Dual-Licensing Linux Kernel with GPL V2 and GPL V3
Quoted 15 lines · Dave Neuer
Quoted 6 lines · Alexandre Oliva
Seriously, looking only at the downside of anti-tivoization (tivoizer might turn us down), without even acknowledging that, should the tivoizer change practice and respect users' freedoms, you'd be able to get far more contributions from all those users, is typical minimax strategy.
Quoted 6 lines
Every time you enable someone to disrespect other users' freedoms WRT your software, you cut yourself out of some contributions that user could make. Even if you completely disregard the moral and ethical aspects of software freedom, the open source mentality inherently depends on the notion of respect for others' freedoms. You only reap the benefits of open source when the user gets the freedoms respected.
Quoted 2 lines
Alexandre, while I backed you up on the whole "spirit of the GPL hasn't changed" thing, I think you are wrong here.
Thank you, this is the first time someone voices a coherent argument that doesn't completely dismiss the potential benefits of prohibiting tivoization, and doesn't focus only on the potential downsides of such measures. Even though you come to a different conclusion than I did, I thank you very much for not pretending the point does not exist at all, or phrasing arguments that don't counter the point in any way.
Quoted 6 lines
we undoubtedly get more back in the form of software contributions from paid developers of TiVO-like companies than from the very few end users with the skill to hack the software or the inclination to aquire said skill. The simple fact is that most end users of most electronic devices don't care about the Freeness of the software, they care if their device works.
This is also true of computer users in general. So what is it that makes this case different?
Why should we ask vendors to pass on the corresponding sources to users, if most users don't care?
Why should we ask vendors to not use patents to stop a user from modifying and sharing the software they distribute, if most users don't care?
What is it that makes the reasoning different for the particular case of vendors using technical tricks to stop a user from enjoying the benefits of modifications she could make herself, which could then very well benefit the entire community and even the vendor?
Quoted 3 lines
If you disregard the ethical dimension, I think it's hard to argue w/ a straight face that Linus' stance is wrong from a pragmatic standpoint.
Maybe I'm wrong, but I hadn't seen arguments that even pointed in this direction, so the complete dismissal of the upsides came off as selective attention.
Let's see how this proceeds.
Quoted 6 lines
The problem is that the people saying "just don't buy TiVOs" know full well that because the number of end users who care is so small, they have _no_ economic power to change the situation, and that's why people who do consider this an ethical issue want to leverage the power the copyright holders have both legally and by virtue of their expertise to force the hardware vendors to cooperate w/ end users.
Yes, your reasoning makes sense to me. I can't speak for the FSF's intentions, but I have no reason to disbelieve it's moved by ethics, rather than by some plot to hurt TiVo.
Signature · 4 lines
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}Quoted 7 lines · Alexandre Oliva
Quoted 5 lines · Greg KH
From the very _beginning_ of the v3 process the kernel developers have showed their objection to that section of the license, and we were told, to our face, with no uncertian terms, that it was going to stay, in one form or another, no matter what we thought or said about it.
Quoted 7 lines
This sounds about right.
Quoted 2 lines
So, why would we want to waste our time filling out web forms after that?
If you're adamantly favorable to permitting any form of Tivoization whatsoever, don't bother.
For the record, I completely feel that what Tivo did was both legally correct[1], and the correct thing to do for their system, and would fight _very_ hard any attempt to change the Linux kernel's license that would prevent this usage model.
So I will not bother anymore.
greg k-h
[1] The FSF lawyers also agree with this.
Quoted 6 lines · Greg KH
Quoted 5 lines
Quoted 2 lines
So, why would we want to waste our time filling out web forms after that?
If you're adamantly favorable to permitting any form of Tivoization whatsoever, don't bother.
Quoted 6 lines
For the record, I completely feel that what Tivo did was both legally correct[1], and the correct thing to do for their system, and would fight _very_ hard any attempt to change the Linux kernel's license that would prevent this usage model.
So I will not bother anymore.
Linus has spoken, Greg K-H has spoken, many other people spoken --- and yet Alexandre keeps on speaking, and speaking, and speaking....
It's pretty clear no one is convincing anyone, and that everyone understands their position, and are happy with it, and so all we're doing now is wasting bandwidth.
Can we please end this thread?
Please?
- Ted
Quoted 4 lines · Greg KH
Quoted 3 lines · Alexandre Oliva
If you want your opinions to stand a chance to make a difference, the right place to provide them is gplv3.fsf.org/comments, and time is running short.
[...]
Quoted 2 lines
So, why would we want to waste our time filling out web forms after that?
In case anyone was wondering if the FSF is genuinely interested in feedback - I went and made some comments on the draft, and they appear to no longer be there a few days later. Thanks for the invitation Alex.
Bernd
Quoted 2 lines · Bernd Schmidt
I went and made some comments on the draft, and they appear to no longer be there a few days later.
This would be very bad. Please let me know what they were about and I'll try to figure out what happened.
Did you by any chance file them against an earlier draft? Those (for obvious reasons) no longer appear against the current draft, but they're still accessible by other means.
Signature · 4 lines
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}Quoted 3 lines · Alexandre Oliva
Can you please acknowledge that it doesn't, such that I can feel I've fulfilled my goal of dispelling the myth that the GPLv3 changes the spirit of the GPL?
No. I don't do metaphysics. This thread alone has shown that the notion is not well-defined *at* *all*, to the point of being useless and seriously misleading. I.e. the phrase about similar spirit should be replaced with something far more explicit and very, very hard to miss. I don't think you need more proof that people *do* interpret it in very different ways, with quite unpleasant results.
Quoted 5 lines
Quoted 2 lines
GPLv3, with your involvement in its development or not, sucks rocks, thanks to what you call anti-tivoization section.
Is it correct to say that you share Linus' opinion, that the only problem with the GPLv3 is the anti-tivoization provision?
No. If you want a basic splitup by sections compared to GPLv2,
1 - at least not better; attempts at being precise
end up creating a no-common-sense-land *and* turn out to leave serious unanswered questions in that area.
2 - no opinion on actual changes
3 - more or less an improvement
4,5 - about on par with v2, modulo wording in (5)
6 - much worse
7 - if I want to give additional permissions, I don't
want them stripped, for fsck sake! There is a bog-standard mechanism for _that_ (dual-licensing),
thank you very much. I.e. that section looks like a pile of dishonest PR games, pardon the redundance.
8 - on par
9 - on par, modulo piss-poor attempt to define "modify"
backfiring here (e.g. prelinking constitutes modification according to it, so does running rdev(8), etc., etc.)
10 - no opinion on actual changes
11 - improvement
12 - on par (aside of basic bad writing, but there are
much worse problems *not* with wording, so that's not interesting)
13 - special-case kludges are fun, aren't they (specifically
"linking"?), but in any case, that's secondary. FWIW, I'm not fond of ideas behind Affero, so if anything, that's a point against v3.
14 - ... and thank you very much for keeping such a lovely
source of periodic clusterfucks in v3 as well. I think it's painfully obvious for everyone in this thread that reference to "spirit" is a recipe for massive disagreements down the road. If you want the words you are using to be interpreted your way, use ones that have commonly agreed upon meaning. The measure is "do other people read it differently?", not "how sure I am in deriving the meaning I want from the words I've used?". Related problem is that version choice rules _must_ be stated in maximally unambiguous and hard to miss way. Look through Bernd-produced parts of this thread and you'll see the reason why it is needed.
Moving that into terms and conditions is a good step, but it's still not enough. E.g. you really want to be explicit on the form (in)sufficient to specify the version of license.
the rest on par.
Overall: definitely worse than v2. v2 + (3) + (11) would be an improvement, provided that v2 section 9 is cleaned up.
Quoted 4 lines
To make this more concrete, if there was a hypothetical GPLv2.9, consisting of GPLv3dd4 minus the "installation information" requirements for user products, (i) Would you consider it a better license than GPLv2?
Negative, see above
(ii) Better for Linux?
Negative, for kernel as well as for userland
(iii) Enough to go through the trouble of switching?
See above.
In other words, I don't see any chance for v3 to be a good choice for anything I write, kernel or userland. If I end up sending patches to v3 projects, I'll put the patches under BSDL and let them convert on merge.
Note that this is *not* about the problems with wording; those also exist, of course (_that_ is a final draft?), but that's a separate story and it interests me only inasmuch as it is caused by inherent problems with meaning of section in question.
Quoted 2 lines · Al Viro
I.e. the phrase about similar spirit should be replaced with something far more explicit and very, very hard to miss.
This is a very good idea. Would you please file it at
http://gplv3.fsf.org/comments?
Quoted 2 lines
I don't think you need more proof that people *do* interpret it in very different ways, with quite unpleasant results.
Agreed.
Quoted 4 lines
Quoted 2 lines
Is it correct to say that you share Linus' opinion, that the only problem with the GPLv3 is the anti-tivoization provision?
No.
Thanks for your detailed analysis. I wish I knew what to do with it. (I decline impolite suggestions, thanks ;-)
Would you like me to put you in touch with Richard Fontana, one of the lawyers involved in GPLv3, that's regarded as a legalese compiler, to discuss your issues about wording with him? Or would you rather file them (with a bit more detail) at gplv3.fsf.org/comments?
Aside from wording issues, which appear to dominate your comments, is it fair to characterize that your objections to GPLv3 are anti-tivoization provisions (strong) and Affero compatibility (weak?)?
I'm setting wording issues aside because these are easier to fix once the problem is understood, rather than ideological differences, that would probably be pointless to attempt to fix.
Quoted 4 lines
7 - if I want to give additional permissions, I don't
want them stripped, for fsck sake! There is a bog-standard mechanism for _that_ (dual-licensing), thank you very much.
additional permissions are indeed a form of dual-licensing, but one that doesn't require one to create a copy of the GPL and add the additional permissions to that copy. Yes, it could be accomplished with dual-licensing terms such as "you can follow the terms of the GPL, with the following additional permissions".
I don't quite see the point of criticizing this. This is more informative than anything else.
The meat here is really in the few additional restrictions, and the provisions to combat the practice of adding restrictions on top of the GPL and claiming the software is available under the GPL, which has made for a lot of confusion over time.
Thanks a lot for your feedback.
Signature · 4 lines
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 message that is not in the archive (sent off-list or not yet received): [email protected]
Quoted 6 lines · Linus Torvalds
Quoted 3 lines · Alexandre Oliva
They're based on the Free Software definition, that establishes the four freedoms that the GPL was designed to respect and defend.
The GPL is a software license, *independent* of that thing.
One more time, I'm not talking about the license (the legal terms). I'm talking about its spirit. It's encoded in the preamble, that refers to "free software", which can't possibly be defended as meaning anything but what the FSF itself defined in the Free Software Definition.
Is this some form of mental block that stops you from realizing that the spirit of the license is pretty much all I've been talking about here, and that I've already said that at least 20 times in this thread?
Why do you insist in bringing the legal terms back into this discussion about the spirit of the license?
What are you trying to accomplish, other than generating more confusion and pretending that you have a strong point?
Signature · 4 lines
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}Quoted 2 lines · Alexandre Oliva
One more time, I'm not talking about the license (the legal terms).
Ok. Then go away.
Everybody else just cares about the legal reasons.
The "legal terms" is the only reason a license *exists*. That's what a license *is*, for crying out loud!
If you don't care about the legal side, go and read the free software manifesto. That's the paper you're really arguing about.
If you want to argue about the GPLv2 *license*, then you'd better start caring about the legal issues. Because that is what the license is: a _legal_ document.
Linus
Quoted 7 lines · Linus Torvalds
Quoted 2 lines · Alexandre Oliva
One more time, I'm not talking about the license (the legal terms).
Ok. Then go away.
Everybody else just cares about the legal reasons.
That's false, and the reason I know it is that, if this was true, I probably wouldn't even have got into this debate.
