't like DRM" and when the GPLv3 appears it has language that makes those violations of the license. Just like people have started using "GNU/Linux" or "GNU+Linux" to refer to Linux - a big voice spoke and said "It should be GNU/Linux" and it happens. (Not that I really have anything against that - Just because I don't like the license doesn't make disqualify me from liking something that uses the license. And I doubt I've "confused" anyone - where I The word "Computer", in the manner I used it there, means "General Purpose Computational Device". A "TiVO" is not, and has ...
BTW: don't they sell their hardware (as well)? I think it should be easy to replace the ROMs (EPROMs? flash ROMs?) using some diagnostic clip and/or JTAG. Unless the CPU itself verifies ROM signatures, they shouldn't matter. -- Krzysztof Halasa -
Have you never wanted to improve any aspect of the software in your cell phone? In your TV, VCR, DVD player, anything? In the microwave I was not talking about installing another OS, I was talking about making changes to the OS. As in, improving one particular driver, Sure. Such that I can watch shows while wasting time in public And the fact that TiVO can be, and has been modified remotely to add restrictions on what users could do, means nothing you do with it is safe. You, and everything you've recorded with the TiVO, are at the What would their consuming GPL software buy us, if they won't respect users' freedoms, which is the very reason behind the GPL? Heck, if they don't want to play by the rules, that's up to them. But then they shouldn't use the software at all. Yeah, I wish they'd rather play by the rules, but if they don't want Don't worry about parrotting here, he hasn't provided that explanation yet ;-) Please give it a try. BTW, what license is Linux licensed under? It's GPLv2 plus userland Have you ever wondered *why* he doesn't like them? Could it possibly be because they harm the goal of his life, which is No, no, you got it wrong. Linux is the kernel. GNU was the nearly-complete operating system it fit in. GNU+Linux is a complete operating system. And you don't have to believe me, believe Linus, the initial author of Linux: http://www.kernel.org/pub/linux/kernel/Historic/old-versions/RELNOTES-0.01 Although linux is a complete kernel Sadly, a kernel by itself gets you nowhere. To get a working system you need a shell, compilers, a library etc. These are separate parts and may be under a stricter (or even looser) copyright. Most of the tools used with linux are GNU software and are under the GNU Err... Last I looked it was a bunch of general-purpose components, packaged in a way that made it not look like a general-purpose 2. You may modify your copy or copies of the Program or any portion So does the ...
Nope. I've been tempted several times, but decided that the extra bits I'd Ah, well... In the case of "Windos" and other proprietary OS's I try to educate people and get them to switch. I don't, personally, have any computers that run Windows (and I switched my Palm back to PalmOS because it wasn't getting the same performance under Linux - which rather surprised me. Under the US Copyright law I'm not sure that making a "second copy" like that As has been noted in their TOS and the licenses for the hardware from the start. The FSF itself explicitly reserves the right to change the GPL at any time - which is no different. (when you remove all the bits explaining the I'm not referring to companies that are embedding GPL'd software in their products. The companies I'm referring to are the ones that would like to use GPL'd software internally. A lot of them would probably have private modifications that would never be distributed - and under the GPLv2 it is clear that you can keep modifications private as long as you don't distribute them. "Pushing them away" means that they'd not do that because they would be concerned that the license will change under them in such a way that even those private modifications need to be released to the public. (and don't try to argue that even though those modifications are truly private (to the company) they should be released anyway to comply with the "spirit" of the license. It is made clear that it isn't by the text of the license But he has. Whether you have accepted that his explanations are valid or not The kernel itself is GPLv2 (only). Individual components - even individual files - have other licenses or retain the "any later version" clause. (Someone pointed out, earlier in this thread, that there is GPLv1.1 code in Not really. I've always figured he had reasons similar to mine for not liking DRM. As to his dislike of "Tivoization", well, that I've always attributed to the fact that someone at that ...
Good. So I presume you'd tell them to switch away from a turned-proprietary GNU/Linux operating system as well, right? So, again, what do we gain if companies abuse the GPL and disrespect Even if you delete the "first copy"? Actually, I thought fair use in US entitled you to make a backup copy. So the copy in your TiVO would be your original, and the external copy If it is used to disrespect the inalienable freedoms associated with the GPL software in the device, it seems like a license violation to Actually, it's completely different. If the FSF revises the GPL, the old version remains available for anyone to use for any new software, and all software released under the old version remains available under that old version. In contrast, your TiVO may get a software upgrade without your permission that will take your rights away from that point on, and there's very little you can do about it, other than unplugging it from This would not only change the spirit of the license, but turn it into a non-Free Software license. And then, again, the license can't possibly be changed from under them. A new revision of the GPL would only affect software licensed under that new revision. If you already got it under an earlier revision, you know what you got, and nobody can take that away from How could you possibly come to the conclusion that forcing anyone to release private modifications would be in compliance with the spirit His explanation is based on a reading of the license that doesn't match what its authors meant. I guess the authors know better what they meant the spirit of the license to be than someone else who studied it a lot but that until very recently couldn't even tell the Referring to Linux as GNU/Linux would be wrong, because Linux is the kernel, and that's unrelated with the GNU operating system. It's the combination of them that forms GNU+Linux. And it's referring to this combination as Linux that is wrong. I'm sorry that I got the ...
If that happened I'd be lost. I've tried the various BSD's and found they had problems with hardware support and getting a new version of the BSD kernel to compile and boot is something of a black art. As much as the US "Declaration of Independence" and other sources want people to believe otherwise there is no such thing as "inalienable rights" or "inalienable freedoms". In this case I have been unable to find this "inalienable freedom" to run custom versions of software "on the same machine" that you received the original copy on anywhere before the GPLv3 - and even then it isn't explicitly clear. There is no restriction on your right to modify, copy, distribute or run the software as provided by versions of the GPL prior to version 3. If this "run modified copies on the same hardware you received the original on" *IS* the "spirit" of the license, then why isn't it stated anywhere before GPLv3? (After all, the FSF has have 20+ I'll grant you that. But, at this point, where can I find a copy of the GPLv1 And because its a device that connects to their network - and TiVO isn't a telecommunications company - they have the right to upgrade and configure the True. But that doesn't save them from lawsuits trying to force them to obey the terms of the new revision even though they received the software under an I was trying to be sarcastic and inject a little humor here. Guess I should And his interpretation is no less valid than that of anyone else. In fact, after a recent conversation with a couple of lawyers that I know, I can state Then you're lucky. I've had a lot of people say something similar to the following: "Oh, I've heard about that. So which version of the GNU-Linux Yes, it does. While pronouncing the '/' or '+' sounds a bit odd it does get the point across that it's the GNU userspace running on top of the Linux As I've stated before - I can find nothing in the history of the GPL or the FSF that makes the "on the same ...
Look again, it's already happened in the TiVO and other devices. The software that ships in them is no longer Free Software. Consider a new microprocessor. Consider that Linux is ported to it by the microprocessor manufacturer. Consider that the manufacturer only sells devices with that microprocessor with TiVO-like locks. How exactly can you enjoy the freedoms WRT the GPLed software you got from the manufacturer? Now consider that you have a single computer, and that's built by TiVO. How exactly can you enjoy the freedoms the author meant you to have, For the same reasons that the pro-DRM laws weren't mentioned before, and the patent retaliation clauses weren't mention before: these specific cases hadn't been studied, only the general idea of In the program you received under GPLv1. Hey, you said there was code under GPLv1.1 in the Linux tree. Then, there should be a copy of GPLv1.1 in there, otherwise AFAICT the In what sense does the GPLv3 make this particular point any less Nothing saves anyone from silly lawsuits. This one would likely be laughed out of court in no time. Anyone worried about this should also be concerned about their neighbor suing them for copyright infringment every time they set their stereo loud enough for the neighbors to listen and be annoyed. (Hint: only the copyright holder Interpretation as applied to the legal terms, yes. As for the spirit of the license, the authors ought to know better than anyone else what they meant. Sure, other interpretations might lead to different understandings as to what the readers *think* it means, but that Put the considerations above, about a single computer or a uniformly-limited computing platform, and you'll see that this "on the same hardware" argument is just a means to deny people freedom. If I could stop you from running modified versions on one piece of hardware, then I could on two, and 3, and then soon it's all of them, If you replace a component in the ...
So now the copy of the GPL v2 isn't good enough for the GPLv1.1 code? Maybe that code said 'or later' in the license and hence someone added it to a GPL v2 project since that sounds perfectly OK. -- Len Sorensen -
Where did that GPLv1.1 nonsense come from? There is no GPLv1.1 code in the tree. By the time I selected the GPL for the kernel license, the GPLv1.1 had long since been discontinued. The kernel was *never* GPLv1.1-only compatible. That's just total nonsense. There was indeed a kernel license before the GPLv2, but it wasn't the GPL at all, it was my own made-up thing. Appended here, for those who are too lazy to actually look up and check the original Linux-0.01 announcement. Linus --- This kernel is (C) 1991 Linus Torvalds, but all or part of it may be redistributed provided you do the following: - Full source must be available (and free), if not with the distribution then at least on asking for it. - Copyright notices must be intact. (In fact, if you distribute only parts of it you may have to add copyrights, as there aren't (C)'s in all files.) Small partial excerpts may be copied without bothering with copyrights. - You may not distibute this for a fee, not even "handling" costs. -
A hundred or so messages back someone stated that the parport driver in Linux is GPLv1.1 - however, on checking on this statement for myself I've found that there is no statement about it being v1.1 and, in fact, given that Linux itself is GPLv2 there is no possible way any code covered by GPLv1.1 can exist. -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
Wrong again. If a piece of code was merged into the kernel with a GPL v1 "or later" license then it still has a GPL v1 "or later" license. The "or later" makes it compatible with the v2 code but does not change the fundamental copyright on the original work that was combined. Thus if you could identify specifically a GPL v1 work within the kernel you could use that GPL v1 work as per GPL v1 providing you didn't mix it with v2 code. If I take a public domain book and create a derivative work from it the original work does not magically become restricted. Alan -
Probably a misinterpretation - there are comments in the parport driver=20 mentioning the GFDL version 1.1. If you just grep through, you might think= =20 it's GPL version 1.1 (but the code is really v2 or later). =2D-=20 Bernd Paysan "If you want it done right, you have to do it yourself" http://www.jwdt.com/~paysan/
If it exists and it's 1.1-only, I believe it wouldn't, but IANAL. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
In *YOUR* opinion and by *YOUR* definition of the term. Yes, I have seen some evidence that TiVO hasn't made some of the modifications they made public - doesn't mean that they won't, just it hasn't *YET* been done. (Not that I'm so omniscient I can say, definitively, whether they will or won't - or even that they haven't done it already). The same as I would with a TiVO. I have the right to copy, modify, distribute and run the code - even if I can't do any of those things on the hardware the Simple: I don't buy it. Each and every piece of hardware I buy has a rather laborious research process before I actually spend the money on it. This makes it a certainty that I can use the hardware in the manner I want without problems like your hypothetical. Whats worse - forcing your morals and ideals on someone or giving them the same freedom of choice you had? Before you answer remember that that is *EXACTLY* what is being done with GPLv3. With GPLv2 and prior there was a simple guarantee that every "Licensee" had exactly the same rights. With GPLv3 you are forcing your ethics and morals on people - and isn't this exactly what the Roman Catholic Bzzt! Wrong! The reason is that it wasn't necessary - at all. It still isn't, but a group that feels modification == replacement wants it to be, so it has suddenly become necessary. (Note that anti-DRM stuff *IS* good - DRM is part Ah, but I never said I had a GPLv1 program. If GPLv1 is still valid and available I should be able to find a copy of it *RIGHT* *NOW* to license a new project if I want to use GPLv1 as its license. So your logic is again Yes, they do. It isn't a right they have as "copyright holders" - in fact, it isn't a part of their rights under the copyright at all. It's a part of their Never claimed it was less obscure, just that you've usually got a board-room filled with middle-aged men that might have problems agreeing that it is a Yes, but the fact that it would cost money to get the ...
I fail to see the distinction you're making between GPLv2 and GPLv3. AFAICT, with GPLv3, there still is a simple guarantee that every licensee has exactly the same rights. Sure, GPLv3 follows the spirit of the GPLs more strictly than GPLv2 possibly could. How is that "forcing ethics and morals" any more than I thought you had a copy of Linux and, per what you'd said before, there was GPLv1 code in it. I was just trying to make it easy for Again, how are these arguments against GPLv3? They apply equally to You're talking about the legal terms. The spirit of the license is a very different matter. It can guide the interpretation of the legal terms, but the author is at a better position than anyone else to know Well, then, lock down the software. Make it irreplaceable, even by yourself. Problem solved. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Because GPLv2 doesn't enforce limitations on the hardware a GPL'd work can be put on. It doesn't make artificial distinctions between "Commercial", "Industrial" and "User". What it does is *ATTEMPT* to ensure that nobody receiving a copy of a GPL'd work has the same rights as any other person that gets a copy. GPLv3 gives people *additional* rights beyond those. In the "TiVO" case it forces somebody releasing a HW platform to grant *additional* rights if they are going to use software covered by the GPLv3. The reason for forcing the giving of those additional rights is "the FSF and GPLv3 committees think that what TiVO did is 'morally and ethically' wrong, so were are enforcing our morals and ethics". Note that these are the rights that TiVO got from the GPLv2: 1) The ability to make copies of Linux 2) The ability to modify Linux 3) The ability to distribute Linux *NOTE* that those are the rights *GUARANTEED* by the GPL - despite what anyone *WISHES* it to say, or what the "Intent" or "Spirit" of the license may be, those are the only guaranteed rights. In shipping their devices with an "object code" version of Linux on them they exercised their right to perform such a distribution, granted under section 3 of the GPLv2. They made modifications to the Linux so it functioned properly on their devices, as allowed by the GPL. They have made numerous copies of Linux, as allowed by the GPL. And, as required by the GPLv2, they made the source code form of their changes available. In fact, they went beyond the requirements of the GPL (which only requires you make the source available to people you have given an "object code" version to) in making it fully available to the public *AND* in contributing those changes back to Linux. What rights did they give to "downstream" recipients of the "object code" version? *EXACTLY* those they received from the GPLv2. What rights do they have as creators of a *PROPRIETARY* hardware platform: 1) The right to restrict what ...
Doing the e-mail equivalent of yelling about this will not change the fact that people who think Tivo did something wrong -- legally and/or morally -- consider DRM locks on a piece of software to be part of the "work based on the Program" that is governed by the GPL. The fundamental reason for this is that neither the executable code nor the digital signature serves the desired function alone. The user received a copy of the executable for a particular purpose: to run the program on a particular platform. With DRM signatures, only the combination of program and signature will perform that function, and separating the two based on strictly read legal definitions is risky. The question of whether DRM signatures are covered by the license must be resolved before one can determine whether Tivo gave "*EXACTLY*" the same rights to object-code recipients as Tivo received. GPLv2 is worded such that the answer to this does not depend on whether one is in file A and the other in file B, or whether one is on hard drive C and the other is in flash device D, as long as they are delivered as part of one unit; it *might* matter if, say, one is received on physical media and the other is downloaded on demand. (Linus likes to say that FSF counsel thinks that Tivo did not violate GPLv2. I suspect the actual situation is that FSF counsel believes that there is no case law on point, and that it could go either way, making it improper to publicly claim that Tivo violated any copyright. Until a court rules on a close-enough case, the question of whether GPLv2 covers DRM signatures remains open. In the mean time, it makes more sense for the FSF to issue a new license that squarely addresses this -- such as the GPLv3 -- and persuade as many developers as possible that using it is the best way to protect free software.) Michael Poole -
Sorry, I wasn't trying to "yell" - just provide a note that at that point I All I've done is get a little annoyed that, despite evidence that it isn't legally wrong - at least under the laws I am most familiar with - people continue repeat that it is. I can't argue that it isn't "morally" wrong. While it may not be against my morals, it could be against those of another person. It has never been my I have read the GPLv2 at least three times since it was pointed out that I had forgotten part of it. At no point can I find a point where Tivo broke the GPLv2 requirement that they give the recipients of the object code the same In examining the GPLv2 and the situation from a strictly factual basis I can believe Linus' statement fully. The facts are as I stated them in a previous mail. -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
I am trying to reconcile your responses to those two paragraphs. If the DRM signature and program executable are coupled such that they are not useful when separated, the implication to me is that they form one work that is based on the original Program. This is beyond the GPL's permission for "mere aggregation". If they are one work, and the original Program was licensed under the GPLv2, the combined work must also be licensed under the terms of the GPLv2. The input files required to generate a DRM-valid digital signature are part the preferred form for modifying that work. If those bits are not distributed along with the rest of the GPL'ed work, the distributor is either not giving the same rights to the end user, not distributing the source code for the work, or both. Which is it? Michael Poole -
So you want to make things like a 160-bit SHA1 hash of a binary be a "derived work" of that software? Trust me, you *really* don't want to go there. It's an insane legal standpoint, and if you were right, we'd be in a *world* of trouble. Think about something as simple as security software that creates filesystem checksums for verifying the integrity of the filesystem, and protects against tampering. Do you *really* want to claim that the SHA1 checksum of your "oracle" binary is owned by Oracle, and you need to have a special license to copy that checksum around and verify it? Do you *really* want to claim that the RIAA owns the CDDB checksums (well, I guess "feedb", these days) of the CD's that get uploaded for music databases? Do you realize that in your INSANE world, there is no notion of "fair use", and you just tried to extend the notion of copyright so far that you turned your utopia into a total distopia. In other words, anybody who claims that copyright in a program extends to the cryptographic hash of the binary, and at the same time makes a "free software" kind of argument is so damn clueless that it's not even funny. You're arguing for "freedom" by using logic that is the very *antithesis* of freedom. That's just incredibly stupid and incredibly short-sighted. If that were to seriously be an FSF argument, then I would officially lump the FSF as a *much*worse* danger to the free world than the RIAA and the MPAA combined! I seriously doubt you really thought your idea through! Because it goes beyond stupid. Linus -
No. That is why I specified "not useful when separated". I also intentionally avoided the phrase "derived work": the legal definition of derived work is based on entirely different factors. If the signature is one that serves to indicate origin, to detect tampering, or the other things you mentioned, the program's binary is useful when separated from the signature. My objection arises when a functionally equivalent binary -- including advertised functions such as "runs on platform XYZ" -- cannot be produced from the distributed source code. Michael Poole -
Ahh. Ok, that's a totally different issue, and is one where I heartily agree with you. I would actually *love* for the GPL (any version) to have a "guarantee of authenticity", where if you distribute a binary, there has to be some documented way to get *exactly* that binary out of the source code that got distributed. Of course, SHA1's can be used to verify that, although, quite frankly, I'd expect that a simple "cmp" would be the more straightforward approach. So the "verification" can be used both to lock down a particular binary _and_ to authenticate that the binary really came from the source code it was claimed to come from. Of course, in practice, it's actually really nasty to do that verification. Many compilers actually do things like insert date-stamps in the object files etc. So it's probably not all that practical. Linus -
I would hope that this is *required*, somehow, when dealing with medical equipment. I don't think those appliances even have the capacity to build every upgrade from source. None that I've tinkered with do. These things almost need a license of their own. As long as the signing mechanism can't be used to force clinics to pay for the privilege of upgrading free software, that is. It would truly suck if an ultrasound loaded with free software sat in a corner useless because a free clinic could not afford to pay for what they already paid for. If you guys can find a way to make that practical given my above concerns, that would be entirely useful. I hate the fact that this kind of trust is needed because it is so very easily mis-used, but people dying due to hacked IV regulators really wouldn't much care about those politics. I think, also privacy implications for patients. A rootkit in a MRI would be very bad. Regardless, like it or not, kernel code is in or headed for medical devices, so I hope some more brain power is burned on this. Best, --Tim -
How about the combination of the software binary with the hash? Considering that the hash is a functional part of the software, as in, Remember: I don't speak for the FSF, and I don't speak for FSFLA, just like I don't speak for Red Hat. Just like you shouldn't redirect your qualms with the FSF to me, you shouldn't direct your qualms with me to the FSF. That would be very wrong. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
not only that, but it would instantly turn everyone who owns a hard drive or a CD-ROM into a copyright violator: the disks checksums the content of the disk _in a reversible way_. Same for RAID5 and RAID6 techniques. By installing Quake3 on a Windows box one sure does not have permission to create a derived work of Windows and Quake3, right? =B-) a checksum, a one-way hash, or even reversible parity bit(s) that 'mixes' the copies of multiple works together clearly cannot be new work that falls under copyright protection. Firstly, it is not a new work, because a 'work' has to be created by a human - and here the new content was created by a machine. Copyright protection only applies to sufficiently original works created by humans. Secondly, it is _at most_ a new, partial copy of existing works and hence you need the permission to copy all the works in question. (but you needed that permission to create the harddisk anyway) Thirdly, it could be argued that the sha1 is not even a copy, because it is irreversible and hence not even a single bit of the original work can be reconstructed from it hence it cannot even be a 'partial copy of the original'. that's at least 3 robust levels of argument against the insane and absurd notion that the SHA1 key is somehow a derived work of the copy it checksums. Ingo -
Wouldn't you consider the signing key as one of these existing works? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Following your logic it would be a "failure to distribute the source code for a work". However, since the signing is an automated process it cannot generate a "new" work - at least, not under the laws of the US - so the signature itself cannot have a copyright at all. DRH PS: This is the exact same reason that the GPL cannot apply to a Bison generated parser in the US. The "input" file that causes Bison to generate the output can have a copyright, but not the output - no matter what RMS or -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
I do not suggest that copyright subsists in the signature or in the signing key. Whether it does is irrelevant to the signing key being part of the source code (when the signature is needed for the binary to work properly). Similarly, copyright might not subsist in a simple linker script -- its content being determined by the operating system and perhaps the rest of the program's source code -- but under the GPL, the linker script would be part of the source code for a compiled version. Michael Poole -
it is very much relevant. By admitting that the key is not part of the "work", you have lost all moral basis to claim control over it. Cutely "defining it" into the source code just hides what this really is: the key is a "payment" in exchange for the license, which payment goes outside the scope of the software itself. It has no relevance to the software work being "free", it reaches for paymeant beyond the work to advance the FSF's agenda. yes, a copyright license can be used to control other works, it can be used to control the movement of non-copyrightable items as well (such as money), but the GPL always tried to stay out of that kind of business. Where does this "reach out for more resources in exchange for the license" process stop? As the value of free software increases, will the FSF iterate the GPL to ask for more and more consideration for the privilege to license that software? (All in the name of achieving more the linker script is still part of the whole work though - even if that particular element might not be copyrightable in isolation. Likewise, the kernel contains code that is in the public domain - to which copyright protection does not extend either. But you cannot argue that the Tivo 'key' is part of the whole work. It is part of the _hardware_. The Tivo box is a compilation (at most a collection) of multiple works, and allowing the GPL to jump over derivation/modification lines is wrong. The GPLv2 certain doesnt do that land-grab. Ingo -
I have not admitted any such thing. I have said the key and signature do not have separate copyright protection. Variables named "i" in a file are not protected by copyright, but they are very much part of Where in the Tivo hardware is the signing key? There is a related key in the hardware, but that one is not used to generate an integral part of the kernel binary. Michael Poole -
the problem with your argument is that the definition of what constitutes "work" is up to copyright law, _not_ the license writer. I.e. you cannot just cleverly define "source code" to include something unrelated and then pretend that it's all in one work. And that's exactly what the GPLv3 does: it creatively defines the hardware's key into the 'source code' of the software and then asks for that to be provided _not_ because somehow the key derives from the software (it clearly does not), but as a "compensation" for the right to redistribute! I.e. it's trying to extend its scope to some item that is not part of the software. See? Ingo -
Linux is unquestionably a work protected under copyright law. When I compile Linux, copyright law still protects the executable form. This No. The GPL does not care about the hardware's key, as I pointed out in the part of my email that you cut out. The GPL cares about the key used to generate an integral part of the executable form of the GPLed work. The executable does not function properly if it lacks that part. This is exactly the same way in which the GPL cares about the programming instructions in other parts of the source code: if you remove them, the resulting work does something quite different. See? Michael Poole -
it is a false statement on your part that the executable "does not function properly" if it lacks that part. Try it: take out the harddisk from the Tivo (it's a bog standard IDE harddisk), put into a nice Linux PC, mount it, modify a bit in the kernel image header and it will likely still boot just fine on that PC. now if you put the harddisk back into the Tivo, the Tivo's bootloader will refuse to run that modified kernel. So will it (and any Linux bootloader) refuse to load the kernel if you corrupt the compressed format and the gunzip function finds a CRC error. You cannot run arbitrary binaries on hardware without knowing the properties of that hardware. One such property of the hardware might be: "i only run applications that use at most 500 MB of RAM" - because ... the hardware might only have 512 MB of RAM. Another property of the hardware might be: "i will only trust and run applications that match a given signature". Dont buy that hardware if you dont like its inherent limitations! The modification the GPL talks about is about modification of the SOURCE CODE. But if you have a new binary, you have no expectation of being able to run that on a piece of hardware. It might or might not run. (for example if you modified the software to include a 1 GB static array then the software might not work on a system that has only 512 MB of RAM.) go download the Tivo Linux kernel from: http://dynamic.tivo.com/linux/811/linux-2.4.tar.gz modify and build it. Boot it on your general purpose PC. It will quite likely work just fine! Ingo -
Tivo did not program or sell the hard drive to be used in an arbitrary Linux PC. They sold the hard drive to be used in their hardware, with a Linux kernel specifically modified for that. Without the right digital signature, it does not do the same thing: it is *incomplete*. That is eminently a software issue. Hardware limitations -- whether they be RAM size or requirement for a certain digital signature -- are beside the point. The requirement that I "modify a bit in the kernel image header" is also one of the most pathetic cop-outs I have seen. What makes that binary format the preferred form for modification of Linux? Michael Poole -
Ok, try this: take the disk out, remove/replace/modify the signature, put the disk back in, and tell me what it is that fail to run. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
you mean back into the Tivo? That is not support for what you claimed. You claimed the "executable does not function properly" if it lacks that part (and you did not qualify your statement with anything). That was a false statement, because it still works fine in just about any bog-standard PC. A true statement would be: "the modified executable does not function properly _in the Tivo_". It still works fine on a general purpose PC. In fact, you couldnt even modify the binary on the Tivo, because the Tivo is not a general purpose PC, it is a PVR. You'd have to put the disk into a PC to modify the binary. And then you'd have to put it back into the Tivo. So even in this silly example of yours you _already_ have to have a general purpose programmable system where the free software runs fine, and even under your strained and invalid interpretation of the GPLv2, your "rights" to modify the software are very well present on that general purpose system. But you didnt really want to make use of Tivo's free software enhancements, right? Lets face the sad truth: the overwhelming majority of Tivo 'modders' wanted to hack the PVR not to enhance the Tivo, they more likely wanted to watch pay-per-view content without the pay bit and they perhaps wanted to get around service restrictions that the Tivo implements (and through which it funds lower-than-production-cost for the PVR). So the 'rights' you are trying to protect are invented 'rights' of mostly _freeloaders_ in fact. The 'Tivo community' was conjured up after the fact. So even in this supposedly golden and hand-picked DRM example of RMS, the whole story stinks from beginning to end and has all the classic earmarks of detached-from-the-real-world religious extremism in the works ... and the whole effort is totally pointless anyway. Consumers are already voting with their feet against DRM restrictions. So the only DRM victims of the GPLv3 attack measures will be the _good_ uses of DRM. People will be ...
I claimed that. Unless I missed something, Alexandre did not. Ability to run on a standard PC is irrelevant. Tivo distributes the executable for the specific purpose of running on their hardware. Having the signature accepted by the hardware is a critical aspect of the executable. That purpose and function are what make the signature part of the work based on Linux. Courts consider purpose and intent when analyzing actions; except when one has bought the best available legal system, they would not follow your logic. (The role the signature plays in controlling access to a copyrighted work, per DMCA, might also separately identify it as part of the work based on Linux.) If I wished to distribute a kernel with extended functionality from a C file but not the C source files, under your logic I need not give them out -- a user could modify the binary and run it on a general purpose PC. Right? At most it would take clever linker tricks to make the change small enough. As to the suggestion that vendors would use another kernel: I would not mind. A huge fraction of the interesting and useful work in open source kernels happens in Linux (first or only). Using any third party software is a trade-off of what you get versus what you give up. Michael Poole -
I stand """corrected""". It doesn't matter, because the TiVo is where the combination of the executable with the signature shipped, and, see, I didn't talk about modifying the executable, what I wrote about Sony Betamax anyone? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
GLPv2 doesn't: why do you think the digital signature is an integral part of the executable? It can be a totally separate blob, distributed via a separate channel and even stored at a different location than the executable. Does it still look like an integral part of the executable to you then? (unless of course you're trying to argue that the hash itself is a It works just fine given the right environment. The right environment may be some other hardware (without DRM restrictions) or the DRMed device + an authorized digital signature. The digital signature is not part of your executable. Do you honestly believe GPLv2 requires the distributor to provide you with the right environment for your modified copy to "function properly"? I would say it doesn't, but feel free to point me to specific sections which *state* otherwise. AFAICT, GPLv2 is specifically limited to "copying, distribution and modification". How you use (or don't use, or can't use) your modified copy is totally outside its scope. --- fm -
Yes. If I cut a book in half and store the halves separately, does the second half become an independent work? The integral-ness is a function of how the thing is created and how it functions, not how it is stored. If you need part B for part A to execute as intended, then part A is not a complete work in itself. On top of this, in the Tivo case the two are distributed together, and even part of the same file. The GPL does not require a distributor to provide me with any kind of environment. If I get a Tivo kernel image but do not have a Tivo, the GPL does not require anyone to give me hardware. Fortunately, that is not at all my argument. Michael Poole -
Except in this case you're not touching the book at all. If you write a review for a book (much better analogy methinks), then your review is obviously not an integral part of the book even though it's based on its Being an integral part (as in combined or derived work) has nothing to do with usability. There are many other bits and pieces your executable needs in order to function properly (or at all) but that doesn't make your CPU microcode & electricity provider an integral part of the program, does it? Luckily, it doesn't really matter what you or I think that "integral-ness" means, all it matters is how copyright law defines a "derivative work" and whether a cryptographic hash is such a thing. Now It's mere aggregation, but it's totally irrelevant because they could just as easily change their approach. --- fm -
Extremely poor analogy. I do not distribute my review with the book. Someone buying the book is able to use the book just fine (for the purpose for which it was sold) without my review. They need neither my review nor other modifications before the book becomes readable. As Ingo said, you need either the digital signature or other changes No. Those are independent works. They are not distributed to make a certain piece of software function in a particular way or place. The presence of software in a box with CPU microcode is -- at least in No. I explained this before. Try reading the thread and the GPL. I am not sure where people get the (wrong) idea that the GPL only If and when they do, I'll consider the rules that might apply. Until then, it is fairly stupid to try to defend Tivo by saying they *might* do something they currently don't, and if they did, they *might* have a defense that they currently don't. Michael Poole -
But you do (because I say so ; ), and guess what? It makes no Exactly. So what's your difficulty in downloading the Tivo code, reading it and re-using it in your own projects, on your other devices? How is the missing signing key preventing you from doing any of that? Someone buying the book may be free to read it anywhere but if they insist on reading it at your table you may sensibly require they bring a copy of your review with them (to prove their genuine interest ; ). Failure to comply only means they have to read the book someplace else. Can they read the book? Sure. Can they read it at your table? Only if GPLv2 guarantees that the book remains readable. It does not grant you (doesn't even try) the right to execute a modified copy on any particular piece of hardware. Your kernel is perfectly functional on any platform that supports it - it just so happens that the Tivo device does So is a digital signature. Again, are you arguing the digital signature I guess you'll have to explain again because copyright law and its definition of derivative works are the things that make the GPL work: "0.This License applies to any program or other work which contains a notice placed by the copyright holder saying it may be distributed under the terms of this General Public License. The "Program", below, refers to any such program or work, and a "work based on the Program" means either the Program or any derivative work under copyright law". What I can't find though is any reference to "integral parts" or your taken-for-granted right to run a modified copy of the program on the same device used for distribution (or any mention of functionality at all for that matter). Actually: "Activities other than copying, distribution and modification are not covered by this License; they are But you're missing the whole point: the rules are the same, nothing changes! You are drawing an artificial distinction between the two cases and focusing on aggregation, ...
Dude, you're 3 drafts behind. And some people already read GPLv2 like The "compensation" is and has always been "respecting others' freedoms". If the key is used to disrespect others freedoms, as it is by TiVO, then TiVO is failing to keep its part in the deal. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
legal basis, maybe. legality and morality are quite different concepts. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
> Because GPLv2 doesn't enforce limitations on the hardware a GPL'd work ca= IMO this statement expressedly exposes the different viewpoints as used in various factions in this discussion. Without adopting all the details I think I can agree to the above stmt. However I don't agree with the implied msg as I perceive it. In the following I'll try to explain what I mean by the above. I don't know whether what TiVo did actually was allowed by the legal phrases of the GPLv2. I can image it was legally valid but I don't know. But then I'm convinced it was one of the things the inventors of the GPL wanted to make illegal by it -- they may have failed to do so when wording the legal part. I like to remind you of the story with the broken closed source printer driver RMS tried to fix at MIT (if I recall correctly) and the frustration that he couldn't do so that finally made him start the FSF. No customer can fix his TiVo box without the cooperation of the HW vendor. If they refuse there is nothing that can be done. For me this is very much like printer story above. Assuming you (the reader) agree so far: I find it obvious that the GPL was meant to prevent such to be possible. This is what I mean by the "the spirit of the GPL". Living in germany I'm also used to the courts valueing the intention over the exact wording of a contract (a licence after all is a contract). So I _think_ in germany TiVo would have lost a lawsuit if they had tried it. Now for a different PoV: Do I think Tivoisation is bad for the community ? Of course I think it is but your mileage may vary. Anyway, if one considers Tivoisation acceptable then there is no reason to stop using GPLv2. If one wishes to prevent it there are two related questions: =2D does GPLv2 prevent it ? =2D if GPLv2 does not prevent it then how can we change it to achieve that ? To me it seems as if the FSF tends to answer the first question with 'no' and consequently answers the second question with ...
No doubt. However, GPLv2 actually states in clear and concise english that it doesn't cover *anything* but the rights to "copy, distribute and modify" covered works. It actually states that other rights are beyond the scope of the license. That statement, IMHO (and IANAL), obviates any other "intent" the "inventor" of the license may have had by making the scope of the license clear. I own an XBox 360. If it breaks I can't fix it without the cooperation of MS. The fact that a TiVO runs an OS that is licensed under the GPL doesn't change the fact that the situation is *exactly* the same. TiVO breaks? Manufacturer (or someone certified and licensed for the task by the manu) fixes it. XBox breaks? Manufacturer fixes it. My laptop breaks? As long as its under warranty, the manufacturer fixes it *FOR* *FREE* - if it's out of warranty, I pay for the "service" but they still fix it. QED: The "manufacturer must cooperate in or perform the repair" is not some new idea - its actually It might be that you are correct. However, thanks to someone having actually identified the exact scope of the GPLv2 *in* the legally active text of the license the "intent" shouldn't have much weight or bearing. Look at the first sentence of the second paragraph of section 0. "Activities other than copying, distribution and modification are not covered by this License; they And I happen to agree with you. What I disagree with is taking steps to make "bad == illegal". I also have a problem with doing things that force my And neither am I. My whole point in arguing over it has been that, despite what some people want to believe, it isn't violating the GPLv2 in any way, Yes, you have. -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
Surely it's more: bad == go away and don't use future improvements to our software anymore please. ?? *If* you think it's bad (Linus doesn't as far as the kernel goes) then isn't it reasonable to exclude 'bad for the community' from the community? This isn't retroactive - they can continue to use any V2 software they had, they wouldn't be able to use V3 developments. That seems to me to be a very, very reasonable thing to do (and very much *not* bad == illegal IMHO) David PS well, I was just seeing if anyone had fixed my libata/md bug yet but this seemed more interesting. PPS and Tejun has, I'm off... -
I agree. I stated it in the terms I did because Alexandre originally brought "ethics and morality" into the discussion. Not that my word choice is I agree that it is "bad for the community". The impression I've gotten from reading the GPLv3, reading transcripts of interviews with RMS, reading transcripts of interview with Eben Moglen *and* from the FSFLA members participating in this discussion the reason for the anti-tivoization language in GPLv3 isn't "its bad for the community" but "we find it ethically and morally wrong". That being the reason, what they are saying with the language in the GPLv3 is "this is bad/evil and things that are bad/evil should be It is. Consumed a solid six-hour stretch of my day (5PM EDT to 11PM EDT, June Good for you! DRH -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
Well, with the understanding that I don't think that what Tivo did was bad in the first place, let me tackle that question *anyway*. The answer is: Not necessarily. Some people can be "bad" for the community. They may simply be disruptive and not productive at all. They may troll the mailing lists without actually ever doing something good, or they may do other "bad" things. In fact, let's make it *very* specific: let's say that the bad person is a cracker, and specializes in finding security holes, and writing exploits for them, and selling those exploits to spammers. Most of us might agree that that is a "bad" person for the community, no? Now, by your own logic, let's look at what that means for the license. Should we write into our copyright license that you cannot try to find security holes? Would that be a good addition to the GPLv2? Now, I stated that in a way where the answer is obvious: that would be a *horrible* addition to the GPLv2. I think everybody can agree on that. It would be really stupid to say "you cannot look for security holes" just because *some* people who do it are bad. Now, think about that for a moment, and then go back to your question about whether Tivo is bad for the community, and whether being bad for the community should mean that the license should be written to say "go away and don't use future improvements to our software". See where I'm trying to take you? I think that even people who *do* think that what Tivo did was "bad", should think very deeply about the issue whether you should try to lock out "bad uses" in your license. Yes, the answer may be "yes, you should". But I'm arguing that the answer _may_ also be: "No, you shouldn't, becasue it turns out that you might lock out _good_ people too". So in my cracker/spammer example, by trying to lock out the bad people, the obvious (and _stupid_ - don't get me wrong, I'm not at *all* suggesting anything like that should ever be done) license addition ...
in fact there was news in the last week or two about a law in Germany that does exactly this. it outlaws all programs that can be used for hacking systems. David Lang -
I do agree with what you say here. Maybe a summary: Babies, bathwater... My concern is around embedded type systems and maybe even the 'trusted' frameworks etc. I _think_ I can see a completely opensource system that the end user cannot modify _in any way_. Which kinda defeats the point (to me) of opensource. This 5 minute design undoubtedly has flaws but it shows a direction: A basically standard 'De11' PC with some flash. A Tivoised boot system so only signed kernels boot. A modified kernel that only runs (FOSS) executables whose signed hash lives in the flash. Do we (you) _want_ to prevent this? Do we trust in 'the market' to prevent this? Do we use license tools? David -
How hard would it be to reprogramm the flash? -- Krzysztof Halasa -
The flash contains hashes signed by the companies private key. The kernel contains the public key. It can decrypt the hashes but the private key isn't available to encrypt them. So although you can put a new application onto the system, you can't create a signed hash to write to the flash. The kernel only runs the executable if the hash is valid. You can re-write the kernel to avoid this check - but the hardware is Tivoised - so you can't run it. I am not suggesting the kernel should go down the GPLV2 route - I am wondering if this is a viable scenario or one of Schneiers' "movie-plot" threats :) David -
I read it: the flash contains everything from the bootloader to the kernel and file system. The bootloader contains the public key and checks if the kernel/fs are ok. That includes calculating hashes and checking signatures. No encryption/decryption there at all. Right? Then how hard would it be to reprogram the flash, to get rid of all this crap? Or to just put your public key there. Do they at least use BGA type of flash chips so you can't attach a clip and have to use something more demanding? -- Krzysztof Halasa -
Stop trying to technically crack my 5-minute fag-packet design - that's easy and boring :) Tivo have solved this problem - use their solution - but do it on something more general purpose. Help fix it - the point is more "is this feasible". And if it is, "does it matter?" David -
Umm. It may well have been meant by *rms*. But your argument fatally falls
Ehh. The intent that matters is not the intent of the person who authored
the license, but the intent of the person who *chose* the license.
In other words, rms has *no* input on the kernel.
What matters is *my* intent in *choosing* the GPLv2, not *his* intent in
writing it.
But to make it even less relevant: intent really only legally matters when
the legal issues are unclear.
Well, I think it's fairly unquestionable that the GPLv3 does prevent it.
So your second question isn't even really interesting. We know the answer.
Yes, I do agree with that reasoning, but there are *other*, and more
direct, reasons than just the FSF's answer to say that the answer to your
first question is "no".
The fact is, plain reading of the license (which *always* takes precedence
over "intent", even in Europe) simply doesn't make what Tivo did illegal.
You literally have to read the GPLv2 in ways that are obviously not true
to get to any other situation.
For example, Alexandre made the same two mistakes over and over in his
reading when he tried to argue that the GPLv2 disallows what Tivo did:
(a) The right to modify means "modify in place"
This was a point that Alexandre (and others) have tried to make, but
it really is *not* supported by any reality.
First off, the GPLv2 simply never *ever* says "in place". That
wording (or anything equivalent) simply does not exist! So you really
have to add it by "reading" it some special way, and quite frankly,
no such reading is sensible.
I can logically *prove* that such a reading is not sensible by the
two examples I already made clear to Alexandre multiple times:
- Red Hat sends out DVD's with GPL'd software, and thus
distributes copies that CANNOT be "modified in place". So
thinking that "modified in place" is made illegal by the GPLv2
is simply untenable, unless you think ...You're mixing two separate issues: 1- does GPLv3 change the spirit of the GPL? 2- is GPLv3 better than GPLv2 for Linux? The answers may be different, and the reason I got into this debate was to set the record straight on 1. As the discussion evolved (if developing into a flamewar can be characterized as evolving ;-), I realized the motivations for preferring v2 over v3 were not clear to me (and they still appear contradictory to me), so I started investigating that, which is indeed 2., but is not about 3: When you leave an essential portion of the reasoning out, which you repeatedly did, this conclusion is obvious. But it's also obviously Isn't a restriction on in-place modification a further restriction on the permission to modify granted by the license? A further restriction that is not permitted by the license? Again, this is not about ROM, CD-ROMs and other unmodifiable media. In this case, the distributor is not imposing this restriction, it's not selecting the media with the strict purpose of forbidding modification. It doesn't retain the ability to modify without failing to pass it on. This is the key distiction that you repeatedly I agree, and I don't think I've ever claimed otherwise. It's rights as far as the software is concerned, and even this might be pushing it a bit too far. That's why the spirit gives the intuition, but the legal terms are precise in turning that into "no further restrictions", as I'd already explained long before you did. But then, again, the license grants the right to modify, and prohibits further restrictions to it, so I claim that saying "you can modify, just not in place, because I won't let you do it" (rather than because it's impossible), that's a further restriction of a freedom granted by the license, which turns into a license violation. Now we can turn into the debate on whether replacing is modifying, and the conclusion is quite possibly that, in legal terms, it isn't. I don't care. I'm not here to ...
While I raised this argument on the Lunix kernel ML it was not meant to be valid specifically for it. My observation in this thread is that almost everybody discusses different aspects of the same thing and everybody is somehow right. I was trying to "go back to start" and have the look at the overall picture which in this case for me is the question what the GPL's spirit is. Whether and which of it _you_ intended to adopt for the kernel I had That seems to imply that we have to deal with myriads of intended meanings, namely those of all who contributed to the kernel. I'm pretty sure I don't wish to walk that road. If you want to we'll have I beg to differ. By adopting _his_ license you adopted his view. If you don't like that then choose a different license (which obviously you are free to do). It's just not feaseable to have something like "my GPL means a different My second question leaves out whether or not GPLv3 is an acceptable answer to my second question. While the FSF says it is it is by no means clear that I will agree -- all I wanted to do is present the situation as I see i= I disagree and I don't see that plain reading of the license is that obvious w/r to the SHA1 key because from a certain perspective said key is required to create a working modification which I'm entitled to under the GPLv2. I also agree that your perspective has merrit too. I'm simply not sure which of the above is "correct" (as in agreeable from a judge's PoV). Based on that I disagree with your above stmt, at least I don't think your implication every other reading being outright wrong is false. This thread IMO clearly shows that apparently it isn't that clear -- far I object against the word obvious as an obsolute measure. You have all right to consider it obvious from your PoV. My PoV may differ and I strongly claim I do not agree with everything Alexandre wrote but I do agree with some I tend to agree and I didn't like it when it was brought up. But IMO ...
ianal, but fortunately that's not what the law is. The license says what it says, and that is what controls. The intent of the author (of Linus and other copyright holders) is a secondary source of information /if and only if/ any ambiguity of meaning arises (as determined by a judge, not by you or me). But the opinion and intent of RMS (unless adopted by Linus) is quite immaterial. ( there is a legalistic special-case. If any dispute arises over what license the COPYING file in Linux itself (and only that file) is under, then the intent of RMS matters too, but only for that limited matter for that single file. Btw., the COPYING file itself is not licensed under the GPL. ) Ingo -
I agree with the "/if and only if/ any ambiguity of meaning arises" part. I'm sorry I didn't make that clear before. However if that situation arises (i.e. the judge decides there is an ambiguity) then as far as my experience tells me it is the intention of the author (RMS et al in this case) that counts. But I erred before... Best wishes, Michael =2D-=20 Technosis GmbH, Gesch=E4ftsf=FChrer: Michael Gerdau, Tobias Dittmar Sitz Hamburg; HRB 89145 Amtsgericht Hamburg Vote against SPAM - see http://www.politik-digital.de/spam/ Michael Gerdau email: mgd@technosis.de GPG-keys available on request or at public keyserver
I doubt this. In a situation like that the intent of the licensor is what matter, not the intent of the original author of the license. -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
Each person that contributed code to the linux kernel *CAN* have their own interpretations of the GPLv2 as it *APPLIES* to their code. The interpretation that matters when talking about the kernel, a a whole, is Wrong. If I adopt the GPL it will be because of the *interpretation* I give it when reading it. And because *I* am the one then granting the license, it is A plain reading of the license doesn't entitle you to create a "working modification". See the disclaimers of warranty and guarantee - sections 11 and 12 of the GPLv2. They mean that the person *GRANTING* the license doesn't have to make sure that the program will be useful for your purposes, is modifiable to fit your purposes or will even *NOT* damage your hardware when it runs. "THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU." In other words there is no guarantee in the GPLv2 that you will be able to create a working modification. The *ONLY* guarantee that exists, in regards I agree that "too many intelligent people disagree". But this is just human nature. That people believe it says one thing when it doesn't is analogous to But the "modify in place" argument is part of the reasoning behind the claim But the "within reason" isn't there. That some people have inferred that term And here is where it becomes obvious that an inference is needed. *BUT* the argument was that the preamble, as written, states the intent of the license. This is the way it should be. However, the letter of the contract, in this He doesn't. But the fact is that he has claimed that the "give you all the Agreed. -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
btw., still ianal, but the GPLv2 is not a "contract" but a "pure copyright license". A contract, almost by definition is a restriction of rights in exchange for consideration - while if you accept the license of a GPLv2-ed work this act only gives rights that you did not have before. Furthermore when you get source code of free software then there is no "meeting of minds" needed for you to accept the GPL's conditions, and only the letter of the license (and, in case of any ambiguities, the intent of the author of the code) matters to the interpretation of the license, not the intent of the recipient. (while in contract cases both the meeting of minds is needed and the intent and understanding of both parties matters to the interpretation of the contract.) Ingo -
I've been told by several independent sources that it really doesn't matter. The "pure license" argument was born largely for silly reasons: people claimed (a _loong_ time ago) that the GPL wasn't enforceable in the US because in order to be enforceable, something of value has to change hands (in the US, for example, it would be common to "sell" something for a nominal sum of $1 USD rather than to give it outright, to "seal the deal" and make it irrevocable). That's generally considered a specious argument, apparently. In most jurisdictions in the US, a license and a contract are judged to be legally exactly the same thing, and if you don't follow the GPL and have no other contract to show for it, you're in violation of federal copyright law, so whether it is a license or a contract really doesn't matter. So it's true: the GPL just gives you rights, and without it you have no rights (other than fair use ones etc), and blah blah. But the distinction between "license" vs "contract" really isn't a very important one in any I do agree that you can probably use this to say that the intent of the copyright has a stronger position, and that his "intent" thus matters more. But I suspect that the "intent" angle is fairly weak legally to begin with, and if you cannot show that the intent was mutual, it's probably weaker still. So yeah, the intent of the copyright owner arguably might matter more, but quite frankly, I suspect everbody is better off not worrying so much about "intent", and worrying more about the "terms and conditions" part. (I've said several times that intent _matters_, I just don't want people to think that it matters a whole lot). What is pretty clear, though, is that the intent of a third party in the license/cotnract matters not at all. In the case of the kernel, the FSF being such a third party. Linus -
Er, copyright law is federal, contract law is generally state level? So not only does contract law vary a lot more by jurisdiction, but it's enforced by different courts than suits over copyright? (You'll notice the GPL doesn't say which state law holds sway. If it was a contract this would be kind of important.) Also, in addition to the "exchange of value" bit there's "privity of contract" and "informed consent" when dealing with contract, which are cans of worms which can be avoided by Not Going There (tm)... (These were largeish issues in the SCO vs Novell case, involving lots of motions in Utah detailed blow-by-blow on Groklaw...) Rob -- "One of my most productive days was throwing away 1000 lines of code." - Ken Thompson. -
That seems to be a special property of the US legal system. At least I'm not aware of this or a similar distinction in e.g. germany (or most parts of europe AFAIK). Best, Michael =2D-=20 Technosis GmbH, Gesch=E4ftsf=FChrer: Michael Gerdau, Tobias Dittmar Sitz Hamburg; HRB 89145 Amtsgericht Hamburg Vote against SPAM - see http://www.politik-digital.de/spam/ Michael Gerdau email: mgd@technosis.de GPG-keys available on request or at public keyserver
yeah - and from everything i know about this subject the distinction between contract and license is small and more of a technicality - but still, it's a nice touch that the "pure license" argument that the FSF has advanced for a long time (and which it is now more silent about, given the GPLv3's not so pure structure) neatly defeats the common argument: "but, but, when i received the Tivo with GPL-ed software on it the GPLv2 was not intended to be like that, there is a right to run yeah. But the argument goes a bit further: people who chose to _license_ the kernel (by receiving a Tivo for example and downloading its kernel source) claim that _their_ interpretation of the GPL is that of the FSF's and that Tivo ought to follow it. The whole "Tivo is cheating the GPL deal with the end users" line of PR. As far as license interpretation goes there is _no end user deal_ and the 'end user' does not even play in terms of intent - only if she choses to be an active member of the community. That's why i think it's better to talk about a license than a contract. (even though legally, at least in the US, the two are quite close to each other.) so a 'pure copyright license' stresses the point even more that you only really count in the ecosystem if you contribute in one way or another. The system should be and _is_ assymetric towards the actual black letter text of the license and, as a second layer, towards the intent of the people who actually produced this 1+ billion lines of code, documentation, bugzilla entries and other nice works. And that is a thing the FSF is missing sometimes i believe - the "listen to _all_ the people who enabled this cool stuff" part. Ingo -
And this is the beauty of a multi-author project. Even if some authors think that the license permits something, if any of them understands it doesn't, he can try to enforce that WRT his own contributions. So those exploiting the gray areas of the license can still get caught. In Brazil, this is kind of contract/license is called a beneficial That's correct, but with a catch: since the contract or license is chosen by the licensor, in case of ambiguity in the terms, many courts will interpret it in a way that privileges the licensee, regardless of the fact that copyright licenses are to be interpreted restrictively (at least in Brazilian law). And IANAL ;-) -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
--- Hmm. In such a suit, however, the user would not be "the licensee" and would not be a party to the suit - some author would be the plaintiff and would be suing someone for doing something in violation of the license that author granted - that is, the *defendant* would be the licensee who would get the benefit of the doubt... scott -
Yes. And so justice is made. Licensor gets to pick the license, licensee gets the benefit of the doubt. What's the 'however' about? Was this not obvious? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
--- Sorry - I thought you were saying ambiguity would be resolved in favor of the user. If you meant in favor of the licensee (regardless of that limiting the user's rights), then I agree. scott -
I'm sorry, but that's simply bullshit. The GPLv2 does not state that you have to become a slave of rms and follow him in all things, and agree with him. Really. You must have read some other (perhaps unreleased early draft?) version. The GPLv2 says what it says. Not what you (or rms) *wished* it says. You don't enter into contracts and licenses based on wishes and intents. Your view is not relevant. The fact that the "preamble" is not the "conditions" is what's relevant. The preamble is explicitly stated to be *different* from the exact conditions. It's not the real "terms of copying". It's there to explain, it's not there to *be* the license. It's explanatory, but the wording that actually *matters* is the "terms and conditions". And the fact that *you* can mentally add words to it when you read the license (adding a "within reason") has absolutely no relevance what-so-ever. Linus -
Technically what they're holding back is _trademark_ rights, which are a different area of IP law and not addressed by the GPL. (I know you know this, but just for the record...) The five main areas of IP law as I understand them are copyright, patent, trademark, contract, and trade secret. Each of which is a different animal with a different legal foundation and different enabling legislation. The GPL is a copyright license with some language about patents. It is not based on contract law (although that's a common misperception that Lawrence Lessig and Eben Moglen have spent some effort debunking), and doesn't even mention trademarks. So Red Hat isn't saying "you can use some of our copyrights but not others", last I heard all of their copyrights are licensed GPLv2 as a matter of corporate policy. What you can't use is their trademarked name or logo, because they are explicitly refusing to license the trademarks to third parties. And under GPLv2, this is allowed. Rob -- "One of my most productive days was throwing away 1000 lines of code." - Ken Thompson. -
No, technically Red Hat really *does* have copyrights of their own. Red Hat owns the "compilation copyright" on their distribution. That means, for example, that even if they have _only_ open source programs on their DVD image, you still are not necessarily able (without their permission) to set up a "cheap-cd's" kind of operation, and sell their CD-ROM/DVD images for a lower price. So yes, they do own the Red Hat trademark too, but they fundamentally do own copyrights over and beyond those of the individual programs they distribute! Now, I think it so happens that the RHEL DVD's contains other programs than just open source, and that you couldn't legally copy them *anyway*, but that's a different issue. Also, happily, a lot of vendors do not *want* to exercise their copyright in the compilation, so you can go to cheapbytes.com, and you'll find Fedora CD's, OpenSuSE CD's, Ubuntu CD's, etc, and as far as I know, they're all perfectly legal. Exactly because open-source vendors usually don't want to look nasty by limiting the compilation, when they can't I'd not put contract there, but fair enough. But what I was really trying to point out is that there are many different "levels" of copyright. So you can own a "copyright in the compilation" - which just means that you own the details of how you set it all together - _without_ actually necessarily owning the copyrights in any of the individual packages (although you obviously have to have a license to _make_ a compilation of them - but the GPLv2 is one such license). Linus -
They do have copyrights. They license them under the GPL, and afterwards they I agree that they have this right, but that wasn't the rationale they gave in the cease and desist letters they sent out in 2001. Those said it was ok to redistribute, but you can't use their trademarks to promote it when you did so: http://www.newsforge.com/article.pl?sid=01/12/10/2014239&mode=thread [Rummages around for their current policy statement...] The restriction is embodied in their "trademark guidelines and policies": http://www.redhat.com/about/companyprofile/trademark/ If you open the PDF: http://www.redhat.com/f/pdf/corp/RH-3573_284204_TM_Gd.pdf That seems to say that their compilation copyright is also licensed under the I agree they claim compilation copyrights. But they seem to have licensed their compilation copyrights under GPLv2. If they're including GPLed works in the compilation, this may actually be a requirement. (Lawyers would happily fight over this issue for months: asserting a copyright over the aggregation takes the "mere" out of it, don't you think? Is a compilation a derived work of the components that were compiled, at least for the purposes of GPLv2? Can it be "mere aggregation" if you're enforcing a copyright on that aggregation? Does it not then become a larger work with GPL components? I dunno.) You'll notice that back under Bob Young, Red Hat carefully didn't go there, by licensing the compilation GPLv2 and segregating incompatibly licensed content to a separate CD. After the IPO, he retired and different management took over, and introduced Red Hat Enterprise to eat Sun's market[1], and who knows what they're thinking now? I suspect their lawyers still want the GPL-incompatible stuff on a separate CD so they can sleep at night. (Either that, or they just don't assert a compilation copyright, but why give I haven't noticed any specific non-GPL packages removed from Centos. Buried down in their FAQ they say ...
Correct - all the vendors face the problem that there are people out there who want to try and pass crap on using someone elses good name whether its toothpaste, perfume or software (or in the case of RHEL software/services/support bundles) The Fedora mark is thus used to make sure that if you get a Fedora CD, it actually has -Fedora- on it etc. -
This is perhaps the part that's the most interesting. For the very small number of people that _do_ want to change these things (usually at the expense of a voided warranty, in the consumer device case), there's always a way to make these changes, even if you must resort to hardware hacking. Trying to mandate this sort of functionality in the license might make it easier for a few people to get their code loaded, but the vast majority of users have zero interest in anything like this. I don't see how you can claim that the vendor is infringing on your freedom, _you_ made the decision to go out and buy the product knowing that the vendor wasn't going to go out of their way to help you hack the device. In many cases the vendor doesn't even have the option (802.11b channels and certification come to mind, GSM, etc.) of opening things up to the end user, and making changes to the license isn't going to magically change any of this. If you don't like what the vendor has done with the product, you have the freedom to not support the vendor, and to try and encourage people to follow suit. As an example, I simply opted not to buy a tivo since I wasn't able to do what I wanted with it out of the box, rather than opting to rant about it (or coin an idiotic buzzword) much to the dismay of every other person on a mailing list. This was neither something I lost a great deal of sleep over, nor did I at any time feel like my freedom was being eroded. True story. If the vendor's bottom line is measurably impacted, they may even reevaluate their position on supporting device hacking, but it's certainly not going to be through draconian licensing that vendors suddenly decide to play nice. There were certainly enough vendors that followed the letter of the GPLv2 without following the spirit of the license, with varied benefit (especially with consumer device vendors). Imposing additional constraints under the guise of the FSF's current version of "freedom" isn't going to get these sorts of ...
I don't feel this is a very conclusive argument. How many computer users do want to change their OS? I mean not only want to= =20 change the OS in the sense of "apply patches released by Microsoft", but on= =20 their own? Many typical computer users ask for help to "fix their computer"= =20 when turning it off and on again already "fixes it". They would never ever= =20 change the source code of their OS even if they technically could do it -=20 they are not programmers. However, if there is the technical possibility to change the firmware of an= =20 appliance, somebody does it, and often mere users upload these changes to=20 their own device (like the OpenWRT stuff). Let me give one example: My parents own a DVB-T DVR. It was a cheap one, an= d=20 it was cheap because the software is lackluster. Unfortunately it isn't=20 free. Many users of this device complain to the manufacturer about the=20 stability and quality of the software, but with no avail - there haven't=20 been any updates in the last two years. I suppose I would be able to fix=20 the problem, most other users probably wouldn't (and my parents neither).=20 But if I did fix the problem, and provided them with an updated firmware,=20 they would install it on their device. That's the "help your neighbour" right in the GNU manifesto. It's as=20 important as the "help yourself" right, maybe even more. It was the=20 original motivation of RMS to make free software - the frustration of not=20 being able to help his neighbours. He had an NDA to help himself. What people want is software that works. If the firmware of your microwave= =20 or DVR works, you don't care so much if it is free or not. You only care if= =20 it doesn't work, and you feel the urge to fix it (and turning it off and on= =20 again doesn't fix it). That's why people complain loud about ATI drivers=20 not being open, and don't care that much about the Nvidia driver, which is= =20 just as closed, but works. =2D-=20 Bernd Paysan "If ...
But I also made this decision fully aware that the software included in the package was published under a license that said I was entitled to modify it. More than once I purchased a device that claimed to have GNU/Linux software on it, only to find out that I couldn't use the freedoms, because the distributor was infringing the license in Sure. But wouldn't it be nice if the copyright holder could also help in this effort? It doesn't mean the copyright holder has to: s/he can always grant an additional permission, or simply refrain from enforcing this provision of the license. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
And you certainly are free to do so. The vendor ships the product with the binaries, and you get the source as a result. You can in turn modify that source and do whatever you like with it. If the vendor is more proactive, they may have even tried to get all of their changes merged by the time the product hit the market, so they wouldn't be sitting on anything "special" anyways. This however has nothing to do with your ability to apply those changes to the _hardware_. If the vendor doesn't want to, or is unable to support third-party modifications on their product, they have the basic right to make that decision, as you have the basic right not to buy the product if In this example, at _no time_ did the vendor infringe on the license. They haven't given you an easy way to change the hardware, but they're completely compliant both in terms of the letter and the spirit (depending on how they work with the community) of the license. If you're trying to pretend that GPLv2 had _anything_ to say about hardware, you'd be wrong. In such a situation, there'd hardly be a "need" (as you seem to see it) for GPLv3 at all. If you think this bizarre coupling of the hardware/software paradigm is in any way constructive, you're of course welcome to use the GPLv3, but this does not retroactively change the terms of the GPLv2 simply because you saw this as an area that was apparently "lacking". And on the other hand, you're more than welcome to dual-license all of your kernel changes under v2/v3 if you really feel that that's the best way to go, just as I'm welcome to print out and burn the GPLv3 as a symbolic gesture. Simply because some folks have no intention of ever supporting v3 doesn't stop you from using it on any of your own changes. -
Pure GPLv2. Userland exception? Never heard of. Module exception? Perhaps you mean "interpretation"? -- Krzysztof Halasa -
since the latest draft of the GPLv3 now discriminates against some uses (industrial vs commercial I think are the terms used) does it even qualify as a Open Source lincense anymore by the OSI terms? David Lang -
A "User Product" is either (1) a "consumer product," which means any tangible personal property which is normally used for personal, family, or household purposes, or (2) anything designed or sold for The definition is about the hardware, not the software, so it may still qualify. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Sure it does; you received a program (the kernel) and you can modify it. You also received hardware; they don't support modification of that. Nowhere in the license does it say they have to, because the license only covers the program. Or are you claiming that putting software on hardware makes the result a derivative work? I think it falls under the "mere aggregation" clause. What if TiVo had put the kernel in a burned-in ROM (not flash, or on a flash ROM with no provision for reprogramming it)? Would that also violate the "spirit" of the GPL? Must any device that wishes to include GPL code include additional hardware to support replacing that code (even if that hardware is otherwise superfluous)? -
This is an area the GPLv3 tries to clarify and for good reason. Of course these days in the US someone would probably sue arguing that a ROM is protection scheme ;) Alan -
As a PS to the GPL3 comment here is the basic difference ROM - I can't modify the code on the device The creator can't modify the code further on the device Tivo - I can't modify the code on the device The owner can modify the code One is an implicit limitation of the hardware (just like I can't run openoffice on a 4MB PC even though the license gives me the right to try), the other is an artificial restriction. One case is witholding freedom in the GPL sense by one party while keeping it themselves, the other is a limitation of the system inevitably imposed on everyone. -
I've been following this discussion and I find this interesting.
Consider these two cases:
1.) I ship the device back to the manufacturer, they replace the ROM,
and ship it back to me.
2.) I ship the device back to the manufacturer, they load new code
into it, and ship it back to me.
How do these two differ? Or is it now just a question of the ROM
being in a socket? I can't see how the technicalities of how the
hardware is constructed can change the legality of the software.
--
Dan
-
At first glance I think a construct where the manufacturer is obliged to load _MY_ modified software in a timely fashion and at a reasonable price into the device would fit my understanding of the GPL's spirit though this leaves room for the definition of timely... Best, Michael =2D-=20 Technosis GmbH, Gesch=E4ftsf=FChrer: Michael Gerdau, Tobias Dittmar Sitz Hamburg; HRB 89145 Amtsgericht Hamburg Vote against SPAM - see http://www.politik-digital.de/spam/ Michael Gerdau email: mgd@technosis.de GPG-keys available on request or at public keyserver
In the replace/reflash the ROM case its about access to the righ tools - I could do it myself, send it to another company to load my code etc. In the Tivo case its about one company having the ability to make such mods and blocking others from doing so. -
I don't see that they differ. If the software can be replaced, the manufacturer ought to tell you how to do it. It doesn't have to do it for you, it doesn't have to give you the hardware tools needed to do it, but if you're not able to start from the source code and the information provided by the manufacturer and get to a modified version of the software on the device, while the manufacturer could do it, then the manufacturer is locking you in, and therefore you're not free. This is a clear violation of the spirit of the license, even if the legalese might make room for some such misbehavior. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Tivo gets sick of the endless flamewars on lkml, signs a copy of QNX, pushes it out to the hardware. No more Linux on Tivo. You also can't replace that but Tivo can. As I see it the two are completely orthagonal: a) Can anyone but the manufacturer upload new software into a a device without taking extreme measures (soldering a new public-key-containing-chip onto the board) b) Is the software currently installed on a device licenced under a rule which requires the distributor to also distribute source code upon request. Now I think it would reasonable to ask that the source code be able to be built by [same compiler, same flags, same ...] to produce an identical binary to the one running on the device so you can confirm that it's exactly the same code. That's separate from being able to upload a changed binary. Bron. -
What do we lose? Do we actually get any benefit whatsoever from TiVO's choice of Linux as the kernel for its device? Do TiVO customers lose anything from the change from one non-Free software to another? (the Linux binary, as shipped in the TiVO, has become non-Free) -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Sure, if they make any changes or fixes to Linux. Other than that, only the same benefit that Microsoft get from Windows piracy - TiVo employees become familiar with Linux and are more likely to use it and maybe contribute more in another job later. What we don't get is TiVo having a better kernel than everyone else because they've put some work into extending it without giving that work back. I see stuff in arch/powerpc/kernel/ which is Copyright "TiVo, Inc" and more recent stuff in usb/net/asix.c and usb/net/mcs7830.c which is more than I've ever contributed to the kernel, despite making extensive use and even selling services where I ran servers with Linux on them but didn't allow my customers to change the kernel Not particularly, no. Other than maybe some nice features that TiVo gains from being able to use Linux. Bron. -
Now, what if TiVO actually permitted all of its customers to make changes or fixes to Linux, and become familiar with it and become more likely to use it and maybe contribute more later? Would we lose more or gain more? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
a) there's nothing that prevents a Tivo user from changing or fixing Linux completely outside of the Tivo b) the 'interesting' bits that someone would modify the Tivo to change *aren't actually the bits that everyone is kvetching about here* Bill -
But how about inside the TiVO, so as to use Linux and the rest of the GNU/Linux distro put in there for an even better DVR experience? Sure, this might still be accomplished on another hardware platform. But the TiVO already has all the hardware there, and you already have all the software ready to work on it. Except that you can't change it. You'd have to waste time and money just to get to the same status on another hardware platform. What do we gain? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Nothing. But that's not the terms it was licensed under, and no matter what someone may claim about the *spirit* of the license, adding clauses that restrict how you can deploy GPL software for use is a fundamental enough change to the practical aspect of the license that it's no wonder that people will choose not to use it. If the designers of the license are more interested in vendettas against those using the software in a way they didn't see beforehand (come on, explicitly trying to define 'consumer product'?) in order to accomplish pyrrhic victories (people moving to other platforms instead of using your newly licensed code), that's fine, it's their choice. But not everyone will want to follow that choice. Bill -
Do they contribute back any code that makes Linux better? If Tivo doesn't, what about other vendors who may be in a similar situation? Bernd -
If I take the software I received, build it and install it on the same hardware, it won't run. Something is missing in the source code I received, I guess.. If I make changes to the source code, build it, and install it on that same computer, it won't run. How is that being able to modify *that* copy of the program? If TiVo makes the same changes, builds tehm, and installs it on my computer, it will run just fine. How are they passing on the right They don't have to support them. They don't have to help me if it breaks. But if they can do it and I can't, they're failing to comply They can't distribute the program while imposing restrictions on I tend to agree, in this particular case, but IANAL. I don't rule out Then they wouldn't have the ability to change it any more, so there wouldn't be such a right to pass on to the users. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
And they give you the same right that they had, which is obtain free software that you can modify and redistribute. There's nothing in there that says they should give you the tools they used after they received the software, which It does, can't you modify their kernel source? Where does it say you should be The only fear that I have with the whole Tivo saga, is that companies like Dell can use the same thing to say: "Our hardware will only run Company's X distribution of Linux". Do we just hope users won't buy those Dell machines, or do you modify your software license to force Dell to allow custom distributions to run on their machines? Then where do we draw the line of "Software Licenses". -
Wrong, the GPLv2 says:
"For an executable work, complete source code means all the source code
for all modules it contains, plus any associated interface definition
files, plus the scripts used to control compilation and installation of
the executable."
The question is whether this includes private keys.
Different people have different opinions regarding this issue.
cu
Adrian
--
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed
-
IANAL so I won't comment on the legal aspects of TiVo's doing. However it definitely is against _MY_ understanding of the spirit of the GPL. At least to me that's quite obvious. I'm sure you all know the story of the printer driver RMS couldn't fix that reportedly made him start the whole FSF business. Looking at what TiVo did I realize glaring similarities. <disclaimer> I'm in no way related with the FSF. I hereby state I'm not parroting anyone's else position but have come to this conclusion solely on my own. Come on! The whole idea of software is to have it run on some HW. Why would I want to change it in the first place if I can't run it ? If what they did is actually allowed by the wording of the legal phrases of the GPLv2 then that IMO is a loophole w/r to the spirit (as I understand) Would not such a restriction voilate the spirit of the GPL ? Anyway, my simplistic view is: Once it is under the GPL I could change it and actually make the changes work as I see fit. That's what I think my freedom as of the GPL is about. Now all that needs to be done is make sure the legal phrases are such that they convince the judges they actually mean this in court too. Best wishes, Michael =2D-=20 Technosis GmbH, Gesch=E4ftsf=FChrer: Michael Gerdau, Tobias Dittmar Sitz Hamburg; HRB 89145 Amtsgericht Hamburg Vote against SPAM - see http://www.politik-digital.de/spam/ Michael Gerdau email: mgd@technosis.de GPG-keys available on request or at public keyserver
=46orgive my poor mastery of the english language and me letting slip this inconsistency. The first sentence you cited was a general remark IMO valid outside of this context and possibly ill placed as it was. The second sentence pertains the key msg I was trying to deliver and apparently I did a poor job in phrasing it so let me redo it: Why would I want to change the SW targetted for some HW if I can't run the changed version on said HW ? [note that for the TiVo case I possibly would not own or be able to own similar HW being able to run my modified SW; so even some HW would not be triggered either] ^^^^ Remember I'm discussing my understanding of the spirit of the GPL, not whether the legal part actually does give me that right enforceable in court. Here is another stmt which is valid outside of this context AFAIAC: If the GPLv2 does not legally give me the right that I think its spirit gives me then the legal phrases should be changed to achieve that. Whether or not others share my view of what the spirit of the GPL implies is completely theirs to decide and if they differ they likely won't agree on my previous stmt either. Fine with me. And this leads to another observation: IMO this thread is partly fueled by a fundamental mixing of PoVs. Some argue based on their perceived view of the spirit of the GPL and some based on the actual legal phrases in GPLv2 and GPLv3 and whether or how they reflect the perceived spirit. Best wishes, Michael =2D-=20 Technosis GmbH, Gesch=E4ftsf=FChrer: Michael Gerdau, Tobias Dittmar Sitz Hamburg; HRB 89145 Amtsgericht Hamburg Vote against SPAM - see http://www.politik-digital.de/spam/ Michael Gerdau email: mgd@technosis.de GPG-keys available on request or at public keyserver
Maybe this quote will summarize the situation: Judith: [on Stan's desire to be a mother] Here! I've got an idea: Suppose you agree that he can't actually have babies, not having a womb - which is nobody's fault, not even the Romans' - but that he can have the *right* to have babies. Francis: Good idea, Judith. We shall fight the oppressors for your right to have babies, brother... sister, sorry. Reg: What's the *point*? Francis: What? Reg: What's the point of fighting for his right to have babies, when he can't have babies? Francis: It is symbolic of our struggle against oppression. Reg: It's symbolic of his struggle against reality. -- Monty Python's "Life of Brian"
Can they modify the software in their device? It's not the kernel source. That's not where the TiVo anti-tampering machinery blocks modifications. It's about that copy of the kernel that ships in the device in object code. That's the one that TiVo customers ought to be entitled to Where it says that you should pass on all the rights that you have. While TiVo retains the ability to replace, upgrade, fix, break or make any other change in the GPLed software in the device, it ought to pass it on to its customers. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
The GPLv2 makes no real provision for *DIRECTLY* modifying object code. What provisions the GPLv2 has apply to the source code. And no, the end user *SHOULD* *NOT* be entitled to run whatever kernel they like on a TiVO. It was designed with the "install new kernel" functionality so that the TiVO corporation could update the kernel running on the hardware when security problems came up, when bugs were fixed or even when the new It *DOES* *NOT* say "All rights that you have". It says "All rights that are granted you by this license". If every piece of software released under the GPL had *ALL* rights passed on, then *ANYONE* could do the "I'm granting company X the right to use this software outside the GPL for $50,000USD." instead of just the *creator* of the software. DRH -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
You may be right. The spirit says it should, but the legalese may Sure. And that's not what I'm talking about. What I'm talking about is being able to replace, upgrade, fix, tweak, hack, and otherwise modify the program on the machine in the same way I.e., it was designed such that TiVo could modify the installed kernel, but the user couldn't. That's an outright violation of the I suggest you to reboot into memtest ;-) The preamble of GPLv2 says: For example, if you distribute copies of such a program, whether gratis or for a fee, you must give the recipients all the rights that you have. The requirement above applies to licensees, not to the licensor. The licensor doesn't have to pass on all the rights s/he has, s/he only decides to respect the licensee's freedoms, conditioned to the respect of others' freedoms by means of passing on all rights the licensee has over the software. Arguably, one could use this argument to state that any authors of derived works ought to pass on the right to choose the license for the derived work under the GPL, but since (a) the above is not in the legal terms, and (b) the downstream recipients would be bound by the terms of the GPL anyway, and that requires the use of the GPL itself, this would make no difference whatsoever. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
So if I am a sole author of a program and I chose to distribute it under GPL then all recepients will get _all_ my rights, including right to re-license the program under BSD or a proprietory license? Yeah, riiight... Thankfully it is just preamble and not the actual license text. -- Dmitry -
then you're not a licensee, you're a licensor, and these terms don't apply to you. Already covered upthread BTW. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Heh. When you change a GPLed program and pass your changes you are the licensor for the new code. You still have a right and license pieces of the code you wrote under different license but you do not pass that right to recepient of modified work. -- Dmitry -
You are the author of the change, and you can license them however you like. Your change itself is not bound by the terms of the GPL, it is only if it is a derived work of the GPLed work. If your change is not a derived work, you're not bound by the terms of the GPL as far as the change is concerned, so the GPL has no say whatsoever as to how you must release it. If you choose the GPL, then you're a licensor, and the requirements to pass on all the rights you have do not apply. If it *is* a derived work, then you're constrained by the terms of the license, and you can only distribute it under the same license. You don't have a right to offer it under a different license in the first place, so you can't pass this right on. Derived work or not, when you combine that change with the program, then you're bound by the terms of the license, and then you cannot change the licensing terms of the whole program, so you can't pass this right on either. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Ok, consider non-derived work. Because I am distributing whole program I have to do it under GPL. However I still have the right to distribute just the portion that is written by me under whatevel license I want but you as a recepient of GPLed whole do not get this right. IOW I am not passing all the rights _I have_. -- Dmitry -
I see what you mean. IANAL, but I don't think that's how it works. When your work is not a derived work, the GPL that applies to the rest of the program does not make you a licensee, and it only covers your work if you choose to license it that way. And then, you're the sole licensor of that piece of the work. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Please notice this sentence. GPL still influences the way I release stuff (if I want to release the work as whole) but it does not mean So, with regard to TIVO, why are you saying that GPL shoudl affect their hardware (I assume that key check/enforce is done in firmware taht is separate from kernel image)? -- Dmitry -
I'm not. I'm just saying that TiVO, as a licensee of Linux, agreed that it wouldn't impose further restrictions on recipients of Linux on the exercise of the rights granted by the license. So, just like it couldn't use a patent to stop people from modifying or sharing Linux, it can't use the hardware to do that. And if they fail to supply portions of the functional source code in order to prevent modified versions to run, they are infringing the spirit and quite possibly the letter of the license. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Why the hell do you keep saying that? There *are* lawyers who have said that what Tivo did was legal. They were the FSF's own lawyers. So now you're saying "I am not a lawyer, but that's not right". So you're trying to state some legal point, admitting that you're not a lawyer, and admitting that actual real-life lawyers disagree with you? So please explain why the *hell* you would expect us to take your points seriously? Linus -
What I wrote above had ZERO to do with TiVO. Please re-read the message you responded to, and the two previous messages in that sub-thread for the context you snipped out. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
So according to your logic, I can go to Sharp's website and download the GPL source code for their Zaurus. But I don't own a Sharp Zaurus; to keep with your interpretation of the spirit of GPL, they have to give me a Zaurus so that I can run my modifications on the same hardware? -
Sharp can modify the copy of the code in your Zaurus as much as you do, when you don't have a Zaurus. I don't see how you can get to the conclusion that they have to give you a Zaurus, when all you're getting is software. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Note that Harald Welte has already managed to force Siemens to unlock=20 a "tivoized" Linux router with the GPLv2 in Germany. German contract law=20 values intention when the contract has no specific clause that deals with=20 the issue, and in German law, an accepted license is a contract. So the fact that tivoizing Linux is against the spirit of the GPLv2 is a=20 court-proof fact, not just some speculation. What about if your GPL program ends up in a piece of hardware (e.g. a ROM,= =20 or an embedded ROM, or if it's some GPL code from OpenCores, as gate=20 netlist in silicon)? My interpretation is that you need a permission from=20 the author for doing that, unless there's an easy way to replace it with a= =20 modified copy (e.g. if you put the OpenCores stuff into an FPGA, replacing= =20 the configuration PROM would do it). Some people have difficulties with intentions of contracts rather than=20 direct rules. That may be due to different rules in different countries. In= =20 continental Europe, contract law usually bases on Code Napoleon, and=20 there, "good faith" is an important principle (and "good faith" means that= =20 the intention is more important than the actual coded practices). In the=20 roman law that was used before and has survived in countries who didn't let= =20 Napoleon in (like the UK and the USA), it's slightly different. But a=20 contract or a license still is not a program where anything that isn't said= =20 explicitely isn't said at all. =2D-=20 Bernd Paysan "If you want it done right, you have to do it yourself" http://www.jwdt.com/~paysan/
The GPL does not require it to be easy in fact to modify the piece of software. It just requires that you have the right to modify it, that is, that there be no legal obstacles in your way. You are entitled to the source code in modifiable, understandable form. There are no legal restrictions, other than those in the GPL and in the law, on what you can do with it. What you are actually *able* to do, however, depends upon a wide variety of factors way outside the scope of the GPL. By the way, I have a lot of sympathy for the argument that *if* you provide me a binary made from GPL'd code that required a key to produce that binary, I am entitled to that key. The key is precisely analogous to any other piece of source code -- it is mathematically 'combined' and 'processed' by tools to produce the final, distributed executable. If there's some rational basis for a legal difference between a signing key and a header file, I don't know what it is. DS -
Yes it does, section 3: "The source code for a work means the preferred for= m=20 of the work for making modifications to it." It then even lists that you=20 need to provide all the scripts and stuff you use to make it easy for you. Come on, *READ* the GPL, before you argue. =2D-=20 Bernd Paysan "If you want it done right, you have to do it yourself" http://www.jwdt.com/~paysan/
Nice job quoting me out of context. For the record, here is the context, and it addresses your criticism already: "The GPL does not require it to be easy in fact to modify the piece of software. It just requires that you have the right to modify it, that is, that there be no legal obstacles in your way. You are entitled to the source code in modifiable, understandable form. There are no legal restrictions, other than those in the GPL and in the law, on what you can do with it." DS -
Stop right there.