The reason I got into it was that there were false claims about changes in the spirit of the license, implied in Greg KH's e-mail that refers to the long thread and position paper that made these false claims. I got into this debate in response to this reference:
http://www.ussg.iu.edu/hypermail/linux/kernel/0706.1/1904.html
to a thread about an article that stated
[...] solemn trust, as stated in article 9 of GPLv2, only to licence the code under versions of the GPL that "... will be similar in spirit to the present version". We, like all the individual contributors to GNU projects, have taken that trust at face value and accorded the FSF a special role in the Open Source Universe because of it. It goes without saying that any updates to GPLv2 must be completely in accord with the execution of that trust.
before it went on to misleading references to a DRM clause (*), trying to frame anti-tivoization as end-use restrictions and how unacceptable that is in a license, while blessing tivoization itself (an end-use restriction, no less).
(*) meaning the anti-tivoization provisions, in spite of the existence of an unrelated section named DRM in a draft available at that time
And then, the article goes on about FUD of jeopardising patent portfolios, which is *obviously* a misunderstanding, as I wrote in my e-mail.
In response to that intervention, my first one in this subject, you exploded. And then, shortly after your explosion and all the name calling, there was this:
http://www.ussg.iu.edu/hypermail/linux/kernel/0706.1/2235.html
And you say people don't care about the spirit?
Seriously, I'm not even asking for an apology for the name calling. All I'm asking for is that you don't lie about GPLv3 and about the spirit of the GPL.
I know it's hard when you're trying to frame yourself as a victim of a tyranny, but since FSF's offer of a new license, and switching its own projects to it, framing it as "suborn or coerce others to go along with them" (from the article) is utterly false, and an evident misunderstanding of the intent.
And that's not even unusual. See, when Free Software advocates *ask* people to refer to the GNU operating system as such, even when it's combined with the kernel Linux, a lot of people over-react and complain that the FSF is trying to *force* anyone to rename Linux. This is false on two accounts. Linux is the name of the kernel, and the FSF has never tried to rename that. GNU is the name of the operating system, and it's people who confusingly call that operating system Linux that renamed it. Free Software advocates merely ask that people call it by the right name.
I can appreciate that some FSFers take positions that may come off as unfriendly and trigger this kind of negative feeling of being forced to do something. I suspect this is the same case.
But taking it as forcing is just as inappropriate. There's no coercive force to be applied, and no moral grounds to coerce. In both cases, it's a request for people to make the choice of helping defend users' freedoms.
I know you're not going to believe this, because it doesn't fit in your world-view.
My intent in participating in this discussion was really not to convince you. I had a feeling, all the way from the beginning, that this was probably hopeless, even though there was thin hope.
My intent was to enlighten others who were still open to listening and to accepting different viewpoints. From personal feedback I got, I know I've accomplished a lot of that, and I thank deeply all of you who encouraged me to proceed, and who thanked me for the information I added to the debate.
I wish I could say I'm going to step out of this debate now, because it's taken me far more time than I could afford. But since my goal was to counter the spreading of mis-information about GPLv3, and since I am indeed addicted to this kind of discussion, I may end up failing to stay away from the discussion.
Signature · 4 lines
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: I'm writting this while offline and likely to remain so for the next 8 hours or so, so I'll probably miss a bunch of other replies]
Quoted 8 lines · Alexandre Oliva
Quoted 6 lines · Bron Gondwana
Quoted 5 lines · Alexandre Oliva
Quoted 2 lines · Bron Gondwana
because it could easily be argued that they linked the BIOS with the Linux kernel
Quoted 7 lines
Quoted 4 lines
Quoted 1 line
How so?
Er, they installed it in the same piece of equipment, and the kernel couldn't function without it in that work.
I see what you're getting at. You're thinking of a license that doesn't respect the idea of "mere aggregation", right?
No, I'm arguing that it's not "mere aggregation" - the kernel is useless on that machine unless the BIOS is present or replaced with something else with equivalent functionality. I suspect any decent lawyer could make the theory that this made the kernel as compiled on to that machine with specific chipset support selected for that hardware into a "derived work" of the BIOS - especially if the vendor had contributed GPLed code for drivers which interact with their hardware into said kernel.
In fact, particularly if the hardware vendor has also contributed GPL code that interacts on one side of the software/(firmware, hardware) boundard which worked around bugs in said firmware/hardware which they also had the ability to change. The two really are a combined work of which only one part is GPLed.
Ringing any binary kernel module video card driver bells yet? It's really the same thing from the opposite direction - the only criteria is where you fit in the pecking order - hardware manufacturers work around Windows bugs, Linux kernel drivers work around hardware bugs - it's all about who has more to lose if they aren't compatible.
Quoted 3 lines
For starters, this wouldn't evidently not qualify as an Open Source license, and I'm pretty sure it wouldn't qualify as a Free Software license either.
Strawman licence?
Quoted 8 lines
Quoted 3 lines
By using GPLix as part of their boot process along with their non-GPL BIOS, they're subverting the freedoms that the user should have in being able to control the entire boot process.
You're pushing the "freedom to change" too far. Sure, I'd like to be able to do that, and I prefer hardware that lets me do it, but it's not like this BIOS in the scenario you described is being used as a means to stop me from modifying the GPLed software.
Well, yeah - except this is the direction GPL3 takes us, and it's a theory that GPL3 makes more likely to fly in court than GPL2 does - meaning that hardware vendor lawyers lie awake at night worrying about stuff (I'd hate to be a good lawyer - I'd never get any sleep!)
Quoted 4 lines
I have never said that including a GPLed piece of software should grant users the right to modify anything whatsoever in the system, or grant them control over the entire system. Others have, but it's not true, it just shows how much mis-information is floating around.
No, but your interactions with Linus (lazy bums 'r' us) have shown that the logical result of what you do want includes this. It's a lot harder to objectively judge one of these than the other:
a) have they provided the source code to this binary to anyone who asks.
b) have any of the limitations of this piece of hardware been created with the intent of making it more difficult for J. Random Enduser to build modified binaries from said source and have them function correctly.
(b) has much more scope for shenanigans by bad apples on the copyright owner side - and don't pretend that only the hardware vendors are bad guys - it takes all sorts and the idea of a licence is to protect both parties.
Quoted 4 lines
All the GPL stands for is to defend the freedom of the users over the particular program it applies to. You can't impose further restrictions on the user's ability to modify what *that* software does.
Except where they run into limitations of the platform itself, or just plain bugs. Oops. The lawyers will have a field day discovering intent every time J. Random's kernel doesn't do what he wants after he fiddles the code a bit.
Quoted 11 lines
If you wanted to change something else, but this something else is not covered by the license, and is not being used to contradict the terms of the license, well, too bad, you lose.
Quoted 6 lines
b) deny themselves the ability to every offer a patch to said BIOS if bugs are found
Point (b) is also exactly on topic for the discussion of enforcing legal safety obligations in hardware on a peripheral rather than the software drivers.
Quoted 6 lines
Quoted 2 lines
It's requiring that these limitations be placed in a technically inferior location to hack around a legal "bug".
I don't think this last sentence is true. If you implement hardware locks that prevent modification of the software even by yourself, then you're in compliance with the terms of the GPLv3dd4. But IANAL.
I obviously wasn't clear enough. The only way to come into complience with GPL3dd4 is to reduce your ability to fix things or grant everyone else the ability to mess with things. This severely restricts you from doing _anything_ in certain problem spaces due to local laws on the topic, even if you're an otherwise good actor who is making worthwhile source code contributions to the rest of the community.
This would be a lot less of an issue if Linux was a modular kernel (don't shoot me Linus) and you could be allowed to change the bits that didn't touch the regulated hardware's access paths. Messy to control if you're running in ring0 though - you need hardware managed restrictions at some level, and a barrier around the entire kernel is certainly the easiest way to do that.
Bron.
Quoted 4 lines · Bron Gondwana
No, I'm arguing that it's not "mere aggregation" - the kernel is useless on that machine unless the BIOS is present or replaced with something else with equivalent functionality.
That's *not* a valid argument!
I know, I know, it's a common one, but it is *nonsense*.
The thing is, "mere aggregation" doesn't mean what you think it means.
"Mere aggregation" doesn't mean that they cannot depend on each other. It means that they are not *based* on each other in the sense of GPLv2.
In other words, "mere aggregation" is about two pieces that are not derivative works under copyright law.
For example, on a Red Hat DVD, *every*single*binary* on that DVD requires a kernel to run on. And the kernel image itself "depends" on the user programs to actually do something _useful_.
But it's all still "mere aggregation", because they are not related to each other in the sense of being derived works!
So "mere aggregation" is not about intimacy. OF COURSE high-tech products depend intimately on each other. The Linux kernel cannot boot on a PC without a BIOS or something equivalent. You cannot run your graphical environment without a kernel, an X server, the CPU, the memory, the display, the BIOS, the power company (or an equivalent hand-crank) etc etc etc, and these things are all very much dependent on each other to make a "usable system", that has absolutely _zero_ relevance to whether they are "mere aggregation" or not.
So the only thing the "mere aggregation" phrase in the GPLv2 means is
simply: "putting together two or more pieces that are not derived works of each other". That's what that
mere aggregation of another work not based on the Program ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
part of the license is all about. The BIOS is "another work", and it clearly is "not based on the Program". The "aggregation" just means "putting them together", and the "mere" is there just to show that as far as the license is concerned, that is not even interesting!
In other words, the sentence in section 2 about mere aggregation says one thing only: the fact that you lump things together storage-wise does not matter *at*all* for the license. The _only_ thing that matters for the GPLv2 is whether something is a derived work, _not_ whether it is lumped together, and depends on something else.
This is why the BIOS is irrelevant. You can put it in the same flash chip as the kernel: it still isn't a derived work of Linux (and Linux is not a derived work of the BIOS), and the GPLv2 explicitly makes it clear that the license holds no sway over the BIOS, even though it's physically on the same chip.
The BIOS can do whatever it wants to, and the GPLv2 has *nothing* to say about it. The GPLv2 makes that very clear indeed. It is just "mere aggregation", and the fact that they are in the same machine, on the same harddisk, on the same flash rom, or on the same DVD doesn't bring them any closer from a *copyright* angle.
Copyright in general (and the GPLv2 in the specific) isn't about how things are *physically* tied together. It's not even about how things interconnect, and how "important" they are for each other. It is _purely_ about whether they are derived works.
So when the GPLv2 talks about "mere aggregation", it talks about it specifically to say that it does not matter, and that the license _only_ affects the actual derived work!
This is why the BIOS, the hardware, the programs you run under Linux etc etc are all totally irrelevant. The GPLv2 explicitly says that they are irrelevant as far as that license is concerned.
Linus
Quoted 13 lines · Linus Torvalds
Quoted 5 lines · Bron Gondwana
No, I'm arguing that it's not "mere aggregation" - the kernel is useless on that machine unless the BIOS is present or replaced with something else with equivalent functionality.
That's *not* a valid argument!
Linus,
Just take a vote and start tagging files and ignore this needless diatribe. It is was it is, I seriously doubt you will
get all of Linux moved to GPL3 as a monolith, since you will never get concensus. You should fork a GPL3 kernel and let people decide whether their code goes in or not. If they don't want to move to it, new people can contribute new code. Start a 2.8 tree or whatever that is GPL3 only.
Jeff
Quoted 9 lines · Jeffrey V. Merkey
Linus,
Just take a vote and start tagging files and ignore this needless diatribe. It is was it is, I seriously doubt you will
get all of Linux moved to GPL3 as a monolith, since you will never get concensus. You should fork a GPL3 kernel and let people decide whether their code goes in or not. If they don't want to move to it, new people can contribute new code. Start a 2.8 tree or whatever that is GPL3 only.