No. If you were logical (which you are not), you would admit that
(a) physical property is very different from intellectual property (the
FSF seems to admit that when it suits their needs, not otherwise)
(b) They never modified "a copy" of Linux - they simply replaced it with
"another copy" of Linux. The only thing that actually got *modified*
was their hardware!
The first copy didn't "morph" into a second copy. There was no "physical"
software that was molded. They do need to follow the GPLv2, since clearly
they _do_ distribute Linux, but you have all the same rights as they do
with regard to the *software*.
The fact that they maintained some control of the *hardware* (and some
software they wrote too) they designed is _their_ choice.
What Tivo did and do, is to distribute hardware that can *contain* a copy
of Linux (or just about anything else, for that matter - again, there's
a difference between physical and intellectual property).
And their hardware (and firmware) will run some integrity checks on
*whatever* copies of software they have. This is all totally outside
Linux itself.
Btw, according to your _insane_ notion of "a copy" of software, you can
never distribute GPL'd software on a CD-ROM, since you've taken away the
right of people to modify that CD-ROM by burning and fixating it. So
according to your (obvously incorrect) reading of the GPLv2, every time
Red Hat sends anybody a CD-ROM, they have restricted peoples right to
modify the software on that CD-ROM bymaking it write-only.
See? Your reading of the license doesn't _work_. Mine does. What I say is
that when you distribute software, you don't distribute "a copy" of
software, you distribute the _information_ about the software, so that
others can take it and modify it. And notice? My reading of the license
must be the correct one, since my reading actually makes sense, unlike
yours.
And yes, when Tivo distributes Linux, they give ...Err, no. Software, per legal definitions in Brazil, US and elsewhere, require some physical support. That's the hard disk in the TiVO DVR, Per this reasoning, nobody never modifies software. When you open a source file in your editor, you make changes to it, then save it, you're not modifying it, you're replacing it with another copy, and the only thing that actually got modified was the hardware. Maybe look what "modify" means in legal context? Then refer to the GPL: 2. You may modify your copy or copies of the Program or any portion Agreed. But as it turns out they use these checks to stop people from modifying the copy of Linux they ship in the device, and this restriction is a GPL violation because they don't provide information You don't retain that right yourself. When you pass that copy on, you pass it on with all the rights that you have. No problem here. This If you choose to disregard the legal meaning of the legal terms used Yes. That's all I'm saying. You just can't use the hardware to take This is not true. The spirit remains the same: let people modify and share the software. If the binary you got can't be created out of the corresponding sources, something is missing. If it won't run without this missing bit, you're missing functional portions of the source code. This all means the hardware is being used to impose a restriction on modification of the software, which is against the spirit of the GPL, and quite likely against its letter as well. If you don't want it to be so, you can always add an additional permission that clarifies this bit, such that TiVO and you will be happy. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
I'm now intrigued, where are these (Brazilian and US) definitions stipulated, and under what authority? Matt -
In the US, 17 USC 101 (the "Definitions" section of the title dealing
with Copyright) makes this definition:
A "computer program" is a set of statements or instructions to be
used directly or indirectly in a computer in order to bring about
a certain result.
As its purpose is to outline the scope of copyright law, this
definition is made under the authority granted to Congress by Article
I, Section 8 of the United States Constitution.
Michael Poole
-
But where is the part that says it "requires some physical support"? It says what it is; "a set of statements or instructions", how it should be used; "to be used directly or indirectly in a computer", and what purpose it serves; "in order to bring about a certain result", but it doesn't seem to indicate that it "requires physical support" aka needing some physical representation. I suspect this argument boils down to the philosophical debate of whether ideas (in this case software) can be truely devoid of the physical. Matt -
Sets of statements or instructions that cannot "be used directly or indirectly in a computer in order to bring about a certain result" are, for the purposes of copyright law, not software. "A computer" is a physical device. It always has been a physical device, except when "computer" referred to a person who performed computations -- and that meaning fell out of common use 40 years ago. Any suggestion that the requirement to be usable on a physical device is significantly different from "require[s] some physical support" is laughably stupid. 17 USC 102 requires that copyright protection only subsists in works that are "fixed in any tangible medium of expression" -- which obviously includes paper and hard drives, and has been ruled to include volatile program memory (the 9th Circuit's holding to this effect in MAI Systems Corp. v. Peak Computer, Inc. is what inspired the addition of 17 USC 117(c)). If the set of instructions exist only in transmission or in someone's head, they are not protected by copyright law. Michael Poole -
http://www.planalto.gov.br/ccIVIL_03/LEIS/L9609.htm LEI Nº 9.609 , DE 19 DE FEVEREIRO DE 1998. Art. 1º Programa de computador é a expressão de um conjunto organizado de instruções em linguagem natural ou codificada, contida em suporte ^^^^^^^^^^^^^^^^^^ físico de qualquer natureza, de emprego necessário em máquinas ^^^^^^^^^^^^^^^^^^^^^^^^^^^ automáticas de tratamento da informação, dispositivos, instrumentos ou equipamentos periféricos, baseados em técnica digital ou análoga, para fazê-los funcionar de modo e para fins determinados. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Well much as I don't like what Tivo did with only allowing signed kernels to run, I don't see anything in the above that says they can't do that. They let you have the code and make changes to it, they just don't let you put that changed stuff on the device they build. The software is free, even though the hardware is locked down. The GPL v3 really seems to change the spirit to try and cover usage and hardware behaviour, while the spirit of the GPL v2 seemed to me at least to simply be to allow people to copy and change and use the code, and pass that on to people. It didn't have anything to do with what they did with it on hardware. Nothing prevents you from taking tivos kernel changes and building your own hardware to run that code on, and as such the spirit of the GPL v2 seems fulfilled. It covers freedom of the source code and resulting binaries, not of the platform you run it on. The GPL v3 has a much broader coverage of what it wants to control, which to me means the spirit is different. I don't have a tivo, I use mythtv on my own PC. Tivo doesn't force you to buy their hardware after all. -- Len Sorensen -
Not to the software installed in the device. What they do is like an author A who distributes a program to user B under a non-Free Software license, and to user C under a Free Software license. C passes the program on to B under the same license. Now B has two copies of the program. One is free, the other is not. Except that TiVO had no right to distribute the program under non-Free terms in the first place, because it was not the author, and the license it had explicitly said it couldn't impose further restrictions. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
So now you want access to all the software that is installed in their device? Could you explain that please? You do have access to the GPL code that they used. If you buy one of Google's Search Appliance, are you expecting to allow Reread what you wrote here and see the complete lack of logic in your argument. Author A are Linux developers who distribute their software under GPL 2, TiVO gets the software under the same license and distributes it to their end users. They then make the all the changes to the Linux Kernel available to their end users under the same terms that they got from the Linux kernel developers. What freedom did they take away? -
Arguably, if I purchased the device, I ought to be entitled to make changes to it, yes. But that's a distraction I'd rather not get into They prevent the user from installing and running modified versions of the program on the box, while they can still do it themselves on the same box. I guess I must have repeated this at least a dozen times in this thread, so I'll just refrain from repeating this point from now on. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Oh, come on: you're not serious, right? Something indeed prevents me -- the fact that I'm not a hardware manufacturer, I don't have fabs, outsource vendors to provide me w/ designs, ASICs, etc. Nor to I have the money to pay one-off prices for various components if they're even available in batches that small. This argument seems totally disingenuous to me. The GPLv<3 was written in a time when the majority of sotware to which the license was applied was written for general purpose computers. The "user" was the owner of the computer, and Freedom 0 was about letting that user RUN modified copies of the software. Things have changed a lot; we're surrounded by embedded computers, and Freedom 0 seems to strongly imply I should have the right to run modified versions of the Free Software I own on the hardware I OWN. Or is the future of Open Source that you'll be able to hack on free software as long as you work for Intel, Red Hat, TiVO, Google or OSDL? Or own many-thousand-$$ fab printer? Look, I totally respect Linus' and others' position that the license is an inappropriate way to enforce what they feel are hardware design decisions, but can we dispense w/ the silly argument that the intent of the GPL is fullfilled as long as the user is allowed to modify the software where modify means "imagine a world where they'd be able to run" it? Dave -
And what about people who can't modify the Linux kernel? They don't know C. They don't know how to use a shell. They're not familiar with UNIX operating systems at all. Maybe they aren't smart enough to modify kernel code. The GPL is about having the legal right to modify the software and being able to put other people's distributed improvements back into the original code base. It does not guarantee that you will actually be able to modify the software and get it to work on some particular hardware. I certainly understood the GPL as ensuring the right to get the source code so that you could do something else with it. I never understood the GPL to be about getting hardware to do something else just because it ran GPL'd software. DS -
I learned C in part by modifying the Linux kernel and running the modified kernel on hardware I own, and enabling precisely that kind of tinkering is what the "spirit" of the GPL is about, as is quite plain Please don't conflate my endorsement of the "spirit" of the GPL with Alexandre's assertion that the GPLv2 forbids TiVOisation. I don't agree with him. My point is that people arguing that the spirit of the GPL doesn't revolve around the freedom of the end user to modify the software *and* run modified copies seem to be missing the point. Linus gets that, as he said in a previous message, he just doesn't personally care about freedom defined that way. Dave -
Yes I am serious. I wouldn't want to buy any such locked down hardware, but that still doesn't mean that I don't think it fits within the spirit I think it depends on the type of hardware. Certainly I agree some types of hardware really should not allow you to change the code on them due to the potential risks from doing so. Hence if a license starts to get into the grey area that covers such things, it is getting onto some thin ice that is probably should stay off. You risk excluding things you didn't intend to exclude while almost certainly still missing things you would like to have excluded. I agree that for many devices I could buy, being able to change the code on it would be great, and that there generally is no good reason to deny me from doing it, but I don't think it is worth the risk to put such a requirement into the license, and I certainly never read the GPLv2 to in any way imply such a thing. Apparently from what I can see, Linus never read any such thing in it either when he chose to use it. In fact I think you have to already have a very narrow preset view in order to read the GPLv2 in such as It seems many people really do feel that it is fulfilled. They may think it is a stupid hardware design and they may also chose not to buy such hardware, but at the same time they can be perfectly willing to say that as long as the modified sources are provided, that is good enough since further development of the source can be done, never mind what you can do with that particular locked down door stop the code was modified to support. Not everyone views the world through the eyes of RMS. -- Len Sorensen -
So your objection here is that one needs additional resources to do excersise their rights. Well, what about spending time and money to get education to be able to do programming work? Being able to understand C and hardware, etc is also an additional restriction imposed on an average person. Do you advocate that every copy of GPL program should be accompanied with an engineer who would explain how it all works? -- Dmitry -
Come on, again w/ the bullshit. TiVO does not try to prevent me from getting a CS degree, or buying a C reference. They _do_ prevent me No, just that hardware vendors not lock me out of _my_ hardware if they've benefitted from code which was intended to be modifiable by end users. Dave -
Yes please. Can she be spunky as well? ta. -
Well, it is not Tivo alone -- look at http://aminocom.com/ for an example. If you want the kernel sources pay USD 50k and we will provide Well, it is not Tivo alone, a large chunk of the vendors do that. The vendors who actually do it the clean way are just few and can be counted very easily. -
GPLv2 deals with that case, and they can (and should) be sued for it [except that US copyright law is designed for large music companies not people] -
Their argument was that the mentioned sources contain propreitary closed stuff from IBM/AMCC for the PPC 405/440 and or for the NXP (MIPS based) chips. Even if the GPLv2 deals with it, well haven't reached anywhere with it, inspite of talks with them. So most of the users just probably stopped talking sense with them, just like me. Have some of those Amino STB's, the software on it being buggy, including myself many others wanted to fix those bugs, but then people had to pay for their annual support to get the fixes. People who were able to fix also were denied the same since there is no source available. But if you wanted the sources, then you pay for the sources. If you don't pay for their sources, then pay for their Bronze/Silver/Gold Support schemes, where people pay through their nose. For a specific case with which i wanted to attach a USB based device to the box, they stated: we can port in a driver that which exists in the vanilla kernel, to their device but just that they need to be paid for that to be done, eventhough if someone else was willing to do that job, but that wasn't possible because of no sources. In either way, if you buy their devices, it is just that, you keep paying us, if you want your device your work as expected. -
As you probably know, this is not a valid excuse to distribute the software under conditions that disrespect its license. It doesn't mean you can force them to give you the source code, it only means the copyright holder can stop them from distributing the software this way. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Hmm, set top boxes are often rented from the cable company rather than sold. Stupid grey area for sure. At least tivo does give you the Well at least where I work we don't try to lock down the hardware, we do contribute our changes and bug fixes to upstream when it makes sense (and where our changes wouldn't make sense for upstream, they are still clearly included with the sources we have.) If a customer wants a copy of the sources, they will get a nice DVD, although strangely none have asked for one yet. -- Len Sorensen -
I am not talking about the rented aspect, since these STB's are usually Providing the changes back itself is a great thing altogether. -
It also makes sense. If the changes are accepted back, the community at large will keep the changes maintained. Less work for me to do when going to newer code versions later. And even better, it may help someone else out too. A company is likely to like the reduced maintenance burden part, but I think the other part is even better. After all we saved time not having to write everything our selves, so helping others save time seems only fair. The GPL may only require giving the sources to the people who buys the product, but there isn't really any benefit to us in doing only that. -- Len Sorensen -
No. The anti-DRM language is still there, and no, it was never a misunderstanding. Now it's been limited to "consumer devices" (after I pointed out some of the _obvious_ problems with the original language), and the only people who called anything a "misunderstanding" were the ones that tried to point to *other* points in the license altogether (ie there was also a "drm section", which didn't really seem to say anything much at all). Rms calls it "tivoization", but that's a word he has made up, and a term I find offensive, so I don't choose to use it. It's offensive because Tivo never did anything wrong, and the FSF even acknowledged that. The fact that they do their hardware and have some DRM issues with the content producers and thus want to protect the integrity of that hardware. The kernel license covers the *kernel*. It does not cover boot loaders and hardware, and as far as I'm concerned, people who make their own hardware can design them any which way they want. Whether that means "booting only No, it was not the former. And I think the whole "the kernel developers misunderstand the license" crap that the FSF was saying (several times) was very trying to confuse the issue: the FSF knew damn well which part of the license was obnoxious, they just tried to confuse the issue by pointing to *another* part of the license. And this is again the same *disease*. You claim that I "misunderstood" the "spirit of the GPL". Dammit, the GPL is a license. I understand it quite well. Probably better than most. The fact that the FSF then noticed that there were *other* things that they wanted to do, and that were *not* covered by the GPLv2, does *not* mean that they can claim that others "misunderstood" the license. I understood it perfectly fine, and it fit my needs. So tell me: who is the more confused one: the one who chose the license fifteen years ago, and realized what it means legally, and still stands behind it? I don't I have said ...
Not all of us agree with this for the benefit of future legal Agreed - everyone contributed to the kernel based upon the GPLv2. Lots of different reasons, lots of different viewpoints about GPL2 v GPL3, DRM , Treacherous Computing, etc. The commonality is not political, not a grand plan, not a grand unified social agenda but a bunch of people for whom the GPLv2 was an acceptable license for furthering their intentions whether that is education for all, a shared commons or just making a quick buck Alan -
Well, even the FSF lawyers did, but one of the reasons I never wanted to do the copyright assignments(*) is exactly because I think people need to make their own judgments on what the GPLv2 means. In the end, the only thing that really matters is what a judge says (after appeals etc), and the fact is, any license will always have gray areas where people disagree about interpretation. And I actually am of the very firm opinion that a world with gray areas (and purple, and pink, and green) is a hell of a lot better than one where everything is black-and-white. Only lawyers want a black-and-white world. So I would actually *encourage* other people to sue over their GPLv2 interpretations, as they have done in Germany (and as IBM has done in the US). I'd sue based on _my_ reading of it, but hey, while my opinion is obviously always correct, I recognize that I live in a world where not everybody else always sees that. [ (*) Obviously, the *biggest* reason not to do copyright assignments is that they are just a total pain in the ass to do, and cause tons of totally pointless paperwork. So "Linus is lazy and not interested in being a lawyer" is obviously the primary reason for the lack of assignments. I'm just much happier with people owning their own code outright. ] Of course, I also realize that suing people over license violations is a big pain in the ass, and in that sense while I "encourage" people to assert their own copyrights, I would obviously also say that it's almost certainly not worth doing if it's in a "gray" area. But that, in the end, Indeed. And it's _fine_ to even be in it "just to make a quick buck". We do want all kinds of input. I think the community is much healthier having lots of different reasons for people wanting to be involved, rather than concentrating on just some specific reason. For some it's the technology. For some it's the license. For some it's just a thing to pass boredom. Others like to learn. ...
agreed, because you cannot imagine at the beginning all fair uses of your project. It's a good thing that people can use it, thinking "hey, it's not not really. They would lose their job. They need a gray world to get customers, but they want to decide what half is black and what half And I think that for many people (including myself), it's all of these in this order : - something to learn (when you're at school) - something to pass boredom (when you're at school too) - the technology (when you're working on designing new products) - the license (when you finally try to put your products on the market) Regards, Willy -
Or rather they didn't think an attempt to enforce that in the US would prevail (or so I'm told). That's not saying what TiVo did was right, and that's not saying that what TiVo did was permitted by the license. Only courts of law can do that. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Wrong! Anyone with half a brain can make the distinction. What TiVO did is entirely legal - they fully complied with the GPLv2. Note that what they *DON'T* allow people to do is run whatever version of whatever software they want on their hardware. They have that right - its the "Free Software Foundation" and the GPL - regardless of version - is a *SOFTWARE* license. TiVO never stopped people from copying, modifying or distributing the code - what they did was say "The code is GPL'd, the hardware is restricted" - ie: "You can do what you want with the code, but you can only run compiled version of it that we provide on our hardware". Why is that legal? Because TiVO produces the hardware and sells it to you with a certain *LICENSE* - because it does contain hardware covered under any number of patents. That license grants you the right to use the patents - in this case algorithms - provided you comply with the terms of the license. (Just like the GPL gives you the right to copy, modify and distribute GPL'd code as long as you comply with its terms) If you believe otherwise then you are sadly mistaken. Now stop parroting the FSF's worn and tired tripe. DRH PS: Looking at your .sig I guess maybe you can't do that without getting kicked out of the FSF-LA -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
The GPLv2 says:
"For an executable work, complete source code means all the source code
for all modules it contains, plus any associated interface definition
files, plus the scripts used to control compilation and installation of
the executable."
The question is whether this includes private keys.
Different people have different opinions regarding this issue.
If "the complete source code" includes private keys, the GPLv2 requires
them to give any costumer the private keys.
Fact is that Harald Welte did in several cases successfully convince
vendors that private keys are part of the source code if they are
required for running the compiled binary on some hardware.
AFAIK there haven't been any court rulings on this issue, and it could
even be that courts in different countries will decide differently.
cu
Adrian
--
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed
-
If the hardware was designed for the end-user to change the software running on it - including running software that it was never meant to run (ie: a complete webserver on cell phone) - then yes, the signing keys are a part of the source, as the software running on the device is designed to be updated by the user using the provided system. If, on the other hand, the only "software updates" the user is expected to perform are the installation of newer versions of the existing code for "Security" or "Bug Fix" reasons then the signing keys aren't part of the source. I haven't looked into what Harald Welte did, but I'd be surprised if someone Agreed. -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
Are you an idiot, or do you just choose to ignore all proof that doesn't
fit your preconceived beliefs?
The GPL doesn't give someone distributing the software the choice of how
much to limit the freedom of the user.
Either private keys required to run the kernel on the hardware are
always considered part of "the complete source code" or they are never
Harald is in Germany, and he therefore takes legal action against people
distributing products violating his copyright on the Linux kernel
in Germany at German courts based on German laws.
cu
Adrian
--
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed
-
Nope. Merely stating a distinction. Either a device is distributed, like the common PC, that is designed for the user to change and update the software on, or, like the PS2 it isn't designed for that. If I find a way to update my PS2 to run Linux and find that it doesn't want to start the "Linux Firmware" because I'm lacking a signing key... In the case of a device that internally runs Linux (or any other GPL'd software) and wasn't designed for the end-user to change the software running on it then the signing keys aren't part of the source. OTOH, if I sell a PC running Linux that requires the kernel be signed then the signing keys *are* part of the source, since a PC is designed for the end-user to change the software running on it. BTW, nice use of irony with that line. Makes me regret letting my fingers get Never claimed it did. I just wasn't as specific as I should have been when No. It all depends on the use-case. If the hardware is designed for the user to install their own, custom versions of the code on then the signing keys are part of the source as defined by the GPLv2. If, OTOH, the hardware was never meant for the end-user to install custom versions of the software on, then while the signing keys are still *technically* part of the source, in practice they are not. Why? Because in most of those cases the end-user isn't granted the right to install and run custom binaries on the hardware. If the manufacturer provided the signing keys they'd be facilitating the commission of a crime. (call it "Breach of I know this. As I said, I doubt that anyone who tried this in America would have the success he has had. -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
Repetition doesn't let wrong things become true.
Where does the GPLv2 talk about the distinction you are trying to make
based on distributor intentions?
We are talking about the GPLv2 licence text, not about what you would
personally prefer.
cu
Adrian
--
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed
-
The GPLv2 doesn't have to cover this distinction to make it a reality. This distinction is *EXACTLY* the type of distinction a lawyer will make when arguing the point. Yes, it's artificial. Yes, it does appear to violate the GPLv2 - *IF* you read the text of such in a specific manner. However, the GPL, until version 3, *NEVER* guaranteed the right to run a given piece of software on *ANY* hardware - not the hardware it *COMES* on. And please, I repeated myself only because your reply seemed to imply that you didn't understand the statement I had made. Since you have now informed me, in a backhanded way, informed me that my interpretation of your response was wrong, I will not repeat myself again. Also note that I have re-examined the facts, in light of new information presented in this discussion, and have come to the conclusion that devices like the TiVO, in keeping the signing keys private (because of the "Facilitation of a Crime" thing I noted earlier), is violating the GPL, but not in the manner almost everyone is arguing. The violation is, rather, with the clause about the license being null and void in event of laws impacting the delivery of the source. (Because, as I also stated earlier, the signing keys are part of the source. Since, in some cases, the license on the hardware prevents running modified binaries (a reason for the digital signing) companies will keep said keys private - doing otherwise can (and I can assure you that some lawyer will do this) be construed as "Facilitating the Commission of a crime". In this case, it'd be "Breach of Contract" - IANAL, but IIRC, licenses fall under contract law)) -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
Reality check:
Harald convinced companies that they have to provide the private keys
required to run the Linux kernel they ship on their hardware.
cu
Adrian
--
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed
-
In Germany, not America. I should have qualified my statement to make it clear I mean "In America". Sorry about the confusion. -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
> In Germany, not America. I should have qualified my statement to make it = You shouldn't say "America" when you mean the "US". Best wishes, Michael =2D-=20 Technosis GmbH, Gesch=C3=A4ftsf=C3=BChrer: Michael Gerdau, Tobias Dittmar Sitz Hamburg; HRB 89145 Amtsgericht Hamburg Vote against SPAM - see http://www.politik-digital.de/spam/ Michael Gerdau email: mgd@technosis.de GPG-keys available on request or at public keyserver
Sorry, I slipped. I'm still trying to rid myself of the uniquely "US" belief that "America" == "USA". Thanks for the reminder. -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
No, the *real* reality check: The operative words here are "convinced companies" - as opposed to "convinced a judge to rule that private keys are required to be disclosed". (I just checked around on gpl-violations.org, and I don't see any news items that say they actually generated citable case law on the topic of keys...) Harald convinced companies that it was easier/cheaper/faster to provide the private keys than to continue in a long legal battle with an uncertain outcome. If the company estimates the total loss due to keys being released is US$100K, but the costs of taking it to court are estimated at US$200K, it's obviously a win (lesser loss, actually) for the company to just fold. Incidentally, this same logic is what drives the average successful patent troll lawsuit - the sued company will buy a license for $25K, just because they know that fighting the lawsuit will cost $100K and up.
You're off by a factor of 10-50. The usual estimates I've heard from people who ought to know is the minimum ante for fighting a patent lawsuit is $1 million to $5 million. Lawyer time and expert witness time to give the judge a granduate education in the technologies involved is *expensive* (since the judge may be really smart, but most judges have no engineering background to speak of, so you have to explain the technologies involved in terms that make sense to someone with an honors education with a Bachelor of Arts degree). Basically, in the US, you get the best justice money can buy. :-) - Ted -
-jb -- Tact is the art of making a point without making an enemy. -
Here in Germany, the rules at court are roughly "the loser pays
everything including the costs of the winner", so if a big company is
sure they will win at court there's no reason not to go there.
And if they did the effort of using private keys to only allow running
an official firmware, they must have seen an advantage from doing so.
I'm not saying it legally clear the other way round, my statement was
an answer to Daniel's emails claiming it was clear what such companies
do was legal.
cu
Adrian
--
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed
-
Well, the thing is (and I've said this before), a lawsuit is (and _should_ be) very much a last resort. I think that the Open Source community (and the FSF too) is much better off *not* concentrating so much on "legal rules" of what can and cannot be done, and instead spend much more effort on showing people why the whole "Open Source" thing actually works. And in fact, I think that's _exactly_ what Linux has been doing for the last decade! A lot of companies are actually doing the Right Thing (tm). Not because of anybody "forcing" them, but because they have literally bought into the whole "Open Source can do things better" mentality. In fact, the whole "coercive" approach is counter-productive. It makes people dislike you. It makes companies _resist_ open source, rather than see it as a potential ally. And no, I'm not speaking out of my *ss. Anybody who goes back fifteen years and looks at how the FSF was acting wrt the GPL (v2, back then), and how many friends - and enemies - they were making, should see that as a big clue. Linux really *did* change the landscape - for the better (*). By being much less contrary. So look at Intel in the open source space. They're doing well. Look at Sun. They aren't _forced_ to open-source, they see others open-sourcing, and they see that it works damn well. In the "Tivo space", look at Neuros. In other words, we're just *much* better off with a friendly license and not trying to force people to choose sides, than with the rabid idealism that was - and still is - the FSF. The FSF always makes for this horrible "you're with us, or you're against us" black-and-white mentality, where there are "evil" companies (Tivo) and "good" companies (although I dunno if the FSF really sees anybody as truly "good"). I'd much rather just see "individuals" and "companies". They're not evil or good, they are all in it for their own reasons (and their reasons are *NOT* the same reasons they are for me, you, or ...
On Thu, 14 Jun 2007 09:01:32 -0700 (PDT) Linus, If you really believe that then why didn't you choose a BSD license for Linux? You didn't say "completely free, no restrictions attached, people will follow because they'll see it's best, we just won't buy products that use Linux in a way with which we disagree". Instead you chose a license which enforced the so called tit-for-tat policy you think is fair. But people who prefer the BSD license may think you're a moron for forcing your political agenda (ie. tit-for-tat) on users of your code. The point of all that being, you _do_ believe in enforcing restrictions or you wouldn't like the GPL v2. So you draw the line of "fairness" and belief that people will do-the-right-thing somewhere short of the BSD license. Why is it so hard then to accept that the FSF draws the line short of the GPLv2 after having gained practical experience with it since its release? You can argue till the cows come home the belief that _your_ restrictions are more fair, moral and reasonable than theirs. But at the end of the day it's all just a matter of opinion about what constitutes fair and reasonable. You think its a fair trade that you get code back, the FSF think its fair that people can hack and run the code anywhere its used.. It all comes down to the author of the code getting to attach whatever restrictions they choose. Sean -
Because I think the GPLv2 is a great license. And I don't like the FSF's radical world-view, but I am able to separate the license (the GPLv2) from the author and source of the license (rms and the FSF). Why do people always confuse the two? The GPLv2 stands on its own. The fact that I disagree with the FSF on how to act has _zero_ relevance for my choice of license. The BSD license, as far as I'm concerned, is _horrible_ for any project I would use. I have actually released code under it, but never a "project". I've given some code of mine that I don't care about that much to the BSD projects, just because I didn't think that code really mattered, and I thought it would be stupid and small-minded not to let the BSD's use it. But for a project I actually care about, I would never choose the BSD license. The license doesn't encode my fundamental beliefs of "fairness". I think the BSD license encourages a "everybody for himself" mentality, and doesn't encourage people to work together, and to merge. Let me put this in source management terms, since I've also been working on a source control management project for the last few years: the BSD license encourages "branching", but the fact is, branching is not really all that interesting. What's interesting is "merging": the branching is just a largely irrelevant prerequisite to be able to merge. The GPLv2 encourages *merging*. Again, the right to "branch" needs to be there in order for merges to be possible, but the right to branch is actually much less important than the right to "merge". See? So I'm a *big* believer in the GPLv2. I think the GPLv2 is an almost perfect license. That doesn't mean that I have to agree with the FSF on Oh, and some people did and do. And you know what? That's PERFECTLY OK! I think that the BSD license is wrong for me. Does that mean that people who choose the BSD license are wrong to do so? No. For *them* the choices .. but I think that the software ...
Is there anything other than TiVOization to justify these statements? Also, can you elaborate on what you mean about 'giving back in kind'? (I suspect this is related with the tit-for-tat reasoning, that you've failed to elaborate on before) The only thing the GPL demands is respect for others' freedoms, as in, "I, the author, respect your freedoms, so you, the licensee, must respect others' freedoms as well". Is this the "in kind" you're talking about? Or are you mistaken about the actual meaning of even GPLv2? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Do you need anything else? But if by the question you mean "would you think the GPLv3 is fine without the new language in section 6 about the 'consumer devices'", then the answer is that yes, I think that the current GPLv3 draft looks fine apart I've *not* failed to elaborate on that before. Not at all. Just google for torvalds tit-for-tat and you'll see a lot of my previous postings. Trying to claim that this is somehow "new" is ludicrous. In fact, some of the google hits you find are from 2004, *loong* before the current GPLv3 discussion. So your "failed to elaborate" is not a failure on my side. Giving back "in kind" is obvious. I give you source code to do with as you see fit. I just expect you to give back in kind: source code for me to do with as I see fit, under the same license I gave you source code. How hard is that to accept? I don't ask for money. I don't ask for sexual favors. I don't ask for access to the hardware you design and sell. I just ask for the thing I gave you: source code that I can use myself. I really don't think my "tit-for-tat" or "give back in kind" is that hard to understand, is it? And no, it's not a new concept. Neither is the fact that I've never agreed with the FSF's agenda about "freedom" (as defined by _them_ - I have a notion of "freedom" myself, and the FSF doesn't get to define it for me). I don't call Linux "Free Software". I haven't called it that for close to I respect your freedom to design products around Linux. You can do whatever you damn well please - I just ask that you give the software back in a usable form. That's all I ask for. And that's all the GPLv2 asks for. Which is why I selected the GPLv2 in the first place, and why I *still* think the GPLv2 is a wonderful license! So I claim that the "freedoms" that the GPLv2 embodies are *greater* than the "freedoms" embodied in the GPLv3. Linus -
Then would you consider relicensing Linux under GPLv3 + additional I didn't. But I've provided evidence that your prior musings on this topic were wrong. I wanted to give you an opportunity to review your Forgive me if I find this a bit hard, because that's *not* what the GPL says. See, that's not what the license says. The license says what you ask for is respect for other users' freedoms. Nothing whatsoever for you. Only for users. Freedom is in "in kind" payment, and it's not even a retribution, a payback: it's payforward, or paysideways. We don't have to agree on our individual definitions of freedom. But we're talking about a specific license that assigns a specific meaning I can appreciate that you think it's better, but unfortunately it appears to be playing a significant role in confusing your interpretation of the GPL. The GPL is not just about making the source code visible, or even modifyable by others. It's about respecting others' freedoms. No matter how badly you prefer Open Source over Free Software, how badly you'd rather disregard the freedoms in the spirit and in the legal terms of the GPL, you chose a license designed to protect those freedoms, not only the ability to I'm afraid that's not what the GPLv2 says. There's no provision whatsoever about giving anything back. Not in the spirit, not in the legal terms. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
No. I'm not stupid. The GPLv3 explicitly allows removing additional permissions. So anybody who does "GPLv3 + additional permissions" is basically setting himself up for people taking those permissions away. Since the Tivo kind of permission is in my opinion a *fundamental right* (or call if "freedom" if you want), then "GPLv3 + additional permissions" simply is not a viable alternative, since anybody could just decide to make improvements and strip those permissions. The whole notion of "additional permissions" in the GPLv3 is totally pointless, since it's legally *exactly* the same as allowing dual licensing (which a license doesn't even have to spell out: you can dual-license *regardless* of the license!). The reason for the "additional permissions" is just to make the LGPL go away, and become a sub-clause of the GPLv3. If you really thought anything else, you're just uninformed and stupid, and didn't think things through. Linus -
So what? You just refrain from accepting contributions that attempt to remove them, and you'll keep TiVO happy. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
You really aren't thinking, are you? It's not about keeping Tivo happy. It's about keeping *me* happy. That's my primary (only) motivation for a license. And let's go back to why I selected the GPLv2 in the first place, shall we? I want to be able to use other peoples improvements. If they release improved versions of the software I started, I want to be able to merge those improvements if I want to. Your *IDIOTIC* suggestion is explicitly against the whole POINT! By saying that I shouldn't accept contributions like that, you just INVALIDATED the whole point of the license in the first place! Can you really not see that? Linus -
Hmm... So, if someone takes one of the many GPLv2+ contributions and makes improvements under GPLv3+, you're going to make an effort to accept them, rather than rejecting them because they're under the I understand. I assumed you had some trust that people would abide by your wish to permit TiVOization, and that authors of modifications were entitled to make "whatever restrictions they wanted" on their code. Pardon me if I think your position is at least somewhat incoherent. Can you help me make sense of it? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
You *cannot* make GPLv3-only contributions to the kernel. I'm sorry, but that's how it is. You can take some of the code that is GPLv2+ in the kernel, and MOVE IT TO ANOTHER PROJECT, and use them there. But not within the confines of the Linux kernel. Within the Linux kernel, the GPLv2 rules - and "GPLv2+" becomes just "GPLv2", since the GPLv3 is not compatible with v2. This is no different from the fact that we have some drivers that are GPLv2/BSD licensed. Within the kernel, they are GPLv2. But on their own, you can choose to use them under the BSD license, make your changes to them, and release them commercially. And correct - I cannot (and neither can anybody else) then accept those Actually, normally I *do* have such a trust. It's why I have no problem with drivers that are dual-GPL/BSD, and in fact, I've told people that I don't want them to turn them into GPL-only, because that is simply not polite. But I hold *myself* to higher standards than I hold others. And in particular, when it comes to people with a religious agenda, I don't expect them to be polite or take my feelings into account. I expect (from good history) that people with a license agenda will consider the license I'm giving up. I'm moving you to my "flamers" list, so that your emails go to a separate mailbox that I read weekly. I've wasted too much time with you, your arguments don't make sense, and you seem to refuse to even _try_ to understand my position, or respect the fact that my choice of license is MY choice, and that I actually have a brain of my own. Linus -
I can make improvements to GPLv2+ files under GPLv3 (or rather will, So which is it? Do you want to be able to use other people's improvements, respecting the conditions you said they are legitimately I understand this very well. You'd have to get the kernel upgraded to GPLv3 in order to accept the contribution. Likewise for any other contribution under any other GPLv2-incompatible license. So, you see, your statement above, about wanting to be able to use other people's improvements, cannot be taken without qualification. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
You can do that, but you won't be able to distribute those changes along with the rest of the kernel. -- Jeremy Maitin-Shepard -
I know. Neither will Linus. But he says he chose GPLv2 such that he could, and the v2 is better than v3 in this regard. What's wrong with this picture? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
I'm sure it's a rethorical question - but what is wrong is that imho the clause that normally is added to any file under the GPL ("... version 2 or higher") is builtin in as a safeguard: If, for whatever reason, in the future it is discovered that the GPL version X has a flaw that was not forseen - then it can be 'patched' by writing a successor, having that released by the FSF and allow anyone to "upgrade" all GPL-ed software to the new license, as such having avoided the problems of the said (fictious) flaw. Therefore, it seems pretty weird to me that LONG before version 3 was written and released, someone purposely would choose to freeze their software at version 2. Why make it "impossible" to use this safe-guard? ["impossible" because if you get a signature from every author where they transfer the authorship rights to you, you can re-release everything under a different license anyway. However, the kernel has so many authors who never signed anything(?) that this is not possible anymore]. I never knew it was possible to change the "version 2 or higher" into "only version 2", but I am not a laywer and not into licenses at all, and I am sure Linus had laywers look into this, so we can take this as a fact. The result is simple: 1) A lot of files in the kernel are fixed at version 2. 2) Version 3 is incompatible with version 2 (which I also take for granted, having read that in this thread). Many (or at least important) authors of version-2 files do not wish to change the license to one that allows it to be transformed to GPL v3 (ie, add the clause "version 2 or higher". Therefore, this will not happen. The result is that it is impossible to accept/add patches that can not be converted to GPL v2 (ie, which are explicitely version 2 or have the phrase "version 2 or higher"). I think this whole thread has only one purpose: To "test" if the kernel source is indeed - by LAW - immutable and fixed to version 2. Of course, as an unwritten rule, everyone here ...
Because what you call "safe-guard" is just call "idiotic". The fact is, anybody who cares about the copyright license he uses on software (and I sure do care!) should *never* give a blank check to somebody else that they don't trust to change that license for them. In other words, the whole "v2 or later" language *only* makes sense if: - You are the FSF, and you *decide* what "or later" actually means In this case, you obviously don't have to trust "somebody else". You just trust yourself. - You don't care about your choice of license. In this case, you might as well let somebody else make that decision for you, although quite frankly, you might as well use something like the BSD license and let many *more* people make that decision to relicense for you. - ..or you trust the FSF implicitly. In this case, you're not an independent entity, you're just a lackey of the FSF when it comes to the license. If none of those conditions are true, you'd be *incompetent* to leave the "v2 or later". And I'm not incompetent. None of the above conditions holds true for me, so the "v2 or later" would be totally idiotic. Which is why Linux has *never* had that statement for *any* code I have ever written! So the fact that *you* (and the FSF) call it a "safe-guard" by no means What you state above is that you didn't even *READ* the GPL, yet you make some arguments about it as if you had! So let me educate you. You may not want to be educated, but I'll try anyway: - Please *read* the license that you discuss. If you don't read it, there's no point in you making any arguments about it. And by reading it, I don't mean just mechanical "reading" in a technical sense. I mean "read and think about it" - Once you have actually read it, you'll notice that there *is* no "either version 2 of the License, or (at your option) any later version" in the GPL v2 license AT ALL. Really. That phrase DOES NOT ...
That's true. But sec. 9 of the GPLv2 says: If the Program does not specify a version number of this License, you may choose any version ever published by the Free Software Foundation. So, by making the COPYING contain the v2 text, is the author specifying a particular version? If yes, then the sec. 9 provision would be meaningless, since there would be no way to not specify a version number. My understanding is that courts would presume that a license term has a meaning, if it has a plausible reading. And there such a reading: that to specify a version, there needs to be (e.g. in the source files) a statement like, "This file [or work] is licensed under the GNU GPLv2." Corrections, flames, etc. are welcome. -Sanjoy -
Of course the "published under terms of GPL." would do. -- Krzysztof Halasa -
It would do, if you could stop there and say no more. But: You may copy and distribute verbatim copies of the Program's source code as you receive it, in any medium, provided that you...give any other recipients of the Program a copy of this License along with the Program. [GPLv2, section 1] So you have to give recipients the license text from a particular version of the GPL. To make that the only version unde which the work is licensed, you have to add something like "Licensed under the GPLv2". Otherwise sec. 9 says that you offer the work under any version of the GPL, and the licensee can take his or her pick -- even using v1 (!). -Sanjoy -
That's exactly what I meant by "not specifying the version". -- Krzysztof Halasa -
Why do you keep saying "upgraded" to GPLv3? How is it an improvement to move from a small, simple, elegant, and tested implementation to something that's more complicated, less elegant, less coherent, totally untested, and full of numerous special cases? Bumping a version number is not in indicator of quality, and spending over twice as much text to express the same legal principles is not an improvement. So far, you haven't brought up a single reason to use v3 except for a higher version number. (Not that I'm asking you to.) You've just tried to argue that it isn't WORSE than the existing license. Rob -- "One of my most productive days was throwing away 1000 lines of code." - Ken Thompson. -
Ahhh, but so much more entreprisy. I never had realized before that the DailyWTF applied to licenses too. OG. -
Just because it has a higher version number. Honest, no other reason was implied. I'm seriously not trying to push v3 here. I got into this to try to Agreed. Still, some people talk about upgrading from XP to Vista (ok, no numbers here, but you get the idea), just like they talk about Sure, that was not my goal. I wasn't even trying. Good, it's nice when people get the idea of what I'm trying to accomplish. I feared this had been lost in the noise. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
No. Linus and other Linux kernels might *want* to take other people's improvements, but thanks to Richard Stallman's choices for GPLv3, they can *not* legally take other people's improvements without violating the GPLv3 license. That's not their fault, it's the fault of people who wrote the GPLv3 license, promulgated the GPLv3 license, and who is attempting to convince everyone that the GPLv3 license is the only valid license for Right Thinking FSF automatons to use. There are plenty of things that I might *want* to do, that I am legally prohibited from doing. that doesn't change the fact that I might want to do it. The fact that GPLv3 is incompatible with GPLv2 is a tragedy, in the Greek sense. - Ted -
On Fri, 15 Jun 2007 01:24:32 -0400 The _exact_ same arguments are made against the GPLv2 by the BSD folks. Given that many people here defend the GPLv2 over BSD, it's ironic the tone and level of vitriol shown against the v3 and such a willingness to use the same arguments the BSD folks use against v2. Both v2 and v3 enforce some restrictions that people who want to participate must obey. And _yes_ I acknowledge that v3 has _more_ restrictions. But then, v2 has more restrictions than BSD and we're more or less happy with that, aren't we? In fact, many of us believe it's a virtue that Linux has a more restrictive license than that of the BSD's. While this isn't an argument that we should happily accept more restrictions, hopefully it will put things in a cheerier perspective. We're not talking about a fundamental disagreement (ie. no restrictions versus any restrictions); we're simply talking about _degree_ of restriction. There's no problem with people voicing honest disagreement with the v3, but please lighten up a bit on FSF bashing and the Greek tragedy talk. Sean. -
<wry> Would you prefer a reference to Resistible Rise of Arturo Ui? </wry> -
Ah no, it's their fault. The GPLv2 always was clear that there will be some= =20 The GPLv2 tries hard to be compatible with any further versions of the GPL= =20 as possible, by allowing people to choose which license you take, and by=20 making sure that no man in the middle can restrict this choice. If people=20 deliberately select to use "GPLv2 only", who's to blame? RMS? Come on,=20 that's bullshit. It's *Linus Torvalds* who made Linux incompatible with=20 GPLv3, nobody else - ok, Al Viro with his tagged GPLv2 files (and honestly,= =20 I think this is just another Linus misinterpretation about the GPL, and he= =20 really didn't do it, because he couldn't). This thread was fun, but I think all arguments have been repeated often=20 enough. I try to give up. I suggest everyone who has some assertions about= =20 what the GPLv2 does read it through and find the place where it says so.=20 Unfortunately, I haven't seen GPL citations from the Linus-fanboy curve,=20 only suggestions that the GPL "does not say something" which it clearly=20 does. =2D-=20 Bernd Paysan "If you want it done right, you have to do it yourself" http://www.jwdt.com/~paysan/
Not true at all. The GPLv2 leaves it up to the person placing their work under the GPLv2 license that its up to them whether they want the license on their "covered work" to be able to be changed. That the boilerplate includes this clause is pretty pointless - anyone can easily remove the "or, at your option, any later version" clause and render section 9 meaningless as applies Incorrect. Read section 9 of the GPLv2. It's pretty clear that the "any later version" clause is optional. Whats more is that since the modern linux kernel *IS* a "composite work" composed of Linus' original code with changes contributed by other people - Linus retains copyright to the work as a whole. This means that he can license it in any manner he chooses, as long as it doesn't affect the copyrights (or licensing) of the people that have contributed changes. I don't have to go to the US copyright law for this - Linus released Linux under the GPL, others made changes and sent them back saying "You let me have access to your code under the GPL, I've made some changes that make it better. You can have my changes under the GPL." QED: Linus still holds copyright to Linux and can license it in any way he chooses. This is limited because of the license he accepted when adding the changes back to his code. He may have locked the kernel, as a whole, to version 2 of the GPL - but that is his right. There is nothing he has done that has stopped people from having their code included that is still "v2 or Because there has been no need to quote the GPLv2 until it became clear that people were going to keep claiming it stated things it did not. Since then I've started quoting the relevant sections of it. But I agree with you - the thread was fun. And then I realized that the discussion was going nowhere at all. So I'm going to answer the last few messages in my inbox and then start filtering messages with this topic off without reading them. DRH -- Dialup is like pissing through ...
Huh - surely not to files added to the kernel that were written by This is totally new to me - if this is true - I'd really like to be sure! I always thought that it would be necessary to get signatures of each and every contributor before you can change a license of a file. Why do you think that the FSF demands written copyright-transfers with signatures before you are allowed to submit a patch to any of their largers projects? If they - as original copyright holder - could do what you claim - they wouldn't need those signatures. Having signed a copyright transfer for 'future' changes for gprof, libiberty, readline, zlib, gcc, gdb, libstdc++, bfd, dejagnu, gas, and binutils, Carlo Wood <carlo@alinoe.com> -
Actually, yes. Even to those - when they are part of "the whole". I'm sorry, but I've learnt more about copyright law, and talked to more lawyers about licensing that probably most of the rest of the people involved in this discussion have *combined*. And yes, at least under US copyright law, and at least if you see Linux as a "collective work" (which is arguably the most straightforward reading og copyright law, but perhaps not the only one) I am actually the sole owner of copyright in the *collective* work of the Linux kernel. The way "collective works" work, there are two separate copyrights: there is the copyright in the "separate contribution", which is vests ininitally in the author of that contribution (unless he signs over his copyrights, often by virtue of working for somebody else). And then there is the copyright in the "collective work", which would be me. Of course, owning coyright in the "collective work" doesn't actually give me complete control anyway. I cannot relicense things in ways that go against the rules of the individual works. But in a very real sense, yes, I actually do own a certain (*limited*) copyright over even the parts that have not been explicitly signed over to me. And yes, there are other potential ways to describe Linux, and in the end, it doesn't really matter. Because the way the GPLv2 works, it makes it clear that as long as a piece is a part of the whole, it has to be licensed under the GPLv2 and nothing else. And btw, just to make you feel safe - I cannot do anything about that, even if I *do* own the copyright in the collective, because of the limitations on what that colletive work copyright implies (it says that I have the right to reproduce and distribute, but I don't have the right to *modify* except as given to me by the original author!) So don't worry. I *technically* have certain special rights, but I practically speaking gave up most all of those rights by accepting code from others ...
US law is only relevant for < 5% of all people.
How valid would any action based on US copyright law be in other parts
Does this include GPLv2'ed code not intended to be used in the Linux
kernel submitted by people other than the copyright holder for inclusion
in the Linux kernel?
If yes, the FSF has exactly the same rights if taking a GPLv2 driver
cu
Adrian
--
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed
-
I tried to explain that in the case of the Linux kernel, we really don't care, since in the end, what matters is the GPLv2, and I have bound myself to the terms of that license *regardless* of any US law. So yes, US law is only relevant for < 5% of all people, but in the specific case of the kernel, even that US law isn't _really_ all thgat relevant at all, not *even* to those 5% ;) So I *really* hope you took my explanation of why I actually have more rights than others as a nitpicking "legal detail", not as a "I own your very SOUL, bow down before me!" kind of thing. But to answer your question by _another_ nitpicking answer, as the original author, I probably do have some special legal status even in Europe, and probably in other places too. The fact that others *extended* on my original work doesn't take away the special place of original authorship, even if the extended version has a totally different form (ie, a movie based on a book ends up still havign the original author of the book holding special rights - and in fact those rights are in some cases much *stronger* in Europe than they are in the US). For example, Europe recognizes "moral rights" in original authorship, in ways that it is much harder to enforce (if at all) in the US. But as mentioned, since I myself has bound myself to the GPLv2, that really is a pretty damn theoretical argument. When it comes to the kernel, I'm "Primus inter pares", if you wish. In the US sense of "within the compilation", probably yes. Ie that right is tied to *linux* as a compilation. In the European sense of "moral rights", no - that right is very much tied to original authorship (so I, as original author of Linux, have some rights with respect to Linux, but on the other hand, they, as the original authors of some specific code, The FSF does indeed have special rights wrt Hurd, regardless of where the code in question has come from (as long as it came there *legally*, of course!). So when it ...
Of course they do. That is defined by the license. They can include it with HURD, but that doesn't give them copyright to it, just the right to exercise the rights granted by the license. -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
Which is why I have taken everything you said so far for granted as being a fact -- no need to try to convince me of something ;) In my case it suffices if you say that you are sure (and before anyone says: you're nuts - what I REALLY mean is 'for the purpose of this discussion'. For me, it makes no sense to waste more time on discussing such an issue; I won't bet all my money it, of course). The point is: can you, or can't you (legally) relicense the whole kernel tree under the GPLv3 (or GPLv2+GPLv3)? At first I thought that you cannot, because too many (significant) contributors have been involved (and you will never get signatures from them all). Then someone surprised me by claiming that the original author had copyright for everything - even files added by others. To me, this seemed to say: even if those contributors don't like it, the original author can still sell the whole to some company under a proprietary license (also still having the original under the GPL on the net, of course), as he could do when he was the sole author. Linus> Actually, yes. Even to those - when they are part of "the whole". But the rest of your reply made it a bit unclear again. Assume you stopped taking your meds and next week you think that GPLv3 is THE thing for the kernel. Then could you legally, or can't you, go ahead and change the license of the whole kernel to GPLv3? And if you can't, then roughly how many files / authors are stopping you from doing so? If the answer is: I can't. Then I think you're a lucky bastard, and have escaped years and years of discussions with people trying to convince you that the GPLv3 is better ;) I think that the question: can OTHERS "upgrade" the kernel to GPLv3 has been answered clearly now: No they can't. But if you can, you're probably not done with dealing with people who want that to happen. -- Carlo Wood <carlo@alinoe.com> -
No. My special rights do not actually give me those kinds of powers, exactly because I'm bound by my _other_ agreement (namely the GPLv2) to Both are true facts, but the "copyright for everything" is a *separate* kind of copyright, which does not include the right to relicense. It's literally the "copyright in the collective". For examples of the US rules, see USC 17.2.201(c) ("Ownership of copyright" and " Contributions to Collective Works"), which spells out some limited special rights that I have (namely the right to reproduce and distribute). Of course, US law being what it is, the USC is just part of the picture. US law is the strange kind of British law, where "case law" is in many ways more important than the written-down rules like the USC. So caveat emptor! So I have limited special rights in the collective, but those rights are actually in almost every way *more* limited than the rights that the GPLv2 gives to me (the "almost every way" is because quite frankly, I'm not entirely sure about certain special cases. In particular, if somebody tried to _revoke_ the rights to their code under the GPLv2, I suspect that my rights in the collective would protect me from that and allow me to still distribute the code in question, since _those_ rights cannot be revoked, and they are _mine_). And btw: the above paragraph is *way* more legalistic detail than I am at all ready to state as "fact". It depends on too many things, and is largely speculative in nature. But one thing is pretty clear and nonspeculative: *nobody* has the right to upgrade the kernel to GPLv3. Not me, not you, not anybody. Not without clearing it with every single person whose copyright is involved and who didn't already give that permission. So only in the case of some really obscure and unclear situations, I _may_ have more rights than some other people, but trust me, but that is damn murky, and you'd better have a good lawyer state it, not just a programmer who ...
http://www.copyright.gov/title17/92chap2.html#201 There are some really interesting bits in Chapter 1 too... There's no revocation clause in the license. They can't. (SCO would have tried to revoke Caldera's license to its contributions it if it were at all Title 17 chapter 1 section 103(a) seems to lean against this. I think it says that having rights to the collective is conditional on having had the rights Entirely agreed. :) Rob -- "One of my most productive days was throwing away 1000 lines of code." - Ken Thompson. -
They don't. They demand the signature so that some contributor can't change their mind at a later date or even be able to give a proprietary software DRH -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
No. It is clear that you have the *option* of keep upgrading, but it is also equally clear that Linux has always decided *not* to exercise that option, exactly because I liked the GPLv2, not some "future upgrade". I decided that long before I saw the GPLv3. And I'm surprised by people who wonder why I did that. I'm _intelligent_, dammit. That means that I can foresee the future to some degree, at least in the limited sense of what is a likely outcome of my actions. Why are people surprised by the fact that I have foresight? I may be known for being an impolite bastard, but quite frankly, anybody who thinks I'm a _stupid_ impolite bastard must be missing a page. You can disagree with my opinions. You can call me obstinate, impolite, and opinionated. But quite frankly, very few people have ever found me *stupid*. So give me that - I'm not stupid. That means that I actually *can* predict the future to some fuzzy degree, and that people really should *not* be surprised by the fact that I never let the FSF control my choice of There's no "blame". There's only credit. Besides, you are wrong. The *default* for the GPLv2 in the presense of license information is *not* "v2 or later" In order to get "GPLv2 or later", you actually have to explicitly specify it. I just find it sad that so many people did that, often apparently just because they didn't actually read or understand the license. Linus -
Ok, open your local copy of the GPL (you should have a copy in /usr/share/common-licences/GPL if you use Debian). Now read along with me... | GNU GENERAL PUBLIC LICENSE | Version 2, June 1991 Ok, we have that settled, this license describes GPL version 2, specifically, no if's or later's. Bunch of legal mumbo jumbo follows until we reach... | END OF TERMS AND CONDITIONS And we've reached the end of the license. Now it it followed up by some helpful text for the reader... | How to Apply These Terms to Your New Programs ... | To do so, attach the following notices to the program. It is safest | to attach them to the start of each source file to most effectively | convey the exclusion of warranty; and each file should have at least | the "copyright" line and a pointer to where the full notice is found. | | <one line to give the program's name and a brief idea of what it does.> | Copyright (C) <year> <name of author> | | 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. Look, some example boiler plate that is suggested to be placed in the source code _to most effectively convey_ that the code is GPL licensed. So clearly there must be less effective ways, one of which may just be to include a COPYING or LICENSE file along with the distributed source. And the author of this example boiler place is clearly 'talking' to the the copyright owner/licensor (i.e. not the licencee), giving suggestions about adding contact information and even possible commands that could be added to interactive programs. So that '(at your option)' may just be a suggestion to the licensor, or maybe not. But that isn't really here not there. Files that don't have such boiler plate do not convey the terms ...
This argument is backwards. It's because of Linus' choice for GPLv2 that he can't take improvements under the GPLv3. Had he chosen any other GPLv3-compatible license, he could. And the same applies to any other incompatible pair of licenses. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Doesn't matter at all. GPLv3 requires that any project incorporating GPLv3 code be licensed under the GPLv3. Linus is, as he has shown, intelligent enough to know this. The *second* he actually accepted GPLv3 code into the kernel it would either be "change the license or start getting lawsuits for You are making a distinction between "part" and "whole". When separate from the kernel the code can have whatever restrictions the creator pleases. If he has said "I want this in the "official" Linux Kernel" (ie: I want this in Linus' Linux Kernel source tree) then the creator of the code has stated a willingness to abide by Linus' decision about the whole work. It's a moot point, though. The Linux Kernel is licensed under GPLv2, which means that *all* code in it has to be under the same license *and* that no code in it can have any restrictions *NOT* in the GPLv2. DRH -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
With Al Viro, at least, specifying that his code has been released *strictly* It is wrong when you look at the text of the GPLv2 only. When you look at how the "Open Source" community works it is clear that the "tit-for-tat" nature is a reality. No, it isn't mandated by GPLv2, but that is the "spirit" of the GPLv2 that most people who work on Open Source projects follow. If you want a *REAL* and *CONTINUING* violation of the GPL just look at Herr Schillings "cdrecord", in which he places additional restrictions on peoples ability to modify the code with statements in the same such as "You must leave this check in place" or "You have to leave this comment in place" - even when the comment isn't part of the "licensing statement" mandated by the It doesn't. But that it doesn't *MAKES* *NO* *DIFFERENCE* because, in practice, that is *EXACTLY* what happens anyway. Which is not in question here. The objections Linus (and others) have to the GPLv3 may share some specifics with my own objections, but my own are that GPLv2 respects my freedoms in their entirety. GPLv3 restricts my freedoms because one (or more) of the people behind it have a political agenda. (No, that term isn't entirely accurate, but its the best one I have found for the situation. Explanation: We don't like the way the law of one or more munincipalities/political divisions/countries is written and rather than trying to get it changed through other channels, we are going to enforce the way we think it should be by using a related set of laws and the text of a But it does. If you are going to distribute your own, modified version - in any way - you have to make the source available as well. No, it doesn't require that you give back a version if you aren't distributing it, but in that case it hardly matters. DRH -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
BSD licenses encourage forking. Specifically, if a BSD-licensed project becomes significantly commercially valuable, there's an incentive for companies to hire your developers away to work on a proprietary fork. When Sun Microsystems started up in 1982, they hired away Bill Joy to work on a closed version of BSD (SunOS). When Berkeley shut down the CSRG, BSDi hired those developers to work on another closed source BSD variant. More recently, Apple hired people like Jordan Hubbard away from FreeBSD to do yet another fork: MacOS X. The loss of people hurts the original project. With BSD licensed code, companies can say "work on the codebase you love as a day job, and you can still work on the open version at night". Then work them 90 hours/week. Or even "we'll release this code open source after we can't sell it anymore, a year or two from now". And then the deadline never comes, or the codebase is irrelevant by then, or too far diverged to merge. You won't get all the developers, but you'll get enough to cost the open project momentum. BSD is 30 years old and the free version is still a pale shadow of its proprietary forks like MacOS X or the bits of it Windows incorporated. Now think about trying to do that to a GPL project. If you hire the developers away, they have to work on a _different_ codebase. Much less compelling, both for the hirer and the hiree. If you think Linux is compelling enough to commercialize, you MUST do so within the terms of the GPL or not do it at all. You can't do a closed fork and distribute the result. Maybe this means companies aren't as quick to jump on the bandwagon trying to commercialize it, but the project can then grow larger without interference until commercial participation _is_ compelling, on its own It's not political, it's pragmatic. GPLv2 has tangible benefits for project Nobody objects to the FSF putting out new licenses if it changes its mind about what it wants to do. They object to it ...
I'm wondering more and more why you choose the GPL and not the BSD
licence for the Linux kernel...
Companies are violating the GPL and this only works as long as noone
starts taking legal actions against them.
And there are some companies for whom it's better if they can take the
open source code and turn it into some closed thing. There's a reason
And what are the risk of your allies actions?
Consider e.g. that your ally AMD offers legally questionable non-GPL
modules and Debian shipping binaries.
If one of the many copyright holders of the kernel wants to take legal
actions against this suing AMD might simply be out of reach due to
financial reasons.
But legal actions against the maintainers of ftp.<your_country>.debian.org
distributing binaries of these legally questionable modules have a good
chance of success.
The legal risks might not be a problem for a big company like AMD, but
anyone seeking an easy legal win will _naturally_ attack mirrors.
And considering the lucratice "cease and desist letter" business in
Germany, it really seems to be only a matter of time until mirrors will
have to pay (the lawyer costs of the ones sending the cease and desist
Look at AMD.
Look at NVIDIA.
Until someone starts to attack ftp mirrors and other people distributing
the questionable code.
There are so many copyright holders of the Linux kernel (including e.g.
the former Caldera), and if someone wants to spread fear among Linux
users legal actions against people maintaining ftp mirrors or selling
their old Linux distribution at ebay will be the cheap and effective way
cu
Adrian
--
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed
-
Why do people confuse "anti-GPLv3" with "pro-BSD"? What's the logic? The BSD license is not doing tit-for-tat. It doesn't give me anything back. I don't believe in that kind of model. So I'd not use it for my projects. The GPLv2 has a good balance. It encourages tit-for-tat, and it makes sure that the software is kept free. And it doesn't try to force anything else, or play politics. The only thing you have to believe in is "tit-for-tat". The GPLv3 goes too far. It's no longer "tit-for-tat", it's "our software is worth _soo_ much, that we want to force you to behave well, or you cannot use it". I think one of the above licenses are good. The fact that I reject the GPLv3 in _no_ way implies that I should like the BSD license. Both the BSD license and the GPLv3 are flawed - they are just flawed in fundamentally different ways. So the whole question of "why don't you use he BSD license then" is just fundamentally bogus. A license is about a *balance* of things. "Fairness" is not about laissez-faire (BSD) or about total-control (GPLv3). To me, It's about something in the middle, where people give back in kind. And btw, that "to me" is important. Different people have different opinions. That's _fine_. Use the GPLv3 for your projects. Go wild. Use the BSD license. It's your choice. But by the same token, it was _my_ choice (and it was an informed choice) to use the GPLv2. And to then come in fifteen years later and call me "confused" about a license I've chosen is a damn affront to me. I'm not confused. Somebody else may be, but it's not me. Linus -
Let me see if I got your position right: when TiVO imposes restrictions, that's ok, but when others want to find ways to stop it, then it's not. *Now* I'm confused ;-) -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Sure. I think it's ok that Microsoft imposes restrictions too on the
software they create. It's *their* choice.
And I think it's ok for you to impose any restrictions (including the ones
You are indeed totally confused.
It's *ok* to impose restrictions on the stuff you create. Everybody has a
different world-view, and for some it's about making money, for some it's
about something else, and some don't want any restrictions at all.
For me, the GPLv2 was the license I liked. I didn't like the BSD license,
so I didn't choose it. I don't like a license that restricts hardware, so
I didn't choose that.
And I *still* don't choose that.
See? I think the GPLv3 is a *much* inferior license to the GPLv2. It's
better than its drafts were, but it's still doing things I disagree with.
So tell me, why do you think I'm confused about the GPLv3? Why do you
think I should have said "GPLv2 or anything else the FSF comes up with"?
So the only thing I want you to say is:
(a) Linus knows what he is doing, and isn't actually confused.
and
(b) It was my right to use the license of my choice for a project that I
started.
and
(c) I have the right to see the difference between the GPLv2 and v3, and
think that the GPLv3 is the inferior license.
Comprende? MY CHOICE. Not the FSF's. Not yours. Not anybody elses.
Linus
-
Last I looked, TiVO was not the author of Linux. Did you sell out or I think you're confused about the spirit of the GPL, that applies equally to v1, v2 and v3. I think you're confused because you claim the GPL is tit-for-tat, that it encourages/requires (you haven't been consistent) contributions in kind, but the only contribution in kind is respect for the freedoms of others. But then, when measures are introduced to ensure compliance with this twisted tit-for-tat notion, you claim they're wrong, that they escape the spirit of the license. This is why I think you're confused. That said, it is possible that you disregarded the spirit of the GPL entirely, focused on some of the legal terms and decided that was something you wanted for your project. And that it models what you want for your project better than GPLv3 does, because GPLv3 takes the spirit that you disregarded even more seriously than GPLv2. I still fail to see why what it is in GPLv2 that makes it better to satisfy your intentions WRT Linux than GPLv3. I must assume that, when you say "tit-for-tat", you mean something else, and not respect for others' freedoms. If you take the time to explain what it is, then perhaps it will become clear why you consider the GPLv2 a better license to achieve your goals, or perhaps it will show that you're I can't say that yet. Maybe after the points above are sufficiently Until you started accepting contributions from others, yes. BTW, in Portuguese the correct spelling would be "compreende", with a double 'e'. "Comprende" is Spanish, and in Brazil, where I live, we speak Portuguese. But thanks for trying, that's appreciated ;-) -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
You're a moron. I'm the original author, and I selected the GPLv2 for Linux. Tivo accepted that, and followed the GPLv2. Even the FSF lawyers agreed that they followed it. But Tivo *is* the "author" of their own proprietary applications, and it *is* the designer of their hardware. And exactly like I had the right to the choice of license when it comes to Linux, they have the right of choice to license and behaviour when it comes to *their* software and hardware (that is not a derived work of Linux). But you cannot follow a coherent argument, because you dont' *want* to follow it. Because following the logical argument would take you to a place where you don't want to be. I'm not going to bother discussing this any more. You don't seem to respect my right to choose the license for my own code. So one final time: - I chose the GPLv2, fully understanding that the Tivo kind of situation is ok. - the FSF lawyers too have acknowledged that what Tivo did was not a license violation, so I obviously am not confused about the issue: YOU are. - I think that what Tivo did was not only "technically valid" by the license, it was what I *intended* by my choice of license! And you are apparently totally unable to understand - or respect - that I actually made an informed decision that happens to be different from what you *wish* it were. Linus -
Wow, do you remember the date when you first thought of this business While you insist in the nonsensical tit-for-tat argument and "in kind" retributions, I don't think I have much of a choice, because this is not what the GPL is about, this is not what it requires of licensees. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
You know what? I'm intelligent. That's what you call people who see th consequences of their actions. I didn't see the *details* of what all the GPLv2 could result in, but yes, I claim that I knew what I was setting myself up for (in a license way) pretty much from the beginning. Did it take me by surprise how people actually ended up using Linux? It sure did. But has the GPLv2 itself ever surprised me? Not really. I read license, the thing I cared about was that source code be freely available. That was the first license, but more importantly, it was why I started Linux in the first place - my frustrations with Minix, and my memories of how painful it was to find an OS that I wanted to use and work with. (That, btw, was not Minix-only: I actually originally was thinking about literally buying a commercial Unix for my PC too. The price factor kept me away from the commercial unixes, and in retrospect I'm obviously very happy). So my first goal was "source must be available and it must be free (as in beer)". Which my first copyright license reflects very directly. What happened a few months into the thing was that some people actually wanted to make floppy images of Linux available to Linux users groups, but they didn't want to have to actually *fund* the floppies and their work themselves, so they wanted to sell them at cost (which the first license actually didn't allow!). And I realized that the money angle really wasn't what I ever really cared about. I cared about availability, but people sure could get paid for their effort in distributing the thing, as long as the source code remained open. I didn't want money, I didn't want hardware, I just wanted the improvements back. So given that background, which license do you _think_ I should have chosen? And given that background, do you see why the GPLv2 is _still_ better than the GPLv3? I don't care about the hardware. I'll use it, but it's not what Linux is all about. Linux is about ...
GPL won't get you that. You want a non-Free Software license. It will only as long as people play along nicely and perceive the No. Honestly, I really don't. Even when I try and look at it from your perspective, that you described very beautifully in the rest of the message that I snipped, it's still a mistery to me why you think permitting Tivoization could possibly be advantageous to your project. What is it in the anti-Tivoization provision that gets you any less improvements back? If anything, I'd think that, by not permitting TiVO to prohibit users from running modified versions of your code that they don't authorize themselves, these users would do *more* than TiVO alone ever could, and if a fraction of them contributes something back, you're way better off. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Perhaps there is no benefit in permitting "Tivoization". But at the same time, perhaps there are benefits in not preventing "Tivoization" in Tivo has provided some code changes and improvements to Linux. If they had been totally unable to use Linux due to the license, they would probably have used vxworks or BSD or something else, and Linux would have gotten nothing back. So the Linux source code improved and other Users of the Tivo hardware would be able to do more, sure, but then again, actualyl, maybe not. After all if it ran vxworks or bsd, the user still wouldn't be able to do anything about it. The end result is the same. The answer is also still the same: Don't buy a tivo if you want to change what it does, because it doesn't let you do that. -- Len Sorensen -
Why do you bother sending out emails that just show that you cannot read or understand? I want not just the code *I* write to be freely available. I want the modifications that others release that are based on my code to be freely available too! That's what the whole "tit-for-tat" thing was all about! Doyou even understand what "tit-for-tat" means? Should I use another phrase? Do you understand the phrase "Quid pro quo"? You are living in some alternate world. The GPLv2 gives me exactly what I looked for. Yes, people can do improvements in private, and by keeping them private they'll never need to release them to anybody else. Big deal. I don't care. By keeping them private, I never see the end result anyway, so they Yeah. So stop bothering me then. Go cry on somebody elses shoulder. Just accept the fact that I'm a grown person, in full control of my faculties, and that I'm perfectly able to make my own judgements, and that I don't need to follow the FSF blindly. And it doesn't even matter if you don't understand me. That is, as I've said, _your_ problem. I've done my best to explain to you, but if you are so limited that you cannot understand that other people have other opinions than yours, there really is only so much I can do for you. Go away. Linus -
Because I can't divine what's in your mind, and if you don't make the With the exception of those who choose not to distribute their changes. Or who choose to distribute their changes to people who are Yes. I even wrote an article about that. Yes. It's there in the article as well. The difference is basically in attitude. Tit-for-tat is adversarial (equivalent retaliation), And in what sense the GPLv3 anti-Tivoization clause doesn't? In what sense does it give you *less* of what you want? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
I don't. Could someone please explain it. Specifically: 1) What is "tat"? 2) How can I get some? 3) Where do I go to trade it in? DS -
4) is it legal to consume it in my country? 5) should I have a designed driver when I do? -- |_|0|_| |_|_|0| |0|0|0| -
6) Is that allowed to be a binary-only driver or does it have to come with source code? -
This is the main reason I dislike GPLwhatever: there is no notion of "orginal author". You might have written 99% of the code, that doesn't matter. You have no rights whatsoever once you release something under the GPL (no more than ANYOne else). The GPL is nice for the community, and for the users - but very, very bad towards it's authors (taking all and every right you might have). If John Doe wants to re-release the whole kernel under GPLv3, then all he needs is a website and some bandwidth. -- Carlo Wood <carlo@alinoe.com> -
Every literary work (including thus software) has an author, and that Wrong. The author has a collection of rights which vary by jurisdiction but which are primarily governed by the Berne Convention and its sequels And a very good lawyer (oh and a GPL3 as there isn't one yet...) Alan -
Like, they can release/sell the whole thing under some arbitrary other license at their choice. But once you license it with the GPLv2, then you can't stop anyone else (who got it under that license) from using the code under that license anymore, as such it doesn't matter that You actually had me check the license of the linux kernel :p But really - it has this paragraph that I was refering to in most of it's source files: * 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. I might be wrong, but I always thought that that meant that John Doe is free to redistribute the software under version 3 of the License, as published by the FSF. There are source files in the kernel without this phrase, ie - they just say: * This file is released under the GPL. But then the paragraph from COPYING kicks in, reading: Each version is given a distinguishing version number. If the Program specifies a version number of this License which applies to it and "any later version", you have the option of following the terms and conditions either of that version or of any later version published by the Free Software Foundation. If the Program does not specify a version number of this License, you may choose any version ever published by the Free Software Foundation. Any, 'any version' probably includes version 3 as well. Finally, there are file that don't mention the GPL at all, for example kernel/sys.c just says: * Copyright (C) 1991, 1992 Linus Torvalds But - if it weren't GPL-ed then that would be a violation of the GPL-ed of the rest (Nevertheless, I think the license header should be added I really don't like license discussions - and after reading in the mailinglist FAQ that license posts are taboo here - I was partly annoyed, partly amazed to see this HUGE flood of mails ...
This is true of most licences. Ask musicians about trying to get their For those marked parts yes - thats the authors choice. Some of the kernel is dual licensed BSD even so you can use that bit for all sorts of stuff. Again authors choice, some authors wanted to share code between Linux and other projects. I believe you can buy proprietary licenses to reiserfs too. Some authors like GPLv2 or later, some don't trust the FSF, some will decide once GPLv3 is out, some couldn't care etc.. as the kernel doesn't do copyright assignment all these wishes are respected and that is how it should be. Alan -
Read the COPYING file more closely, and realize that "the Program" has always specified a version number of this license. It used to include it just by virtue of having the COPYING file *itself* be included (and that's v2), but since some people felt that was unclear, the COPYING file has this language pretty visibly at the top: Also note that the only valid version of the GPL as far as the kernel is concerned is _this_ particular version of the license (ie v2, not Not for the kernel. Exactly because the kernel _does_ specify the version. So the sequence you quoted is a non-issue. In other words: any file that does not *explicitly* say that it's "v2 or later" is v2 only. Linus -
You retain the copyright, and in particular the right to relicense. Only if you make the mistake of including the "or any later version" phrase do you allow others to redistribute the work under a different version of the GPL. Although this provision may seem slightly convenient to authors, its effect is to grant a very large amount of relicensing permission to the FSF. It almost certainly doesn't make Well, he also needs one tiny little extra thing: the permission of every copyright holder in Linux. -- Jeremy Maitin-Shepard -
You mean if the original author gets hit by a bus and their estate gets sold to SCO they can't revoke our rights to the code? How is this a down side? And you do have more rights than anyone else: as the copyright holder you can issue other licenses, and you have standing to sue to enforce the code. (If nobody else has a copyright on the code, they don't have standing to sue to enforce the license terms.) (Right now, nobody EXCEPT the FSF has the right to sue somebody to enforce the license terms on something like gcc. Do you find that a comforting thought?) Rob -- "One of my most productive days was throwing away 1000 lines of code." - Ken Thompson. -
Have you ever signed a copyright transfer agreement to the FSF? Obviously=20 not, because then you wouldn't utter such nonsense. The agreement reads=20 that you transfer a non-exclusive right to the FSF to distribute the code=20 under GPL (versions of your choice, they have this right anyway, but making= =20 it explicit is always good), and the right to enforce the license. You=20 still have the right to relicense the work as you like. You also have the=20 right to enforce the license yourself, or to transfer that right to=20 somebody else like gpl-violations.org. The FSF even doesn't require to=20 transfer copyright if you make a GNU project, but if you don't, the FSF=20 won't help you (because they can't). They make very obvious promises about what they care ("four freedoms"), and= =20 that they will be very consistent in doing so. So far, all track records=20 have proven that they indeed are very consistent in doing so - the main=20 controversy here is not whether the FSF protects the "four freedoms", but=20 whether these four freedoms are the right goal, and if they really should=20 try so hard to protect these four freedoms. This part of the discussion is= =20 fully acceptable, what's not acceptable is that the Linus-fancurve claims=20 things the GPL sais which it doesn't (like "tit-for-tat") or doesn't say=20 which it does (like section 6 - direct license from the licensor, and in=20 cases like Linux where no copyright transfer agreements whatsoever exist,=20 these are the individual contributors). Or that Linux 0.something was=20 already under GPLv2 only, when GPLv2 clearly says that there may be=20 updates, and when you as author don't say something, you are allowing users= =20 to update if they like. The last point IMHO makes clear that my interpretation of the comment is=20 valid: This is a commend made by Linus Torvalds, as how he understands or=20 misunderstands the license text. It's not even something you can take as=20 legal advice, because Linus is not a ...
Ever heard of the Open Invention Network or the Software Freedom Law Center? And since when did the FSF become an ally of Microsoft? The rest of us seem to think the Novell deal's a minefield under GPLv2, and would just LOVE Microsoft to reopen the antitrust can of worms by suing somebody over Linux (they're still a convicted predatory monopolist, only the sentence part was defused)... And yet GPLv3 was DOA until the Novell deal happened and suddenly the FSF had this huge thing. Reminds me of everybody rallying behind Bush after September 11 to support abominations like the Patriot Act. It's sad how fear makes people stop thinking. Rob -- "One of my most productive days was throwing away 1000 lines of code." - Ken Thompson. -
So if Linus wasn't entirely clear to begin with that he wanted GPLv2 only, then that is just too bad? Well why not then say that if the GPLv2 didn't say that what Tivo did was bad, then too bad, you can't No the GPLv3 is not the only solution. Perhaps a GPLv2.1 that is actually in the spirit of the VPLv2 with just enough changes to fix the real problems, that may in fact be a solution. The GPLv3 is very much not the only solution. A totally different kernel specific license may even be an option. There are many options, most of them just happen to be pretty hard. -- Len Sorensen -
Of course they can - as any politician. What, do you really think that Bernd wakes up, looks in a mirror and spits in disgust? I very much doubt that. Here's how that kind of stuff works: to be a successful politician one needs to have ideals. The more, the better. Then there always will be an ideal advanced by action one wants to take. Experienced politician is one that will be able to pick those automatically. And feel damn righteous at that. Of course they can have it both ways - just say something like "we work for the greater good of users", "what advances our goals helps the interests of users", "we'll read both cases in a way that helps the interests of users", "we are consistent in helping the interests of users". And at that point any talk of inconsistencies will be deflected by the highly moral considerations. End of story. Sprinkle with generic constructs making the critics easier to ignore ("they do it because they are Bad"/"they do it because they had been seduced by Evil"/ "they just parrot somebody"/"they do it because of <random psychobubble>"/ "they do it because of wrong ideology"/"they do it because of ego") and enjoy. -
Me, I agree that hardware shouldn't lock users. And since I'm one of those evil european socialdemocrats, I may go as far as to think that there should be laws that *forbid* selling such hardware. But I think that all this iss a *hardware* issue. It seems to me that lot of people at the FSF wants to regulate the hardware industry using the influence of free/open software in the computing industry and the "V2 or later" phrase from the GPL. But the fact is that free/open source runs on _top_ of hardware. You don't control hardware, you only control the things that are built on top of your software, not the parts you use to build your software. And the FSF is trying to control the design and licensing of hardware throught the influence of their software. And I think it's wrong. I'm all to forbid hardware that imposes restrictions on hardware, but software licenses are NOT the way to make it. That's a task for a "Free Hardware Foundation", not the FSF. What the FSF is trying to do is EVIL. It's not about free software, it's not about freedom, it's about the FSF trying to have to much control over things that they shouldn't even try to control. I think that the FSF can do a terrible damage to free/open source with such stupid ideas. I wouldn't even be surprised that some jugde rules that a software license that tries to 'control' hardware is invalid -
Amen. And btw, opencores.org does actually exist. I don't even think open hardware is a big issue: the worry-warts about hardware are likely wrong, and hardware today is a lot more open than it used to be even just a decade ago. You can much more easily design your own (FPGA's are cheap and powerful), and yes, it's more complex today, but that's actually an argument _for_ openness rather than against it (open processes work better in complex environments!). The real issue is "open content", and we do actually have various organizations that support that in particular. I would heartily encourage people to get involved with the Creative Commons, and the EFF, and I think Larry Lessig is a really smart and articulate person, who you should I wouldn't go that far (although, in the heat of the moment I probably _have_ gone that far. Oops ;). I don't think the FSF is evil. They're just too single-minded, and look too much at one issue, and only care about the one thing they care about, and in the process, they tend to have a really hard time seeing the other side of the coin. They define "freedom" one way, and by defining it in a very particular way, they miss the fact that what is "freedom" to them is not "freedom" to somebody else. They have a very particular agenda, and in having that agenda and a very strict view of how the world should look (according to the FSF), they dismiss the fact that other people have _other_ agenda's, and see the same world totally differently. And I think that kind of single-mindedness is silly and counter-productive. I literally think that the GPLv2 has worked so well exactly because you can strip it of its high-falutin' morality and the FSF Kool-Aid, and just see it as a "tit-for-tat" license. It allows everybody to see that the work they put in (into the _software_) is protected, and people cannot make improved versions of that software and distribute those improved versions without giving you the right ...