*snicker*
Yeah, Linus - when do you finally understand that nobody WANTS your GPL3 kernel? :))
ROFL
Signature · 1 line
Carlo Wood <[email protected]>
Quoted 11 lines · Linus Torvalds
Quoted 4 lines · Bron Gondwana
No, I'm arguing that it's not "mere aggregation" - the kernel is useless on that machine unless the BIOS is present or replaced with something else with equivalent functionality.
That's *not* a valid argument!
But it's all still "mere aggregation", because they are not related to each other in the sense of being derived works!
Quoted 8 lines
So "mere aggregation" is not about intimacy. OF COURSE high-tech products depend intimately on each other. The Linux kernel cannot boot on a PC without a BIOS or something equivalent. You cannot run your graphical environment without a kernel, an X server, the CPU, the memory, the display, the BIOS, the power company (or an equivalent hand-crank) etc etc etc, and these things are all very much dependent on each other to make a "usable system", that has absolutely _zero_ relevance to whether they are "mere aggregation" or not.
Ok, can I please rewrite my argument to:
"The hardware manufacturer has built a custom BIOS and also written Linux kernel support for said BIOS. They have released the kernel drivers under GPL as required, but have not released the code to the BIOS, instead just releasing the interface documentation. The BIOS didn't exist before, and as they only intend to run Linux on the device, the BIOS design was heavily influenced around working well with Linux."
Actually, we don't know that last bit, maybe they created the BIOS in a total vacuum and then wrote the Linux kernel driver later. Maybe not.
Anyway, I think I've wound up arguing two sides of the same argument, oops.
Bron.
Quoted 11 lines
Ok, can I please rewrite my argument to:
"The hardware manufacturer has built a custom BIOS and also written Linux kernel support for said BIOS. They have released the kernel drivers under GPL as required, but have not released the code to the BIOS, instead just releasing the interface documentation. The BIOS didn't exist before, and as they only intend to run Linux on the device, the BIOS design was heavily influenced around working well with Linux."
Quoted 8 lines
Actually, we don't know that last bit, maybe they created the BIOS in a total vacuum and then wrote the Linux kernel driver later. Maybe not.
Anyway, I think I've wound up arguing two sides of the same argument, oops.
Bron.
Why does it matter whether the BIOS was "heavily influenced around working well" with Linux? Your argument is arguing about things that have nothing to do with anything. Your comparing things before you have any idea what the right criteria for the comparison is.
Do you understand that all that matters is whether the BIOS contains significant portions of the Linux code base? *Nothing* else matters. Everything else will leave it an independent work and one that the authors of Linux have nor right to claim any ownership of or control over.
I maintain a private application that has huge amounts of code that are heavily influenced around working well with Linux. All the epoll code, for example, in this code base meets that criteria. That doesn't mean it *resembles* the Linux code in any way. It doesn't mean the Linux folks have any right to tell me what I can and can't do with that code.
So you're arguing two sides of no argument at all.
DS
Quoted 1 line · David Schwartz
So you're arguing two sides of no argument at all.
Yeah, pretty much. I take back my arguments in the previous couple of my posts up this thread. They don't actually hold together! Sorry for wasting your time correct me.
Bron.
Quoted 11 lines · Linus Torvalds
Quoted 4 lines · Bron Gondwana
No, I'm arguing that it's not "mere aggregation" - the kernel is useless on that machine unless the BIOS is present or replaced with something else with equivalent functionality.
That's *not* a valid argument!
I know, I know, it's a common one, but it is *nonsense*.
Further to my other response on this (yeah, I know, I should think first).
Where the BIOS author and 'work aggregator' are different organisation with no shady backroom links (other than the usual industry cabal(TINC) of course) then it's clear.
When they are the same organisation then the derivedness state is a lot less clear and more "discoverable", leading to a higher risk of ambush by litigation.
This isn't specific to any particular licence, but it's something that the "intent" theory of hardware limitations being a GPL3 violation makes extra dangerous, because that clause can be used as a hook to drag a claim through summary judgement (IMHO, IANAL, etc)
Bron ( mostly arguing the same things that you are Linus, I think, but
I didn't clarify that I was writing from a devil's advocate position in an alternative reality where Tivo was illegal purely due to "intent" )
Quoted 6 lines · Bron Gondwana
I obviously wasn't clear enough. The only way to come into complience with GPL3dd4 is to reduce your ability to fix things or grant everyone else the ability to mess with things. This severely restricts you from doing _anything_ in certain problem spaces due to local laws on the topic, even if you're an otherwise good actor who is making worthwhile source code contributions to the rest of the community.
I don't know any law that requires tivoization.
There may be laws that require certification or limitations on the user. Manufacturer giving up the ability to make modifications would address this, or *perhaps* arranging for user and manufacturer to each hold half of the key needed to run a modification (which might comply with the GPLv3dd4 terms, IANAL).
There may be business models that require the ability to make changes. Then it's fair to enable the user to make changes as well, such that they don't become dependent on the vendor, or even have their 1st-generation TiVo boxes left out in the cold for a while when the US changes the DST rules again ;-)
Signature · 4 lines
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}Quoted 8 lines · On Saturday 16 June 2007 21:54:56 Alexandre Oliva wrote:
Quoted 6 lines · Bron Gondwana
I obviously wasn't clear enough. The only way to come into complience with GPL3dd4 is to reduce your ability to fix things or grant everyone else the ability to mess with things. This severely restricts you from doing _anything_ in certain problem spaces due to local laws on the topic, even if you're an otherwise good actor who is making worthwhile source code contributions to the rest of the community.
Quoted 7 lines
I don't know any law that requires tivoization.
There may be laws that require certification or limitations on the user. Manufacturer giving up the ability to make modifications would address this, or *perhaps* arranging for user and manufacturer to each hold half of the key needed to run a modification (which might comply with the GPLv3dd4 terms, IANAL).
It doesn't. The GPLv3 (dd4) makes that very clear. See the quote below.
"Installation Information" for a User Product means any methods, procedures, authorization keys, or other information required to install and execute modified versions of a covered work in that User Product from a modified version of its Corresponding Source. The information must suffice to ensure that the continued functioning of the modified object code is in no case prevented or interfered with solely because modification has been made.
DRH
Quoted 5 lines
There may be business models that require the ability to make changes. Then it's fair to enable the user to make changes as well, such that they don't become dependent on the vendor, or even have their 1st-generation TiVo boxes left out in the cold for a while when the US changes the DST rules again ;-)
Signature · 1 line
Dialup is like pissing through a pipette. Slow and excruciatingly painful.
Quoted 9 lines · Daniel Hazelton
Quoted 5 lines · On Saturday 16 June 2007 21:54:56 Alexandre Oliva wrote:
There may be laws that require certification or limitations on the user. Manufacturer giving up the ability to make modifications would address this, or *perhaps* arranging for user and manufacturer to each hold half of the key needed to run a modification (which might comply with the GPLv3dd4 terms, IANAL).
It doesn't. The GPLv3 (dd4) makes that very clear. See the quote below.
You left out the relevant bit:
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).
Signature · 4 lines
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}Quoted 8 lines · On Sunday 17 June 2007 00:19:49 Alexandre Oliva wrote:
Quoted 7 lines · Daniel Hazelton
Quoted 5 lines · On Saturday 16 June 2007 21:54:56 Alexandre Oliva wrote:
There may be laws that require certification or limitations on the user. Manufacturer giving up the ability to make modifications would address this, or *perhaps* arranging for user and manufacturer to each hold half of the key needed to run a modification (which might comply with the GPLv3dd4 terms, IANAL).
Quoted 7 lines
Quoted 1 line
It doesn't. The GPLv3 (dd4) makes that very clear. See the quote below.
You left out the relevant bit:
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).
Ah, but giving the user half the key doesn't mean they still don't have access to the entire key. QED: Giving people half the key won't cut it under the GPLv3 (dd4)
DRH
Signature · 1 line
Dialup is like pissing through a pipette. Slow and excruciatingly painful.
Quoted 13 lines · Daniel Hazelton
Quoted 12 lines · On Sunday 17 June 2007 00:19:49 Alexandre Oliva wrote:
Quoted 8 lines · Daniel Hazelton
Quoted 5 lines · On Saturday 16 June 2007 21:54:56 Alexandre Oliva wrote:
There may be laws that require certification or limitations on the user. Manufacturer giving up the ability to make modifications would address this, or *perhaps* arranging for user and manufacturer to each hold half of the key needed to run a modification (which might comply with the GPLv3dd4 terms, IANAL).
It doesn't. The GPLv3 (dd4) makes that very clear. See the quote below.
You left out the relevant bit:
Quoted 7 lines
Quoted 3 lines
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).
Ah, but giving the user half the key doesn't mean they still don't have access to the entire key. QED: Giving people half the key won't cut it under the GPLv3 (dd4)
I meant really giving, rather than giving a copy, or giving the original and keeping a copy.
You could make it require a pair of signatures, one from the vendor, that the vendor keeps, one from the user, that the vendor never sees, too. Like some bank PINs, it gets generated, used to generate some hash (the signature for the initial installation), printed in an envelope for you and stored in the package along with the machine. Or something like that.
Signature · 4 lines
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}Quoted 9 lines · Alexandre Oliva
Quoted 3 lines · Daniel Hazelton
Ah, but giving the user half the key doesn't mean they still don't have access to the entire key. QED: Giving people half the key won't cut it under the GPLv3 (dd4)
I meant really giving, rather than giving a copy, or giving the original and keeping a copy.
Quoted 6 lines
You could make it require a pair of signatures, one from the vendor, that the vendor keeps, one from the user, that the vendor never sees, too. Like some bank PINs, it gets generated, used to generate some hash (the signature for the initial installation), printed in an envelope for you and stored in the package along with the machine. Or something like that.
Wow, and I thought losing a microsoft "certificate of authenticity" and associated key was a pain. Ouch. Talk about your paper "dongle".
Quoted 1 line
I don't know any law that requires tivoization.
In the USSA it is arguable that wireless might need it (if done in software) for certain properties. (The argument being it must be tamperproof to random end consumers). Obviously an electronics graduate can tamper with hardware ones just as well
The USSA also has interesting rules about some other technologies including GPS on national security grounds.
Alan
Quoted 5 lines · Alan Cox
Quoted 1 line
I don't know any law that requires tivoization.
In the USSA it is arguable that wireless might need it (if done in software) for certain properties. (The argument being it must be tamperproof to random end consumers).
But this is not tivoization.
Tivoization is a manufacturer using technical measures to prevent the user from tampering (*) with the device, *while* keeping the ability to tamper with it changes itself.
(*) tampering brings in negative connotations that I'd rather avoid, but since that was the term you used, and the term "modifying" might bring in legal-based technicalities such as that replacing isn't modification, I just went with it.
So, given a proper definition, do you know any law that requires tivoization?
Taking it further, do you know whether any such law requires *worldwide* tivoization, as in, applying the restrictions in the law even outside its own jurisdiction?
Signature · 4 lines
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}Quoted 9 lines · On Sun, 17 Jun 2007 15:33:33 -0300 Alexandre Oliva <[email protected]> wrote:
Quoted 5 lines · Alan Cox
Quoted 1 line
I don't know any law that requires tivoization.
In the USSA it is arguable that wireless might need it (if done in software) for certain properties. (The argument being it must be tamperproof to random end consumers).
But this is not tivoization.
It most definitely is
Quoted 3 lines
Tivoization is a manufacturer using technical measures to prevent the user from tampering (*) with the device, *while* keeping the ability to tamper with it changes itself.
That accurately describes the FCC wireless rules. As a vendor I am perfectly allowed to tinker, but the product must not be "end user" modifiable. (and some vendors take this to mean binary only, despite the fact end users generally aren't able to modify code and EE students are just as able to modify electronics as we are code).