It's not. It's only working to ensure recipients of the Free Software Can you explain to me how it is that the Tivoization provisions (the only objection you have to GPLv3) conflict with this? (nevermind our disagreement as to whether "tit-for-tat" applies to either GPLv2 or GPLv3) -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Is it really that hard to understand? GPLv2 applied only to works people chose to place under that license or to works that contain so much code that someone chose to place under that license that they are legally considered a derivative work. GPLv3, on the other hand, attempts to extend control over works that don't contain any code that anyone ever chose to place under the GPL. This is a night and day difference. The GPLv2 stands within the legal scope of copyright. If I create a work, I have some rights to control that work. If you create a work *based* *on* *my* *work* I can retain some rights over how this new work is used because it actually *contains* parts of my work in it. The GPLv2 makes no attempt to exercise any control over anything else. The GPLv3, however, attempts to leverage copyright control to restrict what can be done with things completely outside the covered works. DS -
^^^^^ Exactly what has been said to you the whole time, but you still refuse to accept that. If Linus develops and runs his code on a PowerPC and I struggle to install the code that he has released for me to modify and share on a PowerPC (maybe because I'm an idiot). Should I create a license with a Linusation term, because he is evil he runs his code on a PowerPC and I can't? -
Depends. In this hypothetical scenario, what did he do to stop you from installing and running the modified version (or even the pristine version, merely recompiled) in your PowerPC? And why did he do it? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Those may be the intentions, but I claim that your statement is false. The anti-tivoisation FSF movement is not "working to ensure recipients of the Free Software can modify and share the software". They can't, because the fact is that hardware vendors can NOT stop you from "modifing and sharing the software". They only can stop you from running your modifications, which is very different. So this is a _hardware_ limitation. It's pointless to try to address this problem with software licenses. What the anti-tivoisation movement is trying to do: "If you are a vendor of tivoized hardware you must give your users whatever information is needed to run modifications of their software" How it works in the real world: "You can't run this software in hardware that doesn't allow to run code modifications of this software" So while the anti-tivoisation movement is trying to limit hardware design/licensing, the fact is that what you are restricting is not the hardware, but the _software_, in a way very different from the 'restrictions' that the GPL has when compared with the BSD ie: in a way that doesn't benefit freedom or contribution of code. Because your users already can modify and share their code regardless of what hardware they're using (even if they can't run their modifications), you're just adding pointless prohibitions. -
Tangent, but that could in fact quite easily be construed as saying that gay people aren't good which I hope is not the point you are making :-/ Rene. -
I certainly read that as 'trying to force people to be good is just as crazy as trying to force people to not be gay'. Some people are good, and some aren't (no idea why), and similarly some people are gay and some aren't (again, no idea why). Neither can be changed by declaring that it must be changed. I always love Linus' analogies. :) -- Len Sorensen -
Yes, just my sense of humour, I'm afraid... ;-) Rene. -
I'm sorry if I gave anyone that impression. My point was that it would be pointless to argue the case in the US because here it really is, usually , "buy the best justice for the money". -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
Or do what BusyBox and uClibc did (on the advice of Pamela Jones of Groklaw) and sign up with the the Software Freedom Law Center so they can enforce your copyrights for you. Didn't cost us a dime, and they were ok with GPLv2 without the "or later" clause... Rob -- "One of my most productive days was throwing away 1000 lines of code." - Ken Thompson. -
Have you ever installed GNU/Linux on a PC "Designed for Microsoft Windows"? And distributing the GPLed software under this restriction is quite ;-) :-P :-D -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Come on, GPL is software licence, the hardware isn't part of the equation. One can argue that keys are or aren't part of the source (= that digital signature is or isn't part of the executable) but it's totally independent of any hardware and its purpose. For example, it doesn't matter if the signature is merely for checking file integrity (and any signature would do) or if it's for restricting users from running something. -- Krzysztof Halasa -
And if Tivo did sell their crap in Germany, I bet, Harald had brought them= =20 down years ago (as he did in the "tivoized" Siemens router case). But Tivo= =20 doesn't (they started in the UK, and stopped doing so right after Harald=20 unlocked that Siemens router ;-), and in the US, courts may think=20 different. Or they rely that there simply is no Harald Welte in the US, who= =20 goes after the violators. =2D-=20 Bernd Paysan "If you want it done right, you have to do it yourself" http://www.jwdt.com/~paysan/
See also the September 29, 2006 entry where we set up an email address to forward license violation reports directly to them so we wouldn't have to deal with any of it. I'd say this "hasn't cost me a dime", but I believe I'm on my third stamp. (I also note that I'm not busybox maintainer anymore and Erik isn't uclibc maintainer anymore either, but since it's our copyrights they're basing the enforcement actions on, they still bounce an email off us every few months.) Rob -- "One of my most productive days was throwing away 1000 lines of code." - Ken Thompson. -
One of the distinct advantages of German law over the US system (and to a large extent the UK system its based upon) is that German law favours That in theory shouldn't happen as the conventions on copyright are supposed to stop that mess occuring. -
Is there any way how this would be resolved?
I can easily imagine that two courts, no matter whether they are in the
same or different countries, would decide differently in grey areas like
non-GPL modules or the GPLv2 and private keys.
If the two courts are in the same country there's usually a higher court
above both that can resolve this. But what if let's say the highest
court in the USA and the highest court in Germany would disagree on such
a matter?
cu
Adrian
--
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed
-
It would be very unusual for both cases to proceed in parallel. The one court will then strongly consider the decision of the other. -
Upgrade the license so as to provide guidance as to the intent of the authors, such that the disagreement doesn't happen again. If there's room in each country's laws to fix the problem, that is. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
I don't think that's an option.
Consider the question of whether non-GPL kernel modules are legal at
all and the number and different opinions of Linux kernel authors.
Plus the general question whether any "upgrade the license" would be
valid in all jurisdictions ("GPL version 2 or any later version"
as licence might make it possible, but in all other cases I'd have
serious doubts).
cu
Adrian
--
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed
-
No. That's the question as the FSF would like to frame it. But the real fact is that it *not* the right question. You can install Linux on a Tivo all you like. Take out the harddisk, install your own version of Linux on it, and put it back in. That's pretty much how Tivo installs Linux on the things too, afaik, although they don't need to take the disk out (since they just assemble it). No magic needed. In fact, no keys needed. Now, maybe the hardware/firmware knows to expect a certain SHA1 on that disk, that's a different issue. Tivo could even tell you exactly what the SHA1 they are checking is. Maybe they have a list of SHA1's, and maybe they have a way to upgrade THEIR OWN FIRMWARE with new SHA1's, and they could still tell you all of them, and be very open. And you could actually replace their copy of Linux with another one. It would have to have the same SHA1 to actually start _running_, but that's the hardware's choice. See? No private keys needed. No magic install scripts. It really _is_ that easy. Of course, using private keys, and signing the image with them is possibly a technically more flexible/easier/more obvious way to do it, but in the end, do you really want to argue technical details? But I think the whole thing is totally misguided, because the fact is, the GPLv2 doesn't talk about "in place" or "on the same hardware". So take another example: I obviously distribute code that is copyrighted by others under the GPLv2. Do I follow the GPLv2? I sure as hell do! But do I give you the same rights as I have to modify the copy on master.kernel.org as I have? I sure as hell DO NOT! So by the idiotic logic of "modifying in place", I'm violating the GPLv2 every time I'm makign a release - because I make Linux available, but I don't actually give people the "same rights" to that particular copy that I have! Oh horrors of horrors! You need to make a _copy_ of the thing I distribute, and then you have the same rights I have to ...
No. The FSF actually does *not* want to take this position. That's why it chose the formulation of Installation Instructions. It doesn't share my view that the keys needed to sign a binary in order for it to That's the hardware imposing a restriction on modification of the software. It doesn't matter how elaborate the excuse is to justify denying users' freedoms: it's against the spirit of the GPL, and the That's an interesting argument. People don't get your copy, so they're not entitled to anything about it. When they download the software, they get another copy, and they have Yes. You see how TiVO is different? It is modifyable, and I actually Indeed, it's something bigger. But this doesn't exclude the smaller There is a difference. But the GPL doesn't limit itself to the former. It explicitly talks about "copies". -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
And? There is *absolutely* *nothing* in any version of the GPL *prior* to 3 that says that hardware cannot impose restrictions. What the GPL *does* say is that you can't "add additional restrictions to the license" - (IMHO) a piece of hardware having a restriction isn't an "additional restriction added to the license". As well, as Linus stated, there is nothing *anywhere* - AFAICT, not even in GPLv3 - that says that you have to be able to run the software "in place" or "on the same hardware". If a hardware manufacturer - like TiVO - uses GPL'd code in their product - and complies with the terms of the license - they aren't required to allow you to run modified code on that hardware. Without it mentioned anywhere in the GPL *OR* the assorted writings of RMS (who founded the FSF and wrote the original GPL) that "modified software must be able to run on the same But you get the TiVO corporations copy of the software? I smell a logical I don't. You don't get the TiVO corporations copy of the software. You get your own copy, with all the rights that TiVO had when receiving the software. The right to install and run the kernel in the TiVO device is independent of the rights to copy, modify, distribute and run the software. (because the GPL never guarantees you the right to run the software on a particular piece of hardware.) DRH -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
It's not that the hardware is deciding to impose restrictions on its own. It's the hardware distributor that is deciding to use the hardware to impose restrictions on the user. Seems like a violation Not quite. It's more general than that: You may not impose any further restrictions on the recipients' Yes. The customer gets the copy that TiVO stored in the hard disk in the device it sells. And it's that copy that the customer is entitled to modify because TiVO is still able to modify it. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
You *still* haven't figured out the difference between "the software" and "a particular copy of the software", have you? What's your problem? I doubt you're really stupid, so I think your problem is that if you admit that "the software" is something *different* from "a particular copy of the software", you realize (perhaps subconsciously) that your arguments do not make any sense. So you do not allow yourself to think clearly about the matter. So let's look at that "section 6" that you talk about, and quote the relevant parts, will we: You may not impose any further restrictions on the recipients' exercise of the rights granted herein. and then let's look at Red Hat sending me a CD-ROM or a DVD. Now, Red Hat clearly *did* "further restrict" my rights as it pertains TO THAT COPY ON THE CD-ROM! I cannot change it! Waa waa waa! I'll sue your sorry ass off! See the issue? You are continually making the mistake of thinking that the GPLv2 talks about individual copies of software. And I'm continually having to point out that that is a MISTAKE. And it's OBVIOUSLY a mistake, because your reading is nonsensical. If you think that Tivo does somethign bad, then hat Red Hat does is the same badness, thousads times over! I strongly suspect Red Hat has shipped a lot more CD-ROM's than Tivo has shipped boxes! So let me iterate AGAIN: - the rights that the GPLv2 gives *cannot* be about "the particular copy" that you send, since that would be INSANE. Red Hat sends lots of copies of software that are NOT MODIFIABLE! - ergo, the rights about "the software" in the GPLv2 must be about something else. See? Your argument about "individual copies" simply DOES NOT MAKE SENSE! And your point is? Nothing. The rights granted are the rights to "distribute and modify the software". But by "the software", the license is not talking about a particular *copy* of the software, it's talking about the software IN THE ABSTRACT. In other ...
I have. And so has GPLv2, look: 2. You may modify your copy or copies of the Program or any portion Red Hat is not stopping you from making changes. The media is, and that's not something Red Hat can control. Compare this with the TiVO. TiVO *designs* the thing such that it can still make changes, but customers can't. That's the difference. TiVO is using hardware to "impose further restrictions on the recipients' exercise of the rights granted herein", and this violates It does. You're making the mistake of thinking that it doens't. And even in the legal terms that you claimed to have understood so More specifically, some of the rights are: copy and distribute verbatim copies of the Program's source code as you receive it modify your copy or copies of the Program or any portion of it, thus forming a work based on the Program, and copy and distribute such Please read it again. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
On Thursday 14 June 2007 14:35:29 Alexandre Oliva wrote: TiVO isn't stopping you from making changes - the *media* is. (in this case the "Media" isn't even doing as much as a CD-ROM does. The only thing a TiVO No, it isn't. Look at any motherboard. The Bios on the last three or four motherboards I've purchased check for a digital signature on the Bios updates. The motherboard manufacturer can make changes, but the customer No, they don't. The GPLv2 makes no provisions for you being able to execute a modified copy of the code on the same media or hardware that you received it on. The fact is that claiming it was "the spirit" doesn't matter at all - this isn't philosophy you're arguing, its *LAW*, and in law, if it isn't And where does it say that you even have the right to run the "work based on the Program", or even a self-compiled copy of the "verbatim copy of the code" Done. Section 3 of GPLv2 covers the right to distribute "object code" forms of a licensed work. At no point does it even *mention* that, if the object code form comes on a device capable of executing it, you have to give the right to execute a modified form of the work on the same platform. If this has been the "intent and spirit" of the license from the beginning, it should be there somewhere. DRH -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
TiVO made it so, that's the difference. I'll give you that it's not so much about making changes per se, or even installing them, as it is about running the modified versions for That's exactly what makes for the difference between the spirit and It doesn't. The license can't demand the software, or modified versions thereof, to run. The only thing it can demand is that 2. You may modify your copy or copies of the Program or any portion I think you're missing what 'spirit' means. It's guidance, it's not the legal terms. And it's precisely because the implementation (the legal terms) failed to meet that design (the spirit, encoded in the preamble) that the license needs patching. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
By your reasoning it doesn't even matter. I own the hardware, I should be able to change the BIOS to *any* chunk of code I want. Do you see the fallacy here? You're making an artificial distinction based on And the reason behind this is all "ethics and morals". In other words, you are forcing those "ethics and morals" on others and hiding it by giving it a different name. Wasn't it Shakespear who said: "What is in a name? A Rose by any other name No, it doesn't. There is no requirement in the license in question that makes a persons ability to run the program on any given piece of hardware. What it If the intent of a law (or license) is to do A but it doesn't say that, then how is the intent to be known? Your answer: Ask the author. Question: how can we be absolutely certain that the authors intent *hasn't* changed since the law (or license) was written? *ONLY* answer: It is impossible. Conclusion: Unless the intent is clearly spelled out at the time the law (or license) is written, or is available in other writings by the author of the law/license from the same time period as the law/license then it is impossible. Question: How do you know what the "spirit" of a license is? Your Answer: Ask the author. Question: How do we know that the Author hasn't changed their mind about the "spirit" of the license since it was written? *ONLY* Answer: See the answer to the parallel question about "intent". Now that I've knocked down your "Intent" and "Spirit" straw-men you have no way to argue that the GPLv3 is written with the same "spirit" as the GPLv2. DRH -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
What matters to me is that, when the GPL says you can't impose further restrictions, then you can't, no matter how convoluted your argument There was never any attempt to hide that this was what the Free Software movement was about, and that the GPL was about defending these freedoms. Sure, it has other advantages. But the goal has always been the same, No, you interpret based on what the author wrote then. You read the preamble, and any other rationales associated with the license or law. I don't know how it's elsewhere, but in Brazil every law has a rationale, and that's often used to guide its interpretation in courts, even though the rationale is not part of the law. If the author realizes what he wrote was not enough, or it got misinterpreted, author his text, and then whoever feels like it and is entitled to adopts the revised version. In the GPLv2=>v3 case, all that needed revision was the legalese. The preamble has barely changed. This is a strong indication that the Is there anything not clear about freedom #0, in the free software definition, alluded to by the preamble that talks about free software in very similar terms? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Convoluted? Not in the least. Every example I have given has been an example of the application of your logic. If my examples are convoluted, then, QED, I'm not trying to change that. My point in making that statement is to prove that the FSF is doing exactly what the Spanish Inquisition did, what every "Communist Revolution" has done and what Hitler did. Saying "My ethics and morals are better than everyone else, so I'm going to force everyone else to have my morals and ethics". That the FSF isn't doing this through force of arms or threat of violence just shows how sophisticated people have really become in the sixty years that have passed since Hitler - they now use threat Really? Well I must say I'm surprised at the sudden change of heart. I have several mails here in which you have either said "You ask the author" or that Show me where in the preamble that this issue of "it must run on any given piece of hardware" or even less generally, "it must run on the hardware it came on" is even *hinted* at. You wont find it. Nor will you find any mention of anything of the sort in the publicly available writings of RMS. But let me go re-read the GPLv2 preamble again and see if it even hints at this issue... oh, wait, I read it earlier and didn't see anything that hinted at this. So I can safely conclude that no lawyer or judge would find it when interpreting the license. QED: The Tivo clause of GPLv3 causes it to break spirit with the GPLv2. (And, by the way, if the FSF decided to release a GPLv4 that had an active section that said "You must turn over all copyright rights to a work released under this license to the FSF" it wouldn't "break spirit" with the GPL (v2 or v3). Why? Because *both* contain the following paragraph: "We protect your rights with two steps: (1) copyright the software, and (2) offer you this license which gives you legal permission to copy, distribute and/or modify the software." By your logic it is the *intent* of the FSF to ...
I didn't say your arguments were convoluted, and I know I didn't mean to say that. But I've heard enough arguments about excuses to escape the obligations of the GPL (and other licenses and obligations, FWIW) to know that such arguments can get very convoluted. That said, I was actually trying to quote Eben Moglen, who once spoke about this, but the word he used was "elaborate", not "convoluted". It's no change. You interpret what's there. If it's clear, good. If there's a dispute, you have to ask the author, only s/he knows what Why is the burden of the proof on me? You show me where it says "one may impose restrictions on what particupar pieces of hardware the program can run", to override the Can't. These terms wouldn't apply to the copyright holder (the only Erhm... How did you get the (completely flawed, BTW) impression that tivoization was all GPLv3 was about? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
And as I have hopefully given good proof for, asking the author is not a good solution. The author can change their mind about their intent at any point in It's "pass on all rights granted under this license". If I had to pass on "all rights I have" I'd have to pass on my right to change the license on my code. Since that isn't a right I'm obligated to pass on - and you could never It's part of the preamble, in which the "We" refers to the FSF. If the preamble determines the "intent" and "spirit" of the license, then part of the "intent" and "spirit" of the license is collective aggregation of all I've looked through the GPLv3 and "tivoization" and DRM are the only things that are functionally different. In reading the GPLv3 *again* today I got the impression that there are more restrictions than grants of rights. DRH -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
http://fsfla.org/svnwiki/blogs/lxo/draft/gplv3-snowwhite discusses each one of the significant changes (and some of the insignificant ones) and shows why each one of them is more "tit-for-tat" than v2. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
No, by this twisted logic Tivo *cannot* modify that particular copy any more than you can. They can modify *another* copy (just like you) and they can *replace* the copy in your device with the new version (unlike you). So your entire logical construct does not stand because this is not (cannot be) about modifying a particular copy (how would you do that anyway? hexedit the binary blob in place?) but about the ability to deploy the software on a particular platform. --- fm -
Again, replacing is one form of modification. What do you think you do when you save a modified source file in your editor? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
No, it's not: replacing does not create derivative work. Modification does. You've chosen to attach a physical dimension to "program copy" and I'm arguing that even under this distorted line of reasoning you can't * Tivo takes public sources, modifies them and builds a brand new blob * Tivo installs this new copy on the device, most likely side-by-side with the old one - notice how the new copy is derived from public sources and has absolutely nothing to do with the old version (heck, it can be a totally different kernel for what it's worth) * Tivo deletes the old copy from the device It seems pretty obvious that the only right Tivo is withholding is the right to install new versions on the device - they never do (and really never could) "modify" the physical copy on your device (which is your Don't skip the part where the in-memory version started as an exact copy of the original being replaced. Notice the difference? ;) --- fm -
Thanks. Good point. This convinces me that this doesn't work as a legal argument under copyright. I still stand by my understanding that this restriction violates the spirit of the license. And since the specific implementation involves creating a derived work of the GPLed kernel (the signature, or the signed image, or what have you) and refraining from providing the corresponding sources to that derived work (the key and the signature "build scripts"), I still think this specific case is a violation of the letter of the GPLv2, Actually, no. They withhold the right to run versions that they don't authorize themselves. Back when GPLv2 was written, the right to run was never considered an issue. It was taken for granted, because copyright didn't control that in the US (it does in Brazil), and nobody had thought of technical measures to stop people from running modified copies of software. At least nobody involved in GPLv2, AFAIK. The landscape has changed, and GPLv3 is meant to defend this freedom Qualifying it as the main argument is a bit of an exaggeration. I have a number of different arguments. The one about incomplete Sorry, I really don't follow. Both versions of the kernel binary also started from a common source ancestor. Were you trying to make a distinction on these grounds? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Wait, a signed filesystem image that happens to contain GPL code is now a derived work? Under what sort of interpretation does *that* occur? (This pretty much throws the 'aggregation' premise in GPLv2 completely out.) Bill -
On Thu, 14 Jun 2007 19:45:08 -0400 Perhaps the FSF will in future remember to pack a copy of the GPL in each of its md5sum files on the mirror if this is a derivative work, and modify the bittorrent protocol to include a copy of the GPL in the seed files 8) Alan -
I realize you're joking, but for the benefit of anyone who might not understand how this works: A derivative work must, first and foremost, be a work. If it's not a work, it's not a derivative work because a derivative work is a type of work. Aggregations of multiple works, when that aggregation is performed in an automated way, are not works. Even if I compile and link a bunch of source code, provided there is no creative input in the compile and link process, the result is not a work for copyright purposes. It is simply an aggregate of the bits of source code. The gist of a compilation or derivative work is the creative selection and modification process. If someone argues that a program is a derivative work of a header file it was compiled with, he is probably just being sloppy. The resulting executable contains the header file combines with other works. Of course, a source code file that is designed based on the contents of a header file may be a derivative work of that header file, but that would be because the human who wrote the source code file used bits of the header file in the source code itself. It would not be because the compiler combined them -- such an automated combination has no creative input and so cannot produce a work, and so cannot produce a derivative work. This is grossly oversimplified, but should give you the idea. DS -
Is the signature not derived from the bits in the GPLed component, as much as it is derived from the key? Isn't the signature is a functional portion of the image, i.e., if I Not really. It could take some explicit distinguishing between functional and non-functional signatures, but that's about it. GPLv3 chose a different path to make this clarification. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
OK. Let's take this to the simple and logical conclusion. A signed filesystem image containing both GPL and non-GPL code. From your point A, this is a derived work. Let's read the license... 2. b) You must cause any work that ... is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License. ... But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it. and yet later: In addition, mere aggregation of another work not based on the Program with the Program (or with a work based on the Program) on a volume of a storage or distribution medium does not bring the other work under the scope of this License. Pick one. They can't both be valid. Moreover, this interpretation means that Red Hat (and pretty much any other Linux distributor) should close up shop, as that's what we've been doing for years. Bill -
I claim the signature is derived from the GPLed bits, yes. Whether that's a derived work, in the legal sense, I'm not qualified to say. And I claim that, in the case of TiVO, it is not only a functional piece of the system that's derived from GPLed code and missing the corresponding sources, but also it's being used to impose restrictions on the exercise of the freedoms that the GPL is designed to protect. And these conditions are what make it a bad thing, and that deviate, if not from the legal conditions, at least from the spirit of the license. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
it's also derived from the non-GPLed bits as well. so if it were a derived work in a legal sense (nessasary for your argument to have any legal meaning) then it's now illegal to make and distribute a checksum of a CD that contains software with incompatible you keep claiming this, but other people claim you are wrong. what good do your claims do (why are your claims about what's in the spirit of the license and what's not any more valid than anyone else's?) David Lang -
It is not necessary for the end item to be a derived work in order for the GPL to apply. A literal copy is not a derived work; a translation is not a derived work; an executable version of a program is not a derived work of its source code; and so forth. What is necessary is that the "work based on the [GPLed] Program" be more than a mere aggregation of the GPLed component(s) with non-GPLed components. The fact that part of the work-as-a-whole is a descriptor of the GPLed part does not mean all descriptions the GPLed part are governed by the GPL. The critical factor is that the GPLed part will not function properly without the DRM signature. Michael Poole -
Actually, you can't copyright, trademark, or patent a number. In order to copyright something it has to have some creative element. You also can't copyright (or trademark) book titles. So no, last I checked you can't copyright an MD5sum or SHA1sum. I vaguely recall somebody dredging around for the smallest thing there was a legal precedent explicitly affirming you could copyright it, and it was a haiku. So they put an uncompressed ascii haiku in their protocol... Now if you sign the executable binary, then the binary (as a whole) is a derivative work of your copyrighted code etc. ad nauseum pluribus unum and so on. And THAT is due to Apple vs Franklin in 1983: http://www.internetlegal.com/impactof.htm Before which copyright was only guaranteed to apply to source code, not to binaries (which are basically big numbers). That's why everybody distributed source code before then: it was the only thing they knew you could enforce a copyright on... IBM's "Object code only" initiative happened around the same time... http://landley.net/history/mirror/ibm/oco.html Along with the AT&T breakup commercializing Unix, the launch of the GNU project, and the general rise of "shrinkwrap" software. (There's this marvelous book called "Legal battles that shaped the computer industry" by Lawrence D. Graham, devotes a few pages to Apple vs Franklin. Franklin honestly didn't think Apple's binary ROMs were copyrightable. Just as in a 1980 interview with Bill Gates, he couldn't stop somebody from printing a book with an annotated printout of the TRS-80 ROMs Microsoft had a copyright to. He sounded kind of pissed about it, actually. Also young and whiny: Transcript: http://slashdot.org/features/00/01/20/1316236.shtml Audio: http://landley.net/history/mirror/ms/gates.mp3 ) Rob -- "One of my most productive days was throwing away 1000 lines of code." - Ken Thompson. -
Agreed. And this counter-argument of yours is a distraction. I was careful to not talk about "derived work". Please read it again under this clarification (that I'm pretty sure I'd already made before, but it's getting hard to keep track of everything in this thread ;-) -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
"Is the signature not derived from X as much as it is derived from Y." "I was careful to not talk about "derived work"." I'm going to stop feeding the troll now... Rob -- "One of my most productive days was throwing away 1000 lines of code." - Ken Thompson. -
The one that speaks English, not Legalese. IANAL. Last I looked it up, "derived" was a plain-English word. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
On Thursday 14 June 2007 17:39:32 Alexandre Oliva wrote: Not entirely correct. If TiVO is making a change to the binary to include the signature, then it *could* be considered a derivative work. If the signature is stored in another place - say a bit of Flash or a separate file on the disc - then there is no way for it to be considered a derivative work. (Under US law, IIRC and I I've interpreted it (and the related cases) properly then the change would have to be to the source of the program for it to be considered a "derivative". But, as you say often and I should make clear And this is relevant to a software license in which way? In particular how is this relevant to the GPL, which has always *only* guaranteed access to the source if you have access to the binary, the right to distribute your own versions and the right to modify the code. Since the "right to run code" was never guaranteed it *cannot* be a violation. It might be in conflict with what RMS intended when he wrote the first version of the GPL and in conflict with the intent of the people that contributed to GPLv2 but that doesn't matter. However, I will not use (or recommend) the GPLv3 in its current form because I feel it makes unnecessary restrictions. The fact that you have to "allow additional rights" to make it equal to the GPLv2 makes a functional (and spiritual) difference to me. (Why? Because I'm opposed to "In order to protect freedom X we have to restrict freedom Y. Its happening in the US *RIGHT* *NOW* and I have been doing what I can to fight that. Now the same faulty logic is being applied by Why isn't it in the US? Because the binary form of a program does not and cannot have a separate copyright than the source code. Since it is the *SOURCE* that is actually copyright (mechanical translation cannot create a new work, only a new form of an already copyrighted work) guaranteeing the "right to run" is pointless. And you are wrong about that "Nobody thought of it" thing - ...
But since this elusive "spirit" is subject to everybody's interpretation of the preamble, you must surely admit that it remains just a matter of On that particular piece of hw, yes. But who's granted you the right to *run* your modified copy *there* in the first place? GPLv2 explicitly steers clear of anything "other than copying, distribution and Then you agree that GPLv2 does not protect your freedom (taken for granted) to run a modified copy on any particular device, or am I Exactly: they have a common ancestor, they are both derived from it. But there's no ancestry relationship *between* them (unlike your edited file example) so you cannot argue that one is a modification of the other. Hence, Tivo is not really *modifying* the copies it distributes with the device - they're *installing* brand new copies instead. They also choose not to offer everybody the same privilege :-| Does this go against the intent of the GPLv2 authors? Probably. Does it go against the letter of GPLv2? Apparently not. Does it go against your/some people's interpretation of the GPL "spirit"? Obviously. Does it go against everybody's interpretation? Obviously not. --- fm -
IANAL, but AFAICT it doesn't. Still, encoded in the spirit (that refers to free software, bringing in the free software definition), is the notion of protecting users' freedoms, among them the freeom #0, to run the software for any purpose. That's why I believe it's in the spirit of the license to defend this freedom. And that's why lawyers in Brazil believe that, even though the GPL does not affirm the right to run the software, it fits the bill, because, under the light of the preamble, the free software definition, and the US copyright law, it should be interpreted as an Got it. That's bad. :-( -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
And where in GPLv2 is "Freedom #0"? As a simple matter of fact, the *only* activities covered by the GPLv2 Then they have made a bad decision. While it can be argued that "the right to run the software" is guaranteed, the truth is that the license is very clear about what it covers. That's *DIRECTLY* in section 0 of the license. If someone has interpreted it to cover something besides what it explicitly states then it has been badly interpreted. In case you don't remember, GPLv2, section 2, paragraph 2: "Activities other than copying, distribution and modification are not covered by this License; they are outside its scope. The act of running the Program is not restricted, and the output from the Program is covered only if its contents constitute a work based on the Program (independent of having been made by running the Program). Whether that is true depends on what the Program does." In other words, the license cannot be sanely interpreted to cover *execution* of the program. Yes, it says that the *license* doesn't restrict you from running the program, but that *DOESN'T* matter, because the opening sentence says: "Activities other than copying, distribution and modification are not covered by this License; they are outside its scope." QED: The intent of the license is clear and it is to guarantee those three stated rights. -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
Unless I have explicitly installed linux myself in the box, I have received the binary from them, so it can fall in the distribution case. -- Glauber de Oliveira Costa. "Free as in Freedom" http://glommer.net "The less confident you are, the more serious you have to act." -
Sorry if you missed the rest of the discussion, but the above statement was a rebuttal of the "The GPLv2 intended to guarantee me the right to run the software on any given piece of hardware" argument that has been used as the justification for the addition of the "tivoization" language to the GPLv3. As I stated, I fail to see how "running" the program is, in any way, intended by the license, since it *explicitly* states that it only covers "copying, distribution and modification". The exact place where it does that is "Section 0, paragraph 2, first sentence". I'll quote it here again: "Activities other than copying, distribution and modification are not covered by this License; they are outside its scope." DRH -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
It may sound like thin evidence for someone arriving from Venus today, but the preamble talks about "free software", some passages clearly imply that software under this license is "free software", the license is published by the Free Software Foundation, and the Free Software Foundation has a published definition of Free Software that establishes the 4 freedoms. The freedoms defined there resonate very strongly with the freedoms/rights that the license talks about. I hope this is enough evidence to convince you that this is the intent. The only of the freedoms that's not explicitly mentioned in the preamble, the freedom to run the software for any purpose, is mentioned in the legal terms as unrestricted, which is very much in line with freedom #0, but is outside the scope of a copyright license because running the program does not require copyright permission. I'll give you that the preamble doesn't make it clear that the license is purported to defend freedom #0 too. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
And that doesn't matter. In the context of the GPLv2 the only legally active
parts *ARE* in the GPLv2, under this heading:
GNU GENERAL PUBLIC LICENSE
TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
No other text published by the FSF has any legal bearing on the GPLv2 *except*
when it is the FSF that holds copyright to the work and has placed it under
the GPLv2. When I place a program under the GPL, it becomes *my*
interpretation of the license and those "published texts" I might have (that
Nope. Because the intent of the author of the license is worth nothing. The
intent of the person who has placed the code under the GPL is, however, worth
In the context of the license it means that the copyright holder is placing no
restrictions. And that is because they legally *cannot* - as you agreed. But
the copyright holder *can*, OTOH, make it clear that they wish for there to
And the preamble carries no legal weight.
DRH
--
Dialup is like pissing through a pipette. Slow and excruciatingly painful.
-
Doens't matter for what? To indicate what the Linux copyright holders meant? Sure it doesn't. I never claimed it did. To indicate what the authors of the GPL meant? To indicate the spirit of the license they wrote? Yes, it matters a lot. And the latter is what my participation here is all about: to show that the spirit didn't change at all. Until you acknowledge and understand this, I should refrain from answering your other postings. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
GPLv2 section 0 says: Activities other than copying, distribution and modification are not covered by this License; they are outside its scope. The act of running the Program is not restricted, ... The license does not cover running of the program. It doesn't restrict it, but it doesn't cover it. Claiming otherwise is turning the GPL into yet another dreaded EULA. Nowhere does the GPLv2 define modification as "modify and run in place". The Preamble emphasizes sharing; hardware is a fixed object and can't be shared in the same fashion as software. Also, GPLv2 section 2 includes: In addition, mere aggregation of another work not based on the Program with the Program (or with a work based on the Program) on a volume of a storage or distribution medium does not bring the other work under the scope of this License. TiVo's firmware (and any restrictions it may carry) is not affected by the GPLv2. -
For many juridisctions loading from disk into memory is copying and in some from memory to CPU cache a second copy. This is one reason as I understand it GPLv3 talks about "conveying" - to avoid that mess and Really irrelevant to the discussion. Tivo's firmware is up to them. Whether the resulting system permits them to include GPLv2 software with it is what matters. Alan -
Hmm... This is interesting. Let me sidetrack a little bit. Who would be held liable should the copy not be authorized by the copyright holder? The designer of the hardware? The seller? The person who powerer the computer on? The author of the boot loader (if +1 It all boils down to whether they're keeping their promise to not impose further restrictions on the exercise of the rights granted in the license. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
The person copying the software (read: the poor user) - of course with no doubt. Is the manufacturer of a knife or rifle or car or brick responsible if I And to solve the above "legal" problem, (at least in .at) it is explicitly legal (and stated in the local law) to "copy" software for personal use as long as you don't pass it on to others. And this includes (of course) e.g. backup copies on DVDs (and remember, you are not allowed to give them away, just store them in your desk). Bernd -- Firmix Software GmbH http://www.firmix.at/ mobil: +43 664 4416156 fax: +43 1 7890849-55 Embedded Linux Development and Services -
Umm. I notice how you must have known how *idiotic* your response was, because you snipped away the part where I talked about Red Haty distributing CD-ROM's. In other words, Red Hat distributes copies (and yes, you *get* that copy), and you cannot modify that copy that you got. So the "right to make changes" _must_ be separate from the actual copy of the image. And don't get fooled by the "all the rights that you have". That _obviously_ and clearly talks about "the program", which in turn equally obviously and clearly has to be about something bigger than the "one copy", since the GPLv2 requires you have the right to change it. So you edited out the part where I talked about CD's. That's the proof that your reading is untenable, because obviously you cannot change the program on the CD: you got a copy, but the right to make modifications Hey, I'm not exactly known for being polite. I tell it how I see it, and I You keep on harping on that "modifyable", but no-where in the GPLv2 is that an issue. I claim that it *cannot* be an issue, since CD's are obviously ok. So the "modifyable" part is a totally new thing to the GPLv3. You cannot use that as an argument that the GPLv3 didn't change things, that's a circular agument: "the GPLv3 says so, so thus the GPLv3 is in the same spirit as the GPLv2". Doesn't make sense. The fact is, the GPLv3 does fundamentally new things. Things I didn't sign up for (and things that nobody _else_ signed up for either) when I chose the GPLv2 for the kernel. The fact that some people would like to change the kernel license to GPLv3 is no different from the fact that some other people would like to cgange the kernel license to the BSD license. Those people who have argued for using the BSD license, btw, argued so in the name of "freedom". No different from you. Do you think they were right? If so, why the hell do you think _you_ are right? So here's what it fundamentally boils down to: - do you admit ...
Compare the preambles of v2 and v3 and you'll understand why the But individual freedom, rather than community freedom. Because, like you, I'm always right, even though not everyone agrees Yes, of course. The new legal terms are answers to new threats to the freedoms depicted in the preamble, that didn't exist or hadn't been The thing is I'm not arguing that point. I'm disputing that there was a change in the spirit of the license between v2 and v3. Heck, a mere 48 hours ago you couldn't even tell the spirit from the legal terms. I still think v3 will serve better any Free Software community, because it will push away the abusers that contribute little, or turn them into cooperative or at least harmless participants, that further enable the active participation of their downstream users. This would enable wider participation under the same 'tit-for-tat' conditions that you attribute to GPLv2. It appears to me that the only significant point of contention remaining is the issue of Tivoization. If you feel so strongly about permitting Tivoization, even though it denies the freedoms that the original spirit of the license you chose says they are entitled to have, you can make this provision by means of an additional permission Yes, indeed. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
That's TOTALLY IRRELEVANT! There is no language in the GPLv2 (only in the GPLv3 drafts) about "same upgradability as third parties". You're arguing a point that DOES NOT EXIST in the GPLv2. The GPLv2 talks about specific rights, like the ability to make changes and distribute things, and says that you have to give downstream all those same rights. And I've pointed out to you (now about five times) that those rights CANNOT be able "in-place", since even Red Hat does not actually give you No. It does not. I have extra rights as a copyright holder, and that "the rights you have" are as they pertain to the software under the GPLv2, not as it pertains to the physical device, or outside the GPLv2. For example, for any code that I have full copyright over, I have rights that you DO NOT HAVE! I have the right to re-license it under some other license. The fact that I pass on a copy of the software to you under the GPLv2 does *not* give you those rights, but that's not even what the GPLv2 asks for! The GPLv2, when it talks about "passing on the rights", talks about the rights you got *per*the*GPLv2*. Any other reading is nonsensical, since the copyrigth owner *always* has more rights than a licensee! I legally literally *couldn't* pass over all the rights I have to my software! If you read the GPLv2 as meaning that I have to, you are mis-reading it. It's that simple. Anyway, I'm not interested in continuing this flame war. The fact is, the license for the kernel is the GPLv2. And I think it's a superior license. As such, I'd be a total moron to relicense the kernel under what I believe is a worse license. So if you want to argue that I should re-license, you should argue that the GPLv3 is better. And quite frankly, you haven't. Linus -
The spirit gives the intuition of "passing on all the rights". The legal terms have to be more careful about that, to avoid the very situation you're debating, so they state "you can't impose further restrictions on the exercise of the rights". No dispute about that, and this is irrelevant to this point. I've already responded and clarified this point 2 or 3 times in this thread. Do you need me to find a URL for you? It was in respose to In fact, I haven't even tried. So far, I've merely been trying to show that it still follows the same spirit, dispelling the muth that it doesn't, and trying to understand why you think GPLv2 is so much better, which I think is related with tit-for-tat and retribution in kind. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Maybe half a brain can, but anyone with a whole brain can assure you its The keys required to make the code run with the hardware are part of the software. The license requires the software and relevant scripts etc are included. Thus there is a very good argument that the keys are part of the software. And since there is no court ruling to high enough level in the USA, UK or any other jurisdiction on that it remains a matter of opinion. Tivo may control the hardware but the authors control the software (via the GPL), and subject to the limits of what may be specified by a copyright license (as opposed to contract) can make such demands as they You can't patent algorithms either Alan -
Good argument, but I'll stand by my interpretation of the law, the GPL and the situation until there is solid proof that a signing-key is part of the source code. Doubly so because the language of the GPLv2 makes it clear that "all relevant scripts, etc" are only needed to build and run the "covered work" - not for proper installation of it. (and, in the case of a TiVO, the signing keys are part of the installation, not the running or building. Besides needing the proper signing key, the kernel in a TiVO is run the same as any Agreed. However, AFAICT, TiVO meets the provisions of the GPLv2 - they make the source of the GPL'd part of their system available. (And I'm not going to get into arguments over whether kernel modules are "derivative works" or not, since those invariably end up with "They aren't, even though we think they Then explain the patents on the MP3 algorithm, the LZW algorithm, etc... Those patents are real and while the LZW one may have lapsed, still relevant. -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
Is installation not a precondition for running? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
If a company sells you hardware that includes a ROM that contains GPL'ed software, are they in violation of the GPL if they don't include a ROM burner in the hardware? Or are ROM burners like compilers, where you have to supply your own?
this requirement does not apply if neither you nor any third party retains the ability to install modified object code on the User Product (for example, the work has been installed in ROM). -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Do they have to provide a ROM burner if the ROM is socketed rather than soldered into place?
Of course not. They just can't impose restrictions on your obtaining a ROM burner and doing the work yourself. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
do you realize that you have just admitted that the Tivo is perfectly fine and legal? because you can solder off the ROM from the Tivo and can put in a new ROM with another bootloader that does not check the SHA1 key. Tivo puts no restrictions on you to obtain a ROM and a ROM burner and do this work yourself. (they dont help you either, but you just conceded in another thread that the hardware maker does not have to go out on his way to help you in your software modification efforts.) Ingo -
Then you've committed an offence because of the SHA1 key removal. Tivo deliberately create a system where removal of the ROM is an offence (sometimes in criminal law) so that argument doesn't hold water. -
Who cares about whether the module is a derivative work? That's only relevant when you distribute the module as a separate work. When you ship a combined work including both the kernel and the module in question, it's a _whole_ lot easier to interpret the GPL. -- dwmw2 -
Agreed. I said I wasn't going to argue about it because there *ARE* distinctions that the law makes and the GPL ignores. You can't have it both ways. If the module is distributed *with* the kernel *SOURCE* then it doesn't matter if it's a derivative work or not, because it becomes covered by the kernels license. If it's distributed with the kernel *binaries* then it is covered by its own license. In that case the only reason you'd have a right to the source is if the module is considered a "derivative work". DRH -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
Not necessarily. I'm not entirely sure where you got that idea from. If the module is distributed 'as a separate work', _then_ what you say is true: the only reason you'd have a right to the source is if the module is considered a 'derivative work'. But when you distribute the same module as part of a whole which is a work based on the kernel, the distribution of the whole must be on the terms of GPL, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it. The words you used were 'with the kernel', which could actually mean either of the above. In the case of embedded Linux-based firmware though, it's definitely the latter. It's a coherent whole, and it contains both the kernel and the module. Thus the GPL extends to each and every part, regardless of who wrote it. Including the module. -- dwmw2 -
Just because two things are bundled together doesn't put them under the same license or copyright. Take a look at the GPL, which specifically mentions that "mere aggregation" does not cause something to fall under the GPL. Not that the GPL can even change the law - in the US copyright law specifically states that "mechanical translation" and "mechanical processes" *CANNOT* create a "new" work. Since the process of compiling source into a binary is, by definition, a *mechanical* process then the binary can't suddenly become covered by a different copyright license than the source code merely because of the medium on which its distributed or the manner in which it is distributed. DRH -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
What's logic got to do with it? It was fairly much a direct quote from You're confused. If I grant you a licence on the condition that you give me money, would you object on the basis that the money is not a 'derived work' of my code? No. It's just a condition of the licence, and you're not allowed to use my code unless you give me money. If I grant you a licence on the condition that you sacrifice your first-born son to Satan, would you object on the basis that your son is not a 'derived work' of my code? No. It's just a condition of the licence. If you don't do it, you don't have the right to use my code. (You may be able to get me locked up, but you still don't get to use my code without a licence). If I grant you a licence on the condition that you release _everything_ you write this year under the GPLv2, would you object on the basis that your code is not a 'derived work' of my own? No. It's just a condition of the licence, which you choose to accept or not. If I grant you a licence on the condition that anything you release in _combination_ with my code must also be released under the GPL, would you object on the basis that you code is not a 'derived work' of my own? No. Again, it's just a condition of the licence. If you don't want to obey the licence, you don't get to use the kernel in the first place. Talking about how your code can't possibly be a derived work is just a red herring. The GPL explicitly talks about works which are 'independent and separate works in themselves', to which the GPL does not apply 'when you distribute them as separate works'. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it. It's your choice -- you're not _forced_ to use the kernel, and you're not _forced_ to distribute a product which ...
Hrm... Perhaps I misread your post originally. Let me read it again and see if I didn't encounter a parsing error somewhere... Nope. Error of omission. The text you cut changes the meaning of the passage in its entirety. Here, I'll quote it, in it's entirety: These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it. In other words, it applies to *SECTIONS* of the code, not to individual object code files. This is why kernel modules can have their own, separate license from the kernel. It isn't until the code is shipped as a *standard* part of the kernel that it has to be GPLv2. (Dynamic Linking, being a totally mechanical process, cannot create a derivative work under US copyright law, so please, don't try that old saw) What this means is that it doesn't matter that a non-GPL module is shipped, in "object code" form with the "object code" form of the linux kernel it is designed to interface with - it *still* doesn't become automatically covered But you obviously are. After all, what does this have to do with whether the Again, what does this have to do with your apparent belief that me putting a binary of a kernel module that isn't GPL'd on a disc with the Linux kernel And it also says: In addition, mere aggregation of another work not based on the Program with the Program (or with a work based on the Program) on a volume of a storage or distribution medium does not bring the other work under the scope of ...
You're interpreting 'sections' to mean individual linker sections? And you think it's talking about distributing those 'sections' "as separate works"? Despite the fact that all the rest of the language in the document is high-level and doesn't even mention _linking_? The GPLv2 doesn't "magically" change the law, and has no need to. The above is just a demonstration that a licence can have conditions which involve things _other_ than derived works. I can release code under a 'viral' licence which requires you to release _everything_ you write for a whole year under that same licence. You don't _have_ to obey, and there's no suggestion that your own code would be 'derived' from mine -- but if you don't follow the conditions of the licence, then you don't get permission to use my code. There's no It's just a demonstration that a licence _can_ make requirements about non-derived code. You seemed to be making two bogus claims -- first that it _can_ not, and then that the GPL _does_ not. I've dealt with the first; let's look at the second, leaving aside your weird digression about bison... (hint: it's about the code stubs in the output which weren't _produced_ mechanically in the first place; they Yes. That's why I said 'not necessarily' rather than 'no'. If it just happens to be on the same hard drive / tape / CD-ROM that's not important. The important question is whether it's distributed 'as a If by 'correctly' you mean I should interpret 'sections' to mean linker sections, then you're right -- I really can't bring myself to read it that way. Of course, you're not actually _wrong_ until/unless it's I'm certainly confused as to what you're talking about _now_; it seems to make little sense. What do you mean by 'this' in the context of 'is this what the process of making a module does?'? And how is it relevant? You said that modules aren't derivative works. I said 'Who cares?' because there are far more obvious reasons why the module would be under GPL, because of ...
And it's your own, twisted interpretation. When I say "Sections" I mean "portions of the source code", ie: files, functions, etc... In that manner the person who wrote that function, file, etc... (provided it meets the "minimum of artistic expression" required by US copyright law) has copyright to it. If it doesn't become compiled into - ie: linked into as a *standard* part of the build - the final executable the license on the final executable *cannot* have any effect on the stated code. QED: a kernel module, like "i8042.ko" *can* and *does* have a separate copyright *and* license from I never said it couldn't - a license can do whatever the hell it wants. What I said was that the license on one copyright work *cannot* just magically change the license on another work. The change of license *requires* the person holding copyright to *agree* to the change of license. And no, the GPL *DOES* *NOT* have the requirement that a non-GPL'd work included on the same medium - for distribution or otherwise - must change its About bison: doesn't matter. The code that they are included in *is* mechanically generated, as guided by the input file. QED: The output of Bison is a mechanical translation process *exactly* like the compiling of a C And the GPL cannot define, on its own, what a "Separate Work" or a "Coherent Whole" is. That is defined by the relevant parts of copyright law. QED: The passage is largely irrelevant - if not, then the FSF claim that linking to a GPL'd library means your program is "magically" now GPL'd. It would also mean that every Linux live-cd that includes a non-GPL program is violating the Nope. To mean "source code files" or "functions in the source code". "Linker The confusion is probably because you assumed I was an idiot who didn't have a clue what he was talking about. "This" means, if you really don't understand, "combining a GPL'd work with code of your own". So, does creating a kernel module that needs to be loaded into ...
Actually, I suspect Daniel has read it, and is probably referring to another facet of the license: distribution of two things together does *not* imply that those two things have to both be GPLv2's. The GPLv2 explicitly mentions "mere aggregation". Strictly speaking, it doesn't even *have* to mention it, since it does mention in other places that it only covers "derived work", and "derivation" has nothing to do with "distributing two things together". But it's a good clarification. So you guys are *both* right, for different cases! The issue is simply what you mean by "part of the whole"? If you mean "part of the whole kernel distribution", then yes, the kernel is one work, and it is, in its entirety, under the GPLv2. But if the "part of the whole" is about something like a DVD with the whole being a collection of "mere aggregation", the licenses do not necessarily meld together. Let's say that you're a Linux vendor, and you distribute a DVD with both the Linux kernel binary (and all the normal modules that go with it, that obviously are "part of the whole kernel") *and* say the NVidia proprietary kernel module. Is that the *only* way to read things? No. It's a matter of interpretation, and which "whole" you are talking about. The whole aggregation, or the whole program? And is the NVidia module a "derived work" or not? That's a gray area, and that's really what it hinges on. I personally think it's not, but I know others think it is. Which is why I think you're both *potentially* right. Which one of you is *actually* right will depend on the exact circumstances ;) Linus -
Actually, I don't see where it explicitly states that it only covers derived work. On the other hand, I _do_ see where it explicitly states that in some cases it _does_ 'infect' non-derived works. That's §2, which goes along the lines of ... 'sections...not derived from Program' ... 'License...does...not apply...when...distribute...as separate works'. 'But..same sections as part of whole... License...extend...to each and every part regardless of who wrote it'. As you say, it goes on to _clarify_ that 'mere aggregation on a volume of a storage or distribution medium' isn't what it's talking about here. But there is _some_ class of collective work which combines entirely non-derived work with the original GPL'd Program, and for which the GPL requires that you release your non-derived code under the terms of the GPL. You could argue till the cows come home about precisely what falls Yep, those are two extremes which fall either side of the grey area and A DVD with a Linux distribution isn't _quite_ the same as an arbitrary bunch of gratis software which happens to be thrown together on a disc. A _lot_ of work goes into making that a coherent product where everything interoperates sanely, rather than just such an arbitrary sack of bits. There is at least one prominent North American Linux Vendor who has been observed to claim that the distribution _is_ a collective work and copyrightable in its own right -- which would mean that it _is_ a work based on the Program, and thus that including binary-only modules in it is not permitted. And there are other distributors who've stopped including binary-only modules under threat of legal action (not that If the DVD of the distribution is considered to be a work in itself; a 'work based on the Program', then it doesn't actually matter whether the nVidia module is a derived work or not. Unless you're willing to disregard those two paragraphs of §2 entirely? The case which interests me most is when someone makes an ...
See "Section 0": The "Program", below, refers to any such program or work, and a "work based on the Program" means either the Program or any derivative work under copyright law: so yes, if you grepped for "derived work", you wouldn't have found it. The exact wording used in the license is "derivative work under copyright law". So the very *definition* of the word "Program" is indeed limited by the Hey, I kind of disagree. What is a DVD? It's just a "blob" of a UDF image, potentially containing the Linux kernel. How is that different from a "blob" of some other kind of image (say, a cramfs or similar image) on a rom? What makes UDF so different from cramfs? What makes a DVD so different from a ROM chip? Why would copyright law care about one and not the other? So I really do _not_ think it's at all obvious. Personally, I think it's exactly the same case. Others disagree, but I've never really seen a good *reason* for them disagreeing. Linus -
Yep. And §2 talks explicitly about independent and separate works when they are distributed _with_ the Program, as part of a larger work based The differences are subtle, but they do exist. They're not really about whether it's iso9660 or cramfs; it's about whether what you put on them is a coherent work in its own right or just a bunch of bits which happen to be thrown together onto the same medium. And in the router case, there's little point to its existence without the binary-only module. At least with the DVD it _can_ work without the binary-only module. Although as I said, some distributors definitely It's a grey area, and nobody's 'right' until/unless a court decides. And then only until/unless a higher court contradicts it. The reason I jumped in was to point out that it isn't _just_ about whether the module is a derived work or not. The GPL goes further than that. -- dwmw2 -
i think you are putting too much weight behind the distinction between derivatives/modifications and collective works / collections. The two are very closely related: a derivative/modification is a work that arises out of an existing work, while a collective work arises out of a set of pre-existing works by adding some 'glue' to it (where the glue itself is not a work in itself - the whole thing is the new work). In fact one could argue that a derivative is a collective work based on a _single_ preexisting work! and thus the whole issue of "what is a whole", how strong the "glue" needs to be so that the copyright of the collective work is meaningful on its own and starts to affect every component and requires each of them to be GPL licensed. This largely depends on how deeply a distributor integrates said binary blobs. For example some Linux distributors certainly found it safe enough to ship restricted software on separate medium - even if they happen to be in the same physical package. (which one could attempt to argue to be 'one work'.) That boundary is indeed fuzzy, because life is fuzzy too and the possibilities are virtually unlimited. But one thing is pretty sure: as long as some component is merely put alongside of a larger body of work, even if that component has no life of its own without _some_ larger body of work, that component is not necessarily part of a collective work and does not necessarily fall under the GPL. (unless it falls under the GPL for entirely different reasons: for example it was continuously developed out of internals of the 'larger body of work' and thus became a derivative of the larger body of work.) For driver blobs that are shared between Windows and Linux it would be hard to argue that they are derived from the Linux kernel. Merely linking to some larger body of work does not necessarily mean that the two become a collective work. No matter how much the FSF is trying to muddy the waters with the ...
Not _necessarily_ a collective work. But not necessarily _not_ a I think it's quite clear that the intent of the GPL _is_ to 'muddy the waters', as you put it, and to indicate that bundling stuff together _should_ put the non-derived parts under the GPL too; at least in some circumstances. But still, nothing's true until it's ruled by a court. -- dwmw2 -
yeah - i keep interchanging the two because they are so closely related. (the same goes for modification and derivation - for software the two are quite similar.) Depending on how deeply a distributor integrates a binary blob, it might or might not fall under the umbrella of a collective work, and the GPL (covering other components of the but it's not up to the GPL to define that! Whether something is a collective work is a matter of law (which operates on the specific facts of the case), not a matter of licensing. and that's where the GPLv3 errs: it arbitrarily attempts to "define" some work that can _easily_ be completely separate from the GPL-ed work to be under the scope of "source code". Yes, it can do that legally because its framers knew what they were doing and they did not attempt to implement it as a 'this work belongs to us' thing (which would be misuse of copyright) but as a 'you got to pay with your work for our permission' - but the external communications about this is all false: the pretense that the key in the Tivo case somehow belongs to the GPL-ed work is just bogus. A key _can_ belong to a GPL-ed work, but it does not automatically so. The GPLv3 automatically and unconditionally moves it under the scope of the license and that aspect of the GPLv3 is just wrong and moves the license closer to a Microsoft EULA contract than towards a pure and just copyright license. Ingo -
I think that's a somewhat valid argument, although I'm not really sure whether there is any difference between, say, a Fedora 7 "livecd", and a router with a cramfs filesystem in rom. Both really work the same way, and both really are very much targeted towards a specific hardware platform. Yes, it's true that a small router migth be a more *coherent* hardware platform than the Fedora 7 livecd is, but that's more a factor of the wild and crazy PC hardware culture than of the small router. For example, what about a livecd for PPC-based hardware? Those tend to be much more uniform (read: I think the livecd's generally work on mac clones). So it's a question of degrees of separation. Does it make a difference that some of these embedded images work across a whole range of (rather similar, but still.. not identical) routers? I don't really have any point, except that there is no real *technical* difference, and in many respects the only difference in the end really seems to be about "intended target device or audience" rather than anything else. How can you make hard licensing decisions in situations like that? I don't think you really can. In fact, the OSI rules even forbid making licensing decisions based on things that get rather close to the differences you are describing (both the "not specific to a product" and "license must not discriminate against fields of endeavor"). It's also really really *hard* to make a choice based on a gradual scale. Where do you put the limit? Wherever you put it, it's going to be arbitrary. Is that really a good thing? So I would at least *personally* suggest that people not look into the license for these kinds of things, and also that you really need to have a very specific case, and just basically put it in front of a judge. At some point, *somebody* has to decide in a gray area, and I'm not saying that a judge is really _technically_ any better really to decide the issue, but at least he is ...
I'm inclined to agree. And I'd probably suggest that the Fedora 7 'livecd' would be in violation of the GPL if it were to include the binary-only modules, too. Enough people agree with me that we _don't_ in fact include those modules. And other people have been convinced to Indeed. -- dwmw2 -
The "Program", below, refers to any such program or work, and a "work based on the Program" means either the Program or any derivative work under copyright law: that is to say, a work containing the Program or a portion of it, either verbatim or with modifications and/or translated into another language. (Hereinafter, translation is included without limitation in the term "modification".) In other words, the "that is to say, a work containing the Program or a portion of it, either verbatim or with modifications and/or translated into another language." is a clarification of the terms so that you do *NOT* have to know copyright law. However, the license, being based in copyright law, *CANNOT* change that law without making itself invalid. QED: What copyright law says is a "derivative work" is what matters, not the definition provided in the license. -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
... unless the other parts of the whole fall under the mere aggregation exception, methinks, but IANAL. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
i fully support the notion you articulate, that whether bin-only modules are part of a derivative work of the kernel or whether they are independent works is not an automatic thing at all. The answer is: "it depends, talk to your lawyer". For example i'd say VMWare's ESX bin-only module is likely derived from the Linux kernel and should be distributed under the GPL, but that for example the ATI and nvidia drivers, although being a large PITA for all of us, are possibly independent works. but lets note that this is irrelevant to the Tivo argument. Tivo is not using bin-only modules AFAIK, all their source code is available for download. (their kernel source is totally uninteresting by the way - they have some weird crap IDE controller with hacks that will never go upstream.) Ingo -
I was actually trying to avoid the question altogether. It's not that And thus not affected by the GPL _if_ they are distributed as separate works in their own right. But if you bundle them with the kernel into a Right. It was a digression, which I picked up on because people were talking about derived works in the context of modules again, and missing the point that the most _obvious_ GPL violation with modules doesn't actually involve those modules being a derived work at all. -- dwmw2 -
yeah. Section 2 of the GPLv2 takes a permissive (and IMO correct) approach here: Thus, it is not the intent of this section to claim rights or contest your rights to work written entirely by you; rather, the intent is to exercise the right to control the distribution of derivative or collective works based on the Program. In addition, mere aggregation of another work not based on the Program with the Program (or with a work based on the Program) on a volume of a storage or distribution medium does not bring the other work under the scope of this License. the argument is quite strong that the linking of two independent works is "mere aggregation" as well. (as long as they are truly separate works) Ingo -
You think so? If even linking was considered 'mere aggregation on a volume of a storage or distribution medium', then when would the 'But when you distribute those same sections as part of a whole...' bit _ever_ apply? It _explicitly_ talks of sections which are independent and separate works in their own right, but which must be licensed under the GPL when they're distributed as part of a larger whole. I don't see how we could hold the view that _even_ linking is 'mere aggregation on a volume of a storage or distribution medium', without conveniently either ignoring entire paragraphs of the GPL or declaring them to be entirely meaningless. Of course, that doesn't mean that a court _wouldn't_ do that. Given enough money, I'm sure you could get US court to declare that the world is flat. But it doesn't seem to be a reasonable viewpoint, to me. Or a likely outcome. -- dwmw2 -
as long as it's not distributed in one collective work, where is the problem? A driver could be argued to be part of a mere compilation of works (not part of a collective work), or just two separate works. But i'm not that cynical about US courts. Ingo -
As long as it's not distributed "as part of a whole which is a work based on the Program", there's no problem. You seem to be suggesting that even linking the Program together with other stuff doesn't create a 'work based on the Program'. You seem claim it's "mere aggregation on a volume of a storage or distribution medium". Am I understanding you correctly? Is there _anything_ which you admit would actually constitute a 'work based on the Program', when that work wouldn't have been be a derived work anyway? Or do you claim that those whole paragraphs of the GPL are just meaningless drivel, when they explicitly make reference to applying the GPL to works which would _normally_ be 'considered independent and separate works in themselves'? If your interpretation of the GPL means that those paragraphs don't make any sense at all, then I feel your interpretation may be suspect. -- dwmw2 -
Correct. Linking does not create a "work based on the program" because linking does not create a work. Only a creative process can create a work. These terms are synonymous. And neither of them can apply to something that The license is just clarifying copyright law. Even if it intended to do They make perfect sense. They're clarifications of copyright law to help people who might not be familiar with the law understand what their obligations under the license are. All of those sections are reasonable explanations of what a "derivative work" is. It would be extremely strange to try to parse them for subtle differences. In any event, even if the GPL said "if you ever look at any source code to a GPL'd work, the FSF owns everything you code after that", it wouldn't matter. The GPL can't set its own scope. Copyright law, and the definition of a derivative work, set the GPL's scope anyway. The GPL only makes sense if you understand "mere aggregation" to mean 'as opposed to creative combination'. DS -
By the way, the unfortunate answer to the question of what the default position is when contributions to a collective work are received without explicit license, at least in the United States, is: "In the absence of an express transfer of the copyright or of any rights under it, the owner of copyright in the collective work is presumed to have acquired only the privilege of reproducing and distributing the contribution as part of that particular collective work, any revision of that collective work, and any later collective work in the same series." -- USC 201(c) That is, as I understand the law, if you receive a contribution to a project without any specific license in that contribution, it works like this: 1) You automatically receive a license to use that piece as part of that project by virtue of the fact that it was contributed by the author to that project. (Because 210(c) says so.) 2) If the contribution is itself a derivative work of a GPL'd work, then you receive a GPL license. (Because the GPL says so). So it would be very unwise to add a contribution that wasn't itself a derivative work without clear indication from the author that the contribution is offered under the license you need. I had assumed no law set a default, and therefore the default would be the project's license. THIS IS INCORRECT. 201(c) sets the default, and it's the wrong one. This means that contributions of non-derivative works to GPL projects should not be added to the project unless the author specifically licenses that piece under the GPL. I would not consider it safe to assume that the fact that the author knowingly contributed the work to a GPL'd project is sufficient to change the 201(c) default. What's worse, section 203 appears to grant the author various ways, by law, to *terminate* a license grant. This termination removes the ability to create subsequent derivative works. Ouch. http://www.copyright.gov/title17/92chap2.html I sure hope I'm misunderstanding ...
Nope. In fact, "work based on the program" is so unclear that it means that if I wrote a book about the creation of the Linux Kernel that is entirely original - containing nothing that is copyright someone else - I would have to release it under the GPL simply because it is a "work based on the program". Is it okay to make that demand? I don't think so. But that is *exactly* what it means. And it is "the GPL applying itself to works which would normally be considered independent and separate works in themselves". -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
Are you playing Linus' sheeple and parroting his lines just to make a Don't worry, I'm not speaking even for FSFLA, and I'm entitled to my own opinion. I haven't consulted other FSFLA members about this. This is all my own personal opinion. It just so happens that I'm very closely involved in the process, I've spent a lot of time thinking about it, and I happen to share a similar moral and ethical background with others involved in the process, so I arrive at similar conclusions. And then, I influence the process myself, so it's not like some of the arguments I brought up here weren't taken into account while creating the GPLv3, and adopted by its other proponents. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Nope. I'm just tired of giving proof after proof that you're wrong and having Certainly. I never said otherwise. What I stated and then *implied* was that you are repeating the same false logic over and over again trying to make people believe that it isn't borked and that that false logic is exactly the Where I am examining the facts and drawing a logical conclusion. That it Okay. Still doesn't explain why you have argued that the GPLv3 doesn't attempt This is no surprise - I had a feeling this was the truth. Not that it changes my opinion at all. As I've said, I have never liked the GPL at all, but v2 is the best that exists - even though I've put together custom licenses myself, none of them have had the number of lawyers look at them that the GPLv2 has had. DRH -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
It doesn't cover hardware, in the same way that it doesn't cover patents, and it doesn't cover pro-DRM laws. It merely arranges, as best as we've managed a copyright license to do, that they can't be used as excuses (or tools) to disrespect the freedoms that the GPL demands all licensees to respect for other users. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Consider this scenario: Small company A is manufacturing a new WiFi router. They decide to have it run HURD as the OS. In complying with the GPLv3 they supply the signing keys and everything else needed to install a new kernel on the hardware. User B buys the router and modifies the kernel so it drives the WiFi to an output power twice that which it is licensed to carry. FCC finds out and prosecutes User B for violating the regulations. FCC then pulls the small companies license until they change their hardware so the driver can't push it to transmit at a higher power level and levies a fine. Small company A loses the money paid on the fine, has to recall all the devices that can be modified (through software) to break the law at a massive cost *AND* has to redesign their hardware. The total cost drives the company into bankruptcy. Small companies C,D and E, in order to avoid the fate of small company A, purchases a license for proprietary OS "F" to drive their new hardware. Net loss: A lot of the users and publicity that "Free Software" used to get, because GPLv3 contains language that opens the companies to lawsuits that they wouldn't otherwise face. Which is better: Growing the base of installed GPL covered software, or "ethics and morals" that demand the language that has been added to the GPLv3 ? Personally I'd like to see proprietary software driven into a very small "niche" market or entirely out of existence. However much I want this to happen, I cannot be anything *BUT* scared of the GPLv3 simply because I see it creating massive problems - and all because of a *small* portion of the new language it contains. It has taken almost 15 years for "Free Software" to make a dent in the market, and, IMHO, a lot of that is both Linux and the "holes" in GPLv2. DRH -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
You appear terminally confused. The purpose of the GPL as defined by its authors is not commercial success, world domination or making zillions of dollars - it is keeping the software protected by that license "free" in terms of liberty as measured against the set of freedoms to run/modify/etc they discuss in the licence document. The fact this is a good license for making zillions of dollars, producing good software and the like is either incidental or a logical result of the protection of freedoms depending upon which views you believe. -
that's fine, but the fundamental question is: where is the moral boundary of the power that the copyright license gives? The FSF seems to believe "nowhere, anything that copyright law allows us to achieve our goals is a fair game" - and the GPLv3 shows that belief. I dont subscribe to that view. I think the proper limit is the boundary where the limit of the software is - because that's the only sane and globally workable way to stop the power-hungry. I.e. the information we produce is covered by the rules of the GPL. It might be used in ways inconvenient to us, it might be put on hardware we dont like (be that a Tivo, a landmine or an abortion instrument) but that does not change the fundamental fact: it's outside the _moral scope_ of our power. Whether some jurisdictions allow the control of _other_ information via our information is immaterial. If a jurisdiction allows the control of hardware that is associated with our software, so what? If a jurisdiction allows the controlling of various aspects of movie theaters that happen to play copyrighted movies, does it make it morally right? Ingo -
Assuming a democratic state then the laws of the land ought to reflect the 'general will' (if you believe Rousseau anyway). They should thus define the boundary ['derivative work' generally ] according to the general good Does that question not suppose some positivist absolute morality ? I suspect many would argue that it is moral to do so if the end goal of the controls is moral. You might also want to apply the tests in Fuller's Internal Morality of Law ? I'm not sure Philosophy is on topic for l/k however 8) Alan -
uhm, so if the MPAA and the RIAA pays for another nice piece of legislation that extends the power of copyright owners, do you find it morally justified to use those powers, as long as it's argued to be in favor of some long-term goal that you judge to be moral, even if it results in some "temporary injustice"? i think that could be the main difference in thinking. I argue that the only way to be moral is to be moral _now_, not "later, once this very important fight for the common good is over". I think the moral approach to this is to say _no_ to attempts to extend the license to beyond the "moral scope" of the software we wrote - regardless of what new powers are legislated into the hands of copyright owners. It's naturally hard to do, because giving up power is always hard to do. In other words: we need to apply our concepts of freedom and fairness not only to the end result, but to the means and methods of achieving those end results as well. The end goals are often forgotten, it's the process that matters to the end result. Or in yet another set of words: this concept of morality also happens to be expressed fairly accurately in the thousands of years of 'quid pro quo' concept. (shared amongst many, many cultures on this planet, shared amongst far more cultures than the western 'freedom' concept.) (which concept of quid-pro-quo fairness is likely coded into our brains and into our thinking genetically - because it's a simple and very efficient group survival method.) Ingo -
Turnabout is fair play, and unilateral disarmament is a bad strategy in a mexican standoff? Finding it morally justified to _have_ powers is not the same as finding it morally justified to _use_ powers you have anyway. Lots of companies (like Red Hat) amass defensive software patent portfolios because the patent system is so screwed up. But then, I'm a pragmatist, not an idealist. You can be one or the other and make it work. Mixing the two tends to suck. Being ruthlessly pragmatic in the pursuit of an ideal (as the FSF seems to be doing) has often been a recipe for disaster... Rob -
but the GPLv3 definitely takes action against Tivo. It's not "defensive" in any way. It is outright hostile, it irreversibly cuts off certain people from being to distribute GPLv3-ed software alongside with certain types of hardware that the FSF's president does not like. (who, incidentally, is a mathematician who last wrote significant free software perhaps a decade ago, and who thus must have a great and thorough understanding of how hardware and software works today and who must also have a deep knowledge about what makes the free software community tick.) The GPLv2 never did this kind of restriction _of other works_. Yes, you can use copyright law to control other works and thus (if the affected work is a hardware device for example) to control the _use_ of the free software, but it is _wrong_. The GPLv2 specifically said, in section 0: Activities other than copying, distribution and modification are not covered by this License; they are outside its scope. guess why this section has been completely removed from the GPLv3, without a replacement? Ingo -
That's not true. They can just as well throw the key away and refrain How about other works in which GPLed software is distributed? My guess: First, because it was redundant, given that the license didn't quite discuss other activities. Unless you count say "imposing restrictions on the exercise of others' freedoms" as other activities, even though these are associated with modification and distribution. Second, because GPLv3 does indeed talk about other activities, such as starting lawsuits on patent and pro-DRM grounds, or entering agreements for distribution of software along with limited patent licenses. All of these are still associated, at least to some extent, with modification and distribution, but I guess it was worth clarifying that claiming that such harmful activities are outside the scope of the license isn't a valid excuse to escape the conditions determined by the license. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
uhm, so you claim that my argument is false, and your proof for that is a "non-upgradeable Tivo"?? <sarcasm> That is a _great_ idea. Not being able to patch security holes. Not being able to fix bugs. Not being able to add new features. Makes complete sense. Will be a hit on the market! Every PVR maker will flock from Windows to Linux i'm sure. </sarcasm> really, do you even _read_ what you write? All your arguments so far were instantly debunkable. This is one of the lowest quality GPL discussions i was ever involved in ... furthermore, the fact that the GPLv3 had to add carved out exceptions for the anti-Tivo languge is further _proof_ that the whole idea is absurd to begin with! It's like writing a nice new function to implement something, and then when it shows many design flaws, you'd not just admit that it's flawed and would get rid of it and redesign it, you'd instead pretend that it's fine and you'd carve out a few of the more here you prove that you cannot even read what i wrote. I wrote that this section has been removed from the GPLv3. What relevance does it have that in your opinion this section was redundant in the GPLv2?? It would clearly not be redundant in the GPLv3: it would contradict and _completely neutralize_ most of the crap from the GPLv3 that we are dont you realize that declaring certain types of activities by hardware makers as being "against freedom" is _exactly_ such an activity that the GPLv2 did not attempt to control? I could tell you offhand a dozen more examples of human activities that restrict the 4 GNU freedoms of users _much more_ than the Tivo ever did: for example censure, opression of free speech, out of control climate, dictatorship, campaign financing laws, the WIN32 API and human stupidity. By your argument we'd have to add prohibition against those restrictions of freedom to the license too, right? Your argument still leads to absurd results, even now that you've modified it a few times already ...
Oh, so you think patching security holes, fixing bugs and adding new If you didn't mean "removed from the GPLv3 as compared with v2", I misunderstood what you wrote. The fact that it's redundant is v2 means it is reasonable to take it -ENONSEQUITUR How do these stop a user's exercise of the four freedoms of a piece of I hope you're not saying that my listening to you, recognizing mistakes in my arguments and fixing them up is a bad thing. But hey, at least I'm not modifying my arguments as much as you are! ;-) It's pretty easy to shoot a straw man and claim the original argument was broken. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
--- I know you don't see it that way, but I still find it bizarre that "the right to modify the software" should be construed as "implies the right to modify the device that the software was shipped in". I do agree that it's not a change in "spirit" - I'm sure the GPL authors would have disliked TiVoization 15 years ago as much as they do today, if they had thought about it (regardless of the Stallman interview where he said he didn't care very much about devices). However, whether it is a change "in spirit" or not, it clearly is a qualitative change that substantially changes the rights granted under the license and it's perfectly reasonable for some authors who liked the GPLv2 to dislike and reject GPLv3. scott -
It isn't redundant at all. I specifies the definitions of several terms used in the GPLv2 and also defines the exact scope of the license. If you feel that the definition of the terms and the limitation of scope were redundant For a license to be legally enforceable it must be internally consistent. Without that internal consistency it becomes very easy to circumvent it. The GPLv2's definitions and defined scope - as per section 0 - define the limits of the license and are entirely consistent with the rest of it. What it *isn't* consistent with is the FSF's other "propaganda" and the wants of the And that is their right. However, it appears to a nearly unanimous consensus that it is the truth. It may not be liked by some people, but likes and Yep. I've done it to you on more than one occasion, Alexandre. The part that makes me laugh is that you still haven't realized it. DRH -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
this has to be one of the most bizarre arguments i've read in this thread as of date. Are you seriously questioning the notion that it's a good and legitimate idea for a hardware vendor to make the system fixable, patchable and upgradable? Are you seriously suggesting that for a hardware vendor to be able to offer such a solution, if they are under the unescapable restriction of content providers that the system itself must be tamper-proof, it should not be able to use a GPL-ed kernel at all? Because that is what your arguments lead to, and that is what the GPLv3 implements. In case you didnt notice: RMS _does not want the Tivo to use a GPLv3 kernel_, simple as that, and the GPLv3 achieves that. He wants Tivo to either to go out of business or to go to WinCE or some other OS. Did you ever think about the meaning of the "anti" word in the "anti-Tivo" expression? Hint: it's not some friendly suggestion of cooperation and working together ;) Ingo -
No. I'm questioning why the vendor could keep this privilege to No, and I've already explained how I believe this can be accomplished with the wording in the GPLv3dd4, although IANAL to tell whether that's correct. Just make the tivoization machinery require two keys: one that the vendor keeps, one that the vendor gives to the user (maybe without ever knowing it). Neither one can install modifications alone, but the user can approve modifications by the vendor, and the vendor can approve modifications by the user. This is still not ideal, but it at least doesn't permit the vendor to remove features from under the You haven't really read that bit of dd3 or dd4, have you? Or the various portions of this thread in which I showed your I know you're not stupid, but I can't tell whether you're malicious or just misinformed. RMS does not want TiVo (or anyone else) to disrespect users' freedoms, and installing technical measures to prevent users from adapting the software to suit their needs and running their modifications is disrespecting users freedoms. That he is not opposed to the idea of TiVo using a GPLv3 kernel is easy to see, if you take the time to read the draft instead of spreading false assumptions about it: this requirement does not apply if neither you nor any third party retains the ability to install modified object code on the User Hey, wouldn't this be just tit-for-tat? ;-) -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
So what features has Tivo removed (or threatened to remove) from the GPL licensed parts? I think at some point they disabled the _undocumented_ skip feature in their own proprietary software, and ended up reenabling it when a lot of customers complained. Even if those customers had the ability to replace any GPL licensed parts, it would not have reenabled the feature. And it was an undocumented easter egg type thing at that, it isn't like they widely announced it in their advertising or as a selling point. So Google is using Linux right. What if they remove some feature? (let's pick a randomg one, i.e. phone number lookup) Should I get a keycard for their machine room to fix the problem, or maybe we should use some So if Tivo would allow you to boot your own kernel, but keeps the harddisk encrypted if the booted kernel does not have the right signature? In such a case you can run your own kernel and if you replace the harddrive you can install all the applications you might want. You cannot however use their software, any of the recorded content or obtain any further guide data/service updates. And how is that any different from taking an off-the-shelf PC and booting your own kernel with Tivo's modifications? Or really different from the current situation. Tivo complied in as far as they made GPL licensed code available, you can examine it, modify it, compile it. You just can't use it _in combination with TiVo's own software and service_. I didn't think the GPLv2 covered anything related to use and you have retained all the So they keep the system locked down, but include perl/python/emacs and distribute updates in the form of scripts/source code which are either interpreted or compiled to a ramfs filesystem at boot. Time to add another exception? Jan -
Are you claiming Google is tivoizing something in their internal infrastructure? They're not distributing or conveying that software, so, nothing wrong with that. Or are you talking about their search appliance, which I know nearly The information must suffice to ensure that the continued functioning of the modified object code is in no case prevented or TiVo did not sell me that off-the-shelf PC with the Free Software in it. It (hypothetically) sold me a computer with technical measures meant to restrict my ability to adapt the software it shipped to my own needs and to run it for any purpose, while it can still do that. Intent behind this?: weasel out of the obligations of the license. Anyone, probably even a US court, might very well see it that way. They retain the ability to modify the software, so they ought to pass it on to the user. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
--- This characterization misses something important. For many product devices, like cell phones, the modification is never "behind the user's back", but is done because the user has requested it (to fix a problem or add a new feature). If you go to ROM-based software, the user loses, because problems can't be fixed. For certain kinds of problems the user might be able to get a replacement device, but potentially involving losing any data stored on the device. The FSF's approval of this distinction (ROM versus replaceable) places the FSF's particular principles over users interests, for no particular reason - if the manufacturer believes that it cannot legally allow software modification, all the restriction does is force them either to make the software unmodifiable (which advances freedom not at all) or to use software under a different license (which advances freedom not at all). The result? The user STILL has no freedom to modify the software and the community around the software is diminished. To go back to the "behind your back" claim, the only cases I know where the software is replaced behind the users' back are cases were the updates are done by a service (usually not operated by the device manufacturer) that the user has voluntarily requested (like TiVo program guides or cable system subscriptions), which is generally a cases outside the scope of the license in any case. scott -
Okay, take out the "behind the users' back", it makes no difference. That was just to highlight the frequent evil intentions behind keeping the keys. I wonder if giving half the key to the user and keeping the other half would be enough to satisfy the GPLv3 language while still enabling the Right. But if the manufacturer believes that it can legally allow it, and wants to be able to install, software modifications, then it must decide between giving that up and letting the user do it as well. And this is where the users interests may prevail. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
--- Yes, but in highlighting the possibility of evil intentions you distort the fact that usually there are no such evil intentions... --- I think that's a possibility. I don't see how it's functionally different from the usual case where the manufacturer can't modify the device without the user's consent simply because the user has physical access to the device and the manufacturer doesn't. --- Users benefit from the ability to get software updates, from the manufacturer, to resolve problems, fix security vulnerabilities, and provide updated functionality. --- You're harping on the "cannot legally", which is fine but irrelevant. Whether it's a legal requirement or a business decision, the result is the same - neither forcing the manufacturer to make the device non-updatable nor forcing the manufacturer to use different software benefits anyone. I don't know of interesting cases where the manufacturer makes the device non-modifiable out of sheer bloody-mindedness. I don't believe that the existence of this clause will lead to more manufacturers making their devices modifiable - there are too many other options if they think that non-modifiability is important to them. [Note that I *do* think it's perfectly appropriate that authors who feel that they don't want their work used in such devices should be able to license them in line with that belief. I just don't think it has any practical value aside from making them feel better.] scott -
Which they could have the option to do themselves if the manufacturer I agree. But that's an incomplete picture. It's the other part of the picture, that you left out twice, that is They can do that with GPLv3. And those who don't want to stop this can then add a special permission. And then everybody wins. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
--- I don't think I "left it out". The point is that if the manufacturer is unwilling to give the right to modify, no change in the language is going to cause the user to have that right. scott -
If the alternatives are worse for the manufacturer than letting the user have it, then it will have the intended effects. In the other cases, it won't make much of a difference for anyone else. The question is: how does tivoization help the community (under the tit-for-tat reasoning)? Does it help more than anti-tivoization? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
I don't think you can use "usually" and "fact" together like that. Why is it so bad to account for them since they (do) surface and (could) increase significantly in frequency? For me, the (could) is enough to act upon, regardless of the current likely hood of it happening. Things change frequently. This, unfortunately comes pre-distorted depending on what you believe. All of us are right but we still don't agree. Quite a fluke. That's the problem. Best, --Tim -
--- I agree that it is possible to have different definitions of "evil" and "ethical"... scott -
with the little tiny problem that this is not what the GPLv3 actually implements. Our point from the very beginning: the GPLv3 "outlaws" certain hardware restrictions _even if they are fully legitimate_. Yes, of course, it also outlaws 'bad' uses of DRM. The GPLv3 tries to carve out some known 'good' uses of DRM (to stop the GPLv3 from being _totally_ unusable in vast areas of the marketplace), but that limited opt-in approach can in no way be the right solution (think about it as a whitelist - wouldnt you want to be able to _add_ to that whitelist?? The GPLv3 hardcodes it.) In other words: i dont want the police to start shooting innocent people in the streets, in their pursuit of criminals. Yes, this means criminals have an easier job getting away - but _that_ is the price of freedom! and all these problems of the GPLv3 DRM language derives from the same root issue: RMS is trying to make a manual call about what _technology use_ is 'good' and what is 'evil'. For some of these calls we might even agree. But most fundamentally, a license should _never_ get into the business of trying to 'judge' what _use_ is 'good' and what is 'evil'. As you can see it on this list alone, some people see Tivo's intentions as legitimate, they did the DRM to stay in business but still be able to use free software, employ free software developers, show Mythbox how to do this stuff on Linux, etc. But the GPLv3 completely destroys Tivo's ability to use Linux, were Linux to be under the GPLv3. And by doing that, those contested provisions of the GPLv3 itself become a tool against "freedom". You tried to find a workaround for that, by suggesting the 'dont do security fixes then', 'use a split key', 'use a rent model' solutions, but dont you realize that by suggesting those you are explicitly against the intent of RMS, who wants to _stop_ Tivo from being able to do DRM? Dont you think it speaks volumes of the GPLv3's quality that you have to go out and search for a ...