Alan
Quoted 13 lines · Alan Cox
Quoted 9 lines · On Sun, 17 Jun 2007 15:33:33 -0300 Alexandre Oliva <[email protected]> wrote:
Quoted 5 lines · Alan Cox
Quoted 1 line
I don't know any law that requires tivoization.
In the USSA it is arguable that wireless might need it (if done in software) for certain properties. (The argument being it must be tamperproof to random end consumers).
But this is not tivoization.
Quoted 7 lines
It most definitely is
Quoted 3 lines
Tivoization is a manufacturer using technical measures to prevent the user from tampering (*) with the device, *while* keeping the ability to tamper with it changes itself.
That accurately describes the FCC wireless rules.
AFAIK the FCC mandates not permitting the user to tinker. It doesn't mandate the vendor to retain this ability to itself.
Therefore, per the above, FCC doesn't mandate tivoization.
Is there anything else I'm missing that would show it does?
Signature · 4 lines
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}Quoted 4 lines
Quoted 1 line
That accurately describes the FCC wireless rules.
AFAIK the FCC mandates not permitting the user to tinker. It doesn't mandate the vendor to retain this ability to itself.
In practical terms it does since a recall/replacement in the event of rule changes is a bit impractical
Quoted 1 line
Therefore, per the above, FCC doesn't mandate tivoization.
I'm sure you can find a definition to sort your goals whatever.
mea culpa on the meaning of Tivoization (was: Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3)
Quoted 7 lines · Alan Cox
Quoted 4 lines
Quoted 1 line
That accurately describes the FCC wireless rules.
AFAIK the FCC mandates not permitting the user to tinker. It doesn't mandate the vendor to retain this ability to itself.
In practical terms it does since a recall/replacement in the event of rule changes is a bit impractical
Indeed. But that's not a legal requirement, it's an economic reason.
"But I need to make a profit" or "But I need to reduce costs" is no excluse to disrespect the GPL.
Quoted 3 lines
Quoted 1 line
Therefore, per the above, FCC doesn't mandate tivoization.
I'm sure you can find a definition to sort your goals whatever.
Are you per chance implying that I'm twisting the definition of tivoization?
You know... I now believe that would be correct. I have indeed twisted the definition of tivoization, and I'm sorry about that. Which is not to say that I agree that the FCC or any other law mandates tivoization, or that tivozation is a good thing or that it is permitted by GPLv2. Please read on.
After long conversations with RMS about the section on poisoned apples and tivoization in my draft article about GPLv3 (Corresponding Sources is the name of the section in
http://fsfla.org/svnwiki/blogs/lxo/draft/gplv3-snowwhite) I had come to the conclusion that Tivoization amounted to:
denying the user of the computer the freedom to run modified versions of the Free Software in it, while retain this ability to oneself.
This understanding of mine had been strengthened by my understanding of the wording and the rationale of GPLv3dd3, the wording about technical restrictions in the rationales published along with GPLv2dd2, and the various speeches in which the term was presented.
Nevertheless, I consulted with him and others highly involved in the development of GPLv3 about some of the discussions going on here, and got responses over the past few hours that surprised me. A lot.
So I've just went back to that discussion about my article, and to various other cases in which RMS, Eben Moglen and others presented Tivoization, the rationales, and so on, and I came to the conclusion that I had experienced a subtle but very significant misunderstanding.
I'm now convinced that a more appropriate definition would be:
denying the user of the computer the freedom to run modified versions of the Free Software in it, by not sharing information as to how it could be accomplished.
This difference is very significant, and even more so for this discusion, because it contradicts some of what I claimed before about forms to use GPLed software where regulations require the user to be unable to modify the software.
Let me start with an example: I bought a wireless router some time ago, and it had a GNU+Linux distribution installed in it. No source code or written offer for source code, though.
Now, if I called the vendor next day and asked for the source code, and they responded "sorry, I can't give you that. I threw it all away, such that I wouldn't be able to give it to you.", they would still be disrespecting my freedoms, as well as the license, right?
You see where I'm going? Now, if they gave me the source code, but I still couldn't install modified versions, because they introduced technical measures with the purpose of denying me this possibility, then the inability to modify the program wouldn't be caused by something like a physical impossibility (something like ROM), but rather by an active measure to trample my ability to adapt the program and run it for any purpose.
So, if I called them to ask how to install and run modified versions of the GPLed programs, and they responded "sorry, I can't give you that. I threw it all away, such that I wouldn't be able to give it to you.", they would still be disrespecting my freedoms, as well as the license.
The reasons as to why they'd want to disrespect the freedoms don't matter. It could be "making a profit", "complying with the law", "abiding by contractual restrictions", anything. Imposing restrictions to the exercise of the freedoms is not in line with the spirit of the GPL, as such restrictions render the Software non-Free.
The conclusion? Throwing keys away, or using split-key techniques, as I have suggested as potential alternatives to ROM for compliance with GPLv3 are not meant to be permitted by GPLv3. There might be practical advantages to compromising and permitting these techniques, but that would amount to endorsing disrespect for users' freedoms, and more, betraying those who licensed their works under GPLv1+ or v2+ with an intent to not permit these practices. I don't think FSF is willing to be part of this, and this is how it should be.
As for those who didn't mean the GPL this way, they can always grant additional permissions, or simply refrain from enforcing the license in these cases.
I apologize for my terrible misunderstanding, and for spreading it.
Hopefully this message will reach everyone I have misled.
I've tried to Cc: everyone who'd received copies of my mistaken claims directly from me. If I left you out by accident, please holler ;-)
Signature · 4 lines
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}Re: mea culpa on the meaning of Tivoization (was: Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3)
Quoted 4 lines
Quoted 2 lines
In practical terms it does since a recall/replacement in the event of rule changes is a bit impractical
Indeed. But that's not a legal requirement, it's an economic reason.
Cynical Economists would argue 'legal requirements' are just changes to the cost of the various economic options. Sometimes when I look at Microsoft's approach to various cases it seems they think that way too.
Quoted 3 lines
denying the user of the computer the freedom to run modified versions of the Free Software in it, by not sharing information as to how it could be accomplished.
That's a nice definition but one I can see being sort of abusable depending how you read it. We head ever more into the disposable computer era where as a vendor putting the code on ROM is cheap and upgrades don't matter (throw it away get a new one).
I can tell you how to upgrade it ("you can't") yet I as the manufacturer can issue new units with modified code so I still control it even though it is meant to be "free"
Re: mea culpa on the meaning of Tivoization
Quoted 8 lines · Alan Cox
Quoted 4 lines
Quoted 2 lines
In practical terms it does since a recall/replacement in the event of rule changes is a bit impractical
Indeed. But that's not a legal requirement, it's an economic reason.
Cynical Economists would argue 'legal requirements' are just changes to the cost of the various economic options. Sometimes when I look at Microsoft's approach to various cases it seems they think that way too.
I guess this depends to some point on the kind of penalties you face. If they're only economic, then yes. If you may end up going to jail or some such, I think the picture gets different. But yes, that's a way to see it, and I know we're not alone in perceiving some behaviors that way.
Quoted 3 lines
I can tell you how to upgrade it ("you can't") yet I as the manufacturer can issue new units with modified code so I still control it even though it is meant to be "free"
The GPL has never prohibited the distribution of software in ROM, just like it's never prohibited the fixation of software in CD-ROMs. So explicitly permitting is not a step back in terms of defending freedoms, even if there might be something to do that would advance freedoms in this field.
Anyhow, AFAIK software in ROM is not non-Free Software. That it's impossible to modify/replace/whathaveyou it is not the result of a restriction that someone is imposing on you.
It's the difference between "you can't fly because you don't have wings" and "you won't fly because I've tied your wings". With tied wings, you're evidently not free to fly any more. But if the problem is that you don't have wings, if you're free and sufficiently creative, you may be able to invent baloons, airplanes, rockets et al and overcome the barriers that nature poses for you.
Signature · 4 lines
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}Re: mea culpa on the meaning of Tivoization
Am Montag 18 Juni 2007 02:56 schrieb Alexandre Oliva:
Quoted 5 lines
Anyhow, AFAIK software in ROM is not non-Free Software. That it's impossible to modify/replace/whathaveyou it is not the result of a restriction that someone is imposing on you.
Quoted 7 lines
It's the difference between "you can't fly because you don't have wings" and "you won't fly because I've tied your wings". With tied wings, you're evidently not free to fly any more. But if the problem is that you don't have wings, if you're free and sufficiently creative, you may be able to invent baloons, airplanes, rockets et al and overcome the barriers that nature poses for you.
So, if a manufacturer used a ROM instead of a flash memory with the intention to make software modifications impossible, then it is bad, and when he did it for economical reasons, then it is a "natural barrier"?
Your tied-up-wings comparison is simply not valid. Or, more precisely, you will usually not be able to tell whether you don't have wings or if they're tied. Hardware design decisions of a manufacturer should never be the subject of a software license.
Hans
Re: mea culpa on the meaning of Tivoization
Quoted 3 lines · Hans-Jürgen Koch
So, if a manufacturer used a ROM instead of a flash memory with the intention to make software modifications impossible, then it is bad, and when he did it for economical reasons, then it is a "natural barrier"?
This sounds about right to me.
Intent is very significant, but then, what vendor would justify the choice of ROM as "intent to prevent modifications", if this amounted to copyright infringement?
Vendor would be entitled to the benefit of the doubt as to the motivations in this case, so it would likely be unenforceable anyway.
Signature · 4 lines
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}Re: mea culpa on the meaning of Tivoization
Am Montag 18 Juni 2007 20:55 schrieb Alexandre Oliva:
Quoted 11 lines · Hans-Jürgen Koch
Quoted 4 lines
So, if a manufacturer used a ROM instead of a flash memory with the intention to make software modifications impossible, then it is bad, and when he did it for economical reasons, then it is a "natural barrier"?
This sounds about right to me.
Intent is very significant, but then, what vendor would justify the choice of ROM as "intent to prevent modifications", if this amounted to copyright infringement?
Indeed. That nicely shows how useless any licensing discussion is when it comes to hardware design issues, including "Tivoization".
Quoted 4 lines
Vendor would be entitled to the benefit of the doubt as to the motivations in this case, so it would likely be unenforceable anyway.
Right. If GPL v3 comes out, there'll probably be a new task for hardware development engineers: How to find excuses for hardware that prevents software modifications and how to conceal the true intent.
Hans
Re: mea culpa on the meaning of Tivoization
Quoted 6 lines · Hans-Jürgen Koch
Quoted 2 lines
Vendor would be entitled to the benefit of the doubt as to the motivations in this case, so it would likely be unenforceable anyway.
Right. If GPL v3 comes out, there'll probably be a new task for hardware development engineers: How to find excuses for hardware that prevents software modifications and how to conceal the true intent.
Yup. And then GPLv4 will have to plug whatever holes they find to disrespect users' freedoms. That's how I expect the game to be played.
Signature · 4 lines
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}Re: mea culpa on the meaning of Tivoization
Am Montag 18 Juni 2007 23:18 schrieb Alexandre Oliva:
Quoted 5 lines · Hans-Jürgen Koch
Quoted 3 lines
Quoted 3 lines
Vendor would be entitled to the benefit of the doubt as to the motivations in this case, so it would likely be unenforceable anyway.
Quoted 8 lines
Quoted 3 lines
Right. If GPL v3 comes out, there'll probably be a new task for hardware development engineers: How to find excuses for hardware that prevents software modifications and how to conceal the true intent.
Yup. And then GPLv4 will have to plug whatever holes they find to disrespect users' freedoms. That's how I expect the game to be played.
If you were right and it turned out that way, the whole GPL would become so ridiculous that it won't have any of its intended effects. GPLv17 would contain so many #ifdefs that it'll simply not be readable or understandable anymore.
As far as the kernel is concerned, I expect the game's played by simply keeping GPLv2. And I like it that way.
Hans
Re: mea culpa on the meaning of Tivoization
Quoted 10 lines · Hans-Jürgen Koch
Am Montag 18 Juni 2007 23:18 schrieb Alexandre Oliva:
Quoted 9 lines · Hans-Jürgen Koch
Quoted 7 lines
Quoted 3 lines
Vendor would be entitled to the benefit of the doubt as to the motivations in this case, so it would likely be unenforceable anyway.
Right. If GPL v3 comes out, there'll probably be a new task for hardware development engineers: How to find excuses for hardware that prevents software modifications and how to conceal the true intent.
Quoted 6 lines
Quoted 3 lines
Yup. And then GPLv4 will have to plug whatever holes they find to disrespect users' freedoms. That's how I expect the game to be played.
If you were right and it turned out that way, the whole GPL would become so ridiculous that it won't have any of its intended effects.
How so? The intended effects are to protect users' freedoms, by requiring them to be respected. If we keep on plugging holes as they appear, it will keep close to achieving its intended effects. It's earlier versions of the license that will get more and more distant from it.
Quoted 2 lines
As far as the kernel is concerned, I expect the game's played by simply keeping GPLv2. And I like it that way.
Just think about it... What if, today, some law passed, or some court decision came up, that rendered a significant defense provision of GPLv2 or GPLv3 ineffective?
GPLv4 could plug that, and anyone using GPLvN+ would be able to switch to it immediately. This wouldn't revoke previous licenses, of course, but further developments could be made under the newer license, and at least those could still be defended, and, as time elapsed, earlier versions of the software would become less and less relevant, to the point that the holes in their license also become less and less relevant, until copyright finally expires and they enter the public domain.
The distrust for the FSF led to this very short-sighted decision of painting the Linux community into a corner from which it is very unlikely to be able to ever leave, no matter how badly it turns out to be needed. Let's just hope it never is, or that some influx of long-sighted comes in and introduces mechanisms for the license of Linux to be patched, should this ever be needed. I'm not even talking about GPLv2+, there are many other ways to accomplish this, that I've already mentioned in another posting in another recent huge thread.
Signature · 4 lines
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}Re: mea culpa on the meaning of Tivoization
Am Dienstag 19 Juni 2007 04:46 schrieb Alexandre Oliva:
Quoted 16 lines · Hans-Jürgen Koch
Quoted 14 lines
Am Montag 18 Juni 2007 23:18 schrieb Alexandre Oliva:
Quoted 12 lines · Hans-Jürgen Koch
Quoted 7 lines
Quoted 3 lines
Vendor would be entitled to the benefit of the doubt as to the motivations in this case, so it would likely be unenforceable anyway.
Right. If GPL v3 comes out, there'll probably be a new task for hardware development engineers: How to find excuses for hardware that prevents software modifications and how to conceal the true intent.
Yup. And then GPLv4 will have to plug whatever holes they find to disrespect users' freedoms. That's how I expect the game to be played.
Quoted 6 lines
Quoted 2 lines
If you were right and it turned out that way, the whole GPL would become so ridiculous that it won't have any of its intended effects.
How so? The intended effects are to protect users' freedoms, by requiring them to be respected. If we keep on plugging holes as they appear, it will keep close to achieving its intended effects.
No. Credible licenses should be simple like physical laws. Newton's law is expressed in terms of a single mathematical equation. That's why it's still valid, and you still learn it at school although meanwhile people know that there are limitations to it.
If you come up with a new version of a license every year, you will only weaken it. Please note that quantum mechanics is _not_ such a hole-plugging addition to Newton's law. It's a new simple physical law, expressed in terms of a single simple mathematical equation that contains the old law as a border case. If that were not the case, it would have never been accepted.
If you want a GPLv3, please make it simple and make it contain GPLv2 as a border case. The current draft isn't like that.
Quoted 3 lines
It's
earlier versions of the license that will get more and more distant from it.
No. GPLv2 is a simple set of rights and restrictions that's easy to understand and therefore accepted by many courts all over the world. I cannot see any danger for the code I put under it, at least none that would be mitigated by GPLv3.
Quoted 7 lines
Quoted 3 lines
As far as the kernel is concerned, I expect the game's played by simply keeping GPLv2. And I like it that way.
Just think about it... What if, today, some law passed, or some court decision came up, that rendered a significant defense provision of GPLv2 or GPLv3 ineffective?
The best way to prevent that is to make the license simple and easy to follow. If many important open source software developers have problems with the wording of a license, lawyers and judges will have them, too.
Quoted 9 lines
GPLv4 could plug that, and anyone using GPLvN+ would be able to switch to it immediately. This wouldn't revoke previous licenses, of course, but further developments could be made under the newer license, and at least those could still be defended, and, as time elapsed, earlier versions of the software would become less and less relevant, to the point that the holes in their license also become less and less relevant, until copyright finally expires and they enter the public domain.
Ah, now I get the point. Yes, that's one in favor of a complicated and confusing license. You can lengthen court proceedings until copyright expires...
Quoted 6 lines
The distrust for the FSF led to this very short-sighted decision of painting the Linux community into a corner from which it is very unlikely to be able to ever leave, no matter how badly it turns out to be needed.
I'm neither in a corner nor do I feel the need for a different license. I've got some code in the kernel, and I've got it under GPLv2, and I'm happy with it. It's the FSF that thinks I should see myself in a corner.
Quoted 2 lines
Let's just hope it never is, or that some influx of long-sighted comes in
Kernel programmers are short-sighted? What kind of arrogance is that?
Quoted 2 lines
and introduces mechanisms for the license of
Linux to be patched, should this ever be needed.
You know pretty well that Linus clearly said he would change the license when _he_ thinks it's needed. The point is that _you_ want him to change the license to support _your_ political ideas.
Quoted 4 lines
I'm not even talking
about GPLv2+, there are many other ways to accomplish this, that I've already mentioned in another posting in another recent huge thread.
I partly read that "recent huge thread". Linus elaborated his point of view in detail, and I very much share his opinion.
Hans
Re: mea culpa on the meaning of Tivoization
Quoted 10 lines · Hans-Jürgen Koch
No. Credible licenses should be simple like physical laws. Newton's law is expressed in terms of a single mathematical equation. That's why it's still valid, and you still learn it at school although meanwhile people know that there are limitations to it.
If you come up with a new version of a license every year, you will only weaken it. Please note that quantum mechanics is _not_ such a hole-plugging addition to Newton's law. It's a new simple physical law, expressed in terms of a single simple mathematical equation that contains the old law as a border case. If that were not the case, it would have never been accepted.
This is an excellent example of how engineers tend to mis-analyze legal issues. In law, neither simple wording nor interpretation is so simple or so mechanical as the things engineers prefer to work with.
Take an example: "Thou shalt not kill". Very clear, but also very problematic in that it does not address military conquests (which were apparently approved by that law's drafter), self-defense, or a number of other cases.
There are always grey areas between what is explicitly addressed and what is not. Courts interpret laws and precedent in ways that make life (and license or contract writing) more unpredictable. Wishing otherwise will not make a simple license unambiguous. If those areas of ambiguity are exploited enough, the perceived cost of having a hole will exceed the perceived cost of plugging it.
Michael Poole
Re: mea culpa on the meaning of Tivoization
Am Dienstag 19 Juni 2007 13:50 schrieb Michael Poole:
Quoted 13 lines
Quoted 10 lines · Hans-Jürgen Koch
No. Credible licenses should be simple like physical laws. Newton's law is expressed in terms of a single mathematical equation. That's why it's still valid, and you still learn it at school although meanwhile people know that there are limitations to it.
If you come up with a new version of a license every year, you will only weaken it. Please note that quantum mechanics is _not_ such a hole-plugging addition to Newton's law. It's a new simple physical law, expressed in terms of a single simple mathematical equation that contains the old law as a border case. If that were not the case, it would have never been accepted.
Quoted 3 lines
This is an excellent example of how engineers tend to mis-analyze legal issues. In law, neither simple wording nor interpretation is so simple or so mechanical as the things engineers prefer to work with.
Don't make fast assumptions about how _I_ think. I intentionally chose examples from science, not from engineering. There are grey areas in Newton's Law and in quantum mechanics as well.
Quoted 5 lines
Take an example: "Thou shalt not kill". Very clear, but also very problematic in that it does not address military conquests (which were apparently approved by that law's drafter), self-defense, or a number of other cases.
Pretty clear. I accepted that about thirty years ago.
Quoted 3 lines
There are always grey areas between what is explicitly addressed and what is not.
Yes, you named it: _always_ !! We have to accept these grey areas. Tivoization (to come back to our real topic) is one of those. Alexandre Oliva sees grey areas as holes he needs to plug. That's wrong, and that's what I'm criticizing. You cannot get rid of grey areas by modifying licenses. Maybe you can temporarily shift the grey areas to some other place. But at the same time, you get negative side effects because your license becomes more and more complicated and confusing. I consider this worse than some grey areas.
Quoted 5 lines
Courts interpret laws and precedent in ways that make
life (and license or contract writing) more unpredictable. Wishing otherwise will not make a simple license unambiguous. If those areas of ambiguity are exploited enough, the perceived cost of having a hole will exceed the perceived cost of plugging it.
I completely disagree. A very important factor for the success of a license is that people are familiar with it. Only then can they start obeying it. Have you ever been out there in industry, trying to tell programmers what Linux, Free Software, and the GPL is? You'd notice that they hardly know what the GPLv2 is all about. And now, after I explained it to them, should I tell them that there's a much more complicated license about to come, and that it's going to be changed whenever some FSF people find a new hole?
All this hole-plugging just leads to a license that is so far away from being useful in industrial reality that nobody will voluntarily obey it. This Tivoization stuff is a good example. I was a hardware developer for more than a decade. When I developed _hardware_, I made my design decisions without having to read the licenses of the _software_ I want to run on that hardware. And if you go today and tell a hardware developer that there are some people in the world who want to achieve exactly that, he will probably laugh at you and think you're joking.
Hans
Re: mea culpa on the meaning of Tivoization
In reply to Hans-Jürgen KochQuoted 7 lines · Hans-Jürgen Koch
Am Dienstag 19 Juni 2007 04:46 schrieb Alexandre Oliva:
Quoted 4 lines
The distrust for the FSF led to this very short-sighted decision of painting the Linux community into a corner from which it is very unlikely to be able to ever leave, no matter how badly it turns out to be needed.
I'm neither in a corner nor do I feel the need for a different license.
Yes. Some day you may. And then what will you or anyone else be able to do about it?
Quoted 4 lines
Quoted 2 lines
Let's just hope it never is, or that some influx of long-sighted comes in
Kernel programmers are short-sighted? What kind of arrogance is that?
It's just stating the obvious. The upgrade path is a nightmare.
A long-sighted decision should have established *some* means for a quick fix should it be needed. It didn't have to be GPLv2+. In fact, per the stated goals and general feelings, it probably *shouldn't* be GPLv2+. But cutting any reasonable possibility of fixing a legal problem in the license is short-sighted, yes. It's putting too much trust in the perfection of the license *and* the worldwide legal systems *and* legislators.
Quoted 5 lines
Quoted 2 lines
and introduces mechanisms for the license of
Linux to be patched, should this ever be needed.You know pretty well that Linus clearly said he would change the license when _he_ thinks it's needed.
And what makes you think even *he* can change the license?
Quoted 2 lines
The point is that _you_ want him to change the license to support _your_ political ideas.
I would like him to, yes. But this is besides the point. That I see reasons for an upgrade, and that I'd like such an upgrade, doesn't make any difference whatsoever about the plain fact that relicensing Linux today, to any other license and for whatever reason it was, would be a nightmare, and that this is a consequence of the short-sighted decision of not establishing a relicensing procedure.