Except that it does. Go read it, then come back and admit you were mistaken and spreading lies about it. The vendor must decide between respecting the freedom of the user, or What it does is to seek to carry out its mission (*) of defending users' freedoms. Obstacles that are placed to impede the user from enjoying the freedoms are supposed to not be permitted by the GPL. (*) it seems that understanding "spirit of a license" is very difficult for you; does the term "mission" help you understand what FTR, rent model wasn't me, and it doesn't escape the GPLv3dd4 You misunderstand not only the spirit of the license, but his intentions. Oh, wait! They're the same, that's why. Respect and defend users' freedoms. Repeat after me until it sinks. I know you're not stupid. Why do you pretend to be? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Where is the boundary between hard- and software? I'm employed as hardware= =20 designer, and for this purpose, I write programs in a hardware description= =20 language, which can be converted into hardware through a synthesis=20 software. I write firmware, which is assembled into binary and gets placed= =20 on on-chip memory (ROM or NVM). I've even studied computer science, and=20 electric engineering was just a side-course. I know how transistors work,=20 and how gates are implemented in terms of transistors, but in essense, it's= =20 not that relevant unless you want to do analog circuits. Usually, during=20 the development phase, we put the Verilog into an FPGA, where the=20 configuration file still is obviously "software" in any sense it can be.=20 I've even released descriptions of some parts of the work I do under GPL=20 for people to put it into their own FPGAs. There is no boundary between hard- and software in the sense of that=20 hardware is something fundamentally different. Hardware is just another way= =20 to implement programs, and it uses other languages (but SystemC even looks= =20 quite close to C). If there is a boundary, it's way below the distinction=20 between a Tivo and a PC, because these two basically consist of a=20 processor, some RAM, some flash, a harddisk, and a video driver. What's true: We don't have the moral power to define *where* the software=20 goes, but we have the moral power to define *how* users can change the=20 software when they own the hardware (the physical representation). =2D-=20 Bernd Paysan "If you want it done right, you have to do it yourself" http://www.jwdt.com/~paysan/
this is largely irrelevant to my argument: the FSF is clearly trying to extend the scope of the GPL to restrict the distribution of certain hardware+software combinations. The FSF is not really arguing that the boundary between software and hardware is diffuse. (which btw. it clearly is) The FSF simply wants to be able to say via the GPLv3: "to be able to distribute GPL-ed software, the hardware is required to do this and this". please note an important thing here: "required to do this and this" is the _precise antithesis_ of "freedom". The only significant restriction on freedom the GPLv2 allows is that the covered work (the software) is not to be restricted. And that is a fair deal. Even if any additional restrictions would otherwise be for the "common good" and would further "freedom" in creative ways. If the FSF's argument and approach was correct then it would be fine to add these restrictions to GPLv4: - do not distribute non-GPL-compatible software with GPL software on the same hardware. - send at least 10 free samples of the hardware to the FSF headquarters. (after all true freedom is only achieved if developers are not only allowed to modify the hardware, but are allowed to test it as well, for the freedom of the community.) - donate $10 to the FSF. - spray "Linus sucks because he stole RMS's GNU thunder in the 90s and never gave it back!" graffiti on 3 separate walls in your neighborhood. Each of these items is an additional restriction on either the hardware+software combination that is being distributed or on the person who does the distribution, and each of these items - some abstractly, some more directly - advance a notion of the "four GNU freedoms" in some way. And each of these items has a basis in copyright law and might be legally put into a license and might be enforceable. (ok, probably not the last item ;) think about it, the list of things that one can do via license to "achieve more ...
Most people arguing for the expansive interpretation do not really care what hardware is combined with what software. They care about the ability for the user (in the GPLv2's terms, someone who receives GPL'ed software) to have comparable ability to modify and (re-)distribute the software as the software distributor does. The issue of GPLed software on DRMed hardware applies equally to digital video recorders, where the hardware and software distributor are usually the same, and video game consoles, where they are not. There is no good reason to treat a "GPL-incompatible" hardware platform (for example, incompatible due to restrictions on the keys to generate digital signatures) differently than a "GPL-incompatible" patent area. If a software distributor cannot simultaneously comply with the GPL and his other obligations, he should either not distribute the software or be prepared to face the liability from breaching his obligations. Michael Poole -
Agreed - if you want to take my work you are welcome as long as you contribute back your changes. That's the deal that GPL2 enforces and why it has been so successful. GPL3 is a very different beast with a much wider agenda, which makes it far more difficult to achieve consensus on what it should contain. Personally I would have liked to seen a GPL2.x which fixes some of the issues but stays true to the more limited objective. Alan -
That may be a side effect of the GPL, but it's actually not how the GPLv2=20 works (nor is it the intention). "Contribute back" means upstream. There's= =20 no such provision in the GPLv2, you contribute only downstream. And there=20 are cases where you don't need to contribute at all. E.g. the kernel hacking I'm doing at the moment: I have bought a uClinux=20 blackfin board, for testing my digital audio amplifier. For that, I took=20 one of the blackfin alsa audio drivers, and changed it so that it could=20 talk to my digital audio amplifier. I'm not distributing this software,=20 it's a complete in-house project, so I'm not obliged to contribute back. At= =20 the moment, I'm the only person in the world who has both access to the=20 digital audio amplifier and the blackfin board, so releasing this driver in= =20 that early stage is a rather pointless excercise. I think this above explains fairly well the "misunderstandings" that are=20 appearing here. The GPL is not reflective (tit-for-tat), it's transient. If= =20 there's a loop in the transient propagation, it becomes reflective through= =20 the loop, but not by itself. This was the case in GPLv1, is the case in=20 GPLv2, and will be the case in GPLv3. =2D-=20 Bernd Paysan "If you want it done right, you have to do it yourself" http://www.jwdt.com/~paysan/
And the Linux kernel community has been familiar with this situation all along. It's the bargain the kernel developers struck with each other a decade and a half ago. Now the FSF is coming along and being Darth Vader: "I am altering the bargain. That's not specifically a limitation of the GPL, that's a limitation of copyright law which forms the basis of the GPL. It covers distribution, not usage. GPLv2 eliminates the case where I have a modified binary I contributed to, but can't see the source code of those modifications. This has the pragmatic effect of greatly reducing forking in a project, such as the Emacs/Lucid Emacs fork that inspired the "Emacs license" that became GPLv1. Rob -
1) it can't possibly do that. the Linux license is something that only the Linux developers can decide. 2) I don't know how the FSF is approaching the Linux developers, but what I've been personally trying to do in this infinite thread was mainly to set the record straight that v3 did not change the spirit of the license, like some have claimed. 3) Another thing I've tried to do was to try to figure out why Linux developers seem to consider v2 better than v3 for their own goals. I must admit I failed. The presented reasons don't seem to distinguish v2 from v3 to me, or rather make v3 sound better. It's disappointing that I took so much of everyone's time without achieving any of my goals. I hope it was at least useful or enlightening to some. I'll now try to step out of the discussion, but I guess I'm just as addicted to flames. I don't see that it's getting anywhere, and I don't particularly enjoy the name calling. And then, I was politely invited to go away... -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
What do you expect, when you tried to entertain a legal picture of the
GPLv2 that even the FSF counsel doesn't believe in?
I will state one more time: I think that what Tivo did was and is:
(a) perfectly legal wrt the GPLv2 (and I have shown multiple times why
your arguments don't hold logical water - if you actually followed
them yourself, you wouldn't be using a redhat.com email address!)
(b) not just legally right, but perfectly morally right too (it wasn't
some underhanded "trick" thing - it was following the spirit _and_
the letter of the law)
(c) the only reasonable thing many companies *can* do in the face of laws
and regulations and entities like the RIAA/MPAA.
and you should admit that the fact that the FSF counsel says that it
couldn't sue Tivo in the US, it means that while my standpoint may not be
the _only_ possible one, I'm certainly not "confused" about (a) above.
The (b) and (c) points are not "legal" points, they are about the fact
that quite often, morality and practicality are independent of legality,
and you should never see law as being the *only* thing that matters. So
the reason I bring them up is that it wasn't just "legally ok", they also
had good *reasons* for doing it, and there was no hanky-panky about it!
In fact, I consider Tivo one of the good guys, because they were one of
the few people that had things like the GPLv2 actually printed out and
clearly stated IN THE MAIN PAPER MANUAL. In the very first version of
their box. Without anybody twisting any arms at all.
IOW, Tivo really did everything right. I personally think that they were
even classy about it.
And that's my opinion. THINK about that for a moment. THINK about the fact
that I am the original copyright holder in the main software project they
used, and that I state that as neither having ever gotten paid _or_ owning
any stock what-so-ever in Tivo.
Dammit, if I cannot say that I think what they did was fine, who ...I don't think I made significant legal arguments. My points were about the spirit of the license. That's not legal at all. That's moral and ethical background. Your aggressive response directed at the FSF came as quite a surprise I respectfully disagree, and I know I'm not alone in this assessment. I know other kernel developers agree with it. And they're as entitled You countered one of the various arguments I have, and you failed at that. It was another, quite different argument, that got me to realize it didn't work legally. But this says nothing about I don't think I've ever claimed you were confused about (a). I said we disagreed. That's quite different. If you and all other Linux copyright holders agreed about it, sure. Just like you could all grant it an additional permission, just so Yes. They'd have to give up the ability to update the software, or pass it on to the user. If they can't do the latter, they could still That's because when you talk about why GPLv2 is better, you always talked about virtues that are just as present in v3, and that AFAICT I'm sorry if I come off that way. But we really look at this issue from very different perspectives, and it's difficult for me to try to see it your way. Please cut me some slack here. I'm trying hard, but there is so much noise and so many hard feelings that seeing what the real issues are is not that easy. Thanks for your understanding, -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
In other words, you advocate license for technical programs that causes people to make bad technical choices? Yeah, that's real smart. That's a sign of true intellect, isn't it? Wrong. Linus -
I do place ethical issues over technical ones, if that's what you're asking. And then, why should the vendor have any say on the software that runs on the hardware I purchased from them, after the purchase? Heck, I'd feel *safer* if I knew they couldn't modify the code in my box without my permission. Do you support WGA or the Sony Rootkit? How is this any different? But if they want to keep the ability to change the software in my box, I want that for myself as well. If for no other reason, in case they mess things up. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
s!) Linus, Harald Welte managed to get the keys from Siemens, which sold=20 a "tivoized" Linux router. As typical for German courts, he got the key in= =20 a settlement, so there's no citable court verdict (totally, he AFAIK got=20 two court verdicts up to now). But the way settlements work in Germany=20 suggests that it's likely that the court would have decided that way*. So there is no verdict on this question, but a strong hint that you are=20 wrong, at least under German law (which is not the best money can buy ;-).= =20 And as Tivo doesn't sell their crap into Germany, we can't test it with=20 them. It's you who think the GPLv2 is tit-for-tat, not the FSF, and it's not in=20 the text of the GPLv2; this is your private misinterpretation. The GPLv2=20 is "transitive", i.e. everybody receives the same rights, nobody is=20 entitled to take rights from others (the only person to choose is the=20 author), and this certainly includes technical means (even if it does not=20 explicitely say so), and the GPLv3 is just the same (it's just more=20 readable). All arguments from your side put up are straw men like "hardware= =20 vs. software" and such. BTW: Hardware as it is done today (chips, printed circuit boards, etc.) is= =20 copyrighted as well, as it is much cheaper to copy hardware than to develop= =20 it. *) German judges want to resolve civil cases by settlements. They let the=20 parties pass a few arguments, and indicate which way the judge possibly=20 would decide to help them settle fast. Settlement is way cheaper than a=20 court verdict, so most sane people choose to settle (unfortunately it's=20 often the insane people who go to court ;-). =2D-=20 Bernd Paysan "If you want it done right, you have to do it yourself" http://www.jwdt.com/~paysan/
The FSF have certainly tried to talk to me a bit about it - mostly about some product called GNU/Linux which I had to tell them I'd never heard of At least one important one I think is this: A large number of people contributed to the GPLv2 kernel. They did so on the basis there was an agreement about how the result could and would be used. The GPLv3 changes that agreement, whether for good or bad depends on who you are and what you do. What right does Linus or anyone else have to change the rules and unavoidably harm some of the people who contributed on the basis of the previous licence. Any community project is built around a set of expectations and beliefs encoded in culture, licences, documents and so on. The kernel community was built around GPLv2. A large number of the people involved did so for pragmatic not FSF reasons and are not part of FSF culture. The fact that community isn't interested in GPL3 should not be a suprise, nor should it be seen as it seems you see it to be a failure of the GPL3. GPLv2 is how we've done it, it has been for fourteen years and numerous people have contributed on that basis. Should we kick some of them out of that community because a third party says "new license good". What matters more to the project itself - respect for those who work on it and their beliefs or an FSF attempt to strengthen free software protection ? Thats an "ends and means" type question but I think it explains the fundamental question very well. Alan -
None. It would have to be an agreement between all parties involved. A difficult one, everyone knows. But see, this is a distraction. It doesn't even begin to address the relevant (to me) question: why Linux developers seem to consider v2 better than v3 for their own goals. I can appreciate the difficulties that there would be for switching from v2 to v3. This in itself might be a reason to not even try to switch to another license, no matter how much better it could possibly be. But it doesn't give any hint whatsoever as to why v2 is better than v3. In fact, it simply avoids addressing that point. Now, of course, each individual contributor may have different reasons to be part of the Linux community, and each individual contributor may have chosen v2 or v2+ or any other v2-compatible set of licensing terms for different reasons. I'd very much like to hear (err read), from those who think v2 serves their reasons to contribute to Linux better than v3, why that is so. Thanks, -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Where did you get this impression that GPLv2 enforces this deal? It doesn't, and this is *exactly* why I dispute the claim that GPLv2 The agenda is *precisely* the same: ensure that all users are free to modify and share the licensed software. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Software's the bit that's infinitely replicable at zero cost. Hardware tends not to be. Rob -
There's no "zero cost" for software replication, either. You have to pay fo= r=20 your media, even if today's price for a GB harddisk space is just 20=20 Euro-cents. You have to pay for your bandwidth (and even if it's a=20 flat-rate, the maximum amount of data you can get through is=20 bandwidth*(seconds per month) for one month fee). Hardware is replicated as= =20 well as software, the cost for hardware replication is higher than for=20 software replication, because more things are to do. The basical principle= =20 of producing a CD-ROM and a chip is exactly the same: lithography.=20 You "print" it. A decade ago, ES2 had made chips by direct e-beam=20 lithography, so the offset of the mask costs were eliminated. With an ES2-like process, you could have your "free software CPU", where yo= u=20 design modifications yourself, send the file to the fab, and get your=20 customized chip back for essentially the same price as a non-customized=20 version (supposed all the tool-chain would be free software, and not=20 horrible expensive Cadence/Synopsys/Mentor software). =2D-=20 Bernd Paysan "If you want it done right, you have to do it yourself" http://www.jwdt.com/~paysan/
But see, I'm not talking about getting permission to hack the hardware. I'm only talking about getting permission to hack the Free Software in it. It's your position that mingles the issues and permits people to use the hardware to deprive users of freedom over the software that they're entitled to have. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
You have that permission. You can hack on it all you want. Oh, but you want to hack the hardware to accept it? That's a totally different issue. If so, buy a Neuros OSD box. Really. Go to google, and type in "Neuros OSD". Do it *now*, and then stop wasting our time. You *can* do what you want to do. Linus -
Unfortunatly that device has some closed source kernel drivers, so you don't really have "full" control over the system :( thanks, greg k-h -
No you're not...you're talking about being able to hack the software The software license controls the software. If the hardware has restrictions on it that limit what software it will run, then that is unrelated to the software license. There is nothing stopping you from taking the code for the tivo, modifying it, distributing it, or even running it on other hardware. Suppose I had some machine that will only run microsoft-signed binaries. Would it be at all related to any software license that this machine won't let me run linux? Chris -
Yes. You wouldn't impose restrictions on modifying the software like that, now would you? Even though the GPL says you can't impose As in, the license controls the software. If a patent creates restrictions that limit what you can do with the software, then that is unrelated to the software license. As in, the license controls the software. If a discriminatory contract limits what you can do with the software, then that is unrelated to the software license. As in, the license controls the software. If I send you the source code, but it happens to be protected by a key that only the hardware can decode, and it won't decode for you, then that is unrelated to the software license. True. But TiVO is still imposing further restrictions on how I can modify the software stored in their device, while reserving that ability to itself. This is wrong. This is not "in kind". This is not "tit-for-tat". Tit-for-tat is: if they can, then I can too, and That would be an unfortunate machine to have, but if Linux or some other GPLed software was not shipped in it, then I don't see how this is relevant to this discussion. It's not about the hardware, it's about the software in it, and about passing on the freedoms related with it. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
But that right has never been guaranteed by the GPL. It might have been the *intent* of RMS when he wrote GPLv1 and the *intent* of the FSF when they wrote GPLv2, but intent is worth exactly *NOTHING* in the law *UNLESS* that intent is spelled out. Anyway, as I've pointed out before: replace != modify You can *replace* parts of a program and it will be a modification, you can *replace* components of a piece of Hardware and it will be a modification but replacing one software component of a device with another is *NOT* a modification. Why? Because the hardware hasn't changed at all - the hardware is merely there so the software can perform its job. And since you are *replacing* the *ENTIRE* piece of software, it isn't a modification of the Exactly. However, you are making it about the hardware by making the claim that "replacing a program, in its entirety, with another is a modification". It isn't. A modification is when you replace or change a *portion* of a program. By your logic I could write an operating system that is 100% binary compatible with Linux and I'd be *required* to release it under the GPL, because, even though it *replaces* Linux, it's still a "modification". DRH -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
It's not a straw man. See, I was just showing that there's precedent to ensuring that other tricks can't be used to deny users the freedoms that the GPL is meant to defend. By pointing out this is in the GPLv2, you acknowledge the point I wanted to make. So what is it that makes hardware so different that it can be used as That's the different between legal terms and the spirit. And the I'm not sure I agree with the reasoning here, but I'm already convinced that the argument about modification by replacement won't fly. But then again I ask you: why do you think TiVO is making these hardware locks? What do they want to cause or stop? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
On 15/06/07, Alexandre Oliva <aoliva@redhat.com> wrote: [snip] Why can't you understand that the GPL v2 is a *software* license, it doesn't cover hardware at all. If I take some GPLv2 software, modify it and then distribute it on a CD-ROM and provide the source code as well, then I have complied with the terms of the license. If I take the same software, make the same modifications and distribute the software in a ROM inside some piece of hardware, but still allow people access to a copy of the source code used to build whatever I put inside that ROM, then I've also complied with the license. In neither case can you modify the copy on the hardware (be it ROM chip or CD-ROM), but that's not required by the license. As long as you have access to the source code it's OK. The license says nothing about you having to be able to update it on the hardware. The license only says you need access to the source code. No one is taking away your freedom to change the source or redistribute it or whatever. The only thing locked hardware prevents you from doing is installing modified software on that specific piece of hardware, but that is completely outside the scope of the *software* license. I can't know for a fact what TiVO wants, but I can guess. 1) Maybe they want to prevent you installing modified software on their hardware, then contacting them when you break it, costing them money in customer support etc. 2) Perhaps they don't want to risk being liable if you modify the software on their box in a way that allows you to use it as a means to break the law. 3) Maybe they don't want you to modify the software running on their hardware in such a way as to use the software to obtain intimate details about their hardware that could be used by a competitor to create a product superiour to theirs. All quite valid reasons in my opinion. -- Jesper Juhl <jesper.juhl@gmail.com> Don't top-post http://www.catb.org/~esr/jargon/html/T/top-post.html Plain text mails only, ...
The GPLv2 is a copyright license not a software licence, indeed there is no such thing as a 'software licence'. It deals with the circumstances and manner in which you are permitted (by the author) to make copies of their work, to modify their work and in some cases to perform their work (plus other sundry rights). Copyright law doesn't care whether the object in question is as abstract as computer source code (providing it has been 'fixated' in some form) or a two hundred foot high art installation - or a combination of the two. So irrespective of the whole pointless debate going on you are trying to and all wrong. Look up the owning and controlling interests in Tivo and you'll find the correct reason - stopping you doing evil things like keeping movies you've recorded or uploading them to the internet [which ironically of course is the entire effect of the whole 'convergence' thing] Alan -
Hmm, wouldn't that be my guess nr. 2? A way to use the hardware to break the law... Anway, the whole point of my post was mainly to /try/ and say that the GPL gives you a right to obtain source code for modifications, but it doesn't say anything about being able to run a compiled version of that source on any specific hardware. -- Jesper Juhl <jesper.juhl@gmail.com> Don't top-post http://www.catb.org/~esr/jargon/html/T/top-post.html Plain text mails only, please http://www.expita.com/nomime.html -
On Friday 15 June 2007 02:59:31 Jesper Juhl wrote: And you are correct. It is also clear, thanks to language directly in the GPLv2 itself, that there is no "intent" of the license to cover that situation. DRH -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
You're again confusing legal terms with the intent. The legal terms provide an indication of the intent, but the preamble, along with the free software definition it alludes to, do an even better job at that. That said, the letter of the GPL explicitly says that the act of running the program is not restricted, and that this is outside the scope of the license. It's a copyright license, and per US law, running software is not regulated by copyright; this is not so elsewhere, and local interpretation indicates that the intent of the license is indeed to grant unlimited permission to run the program and modified versions thereof, based on the free software definition. But then, when someone says "I won't let you run modified versions of this software on this hardware I'm selling you", is this not a further restriction on the exercise of the rights granted in the license? And, per the spirit, if the manufacturer can still install and run modified versions of the software on that hardware, is it not failing to comply with the spirit of passing on all the rights that you have? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
And the preamble, not being part of the active portion of the license, has absolutely *ZERO* bearing. Just as it is not the *intent* of RMS, the FSF or *ANY* person (or legal entity) that had a hand in crafting the GPLv2 or GPLv3 which is looked at when determining the "intent" of the license. It is the intent of the person and/or "legal entity" that has placed their work under said license. And that may be what courts in Brazil believe, independent of the fact that the license itself *intentionally* limits itself to "copying, distribution and modification". That legal decision, in fact, may not have been motivated by the actual belief that that was the intent of the license - it could (and, from looking at the situation and available facts, might actually have been) In Brazil, because the courts there have rendered a judgment that the license requires the unlimited running of the covered work. But that is *unfairly* applying a license on a piece of software to the hardware on which it runs. Based on your logic the hardware manufacturer would have to enable people to run code compiled for an entirely different processor. Not that it matters in Not in the least. They have the rights to "copy, modify and distribute" the "source code for a work". That is *EXACTLY* the set of rights they have to the code, and it is *EXACTLY* the set of rights they pass on. The GPL does not apply to any *BINARY* form of the work, except for the fact that you are required to provide the source code that was used to generate the binary. The GPL *clearly* defines "source code" as: "the preferred form of the work for making modifications to it" It goes on to state: "For an executable work, complete source code means all the source code for all modules it contains, plus any associated interface definition files, plus the scripts used to control compilation and installation of the executable." (Note that, since the "signing" of the TiVO kernel is part of the installation ...
No disagreement. You keep forgetting that I'm not here to say what Linux licensing means or doesn't mean. It's not build script, it's just regular source code, indeed. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
it is _you_ forgetting to read what you wrote just 1 mail ago above. _Read_ it: "The legal terms provide an indication of the intent, but the preamble, along with the free software definition it alludes to, do an even better job at that.". Your point was totally bogus, and you have been pointed out that your point was bogus. And your answer - instead of admitting that you were wrong once again (i'm not even asking you to apologize for wasting our time) - to pretend that there is "no disagreement" and to patronize your discussion partner with a "you keep forgetting ..." phrase and a non-sequitor statement? How low can this discussion get? I'm truly amazed ... Ingo -
You still don't seem to get the difference between spirit and letter, and the difference between author of the license and licensor of the software (I guess I wasn't clear about this). When I talk about the spirit of the GPL, I'm talking about how its authors designed it, how they described their motivations in the preamble, in the Free Software Definition, in speeches explaining it, etc. When I talked about the meaning of the Linux licensing above, I'm talking about the intent of the Linux authors when choosing the GPLv2. They aren't necessarily the same. Some Linux authors may have read the preamble and understood something else. Some may have simply skipped it, and focused in the legal terms. Some even understood the legal terms in different senses. But they have all agreed to license their code under GPLv2, whatever the motivations each one of them had, and none of them has to match the spirit of the GPL. Now, the spirit of the GPL, the intent behind its design, is something that may be entirely different. And when I say that GPLv3 didn't change the spirit of the GPL, I'm saying that from the perspective of someone who understands very deeply the philosphy and motivations behind it. Please understand that these are two separate issues. That the spirit doesn't change, and that you don't share that spirit, or that you didn't think that was the spirit, doesn't justify a claim that the revision is changing the spirit. It's not. The spirit is, and has always been, to defend users' freedoms (the 4 freedoms of the Free Software definition), such that Free Software remains Free. The fact that they can be different means that GPLv3 may not match the goals you had when you chose GPLv2 for your contributions. That's unfortunate. But this doesn't mean that GPLv3 is changing the spirit of the GPL. It's merely exposing that your goals don't match the spirit of the GPL. There's nothing wrong about this, GPLv2 is not a bad license, just like the LGPLv2 is ...
why do you start babbling about the GPLv3 when the false statement you made was about the GPLv2 - as visible _very_ clearly in the first quote above?? Again, as a reminder, this point was presented to you (see the quotes above), in the discussion about whether the Tivo is fine by the GPLv2 or not: ' It is also clear, thanks to language directly in the GPLv2 itself, that there is no "intent" of the license to cover that situation. ' to which you replied with this falsehood: ' You're again confusing legal terms with the intent. The legal terms provide an indication of the intent, but the preamble, along with the free software definition it alludes to, do an even better job at that. ' and you have simply been pointed out that what you say is trivially false - the simple legal fact is that the GPLv2 does not "embedd" more than what is its letter and, to the lesser extent that letter may be ambigious, what Linus' (and other copyright holders') intent is and was. Not RMS's external intentions or the "free software definition" you mention. how about just simply admitting that you were wrong about this, instead of putting up (and beating down) yet another non-sequitor strawman argument? Is that really that hard and embarrasing to do? I've yet to see a _single_ instance of you admitting in this thread that "oops, it seems i was really wrong about this point. Sorry.". Or are you one of those perfect humans who are never wrong? ;-) Ingo -
That's false. I've explicitly avoided discussions on whether the legal terms of GPLv2 permit tivoization. What I've done was to discuss whether the tivoization was in line with mission (since you seem to have a problem grasping the notion of spirit of a license) of the license, because some people (yourself included) claimed GPLv3 changed the spirit. That others tried to steer the discussion away from this apparently incomprehensible concept of spirit of the license, to the point of their and your getting utterly confused and making nonsensical claims about inconsistencies in my reasoning, moving it into the legal terms that I was not willing to discuss, is not my fault. My participation here was about intent, about spirit, about mission of the GPL. As I've already made it clear, this doesn't necessarily match intent of people who chose GPLv2 as the license, and that's fine. Copyright holders know the spirit as they understood it, and that's how they meant to license their work. But the spirit (mission) of the GPL is the one the FSF wrote about in the preamble. No amount of "but the legal terms say such and such" or Which shows you don't understand the notion of "spirit of license" (as opposed to intent of licensing, which I AFAIK invented today to try to dispell this confusion), and that the fact that the letter of the license doesn't have bearing about the intent of the author of the I'm not. You are. Really. Until you understand the difference between "letter of license", "intent of licensor" and "spirit of -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
I just can't stand this anymore. Might I submit that nobody can be wrong in a hypothetical? The chances of the kernel changing to GPL v3 are the same as fort knox being ripped off. IM GLAD! This means I can continue to merge. Folks I will summarize myself one last time and stop disrupting your list. I (personally) think enough time has been taken from productivity and I'm becoming awfully concerned about the integrity of Alexandre's ass. He's not here to irritate you, he's here to do what he thinks is right. I admire him for that considering the adversity. (its not a mantra, you can keep reading, I promise!) If I found advantage in a bug In Linux, and you fixed it, I have a choice to not apply the patch so that the bug remains useful to me. The FSF found needs to produce another license to serve all of the people who have interest in the FSF. This does not mean you need to *apply* it. Its there if you want it, but its up to you. Just as I told Linus, This is a kernel, not *&$(*# sed, and its a decision I'm happy was not taken lightly. NOBODY should be influencing you with what you do with your contributions!! Make up your own minds in your own ways or have fun in the parrot cage where you sit in your own shit until someone else cleans it up. Who likes that? I am asking as strongly as one can ask, *please* put this to bed. Continuing at this point can serve no useful purpose and everyone involved shares a common intent to be useful! Can we agree on that spirit? Opinions have been expressed over, and over and over again. Everyone knows what everyone thinks. We're all doing more harm than good. Who hates waste? I do. Best, --Tim -
And you have been given evidence that it does. It may not break the spirit as you, or anyone else that believes that the FSF is the singular fountain of truth, see it. But it does break the spirit as a large fraction of the populace see it. Why do they like GPL3? Because they don't like what TiVO did - they see it and say "but that's not fair", so when they see that GPL3 makes it a license violation they accept GPL3 without understanding the bigger picture. To *anyone* who is considered an "Adult" in *ANY* nation and still wants to scream "its not fair" when something doesn't go their way or somebody does something they don't like I have this to say: Grow up! Life *ISN'T* fair - at Not even. If it's anything it's "input to a program". QED: It isn't covered by the license. Hell, give me a week and access to a signed TiVO kernel and I'll produce a signed kernel that is functionally equivalent - it won't run on a TiVO, but I'll have replicated the "signing script" and process that TiVO uses. It isn't until you start extending the definition of source code in strange ways that the key becomes "source code". Hell, for all *anyone* knows (that isn't employed by TiVO) they could enter the key *manually* - ie: interactive input. DRH -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
That's not true. Again, ianal, etc etc, but: "Intent" *does* matter, and if you wrote down the intent at the time you entered some legal agreement, that actually also has non-zero bearing (as it can be used to _show_ intent more clearly than claiming fifteen years later "but, your honour, I _intended_ to do something else"). BUT! Intent only matters if the contract reading itself isn't otherwise clear! In other words, intent can never *override* the contract. But if there is some actual legally unclear area, the intent can be used by a judge/jury/arbitrator to _guide_ them to the decision. And the better documented that intent is, the better it is in that respect. Intent _after_ the fact is not very meaningful, is it? So the preamble *does* matter legally, it just matters a lot less than the actual legal wording itself. Put another way: the "intent" of the parties can be, and is, used to read the contract, but it's actually a mutual thing for a contract (since you have two or more parties in question - the intent of _one_ party doesn't really matter, if he cannot show that the other party agreed with that intent! And to get back to rms' personal history: _his_ intent, and _his_ personal history, and _his_ other documents don't matter for an agreement that doesn't include him). So as such, a preamble can be used to show the intent of the parties, and thus to guide any reading. Of course, so can "real life". If a certain reading of a contract doesn't make sense (and that has nothing to do with "intent"), then that obviously cannot be what either party really agreed to. Oh, and documented intent can be used for laches, ie it can be a _defense_ against an accusation. If somebody has publicly stated some intent, he can't really complain later when somebody else tried to fulfill his intent, can he? Linus PS. "Intent" can obviously have a _totally_ different legal meaning too. The difference between "murder" and "manslaughter" is ...
yeah. What comes up periodically in GPLv3 discussions as 'proof' of what the GPL means are totally detached statements of the FSF and of RMS, often written a decade _after_ the GPL has been chosen for a license of the Linux kernel. (the whole anti-Tivo line was invented well after the fact.) And those statements have little bearing on the interpretation of the license of GPL-ed works. (unless, of course, the author of a GPL-ed work agrees with those statements and intends them to be his interpretation of the license.) Ingo -
Wrong (again) The pre-amble is incredibly important as is the intent of the license creator and even more so of the author. When trying to solve a dispute the process starts with the legal equivalent of banging the two parties head together in the hope they see sense. If that fails then the legal wording is considered in detail. Where it is ambiguous the surrounding context is considered in order to understand the probable intent of the case. Finally the stated intent of the author is considered in defence (The doctrine of estoppel), and at least in UK law whether their intent was honest (The doctrine of clean hands) Legal disputes almost always end up about the things that are not clear (if they were clear one side would shut up and put up) so the preambles and statements are terribly important when this occurs, along with the context and history. Thus for example the fact Linus has said he believes what Tivo does is ok means he can no longer sue Tivo for doing it. They are relying on his promise. Alan -
[ Damn. I moved you to my flamers list, and then I started reading it. I'm addicted to flaming. Sue me. I really do enjoy it too much. If I didn't do software development, my full-time job would probably be to troll various internet sites and try to set up flame wars. I'm bad, I know. It's an addiction. I'm not proud. ] Actually, they didn't want to lock down the hardware at all. The first versions of the Tivo was really quite hackable - and people started hacking them. They were basically forced to add lockdown by the content vendors. You can call them evil for "caving in", but hey, it was their whole market. They really had no choice. Being a company actually limits you in some ways.. If you don't want to cave in to content providers, use a regular PC and soemthing like MythTV. You will probably also have to use the analogue hole, and will have a really hard time unscrambling digital cable TV signals of your own, but hey, you can see it as a challenge. At least in places where it's not illegal. And yes, there are bad laws in the US. But blaming Tivo for them is ludicrous. And the *laws* won't get fixed by software licensing either, quite the reverse. The GPLv3 will just make free software that uses it *less* relevant in that space, rather than more. For example, I'd rather have some GPLv2'd DVD player software that does *not* come with a de-css key (I can get that key myself quite easily), and that thus gets distributed in a "useless" form, than have a GPLv3'd DVD player that cannot be distributed at all, because it needs the magic unlocking key, and distributing the css key is illegal in some countries. Or if I was an mplayer developer (which I'm not - so I have absolutely *zero* say in the mplayer license - please don't take this as anythign like that), I'd prefer for mplayer to be GPLv2, simply because that way I could see my software in some high-end (legal) DVD players that actually complied with the insane laws that exist. ...
Only in the USA. In most of the world its considered quite normal that you can plug a USB disk into your PVR, save stuff to it and then plug it into your PC. Alan -
They can do that. They will still be able to do that with v3. All they have to do is to throw away the keys that enable themselves Just like v2 hinders their many customers. Are you so sure v2 is better in this regard? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
where does this false sense of entitlement come from? The hardware maker ows you nothing but what is written into the GPLv2. Not more, not less. (In fact, most hardware makers that utilize free software today give back _substantially more_ to the community than the license requires! For example they are currently the largest employers of free software developers - although nothing in the license forces them to do so. Why? Because the economic rules that the GPLv2 creates are healthy.) you are not "entitled" to dictate the hardware's design (or any other copyrighted work's design), even if the license gives you the power to do so. By your argument we'd have to put the following items into the license too: - free on-site training for free software developers about the hardware's inner workings. (It is justified to teach free software the same know-how as in-house engineers of the hardware maker. Without this, users are hindered in their freedom to use and effectively modify (fix) the software.) - free access to all the hardware diagnostics tools that the hardware maker has. (Without that it might be impossible to modify the software as efficiently as the hardware maker's own engineers can do it.) - free samples of the hardware to be sent to free software developers, upon request. (The hardware maker's own engineers have free access to samples. Otherwise free software users might not get the same level of driver support as the hardware maker can achieve.) - free access to the hardware manufacturing equipment. (If i wish to modify the free software in a way that requires more RAM than the hardware has, i need access to the manufacturing equipment to produce a new version of the hardware that can run that free software. The hardware maker has this right and flexibility to modify the software, so i should have that same right too.) see how quickly your argument becomes totally ludicrous, if brought to ...