At this point, the situation is very much like a kernel installed in ROM. Who knows that nobody will ever find security bugs in it? How would you go about fixing them?
Quoted 6 lines
Quoted 3 lines
I'm not even talking
about GPLv2+, there are many other ways to accomplish this, that I've already mentioned in another posting in another recent huge thread.I partly read that "recent huge thread". Linus elaborated his point of view in detail, and I very much share his opinion.
Huh? It looks like you're talking about something unrelated with license patching procedures. I don't think Linus ever responded in that thread to my suggestions of various means to establish a license patching procedure.
Signature · 4 lines
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}Re: mea culpa on the meaning of Tivoization
Am Dienstag 19 Juni 2007 20:32 schrieb Alexandre Oliva:
Quoted 8 lines · Hans-Jürgen Koch
Quoted 6 lines
Am Dienstag 19 Juni 2007 04:46 schrieb Alexandre Oliva:
Quoted 4 lines
The distrust for the FSF led to this very short-sighted decision of painting the Linux community into a corner from which it is very unlikely to be able to ever leave, no matter how badly it turns out to be needed.
Quoted 4 lines
Quoted 1 line
I'm neither in a corner nor do I feel the need for a different license.
Yes. Some day you may. And then what will you or anyone else be able to do about it?
A lot of awful things could happen to you or me _tomorrow_. Are you prepared for everything? I'm not. If I'd try to be prepared for every possible disaster, you'd rightly call me mentally ill.
Even if my code is used on a Tivo-like device, I can sleep well and don't need a different license. These are the _normal_ grey areas that _every_ license or law contains.
Quoted 7 lines
Quoted 5 lines
Quoted 3 lines
Let's just hope it never is, or that some influx of long-sighted comes in
Kernel programmers are short-sighted? What kind of arrogance is that?
It's just stating the obvious. The upgrade path is a nightmare.
Well, maybe. Maybe this is a topic that needs further discussion. But I don't find it very important as we're not in a situation where we urgently need a new license.
[...]
Hans
Re: mea culpa on the meaning of Tivoization
Quoted 10 lines · Hans-Jürgen Koch
Quoted 6 lines
Quoted 4 lines
Quoted 2 lines
Let's just hope it never is, or that some influx of long-sighted comes in
Kernel programmers are short-sighted? What kind of arrogance is that?
It's just stating the obvious. The upgrade path is a nightmare.
Well, maybe. Maybe this is a topic that needs further discussion. But I don't find it very important as we're not in a situation where we urgently need a new license.
Agreed. It could have been discussed years ago, when the clarification on GPLv2-only came up, but it's still not urgent, and hopefully it never will be. It wouldn't be a bad idea to think about it, though.
Just in case it's not clear, this is in no way related with GPLv3. It's just that GPLv3 discussions appear to get more people thinking about relicensing, and then the "impossibilities" of doing it come up.
Signature · 4 lines
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}Re: mea culpa on the meaning of Tivoization (was: Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3)
In reply to Alan CoxQuoted 4 lines · On Sunday 17 June 2007 19:50:41 Alan Cox wrote:
That's a nice definition but one I can see being sort of abusable depending how you read it. We head ever more into the disposable computer era where as a vendor putting the code on ROM is cheap and upgrades don't matter (throw it away get a new one).
I'm waiting for cereal boxes in the store to have a display covering the entire front side which changes every 30 seconds, with all the circuitry to drive it taking up a few square milimeters and a small watch battery to power it for six months ala the blinky LEDs of today.
This should be economicaly feasible in what, 10 years? 15? (There are all sorts of weird problems to solve like coming up with a battery that's not only cheap enough but which you won't get in trouble putting millions of in landfills. Fuel cell might be more environmentally friendly depending on your catalysts...)
We haven't even brushed against the "disposable computer era" compared to what's coming. Most of us will probably live to see happy meal toys capable of running Linux, and that's just a _start_.
Rob
Signature · 2 lines
"One of my most productive days was throwing away 1000 lines of code." - Ken Thompson.
Re: mea culpa on the meaning of Tivoization (was: Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3)
In reply to Alexandre OlivaQuoted 66 lines · On Sunday 17 June 2007 19:11:13 Alexandre Oliva wrote:
Quoted 7 lines · Alan Cox
Quoted 4 lines
Quoted 1 line
That accurately describes the FCC wireless rules.
AFAIK the FCC mandates not permitting the user to tinker. It doesn't mandate the vendor to retain this ability to itself.
In practical terms it does since a recall/replacement in the event of rule changes is a bit impractical
Indeed. But that's not a legal requirement, it's an economic reason.
"But I need to make a profit" or "But I need to reduce costs" is no excluse to disrespect the GPL.
Quoted 3 lines
Quoted 1 line
Therefore, per the above, FCC doesn't mandate tivoization.
I'm sure you can find a definition to sort your goals whatever.
Are you per chance implying that I'm twisting the definition of tivoization?
You know... I now believe that would be correct. I have indeed twisted the definition of tivoization, and I'm sorry about that. Which is not to say that I agree that the FCC or any other law mandates tivoization, or that tivozation is a good thing or that it is permitted by GPLv2. Please read on.
After long conversations with RMS about the section on poisoned apples and tivoization in my draft article about GPLv3 (Corresponding Sources is the name of the section in
http://fsfla.org/svnwiki/blogs/lxo/draft/gplv3-snowwhite) I had come to the conclusion that Tivoization amounted to:denying the user of the computer the freedom to run modified versions of the Free Software in it, while retain this ability to oneself.
This understanding of mine had been strengthened by my understanding of the wording and the rationale of GPLv3dd3, the wording about technical restrictions in the rationales published along with GPLv2dd2, and the various speeches in which the term was presented.
Nevertheless, I consulted with him and others highly involved in the development of GPLv3 about some of the discussions going on here, and got responses over the past few hours that surprised me. A lot.
So I've just went back to that discussion about my article, and to various other cases in which RMS, Eben Moglen and others presented Tivoization, the rationales, and so on, and I came to the conclusion that I had experienced a subtle but very significant misunderstanding.
I'm now convinced that a more appropriate definition would be:
denying the user of the computer the freedom to run modified versions of the Free Software in it, by not sharing information as to how it could be accomplished.
This difference is very significant, and even more so for this discusion, because it contradicts some of what I claimed before about forms to use GPLed software where regulations require the user to be unable to modify the software.
Quoted 3 lines
Let me start with an example: I bought a wireless router some time ago, and it had a GNU+Linux distribution installed in it. No source code or written offer for source code, though.
Just want to point out that, when I read this, my reaction was "But... That is a direct violation of the GPLv2. No specific reading of the license needed."
Quoted 4 lines
Now, if I called the vendor next day and asked for the source code, and they responded "sorry, I can't give you that. I threw it all away, such that I wouldn't be able to give it to you.", they would still be disrespecting my freedoms, as well as the license, right?
Yes, they would. They are distributing a modification - in the words of the GPLv2 "a work based on the work" - without complying with the terms of the license.
Quoted 8 lines
You see where I'm going? Now, if they gave me the source code, but I still couldn't install modified versions, because they introduced technical measures with the purpose of denying me this possibility, then the inability to modify the program wouldn't be caused by something like a physical impossibility (something like ROM), but rather by an active measure to trample my ability to adapt the program and run it for any purpose.
Quoted 5 lines
So, if I called them to ask how to install and run modified versions of the GPLed programs, and they responded "sorry, I can't give you that. I threw it all away, such that I wouldn't be able to give it to you.", they would still be disrespecting my freedoms, as well as the license.
Not even the GPLv3dd4 - because they don't have the information anymore either. If, however, they still retained the information - in any form - they would be violating the GPLv3dd4.
The GPLv2 doesn't make the actions described above - "how to install and run" - a license violation.
Quoted 5 lines
The reasons as to why they'd want to disrespect the freedoms don't matter. It could be "making a profit", "complying with the law", "abiding by contractual restrictions", anything. Imposing restrictions to the exercise of the freedoms is not in line with the spirit of the GPL, as such restrictions render the Software non-Free.
Then anyone using GPLv3'd software to drive WiFi devices, radio (HAM radio) networks, etc - in the US, at least - isn't allowed to do such. US Law makes some provisions of the GPLv3 illegal to comply with. Thanks to section 6 of the GPLv3 that invalidates the rights granted under the license.
Quoted 9 lines
The conclusion? Throwing keys away, or using split-key techniques, as I have suggested as potential alternatives to ROM for compliance with GPLv3 are not meant to be permitted by GPLv3. There might be practical advantages to compromising and permitting these techniques, but that would amount to endorsing disrespect for users' freedoms, and more, betraying those who licensed their works under GPLv1+ or v2+ with an intent to not permit these practices. I don't think FSF is willing to be part of this, and this is how it should be.
Umm... making things more strict will just do more damage. As it is there are restrictions on companies that make them unable to comply with the GPL. What the GPLv3 has done is take away options they might otherwise have had. If one of the goals of the FSF is to force proprietary software into a minority then its just done damage to that goal.
DRH
Quoted 11 lines
As for those who didn't mean the GPL this way, they can always grant additional permissions, or simply refrain from enforcing the license in these cases.
I apologize for my terrible misunderstanding, and for spreading it.
Hopefully this message will reach everyone I have misled.
I've tried to Cc: everyone who'd received copies of my mistaken claims directly from me. If I left you out by accident, please holler ;-)
Signature · 1 line
Dialup is like pissing through a pipette. Slow and excruciatingly painful.
Re: mea culpa on the meaning of Tivoization
Quoted 9 lines · Daniel Hazelton
Quoted 3 lines · On Sunday 17 June 2007 19:11:13 Alexandre Oliva wrote:
Let me start with an example: I bought a wireless router some time ago, and it had a GNU+Linux distribution installed in it. No source code or written offer for source code, though.
Just want to point out that, when I read this, my reaction was "But... That is a direct violation of the GPLv2. No specific reading of the license needed."
Yes. Anyone feels like enforcing the GPLv2 in Brazil? I can even recommend lawyers that speak English reasonably well and are somewhat familiar with the GPL, and I've already tracked the distribution chain back to the initial infringer. Harald is aware of the issue, but AFAIK he's decided not to pursue that yet.
Quoted 6 lines
Quoted 4 lines
Now, if I called the vendor next day and asked for the source code, and they responded "sorry, I can't give you that. I threw it all away, such that I wouldn't be able to give it to you.", they would still be disrespecting my freedoms, as well as the license, right?
Yes, they would. They are distributing a modification
There's no reason to assume it's a modification. They're distributing a copy, and that's enough.
Quoted 9 lines
Quoted 5 lines
So, if I called them to ask how to install and run modified versions of the GPLed programs, and they responded "sorry, I can't give you that. I threw it all away, such that I wouldn't be able to give it to you.", they would still be disrespecting my freedoms, as well as the license.
Not even the GPLv3dd4 - because they don't have the information anymore either. If, however, they still retained the information - in any form - they would be violating the GPLv3dd4.
I'm told by the authors of GPLv3dd4 that this case is not meant to be permitted. I suppose they're going to change the wording, or at least the rationale for it.
Quoted 2 lines
The GPLv2 doesn't make the actions described above - "how to install and run" - a license violation.
This is true. They didn't have any such duty, under the GPLv2.
However, if I figured that out by myself, but found that I was unable to run a modified version because something in there checks for a hash computed over the program I'd like to modify, and refuses to run it because of the hash, then the hash is effectively part of the program, and they haven't provided me with the corresponding sources of that portion of the program.
I know you don't want that to be true, and a court might actually decide your way some day. But until then, your claim that this is permitted by the GPL is just as good as mine that it's not. And I really mean "just as good", since my claim is in line with the stated purpose of the authors of the GPL, and yours is in line with their opinion (according to others, I don't think I've got this straight from them) as to whether the license effectively prohibits this practice.