That, BTW, is perhaps the worst problem with v2 (inherited by v3). WTF _is_ "the spirit of the license" and who gets to decide if two licenses are in the same spirit? As soon as we get to "well, original authors of the license are the final authority on that", we are in the "I've always said ..." country. Look, humans _suck_ at revision control, especially that of our intentions and opinions. It doesn't even require malice, all ancedotes about spouses/mothers-in-law/etc. nonwithstanding. We all easily fall into belief that we had always meant what we mean now; that even if we said something different, it was just a poor wording; that if we had known what we know now, we would certainly had come to the same conclusions we have come to now. "In the same spirit" is just about the weakest requirement in that area. I.e. the most prone to drift, especially when one is an ideologist and thus has severely decayed integrity to start with. Call it a professional disease of crystal ball users - or a prerequisite for playing a visionary, if you will ;-/ -
yeah. I see this as: "RMS does not want to let go of the community". This clause amounts to "power to relicense" _vast_ amounts of free software and this is by far the worst problem with the "GPLv3 process". The GPLv3 process was pretended to be "open", but regardless of what the "GPL comittees" said, in the end it was one person: the president of the FSF (Richard Stallman) who singlehandedly decided what went into the GPLv3 draft and what not. For example he singlehandedly has ignored all the criticism that the the "Tivo" section has received. And note how hypocritic RMS's position is here. Where is that freedom when it comes to the licensing process? Why does RMS have more rights over modifications to the license than all the other free software developers have? Should not he give that freedom to others too? Shouldnt there be a fair and just election, a vote? You know, that democracy thing. And with his current attitude he affects somewhere around of 1 billion lines of free software. Via a license that is just a few hunded lines long. I believe RMS should accept the fact that most of that code was written without people having bought into his ideology, and he should accept _responsibility_ for the power he has acquired by genius or by accident (your choice) and he should try to _understand_ how those people tick - instead of trying to further his own personal agenda. He shouldnt say what amounts to "oh, my original intent was this and that, if you didnt understand it and still wrote code and used the default 'or later' license, it's your damn fault". He should accept that what happened happened, after he wrote 100,000 lines of original GNU code another ten thousand people wrote about a _ten thousand times more_ code. He should also accept that the "open source" community is about many other things, and it is alot more varied than his thinking is. He does not have to _like_ Tivo, but he should try to _understand_ them, and he should be ...
No, you're confusing two very different situations. In the case of TiVO, it's getting out of its way to make sure users can't enjoy one of the freedoms that the license says it ought to pass on. In the cases you mentioned, the company would have to get out of its way to put the other parties on equal grounds. The former is bad, it's against the spirit of the license, it's a further restriction. The latter would be nice to have, but it would be wrong to demand it. You're picturing the difference between blocking the way such that you can't get there, and actually taking you there. What the GPL seeks is just that you don't get in the way. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
hey, that's progress. If you concede this single point then your arguments about the Tivo situation all fall like domino stones. Just the GPLv2 license says no such thing, and you seem to be mighty confused about how software licenses work. the GPL applies to software. It is a software license. the Tivo box is a piece of hardware. a disk is put into it with software copied to it already: a bootloader, a Linux kernel plus a handful of applications. The free software bits are available for download. the Tivo box is another (copyrighted) work, a piece of hardware. so how can, in your opinion, the hardware that Tivo produces, "take away" some right that the user has to the GPL-ed software? Because they distribute the software and the hardware in the same package, and because the hardware (as _ALL_ hardware on this planet) has certain limitations? It was _your_ choice to buy that particular hardware+software combination, with whatever limitations the hardware has. One such limitation of the hardware might be that its color is butt-ugly pink. Another limitation might be that the buttons on it are too small for elderly people to press. A third limitation might be that it's not a general purpose computer and that it's not freely programmable by the end user. Bugger, what did you expect? Why didnt you buy a green PVR? Why didnt you buy a PVR with larger buttons? Why didnt you buy a general purpose computer? Did perhaps the Tivo look like a how about quoting what i wrote and rebutting it specifically if you disagree with it, instead of writing a non-sequitor generality? You are involved in compiler development, so you should have the mental ability to follow logical arguments and you should be able to conduct a meaningful and objective discussion. Lets look at one of the examples i by your argument, the user has some "right to modify the software", on that piece of hardware it bought which had free software on it, correct? By your argument, the ...
You can GPL a new graphical logo you painted on your toilet seat, you can GPL hardware designs. It might not be a good licence for either but it is A Tivo box is a collection of literary works protected by copyright, designs protected by design patents and copyright, names and logos protected by trademarks, functionalities protected by patents and many more things. These are the things that restrict what I may do with it and how I may treat it. The collection of bits of metal and sand aren't really of relevance in terms of licencing. If it was a generic housebrick with none of these things attached then within the law I can do what I like with it including copying it. A book is a copyright work but the copyright is about the literary work and the fact it is on paper is largely irrelevant. What determines your usage rights for those pieces of paper are the literary work it carries not the Except the keys - which may nor may not be required depending upon how a court (not a mailing list) interprets the phrases "The source code for a work means the preferred form of the work for making modifications to it" and "For an executable work, complete source code means all the source code for all modules it contains, plus any associated interface definition files, plus the scripts used to control compilation and installation of the executable." If you ask the legal profession about this seriously the answer you get is bluntly "There is no caselaw I am aware of", which means that nobody knows. Obviously Tivo and their legal counsel have formed an opinion and You can't copyright hardware. Sorry but if you are going to try and have a detailed logical argument you need to start from a rigorous base point. Alan -
a "software license" is a common shortcut for "copyright license for copies of software works". It's a commonly used phrase. In fact it is used by the FSF itself too: http://www.fsf.org/licensing/essays/free-sw.html "To decide whether a specific software license qualifies as a free ^^^^^^^^^^^^^^^^ software license, we judge it based on these criteria to determine yeah - the GPL can be applied to most types of works recognized by If you are into technicalities then you fail to achieve that "rigorous base" by a wide margin. The Tivo box is not "a collection of literary works", it is a piece of matter, that also happens to contain fixated copies of literary (and other) works. The Tivo box is just one copy of those works - it is not "a collection of literary works". (Only if there was just a single Tivo box on the planet then could that box itself be meaningfully called a collection of works - a single and unique "master copy" of a work can be called the work itself.) and that distinction, although fine, is very important. Look at GPLv2 section 0: " 0. This License applies to any program or other work which contains a notice placed by the copyright holder saying it may be distributed under the terms of this General Public License. " the work is not the copy! The work is a more 'abstract' entity. The word "copyright" comes straight from that: the right to create specific copies of the work. And that's another reason why it's nonsensical to suggest that somehow the GPLv2 gives us the right to influence the hardware environment that the copy of the kernel got fixated into. We dont. ( unless that hardware environment too is a copy of a GPL-ed work or it is a copy of a work that is a modification of or derives from a GPL-ed work - but in the Tivo case it isnt. It's a collection of copies of works and derivation does not "jump" from the harddisk to the hardware. ) More down the technicalities road: ...
The physical matter is irrelevant. I am perfectly entitled to own, shape and fiddle with sand and bits of metal. If I wish to remove the software from the tivo, melt it down and cast the result into the shape of an obscene gesture and wear it at the tivo shareholder meeting so be it. At that point it would be my work made from melting the tivo that was the Each copy is an instance of the work. My copy does not change its status, nor its legal situation if someone rounds up every other tivo and melts them down. I guess if you want to be pedantic the Tivo contains "an We have every legal right to do so. I am perfectly permitted to try to grant you the right to reproduce my work only if you pay me $25 and the reproductions are provided in a silver box with flashy blue lights. I am perfectly permitted as author of a work to tell you "no". You as box maker are perfectly at liberty to tell me where to go stick my offer and just not use my work. I can influence your hardware all I like. What I cannot do is influence you in any way if you decide not to take any action involving my copyright. Nor can I through copyright require certain kinds of condition (eg control other works on the same media) as that requires contract law and a proper contract, nor certain things that are deemed to be unlawful by the state (The GPL gives me the right to modify the code to break into the DoD, steal all their secrets and mail them to the Iraqi government, Ok I guess thats a question of level of abstraction, like being "an The Lawyers don't. As experts in their field I generally trust their view on this. Also remember that lawyers assess legality not morality so there Agreed. But GPLv2 has many absurdities such as the way it handles copyight notices. It wasn't designed when GUI apps were the norm, it predates web hosted services and the GPL mobile phone was, I suspect, not on the drafters radar let alone in their pocket. If my toaster is ROM based then it is difficult to argue that ...
I'm not the FSF, and I don't speak for it, but it seems to me that this would be "mission accomplished". The goal AFAIK is not to force people to enable others to hack the hardware or software to their liking. The goal is respect for the freedoms, it's not making it more difficult for others to do what you can and want to do. I guess it also goes under the name "Golden Rule". Others might phrase it as tit-for-tat, or quid pro quo. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
This is insane. You start with a lofty ideal involving "freedom", and when you end up with a meaningless technicality (and in technical terms a change for the worse) you consider it a victory? Yes, I know that this is what happens in politics (look here, our laws had an effect!), but I have more respect for you than to think you fall for these kinds of games. I do not wish to revise my opinion. Bernd -
It accomplishes the mission in that everyone is on the same grounds. Same freedom for everyone. If the vendor tries to keep a privilege over the software to itself, denying it to its customers, it's failing to comply with the spirit of the license. It's really this simple. Is this so hard to understand? The goal is not to push vendors away from GPLed software. If they can't permit modification of the software, that's fine, they can still accomplish this. What they can't do is deny it to customers while they retain it to themselves. This is unfair, this is wrong, and this disrespects users' freedoms. Therefore, the GPL should not permit it. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
How the hell does that improve the situation for users? Alexandre, please realize that you are preaching to non-believers. I realize that you have accepted the FSF credo, but if you want that conversation to go anywhere you have to separate the things you believe in from the things you can rationally explain. Apologetics of your variety is not going to cut it. _Can_ you separate the things relying on your beliefs from the things that can stand on their own? If you can't do that, please stop wasting everyone's time and bandwidth. It's a secular maillist; what any of us might happen to believe in is personal and frankly, none of your damn business. -
Maybe it doesn't. How does it make it worse? Maybe just providing an incentive for the vendor to respect users' freedoms will do the trick, and *some* vendors will do, while those who can't will keep the status quo. I've already explained what the spirit of the GPL is. I've already explained that the anti-Tivoization provision is in line with it. I've already asked in what sense Tivoization makes for a better tit-for-tat, and got no reply whatsoever, rational or otherwise. I have already hinted at why it makes things worse. You don't have to believe what I believe to analyze the arguments rationally, just like I don't have to believe what you believe to analyze your arguments rationally. We may still get to different conclusions as to what is better, if we have different values guiding us. But whatever conclusion you arrive at won't change the plain fact that Tivoization is against the spirit of the GPL, because it is a means to restrict users' freedoms that the GPL is designed to defend. It's really this simple. I'm not trying to convince you of anything other than that the spirit of the GPL is not being changed at all. You don't have to agree with that spirit in order to accept this simple fact. And while people keep on spreading this lie, I'll be inclined to point out that it's false. See, this is not about promoting GPLv3, or "pushing it down your throats", as some have claimed. This feeling is just a symptom of the high rejection for the FSF ideology, that appears to blind so many smart people from rational reasoning on matters that touch the FSF ideology. This is not even about showing that the letter of GPLv2 prohibited Tivoization. My arguments concerning Tivoization were all about the spirit of the license, and unfortunately so many people seem unable to tell the spirit from the letter that they keep on moving the discussion to legal technicalities, and then they shoot straw men and feel happy that they shot an argument. But the ...
Now not even the vendor can upgrade the software in the hardware and fix problems for the user. The user loses. -- Jesper Juhl <jesper.juhl@gmail.com> Don't top-post http://www.catb.org/~esr/jargon/html/T/top-post.html Plain text mails only, please http://www.expita.com/nomime.html -
You're seeing that the wrong way.
The correct response is (and I quote from the manual, pick one talking
point at random):
"This is a great step for freedom, as the users now have exactly the
same rights as the vendors."
"When we talk about free software, we don't talk about 'free as in
beer', we talk about 'free as in buggy and unfixable'"
"You're now at least no less free than anybody else!"
"Oh, except for the fact that those other people still design the
hardware you are using, and the programs you watch. But we have a
plan for that too! We will make the GPLv4 outlaw Disney and Britney
Spears!"
"In order to protect your freedoms, we sometimes have to take some
freedoms away. In particular, the freedom of critical thinking got
revoked last year, because people were just too 'confused'"
"There are no American Infidels in Baghdad. Never!"
There's a long list of those things, but sadly I didn't have time to copy
them all when I sneaked into the FSF main office in my ninja suit under
the cover of darkness.
Linus
-
ITYM "upgraded to Loyalty To The Cause, which is the best antidote against doubt and confusion plaguing the so-called rationalists". -
Assuming the vendor's intent as for patching the software is to help the user. If the vendor doesn't want to let the user do that independently, why should this assumption hold? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
No. You've explained one thing only: that you cannot see that people don't *agree* on the "spirit". You think that there is only one "spirit", and that you own the code-book, and that your spirit is thus the only right one. This is where we started. The same way you seem to think that "freedom" has only the meaning *you* and the FSF give it, and that somehow the spirit of the GPL includes the "four freedoms" that aren't even _mentioned_ in it. THAT IS NOT TRUE. But equally importantly, it's not even *relevant*. Nobody is suing the FSF for contract violation for changing the spirit. Yes, people have brought out the argument that the GPLv3 actually even changes the spirit, and you don't seem to realize that people can have different opinions. You just repeat YOUR OWN OPINION about the spirit over and over again. But even if the spirit changes, so what? The GPL doesn't actually say "same in spirit". It says "similar in spirit", implying that the spirit is "similar". .. and we have already explained to you that it's irrelevant. So let's get back to the *real* issue: - The GPLv2 was ok with Tivo. I really tried to explain to you *why* that was, but by now, I can't be bothered any more. Even if you cannot understand it, just accept it. And if you have a hard time accepting it, just accept the fact that the FSF thinks Tivo cannot be sued, which is just another way of saying "they didn't actually break the license". - *I* think Tivo is fine. Other people think Tivo is fine. Other people have told you they think what Tivo did is fine. Some people have even said that they don't like Tivo, but that they don't think the license should stop Tivo. - The GPLv3 tries to stop Tivo. Instead of mumbling about your spirit and feelings (I need to be a whole lot more drunk before I start caring), how about you look at those three statements, and then admit that you see why the people in bullet#2 think that GPLv2 ...
They don't have to. Just like nobody but you can tell why you chose the GPLv2, nobody but RMS can tell why he wrote the GPL. And the intent behind writing the And that's the point that I'm fighting here. It does not change the spirit. It's still ensuring that Free Software remains Free: respecting and defending the four freedoms defined in the Free It is relevant. It was the point that my participation was intended to address. I guess it is just too hard to accept that an FSFer could not be There's disagreement about this, even among developers of the kernel Linux, and you know it. I know you're always right and I pretend to respect that ;-), but why do you think your opinion should prevail over theirs? Don't you realize that they're as entitled as you are to enforce the license, and in the way *they* (not you) perceive and meant to license their code? And then again, this is not something I'm overly concerned about. I probably don't have enough contributions to Linux for my take on it to make any difference whatsoever. This is not the real issue at all. The real issue, that brought me here and got you to name calling me and the FSFs, is that there were false claims about the GPLv3 that I wanted to dispell, particularly the point about its changing the spirit. The anti-tivoization provisions are in the spirit of the GPL, and so much so that a number A minor nit, but no, it doesn't. It tries to stop the practice of tivoization on programs licensed under the GPLv3. TiVo has a number of choices, and so do other tivoizers, even if they Even though anti-tivoization furthers the quid-pro-quo spirit that you love about v2, and anti-tivoization is your only objection to v3? That's what I don't understand. This is so obviously contradictory to me that it's almost funny, and you've so far dodged my questions about this and refrainied from commenting on this contradiction so much that Yes. That's precisely why I don't understand your stance. Because ...
See, you can't even keep the FSF's "Free Software Definition" and its inherent "religion" out of the discussion. Sure, the FSF can claim that the GPL is intended as a way to "defend" the "Four Freedoms" defined *BY* *THEM*, but unless alluded to in the license, the only bearing it can have, anywhere, is on the "intent" of the license, as seen by the FSF. And if the "ability to run a "covered work" on any piece of hardware" is "freedom 0" then binary distribution is in violation of the "spirit" - I can't run an x86 binary on a PPC. Isn't that a "designed in hardware restriction" that violates Not in the least. They still have to release their changes if they want to sell their devices. Or are you so blinded by your belief that the FSF and RMS It is only *YOUR* opinion that the GPLv3 is the better license. As Linus has No, it reduces their motivation to improve the software on *those* devices. If they like the software enough to actually download the source, they probably also liked it enough to install it on their computer *AND* will modify it to Agreed. The disagreement is about what that spirit is. I feel that its spirit is in the free and open exchange of ideas, as personified by the software people write. I *ALSO* feel that it's spirit lies in the phrase "do whatever you want with the software as long - but if you add your own ideas to it, give them back to the people like your inspiration was given to you." You, the FSF and, apparently, RMS, feel it is about the "Four Freedoms" as defined by RMS. I'm quite sure that my view is much more common among the I agree to the "can't enjoy or test" bits. But I don't believe that it reduces anything. Personally I feel that anything that exposes people to "Free Yes. Because a number of your "facts" are massively flawed. Now, please, you've proven to me that you can't, in fact, do any *objective* thinking about this topic. When you are ready to drop your pre-conceptions and think *objectively* ...
Assuming he actually said that, I have no doubt that it would Of course not. That's what the spirit of the GPL is all about. And Exactly! And since the *Free* *Software* Foundation wrote the license, and documented the goals in the preamble, referring to keeping *free* *software* *free*, it is quite safe to say that this It's not. freedom and ability have two very different meanings. Freedom to run the software for any purpose means that people won't stop you from doing that. It may take you some work, such as porting the software, rebuilding it, etc. But if, at the end of that effort, you find that it will run on your development machine, but not in a machine where the original software runs on, and that's because the manufacturer imposed prohibitions on running unauthorized versions of the software, then the manufacturer of the hardware is very clearly disrespecting your freedom #0 WRT that software. Demanding the ability to run the software for any purpose, without any effort whatsoever, would indeed be nonsensical. (BTW, covered work is a legal term, only present in the legal portion of the license, which I'm actively avoiding, because I'm not a lawyer, and my point is about the spirit. but I'm sure I wrote that before Sure, but that's a different point. They could do that with or without tivoization. The point is that, if they have an issue with the program in the device, and they'd like to improve it, but they find that they won't be able to use their modification to get the device to do what they want, they're less likely to make the change. Now multiply this by all customers, and see how much you're losing by permitting tivoization, assuming that at least some tivoizers would change their minds towards respecting users' freedoms, if faced with You're entitled to have these motivations to release software under GPLv2, or any other license that you believe furthers these goals. But you have no say whatsoever on what intent RMS had ...
The "spirit" is no different than "intent". Different words that mean the The intent of the GPL, as seen by the FSF, *DOESN'T* *MATTER* *AT* *ALL* when the software isn't licensed by the FSF. Or did you forget that part of the This isn't what you've argued before. The hardware doesn't allow me to run the But a PPC binary won't run on an x86 either :) No, I'm in agreement with you here. But I'm smart enough to not buy something I use it because that is the term used in the GPLv2. And since the GPLv3 (dd4) Yes. But you are interpreting Linus' intentions using your own preconceived They complain to the manufacturer, file a report with a consumer watchdog Apply the same logic to my above statement and tell me - how much money does It means that I agree that the GPL is about "respecting and defending I never claimed I did. I was just pointing out that your belief that there is If the platform doesn't allow the running of modified binaries, why would the modifications matter? Sure, TiVO might like them - hell, they might even pay for them - but would anyone else? So modifications for that "closed execution" platform might suffer, but that I agreed with *portions* of some of the statements. However, did I actually I've been looking at this objectively the entire time. That you don't understand that is just more proof that you aren't. DRH -- Dialup is like pissing through a pipette. Slow and excruciatingly painful. -
You're mixing up spirit of license with intent of licensing (or something else, I don't remember exactly the term I invented to try to make this distinction clear) The former is the intent of the FSF for writing the GPL. The latter is the motivations of the copyright holder to license his work under that license. The FSF has no say on the latter. The copyright holders have no say on the former. When people claim the GPL changed its spirit, they're claiming the FSF changed its intent. It didn't. If it doesn't allow you to run the software *because* of a designed-in limitation on the freedom of the end-user, then it's a disrespect for the freedom. I've never changed my position in this regard. Maybe I wasn't clear, or you misunderstood, or the network corrupted the bits For this case. But it's not the case where I claimed it made a Dunno. Not much? I know I complain to hardware manufacturers that ship broken BIOSes, to no avail. It doesn't look like they care, or that it makes a difference. And then, I'm not talking about a case in which the thing is broken (in which case the user might have a real case) Think of improvements I'd like to make, that I probably won't do because the hardware won't let me run it. So you'll never see those contributions I and all the other untivoized users could make, and you If you say so, it must be right, in spite of all objective evidence, Which shows you don't know what the spirit really is. It is, and it has always been, what you agreed above that the GPL was about. What others think the spirit is doesn't affect what the spirit is. It just says what others think the spirit is, and how off the mark they are in their assessment of the spirit of the license (= intent behind I guess this should be pretty obvious that I believe this, yes. I don't know that I can make a general assertion about my believing what the FSF says, but I don't remember having had reasons to You're looking at the ...
Given that a number of people who don't buy into FSF ideology (let's call them "open source proponents" to contrast them with the "free software people") have concluded that the GPLv2 achieves their personal goals, and have chosen the GPLv2 as the license for their projects, I'd argue that the spirit that is embodied in the GPLv2 is actually a larger thing than what the FSF intended, and more inclusive. When these same people now disagree with the GPLv3, it indicates that something has been lost, and the spirit of the _license_ has changed. The _intention_ behind writing the license may or may not have been the same (who can tell, after 20-odd years?), but this is separate from the spirit embodied in the license itself - the latter has, in my mind anyway, clearly been changed. You might prefer to say "clarified", but it comes down to the same thing. But personally, I find the discussion about whether the spirit changed or not somewhat beside the point and not very interesting. What's really going to cause problems is the fact that the actual wording took a turn for the worse. Bernd -
This sounds like a good argument, but it doesn't hold water. Consider this: We manufacture bread toasters and sell them in the market with great success. They're big and bulky. So the engineers work on reducing its size, but in a way they can still fit perfectly a slice of bread. When we launch bread toaster, people complain that this new product cannot toast bagels any more, that we've changed the spirit of the bread toaster. See? Just because you could use it for other purposes doesn't make It just shows that they've never agreed with the spirit of the license in the first place. They just saw it could do something else, and used it for this reason. There's nothing wrong about this. What's wrong is to complain that those who introduced the license with a specific and public intent, and that advancing that intent with a new revision of the license, are changing the intent. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
You apparently do not understand "quid-pro-quo". Another way of stating it might be "same for same". A third way of stating it is "software for software". No, the romans never said that, but I just did, to make it just more obvious that the whole point is that you are expected to answer IN KIND! I do *not* ask for hardware access. I do *not* ask for money. And the reason I'm harping on "money" is that "money" is something *different* from what I give out. I give out software. I don't expect money in return. Money is *irrelevant*. It's allowed (and certainly much appreciated), but it's not required. See? Can you agree with that? Can you agree that that is actually part of what the whole "open source" spirit is all about (I'll avoid the word "free software", since you have defined it so rigorously personally that it makes no sense any more). Now, replace "money" with "access to the hardware", and read the exact *same* sentences again: And the reason I'm harping on "access to hardware" is that "access to hardware" is something *different* from what I give out. I give out software. I don't expect access to hardware in return. Access to hardware is *irrelevant*. It's allowed (and certainly much appreciated), but it's not required. See? Exact same words. Exact same spirit. Just using "access to hardware" instead of "money". You have been showing that you have a really hard time understanding that If you don't understand it after the above, I really can only say: "You are either terminally stupid, or you're not allowing yourself to see an obvious argument, because it destroys your world-view". The latter is very possible. It's a very human thing. It's why apparently a lot of people in the US have a hard time believing in evolution. Are they terminally stupid? Yeah, that is quite possible. But it is also possible that they are of average intelligence, and they just cannot mentally _afford_ to follow the argument - ...
Yes. And this was precisely what meant when I wrote "quid-pro-quo" /me hands Linus a mirror Serious, what's so hard to understand about: no tivoization => more users able to tinker their formerly-tivoized computers => more users make useful modifications => more contributions in kind ? Sure, there's a downside too: no tivoization => fewer contributions from manufacturers that demand on tivoization My perception is that the first easily dominates the second, and so Wrong. It enables copyright holders to decide whether forgiveness is appropriate, rather than forcing them to forgive. Being forced to forgive deception is not tit-for-tat, and it's a losing strategy. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
On 17/06/07, Alexandre Oliva <aoliva@redhat.com> wrote: I have to disagree. Let's say I'm the owner of a company selling some device that uses a GPLv2 OS and some GPLv2 applications to do the job. Let's say that for some reason I don't want the end users of my device to tinker with the software inside my device. Obviously I release the source for any modifications I may have made, but I use the hardware to prevent users from installing modified versions on the device (basically I TiVO'ize the device). Now I think you can agree to these things being positive: - My use of GPLv2 software in this device results in my employees being exposed to open source software at work (who knows, some may even start using such software at home as a result). A good thing. - The source code with my own modifications that I distribute as required by the GPLv2 can potentially be of use to other developers working on other GPLv2 software and those other developers are free to use those modifications. Also a good thing. - When creating marketing material for my device I'll most likely include information about the fact that I'm using WhatEverOS that is GPL'ed as well as other Open Source components. This in turn results in many people becoming aware that such software exist. I have to say that this is also a good thing. - When dealing with hardware companies supplying bits and pieces for my device I'll probably push for components that already have open source drivers, so my partners will find out there is value in having open source drivers for their stuff and hopefully end up supporting that. Yet another good thing. - If I end up being happy with my choice of GPLv2 OS & GPLv2 apps there's a, not insignificant, chance that I'll start helping out with the development of those components or maybe sponsor other developers with money to do so. Again we have a positive benefit. The only downside is that the end user purchasing the device can't install modified versions of the ...
BTW: Another reason a vendor might lock down the device is for security. For example, Juniper routers (which now run a significant portion of the "core" of the Internet) run FreeBSD on the routing engine. They include several GNU software utilities (for example gawk, diff, and gdb). Starting with JUNOS 7.6 (IIRC), end-users can no longer build and run their own binaries on the routing engine. This means that the GPLv2 code cannot be modified in-place (similar to TiVo altough done using different means). The reason is that if there ever is a security hole in the routing engine software (FreeBSD kernel, OpenSSH, etc.), it would be a really bad thing if crackers could load arbitrary software (rootkits, spam software, etc.) directly on Internet core routers. If you think spam zombies on cable modems or DSL are bad, imagine them on 100 megabit links! -
To be fair here, this could also be accomplished by having to flip a
physical switch on the router, especially if you did something funky
like:
[---] push this button for a 5 minute access pass to upload new
software through physical cable port 1.
More complex, but not unreasonable.
Bron.
-
Well, there is no restriction on putting files on the routing engine's storage devices (flash and hard drive); it is running OpenSSH, so scp/sftp work fine, and you can drop to a shell easily. The restriction is that the kernel won't run unsigned binaries. Also, flipping physical switches is pretty much an unreasonable expectation for core router operation. These are often in other locations, sometimes other telcos' central offices (where you have to pay to have "remote hands" do something and then hope they don't screw it up). You can easily go the entire life of a device where the primary operators never physically see the device. -- Chris Adams <cmadams@hiwaay.net> Systems and Network Administrator - HiWAAY Internet Services I don't speak for anybody but myself - that's enough trouble. -
Every server I run is like that, but if something is important enough I can remote control a robot over to push the button for me (actually, I think they implement this under the hood by having a human read the ticket I submit and go push the button for me manually, but that could be my imagination. So long as the button gets pushed the black box is functioning) Bron. -
Not sure if it's a good example, keep in mind that at the first exploitable software bug any hardware DRM breaks apart. But since you made a BSD-embedded example, this shows how the only really important thing is that by using linux instead of BSD, they can't make huge improvements or important security bugfixes to the routing engine, without us being able to incorporate them in our "home firewalls", that's the whole difference with BSD and it explains the spirit of the gpl pretty well and in the end why linux by definition can receive more contributions and in turn be technically superior. Whatever the vendor does with the gpl code is generally up to him, and if it uses the closed approach it'll allow somebody else to sell a "open" router (potentially at an higher price). Economy 101. The worry that nobody will step in and sell an "open" equivalent is a red herring. Infact I wouldn't be so certain that openmoko would exist if the current linux cellphones would be already totally open! Now I know this all probably sounds boring talk, but I think it's much closer to reality than the prospect of a trusted computing and/or DRM apocalypse. -
Your analysis stopped at the downside of prohibiting tivoization. You didn't analyze the potential upsides, so you may indeed come to different conclusions, and they may very well be wrong. It's very human to look only at the potential downside of an action and conclude it's a bad action. But it's more rational to look at the potential upside as well, evaluate the likelihood of each in the grand scheme of things, and then decide whether the potential upside will make up for the And therefore you severely limit the number of end users who might turn into contributors because of self interest in hacking the device False assumption. You can create the device using GPLv3 software in it. So your acccounting of necessary downsides is only one of the possibilities. The other possibility would be to have the program in ROM, of course, which would come with a completely different set of downsides, but that would retain all of the "these things being positive" you mentioned above. And, remember, since you merely don't *want* the end user of the device to tinker with the software, you have the option to do let them do that. And, if you do, they may find in themselves reasons and incentives to change the software in the device, and the improvements are likely to get back to the community and thus back to you. Everybody wins. This is the upside that you left out from your analysis, and from every other analysis that set out to "prove" that anti-tivoization is bad that I've seen so far. It appears that people are so concerned about whatever little they might lose from requiring respect for users' freedoms that they don't even consider what they might win, and that they *would* win if at least some of the vendors were to make an choice more favorable to their users and the community. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, ...
Maybe that's because I don't really see any up sides. As I see it, if we prevent tivoization, then the most likely outcome will be that a very few number of vendors will switch to ROM based solutions or similar (everyone lose, both vendor and user), a few vendors that currently tivoize hardware may open up their hardware but I doubt that will be very many, and the vast majority of vendors will move to *BSD or proprietary software since they simply can't or won't open up their hardware. So no, I don't think there are any upsides. We'll lose a huge number of developers, testers and users inside the business comunity and we'll lose a lot of exposure (like "hey, did you know TiVO actually runs Linux inside? Isn't that cool?)... Gaining a few hobyists at the expense of driving a huge number of businesses away from GPL'ed Just because I come to a different conclusion than you doesn't No, it is not. When I wrote that I meant "don't want" as in "really don't want to since it'll destroy our business" or "really really Most people don't care about hacking their devices, and of the few who do only a subset have the skill and only a subset of those will actually contribute anything back. This is a *small* set of people and gaining that small set at the expense of losing the large number of But do you really expect a vendor to put a device on the market where they also lock themselves out of upgrading it and releasing new For a few select individuals that may be true. But for the majority of Contrary to you, I don't believe any significant number of companies will do that. It's simply better for business to just use other software in that case. -- Jesper Juhl <jesper.juhl@gmail.com> Don't top-post http://www.catb.org/~esr/jargon/html/T/top-post.html Plain text mails only, please http://www.expita.com/nomime.html -
You just don't think they'd prevail over the downsides. *This* is an opinion I can respect, even if it's as much of a guestimate as mine. I'm sure both are highly influenced by personal opinions, wishful Agreed. I didn't say they were. I said they could be. Can you prove they're right? Do you even have any supporting evidence to back your guestimates? Heck, you may even have more than I do. I openly admit mine is mostly theoretical. I extrapolate the initial success of GNU+Linux on the PC environment, due in a large part to the ability for users to tinker with their computers, and expect it not to be so significantly different for other kinds of computers. For sure you'll get a far lower *percentage* of hackers in consumer devices than on PCs, whose users used to be far more technically-inclined and thus more propense to become hackers when GNU+Linux started than these days. But then I think of all of these computer users who helped make GNU+Linux what it is today, and other hackers that hadn't discovered this inclination before because they haven't had access to hackable computers. They could be tinkering with their DVRs, cell phones, wireless routers et al, and bringing the same kind of exciting community development to these kinds of computers. I'm saddened that the major Linux developers are willing to trade all of this (which I openly admit may be just a figment of my imagination, or just a tip of an iceberg) for some professional contributions (good) and some additional exposure that won't do justice to their software (bad), because these users will miss a big part of the picture by not being able to tinker with the software in the Last I looked, I was still human. So no. I try to use logic to reason out such behaviors when I realize they might be in action. But, as the saying goes, logic is a tool we use to justify our intutions. Or, logical reasoning is a tool to make the wrong Depends on how badly they want to use the GPLed ...
Ok, so we're on the same page. You're talking about something totally different. Answer my argument: - I think Tivoization is *good*. - Your license stops something *good*. Ergo: There is NOTHING TO FORGIVE! Your whole idiotic argument misses the point: What Tivo did is *good* in my opinion! Can't you get that through your skull? They gave the software back! Be happy! They *followed* the rules. They *followed* the tit-for-tat. Linus -
No, I'm not. You can say tivoization is *good* however much you like. This doesn't dispute in any way my claim that no tivoization would be *better*, that you'd get contributions from the people that, because of tivoization, don't feel compelled to develop and contribute, because they can't use the fruits of their efforts in the device where But they removed incentive for far more users to do so. So you get fewer contributions than you could without tivoization. "Can't you get that through your skull?" :-) -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Sorry, no we cannot. You seem to not accept that "in my opinion". That's not somethign we can disagree on. My opinion is _mine_ to state. There's no room for disagreement. It's my opinion, and your "agreement" is not optional. So when I say "I think what Tivo did is good", then you'd better just say, "ok, that's your opinion, and I respect you for it". Otherwise you're a douche-bag and an idiot. And once you realize that _I_ think that Tivo did is good, you have to accept that I think that the GPLv3 is the worse license in my opinion. So stop blathering about anything else. Just accept it. Just repeat after me: "Linus thinks that the GPLv3 is a bad license, and Linus is not confused". Don't call me "confused". Don't bother talking about what _you_ think, or what the FSF thinks is the "spirit". Don't say that you cannot understand it. Because if you cannot understand it, the only thing it shows is _your_ lack of understanding. Linus -
Sorry. I stand corrected. I didn't mean to disagree with the "in my opinion". I guess I was too distracted by the beauty of your kind words ;-) Sorry, it takes some effort to take focus away from that. I agree that this is your opinion, and you're entitled to it, no matter how much I disagree with this opinion. I disagree that what TiVo did is good. I think it is wrong on ethical and moral grounds, and I think it is bad for the user, for the community, and quite possibly even for TiVo itself. We can agree to disagree as to our opinions, if you want. -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
That's all I ever asked for. This whole thread started with me saying: I see the smiley, but I hate it how the FSF thinks others are morons and cannot read or think for themselves. Any time you disagree with the FSF, you "misunderstand" (insert condescending voice) the issue. _Please_ don't continue that idiocy. Disagreement and thinking that the FSF is controlling and putting its fingers where they don't belong is _not_ misunderstanding. It's just not "blind and unquestioning obedience". so all I asked for in the first place was that you stop claiming that I had "misunderstood" anything. That's really all I've always asked for: - I chose the GPLv2, and I understand it. - you don't have to agree with my choice, but you *do* have to accept it if you want to work on Linux. Because it's the only license that Linux has ever been released under since early 1992. So as long as you follow the GPLv2 (as a _legal_ license), I don't care if you like it or not. I don't care if you think you are a modern-day Napoleon, or if you are a demented squirrel. I don't care if you are an axe-murderer, or if you make sex toys with Linux. I don't care if your hardware is open or closed. I care about one thing, and one thing only: I care that you respect my choice of license for the projects _I_ started. Nothing more. And it doesn't matter one whit if *you* would have made a different choice. You are not me. You don't hold any power over me, and *your* choices are your own - not mine. Choice of license is personal. Many people think that the BSD license is better than _any_ version of the GPL. Are they wrong? No, it's _their_ choice. Is it relevant for the kernel? No, their preference of license is simply irrelevant. They can choose to accept the license that the kernel is under, or go play somewhere else. I think the GPLv2 is superior to the GPLv3. That is simply not something you can argue against. You can just say "ok, it's your ...
I do. Really. Once the issue about the spirit of the GPL is (hopefully) settled with all concerned about it, my job would have been done if it hadn't been Let me explain why I don't see that you've told me at length why you consider GPLv2 superior to GPLv3. 1. I asked you why GPLv2 is better, and you said it was because it promoted giving back in kind. 2. I asked you what you didn't like about GPLv3, and you said it was Tivoization. 3. Then I argued that, since Tivoization enables tivoizers to remove some motivation for potential developers (= their customers) to contribute, you trade the potential contributions of all those users for the contributions of tivoizers, apparently assuming that all tivoizers would simply move away from the community, taking their future contributions away from your community, rather than moving to a position in which you'd get not only the contributions from the company itself, but also from all their users. This last piece of the theorem that proves that GPLv2 is more aligned with your stated goals than GPLv3 is the one that is missing, and so far you've dodged that portion entirely. That's the 'connecting the dots' that I mentioned earlier. You haven't even acknowledged its existence, going back to points 1. and 2. as if they were enough, as if 3. didn't show a contradiction between them. Now, it may be that 3. is wrong, or that you think it is wrong. But you've never said so, or explained why you think so. You've simply disregarded that point entirely. Do you understand now why I feel you haven't answered the 'why'? -- Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/ FSF Latin America Board Member http://www.fsfla.org/ Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org} Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org} -
Right. The GPLv3 asks you to give back *money*. That's like the Microsoft license agreements. I don't like them either. Oh, and replace "money" with "access to hardware", to make that thing technically correct. But the point is, that's what I don't like about the That's simply not my *reason* for doing "tit-for-tat". My basic reason for "tit-for-tat" was not about "lots of potential developers", but simply because I think it's the right choice for me! Can you not understand that? I simply DO NOT LIKE TO CONTROL PEOPLE! I just want software back. I think it is *wrong* for me to ask for anything else. It's literally my personal "moral choice": I think the hardware manufacturers need to make their _own_ choices when it comes to _their_ designs. I feel that I have the moral right to ask for modifications to the kernel (because I started it), but I *personally* am very unhappy about asking people to also give their hardware access. That's *their* choice. Is that really so hard to understand? I ask you to respect _my_ choice wrt license for my software, but the same way I expect others to respect my choices, I also myself need to respect *their* choices. So to me, it's the hardware manufacturers choice to to select the license for their hardware, exactly the same way it was *my* choice to select it for my software. I believe in basically *one* freedom: the freedom to make our own choices! But if you actually want to discuss "number of developers" and their motications, I actually have another few arguments for you: - I just personally think your math is bogus. I think more people think like I do, than people think like you and the FSF does. But I don't even depend on that. Because: - I think that *technical*quality* is more important than *quantity*. And I think you have already proven a point: the GPLv3 seems to attract people who make the wrong *technica