Quoted 4 lines
Then anyone using GPLv3'd software to drive WiFi devices, radio (HAM radio) networks, etc - in the US, at least - isn't allowed to do such. US Law makes some provisions of the GPLv3 illegal to comply with. Thanks to section 6 of the GPLv3 that invalidates the rights granted under the license.
Actually, this is false. Not only because of the ROM provisions in the GPLv3, but because the law requirements aren't anywhere as strict as the WiFi vendors who want to disrespect your freedoms want you to believe.
Quoted 2 lines
What the GPLv3 has done is take away options they might otherwise have had.
It doesn't. Authors can always grant these options separately if they want to. Authors can always choose GPLv2 if they want to.
GPLv3 is an option for those who want to defend freedoms, even if they don't share the perception that this is a moral and ethical issue. If they're in it only for the self benefits, that's fine, GPLv3 can get them that, even better than GPLv2 could, in spite of the short-sighted claims to the contrary.
Quoted 2 lines
If one of the goals of the FSF is to force proprietary software into a minority then its just done damage to that goal.
That's not the goal.
Signature · 4 lines
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}Re: mea culpa on the meaning of Tivoization
Quoted 10 lines · On Monday 18 June 2007 15:09:47 Alexandre Oliva wrote:
Quoted 8 lines · Daniel Hazelton
Quoted 3 lines · On Sunday 17 June 2007 19:11:13 Alexandre Oliva wrote:
Let me start with an example: I bought a wireless router some time ago, and it had a GNU+Linux distribution installed in it. No source code or written offer for source code, though.
Just want to point out that, when I read this, my reaction was "But... That is a direct violation of the GPLv2. No specific reading of the license needed."
Quoted 5 lines
Yes. Anyone feels like enforcing the GPLv2 in Brazil? I can even recommend lawyers that speak English reasonably well and are somewhat familiar with the GPL, and I've already tracked the distribution chain back to the initial infringer. Harald is aware of the issue, but AFAIK he's decided not to pursue that yet.
I don't know if I have the right. None of the code is mine - the fact that they are in violation of the license is not in question (I trust your word on this), but it is the licensor who has the right to press charges. (I will check with the lawyers and law professionals I know, because the GPL makes no statements about the legal jurisdiction under which violations will be tried. It might be that I actually can file suit under Brazillian law)
Quoted 9 lines
Quoted 6 lines
Quoted 4 lines
Now, if I called the vendor next day and asked for the source code, and they responded "sorry, I can't give you that. I threw it all away, such that I wouldn't be able to give it to you.", they would still be disrespecting my freedoms, as well as the license, right?
Yes, they would. They are distributing a modification
There's no reason to assume it's a modification. They're distributing a copy, and that's enough.
Bad word choice on my part. Of course you are correct.
Quoted 6 lines
Quoted 6 lines
Quoted 5 lines
So, if I called them to ask how to install and run modified versions of the GPLed programs, and they responded "sorry, I can't give you that. I threw it all away, such that I wouldn't be able to give it to you.", they would still be disrespecting my freedoms, as well as the license.
Quoted 7 lines
Quoted 3 lines
Not even the GPLv3dd4 - because they don't have the information anymore either. If, however, they still retained the information - in any form - they would be violating the GPLv3dd4.
I'm told by the authors of GPLv3dd4 that this case is not meant to be permitted. I suppose they're going to change the wording, or at least the rationale for it.
Okay. So its possible to change whats running on the hardware - but even though nobody has the information needed to do it, it's a violation. Hrm... I can see some valid reasoning behind this, but it'll take creative legalese to make sure that things like (EE)PROMS are properly covered.
Quoted 11 lines
Quoted 2 lines
The GPLv2 doesn't make the actions described above - "how to install and run" - a license violation.
This is true. They didn't have any such duty, under the GPLv2.
However, if I figured that out by myself, but found that I was unable to run a modified version because something in there checks for a hash computed over the program I'd like to modify, and refuses to run it because of the hash, then the hash is effectively part of the program, and they haven't provided me with the corresponding sources of that portion of the program.
"Effectively" - yes, that is the perfect way to describe it. And even though it isn't directly part, a situation like that should be covered. (In other words, if this was the way the "tivoization" section was written to make this "effectively part of the work" bit the focus a lot of my objections to it would be nullified. Give me a few hours to work on some solid and unambiguous language and I'll send something your way for review)
Quoted 8 lines
I know you don't want that to be true, and a court might actually decide your way some day. But until then, your claim that this is permitted by the GPL is just as good as mine that it's not. And I really mean "just as good", since my claim is in line with the stated purpose of the authors of the GPL, and yours is in line with their opinion (according to others, I don't think I've got this straight from them) as to whether the license effectively prohibits this practice.
Agreed. As Linus pointed out, we've been arguing over opinions and that's pointless. The only thing to do when someone states an opinion is to nod and accept it.
Quoted 10 lines
Quoted 5 lines
Then anyone using GPLv3'd software to drive WiFi devices, radio (HAM radio) networks, etc - in the US, at least - isn't allowed to do such. US Law makes some provisions of the GPLv3 illegal to comply with. Thanks to section 6 of the GPLv3 that invalidates the rights granted under the license.
Actually, this is false. Not only because of the ROM provisions in the GPLv3, but because the law requirements aren't anywhere as strict as the WiFi vendors who want to disrespect your freedoms want you to believe.
Perhaps. I haven't looked into the specific regulations in over a year, so my memory may be failing me entirely.
Quoted 5 lines
Quoted 2 lines
What the GPLv3 has done is take away options they might otherwise have had.
It doesn't. Authors can always grant these options separately if they want to. Authors can always choose GPLv2 if they want to.
Okay. I think that someone pointed out a problem with the "optional grant" idea, but I can't remember the specifics and don't feel like digging through the 500 or so posts that make up this discussion.
Quoted 5 lines
GPLv3 is an option for those who want to defend freedoms, even if they don't share the perception that this is a moral and ethical issue. If they're in it only for the self benefits, that's fine, GPLv3 can get them that, even better than GPLv2 could, in spite of the short-sighted claims to the contrary.
If this is your opinion, then run with it. My opinion on the matter is the opposite - that the GPLv2 does the job in a better way - but, well, that's my opinion. (and like my mother used to say - "Opinions are like assholes. Everyone has one, and the only one that doesn't stink is your own.")
Quoted 4 lines
Quoted 2 lines
If one of the goals of the FSF is to force proprietary software into a minority then its just done damage to that goal.
That's not the goal.
I didn't say it was "the goal", I said "one of the goals". I'm the first to admit when I'm wrong, but in this case I've read interviews with RMS where he has said that one of the reasons he founded the FSF was to marginalize proprietary software. (No, I don't know where this was - the interview was done several years ago)
DRH
Signature · 1 line
Dialup is like pissing through a pipette. Slow and excruciatingly painful.
Re: mea culpa on the meaning of Tivoization
Quoted 5 lines · Daniel Hazelton
Quoted 1 line · On Monday 18 June 2007 15:09:47 Alexandre Oliva wrote:
Yes. Anyone feels like enforcing the GPLv2 in Brazil?
I don't know if I have the right. None of the code is mine
It would have to be some major copyright holder of core Linux code, or the mips port (IIRC it's mips hardware), or some other driver they're using.
Quoted 3 lines
Okay. So its possible to change whats running on the hardware - but even though nobody has the information needed to do it, it's a violation. Hrm... I can see some valid reasoning behind this,
Yup. Same reasoning as "I threw the source code away", really.
Quoted 3 lines
"Effectively" - yes, that is the perfect way to describe it. And even though it isn't directly part, a situation like that should be covered.
And if you look at GPLv3dd1 or dd2 IIRC, that's how it started. For some reason, the FSF turned it into the more lax (in some senses) installation information for user products in dd3. Maybe they decided that the argument about the signature being effectively part of the executable, and therefore the key being effectively part of the source code, was less likely to be upheld in a court of law than this alternate phrasing. All in all, the effect is the same AFAICT, and the spirit is being complied with.
Quoted 9 lines
Quoted 5 lines
Quoted 2 lines
What the GPLv3 has done is take away options they might otherwise have had.
It doesn't. Authors can always grant these options separately if they want to. Authors can always choose GPLv2 if they want to.
Okay. I think that someone pointed out a problem with the "optional grant" idea, but I can't remember the specifics and don't feel like digging through the 500 or so posts that make up this discussion.
Linus claimed he would then have to refrain from accepting contributions from anyone who removed this additional permission.
I don't see how this is different from refraining from accepting contributions under any other license, except that you can't use license incompatibility to reason it out as an impossibility you established for yourself in just the very same way.
Quoted 6 lines
Quoted 4 lines
Quoted 2 lines
If one of the goals of the FSF is to force proprietary software into a minority then its just done damage to that goal.
That's not the goal.
I didn't say it was "the goal", I said "one of the goals".
I stand corrected. Sorry. It's been a long thread and a long week.
My objection was mainly about the "forcing". FSF's stance is about educating users as to the moral and ethical reasons, such that they reject non-Free Software, while at the same time providing software authors with means to stop others from hurting users, by depriving them of the freedoms they're morally entitled to have.
Others often perceive FSF's tactics as forceful, and I don't deny that this may be justified, based on past interactions with the FSF. That said, I think they've improved a lot, even if they're not perfect (who is?). But the perception and the consequent rejection unfortunately remains as strong as ever.
Signature · 4 lines
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}Re: mea culpa on the meaning of Tivoization
Quoted 9 lines · On Monday 18 June 2007 17:31:47 Alexandre Oliva wrote:
Quoted 4 lines · Daniel Hazelton
Quoted 1 line · On Monday 18 June 2007 15:09:47 Alexandre Oliva wrote:
Yes. Anyone feels like enforcing the GPLv2 in Brazil?
I don't know if I have the right. None of the code is mine
It would have to be some major copyright holder of core Linux code, or the mips port (IIRC it's mips hardware), or some other driver they're using.
That's what I thought. Anyone here feel like pursuing that? It is, directly, a license violation. :)
Quoted 5 lines
Quoted 3 lines
Okay. So its possible to change whats running on the hardware - but even though nobody has the information needed to do it, it's a violation. Hrm... I can see some valid reasoning behind this,
Yup. Same reasoning as "I threw the source code away", really.
Not really. "I threw the source away" violates the license in a very clear manner - they are distributing a "work based on the work" without complying with the terms of the license under which they gained access to the work. However, as I said, I can see some valid reasoning. But its more a part of the "Hey, I paid good money for this thing and I can't use it how I want!" type of reasoning.
Quoted 4 lines
Quoted 3 lines
"Effectively" - yes, that is the perfect way to describe it. And even though it isn't directly part, a situation like that should be covered.
Quoted 8 lines
And if you look at GPLv3dd1 or dd2 IIRC, that's how it started. For some reason, the FSF turned it into the more lax (in some senses) installation information for user products in dd3. Maybe they decided that the argument about the signature being effectively part of the executable, and therefore the key being effectively part of the source code, was less likely to be upheld in a court of law than this alternate phrasing. All in all, the effect is the same AFAICT, and the spirit is being complied with.
But the change has some massive problems. If dd1 or dd2 was clearly and concisely written such that the conditions were not open to a different interpretation without creative re-definition of words then changes would not be needed. (I'm still working on the version I mentioned - give me a bit, writing english in such a way that a lawyer can't twist it to mean whatever they are paid to make it mean is difficult.)
Quoted 10 lines
Quoted 9 lines
Quoted 5 lines
Quoted 2 lines
What the GPLv3 has done is take away options they might otherwise have had.
It doesn't. Authors can always grant these options separately if they want to. Authors can always choose GPLv2 if they want to.
Okay. I think that someone pointed out a problem with the "optional grant" idea, but I can't remember the specifics and don't feel like digging through the 500 or so posts that make up this discussion.
Quoted 7 lines
Linus claimed he would then have to refrain from accepting contributions from anyone who removed this additional permission.
I don't see how this is different from refraining from accepting contributions under any other license, except that you can't use license incompatibility to reason it out as an impossibility you established for yourself in just the very same way.
I think there was more to it than that, but the point doesn't matter. If the license used on contributed code *isn't* completely compatible with the license on the project it can't be used anyway. (doesn't the GPLv3 cover situations like that?)
Quoted 8 lines
Quoted 6 lines
Quoted 4 lines
Quoted 2 lines
If one of the goals of the FSF is to force proprietary software into a minority then its just done damage to that goal.
That's not the goal.
I didn't say it was "the goal", I said "one of the goals".
I stand corrected. Sorry. It's been a long thread and a long week.
Understandable.
Quoted 5 lines
My objection was mainly about the "forcing". FSF's stance is about educating users as to the moral and ethical reasons, such that they reject non-Free Software, while at the same time providing software authors with means to stop others from hurting users, by depriving them of the freedoms they're morally entitled to have.
Hrm... When I first hit the end of this massive sentence I was really confused. Took about five minutes for me to remember that "morally entitled" is based on the morals promoted by the FSF.
Quoted 5 lines
Others often perceive FSF's tactics as forceful, and I don't deny that this may be justified, based on past interactions with the FSF. That said, I think they've improved a lot, even if they're not perfect (who is?). But the perception and the consequent rejection unfortunately remains as strong as ever.
IMHO RMS and the FSF go about things in the wrong manner. Yes, the goal is admirable, but the means by which they are attempting to reach them are not. However, I have no connection to the FSF or RMS and thus cannot dictate anything. What I can do is try to educate people as to why I have the dislike of the FSF and RMS that I do. (and how I can *respect* RMS and the FSF while not liking them :)
In this discussion I've tried to point out things that occur that are at odds with the text of the GPLv3 (as it exists - the "31/5/2007 'Final Comment Draft'). I have also tried to point out that what the FSF and RMS claim is the "spirit" of the GPL cannot be conclusively proven to be the same "spirit" it had when it was written. When it comes to that somewhat slippery term "intent" I think I made my point very clear - ie: the intent of the license is what each licensor believes it to be. (meaning my view of the intent is no less valid than the FSF's view)
Everyone that has been part of this discussion - my personal code of morals will not let me get away without this: "Forgive me if, in the heat of the moment, I offended any of you." (it shouldn't need to be said that I've forgiven the few people that managed to offend me)
Now - this discussion has been a lot of fun and really kept me on my toes. It has been years since I've had a "debate" as lively as this one has been. So thank you all for it. Now please, remove me from the CC: list as, IMNSHO, this discussion has run its course and come to all the resolution it ever will.
DRH
Signature · 1 line
Dialup is like pissing through a pipette. Slow and excruciatingly painful.
Re: mea culpa on the meaning of Tivoization
Quoted 11 lines · Daniel Hazelton
Quoted 8 lines · On Monday 18 June 2007 17:31:47 Alexandre Oliva wrote:
And if you look at GPLv3dd1 or dd2 IIRC, that's how it started. For some reason, the FSF turned it into the more lax (in some senses) installation information for user products in dd3. Maybe they decided that the argument about the signature being effectively part of the executable, and therefore the key being effectively part of the source code, was less likely to be upheld in a court of law than this alternate phrasing. All in all, the effect is the same AFAICT, and the spirit is being complied with.
But the change has some massive problems.
Such as? Is the effect really any different?
Quoted 6 lines
If dd1 or dd2 was clearly and concisely written such that the conditions were not open to a different interpretation without creative re-definition of words then changes would not be needed. (I'm still working on the version I mentioned - give me a bit, writing english in such a way that a lawyer can't twist it to mean whatever they are paid to make it mean is difficult.)
It's very difficult and, worse, it might turn out to be unenforceable. You'd have to count on signing keys being copyrightable, and they are unlikely to be, and on signatures being derived works of both, which is a tough call. The whole idea resonates very well with the spirit of the license, but we need more than that, we need it to be very likely to work. I suspect this is why the FSF has decided to take another route to achieve the same (AFAICT) effect.
Quoted 9 lines
Quoted 4 lines
I don't see how this is different from refraining from accepting contributions under any other license, except that you can't use license incompatibility to reason it out as an impossibility you established for yourself in just the very same way.
I think there was more to it than that, but the point doesn't matter. If the license used on contributed code *isn't* completely compatible with the license on the project it can't be used anyway. (doesn't the GPLv3 cover situations like that?)
I'm not sure what you're asking. GPLv3 covers additional permissions, that are really no different from dual-licensing, so anyone can choose to drop them when combining with works (including their own) that don't offer such additional permissions.
Quoted 9 lines
Quoted 5 lines
My objection was mainly about the "forcing". FSF's stance is about educating users as to the moral and ethical reasons, such that they reject non-Free Software, while at the same time providing software authors with means to stop others from hurting users, by depriving them of the freedoms they're morally entitled to have.
Hrm... When I first hit the end of this massive sentence I was really confused. Took about five minutes for me to remember that "morally entitled" is based on the morals promoted by the FSF.
Yes. And the 'them' after the last comma refers to the users, not the authors (although they can be users too), in case it's not clear ;-) :-D
Quoted 3 lines
Everyone that has been part of this discussion - my personal code of morals will not let me get away without this: "Forgive me if, in the heat of the moment, I offended any of you."
FWIW, I never felt offended by you, but I second your request and extend it to all participants in the thread too, particularly to Ingo, to whom I remember having directed some harsh words.
Signature · 4 lines
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}Re: mea culpa on the meaning of Tivoization
Quoted 11 lines · On Monday 18 June 2007 22:57:20 Alexandre Oliva wrote:
Quoted 10 lines · Daniel Hazelton
Quoted 8 lines · On Monday 18 June 2007 17:31:47 Alexandre Oliva wrote:
And if you look at GPLv3dd1 or dd2 IIRC, that's how it started. For some reason, the FSF turned it into the more lax (in some senses) installation information for user products in dd3. Maybe they decided that the argument about the signature being effectively part of the executable, and therefore the key being effectively part of the source code, was less likely to be upheld in a court of law than this alternate phrasing. All in all, the effect is the same AFAICT, and the spirit is being complied with.
Quoted 3 lines
Quoted 1 line
But the change has some massive problems.
Such as? Is the effect really any different?
I haven't looked at it, in depth, today but one of the problems I saw was the apparent loopholes in the text. No specifics, but I remember thinking "a lawyer would have a field day with this - dozens of ways they could sidestep these issues"
Quoted 7 lines
Quoted 6 lines
If dd1 or dd2 was clearly and concisely written such that the conditions were not open to a different interpretation without creative re-definition of words then changes would not be needed. (I'm still working on the version I mentioned - give me a bit, writing english in such a way that a lawyer can't twist it to mean whatever they are paid to make it mean is difficult.)
Quoted 7 lines
It's very difficult and, worse, it might turn out to be unenforceable. You'd have to count on signing keys being copyrightable, and they are unlikely to be, and on signatures being derived works of both, which is a tough call. The whole idea resonates very well with the spirit of the license, but we need more than that, we need it to be very likely to work. I suspect this is why the FSF has decided to take another route to achieve the same (AFAICT) effect.
Agreed. I'm still stuck trying to keep the language concise and understandable without delving into the descriptive flights of fancy I enjoy. (I write a lot more fiction than I do code, even though I started writing code a long time before I started on fiction)
Quoted 10 lines
Quoted 9 lines
Quoted 4 lines
I don't see how this is different from refraining from accepting contributions under any other license, except that you can't use license incompatibility to reason it out as an impossibility you established for yourself in just the very same way.
I think there was more to it than that, but the point doesn't matter. If the license used on contributed code *isn't* completely compatible with the license on the project it can't be used anyway. (doesn't the GPLv3 cover situations like that?)
Quoted 4 lines
I'm not sure what you're asking. GPLv3 covers additional permissions, that are really no different from dual-licensing, so anyone can choose to drop them when combining with works (including their own) that don't offer such additional permissions.
What I was getting at, here, is that the GPLv3 isn't backwards compatible with GPLv2, because you aren't allowed to remove rights from the GPLv3. Remember, there are rights encoded in the GPLv3 that don't appear in v2. In fact, if you want to use GPLv3 code in a GPLv2 project you have to use GPLv3. For some projects, like the Linux Kernel, the upgrade is impossible to accomplish.
Quoted 6 lines
Quoted 6 lines
Quoted 5 lines
My objection was mainly about the "forcing". FSF's stance is about educating users as to the moral and ethical reasons, such that they reject non-Free Software, while at the same time providing software authors with means to stop others from hurting users, by depriving them of the freedoms they're morally entitled to have.
Quoted 8 lines
Quoted 3 lines
Hrm... When I first hit the end of this massive sentence I was really confused. Took about five minutes for me to remember that "morally entitled" is based on the morals promoted by the FSF.
Yes. And the 'them' after the last comma refers to the users, not the authors (although they can be users too), in case it's not clear ;-)
:-D
Yes. I almost replied "-ENOPARSE" because, when I first read it, I parsed it as "by depriving [the authors] of the freedoms they're morally entitled to have". When my brain finally rebooted after that bought of idiocy I was able to parse it properly.
DRH
Quoted 7 lines
Quoted 3 lines
Everyone that has been part of this discussion - my personal code of morals will not let me get away without this: "Forgive me if, in the heat of the moment, I offended any of you."
FWIW, I never felt offended by you, but I second your request and extend it to all participants in the thread too, particularly to Ingo, to whom I remember having directed some harsh words.
Signature · 1 line
Dialup is like pissing through a pipette. Slow and excruciatingly painful.
Re: mea culpa on the meaning of Tivoization
Quoted 4 lines · Daniel Hazelton
I haven't looked at it, in depth, today but one of the problems I saw was the apparent loopholes in the text. No specifics, but I remember thinking "a lawyer would have a field day with this - dozens of ways they could sidestep these issues"
*Pretty* *please* file comments about the apparent loopholes at gplv3.fsf.org/comments
Quoted 2 lines
What I was getting at, here, is that the GPLv3 isn't backwards compatible with GPLv2,
It couldn't possibly be. The whole point of upgrading the GPL is such that it complies better with its spirit of defending the freedoms, so as to keep free software free. This can only be accomplished with additional restrictions that stop practices that deny users' freedoms.
Relaxing the provisions, a necessary condition for compatibility, wouldn't make for better defenses.
Quoted 2 lines
because you aren't allowed to remove rights from the GPLv3. Remember, there are rights encoded in the GPLv3 that don't appear in v2.
I'm not sure what you mean by "rights" in the two sentences above. You know you can grant additional permissions, so I assume that's not what you mean, even more so because you *can* indeed take them out. Is it "conditions", "restrictions" or some such, that in turn translate into freedoms for downstream users, or is it about the granted rights per se?
Quoted 3 lines
In fact, if you want to use GPLv3 code in a GPLv2 project you have to use GPLv3. For some projects, like the Linux Kernel, the upgrade is impossible to accomplish.
Impossible is a bit too strong. I understand it would take a huge amount of work though, so I sympathize with "it wouldn't be worth it", even if, in my scale of moral values, I'd disagree.
Signature · 4 lines
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}Re: mea culpa on the meaning of Tivoization
Quoted 8 lines · On Tuesday 19 June 2007 02:10:02 Alexandre Oliva wrote:
Quoted 4 lines · Daniel Hazelton
I haven't looked at it, in depth, today but one of the problems I saw was the apparent loopholes in the text. No specifics, but I remember thinking "a lawyer would have a field day with this - dozens of ways they could sidestep these issues"
*Pretty* *please* file comments about the apparent loopholes at gplv3.fsf.org/comments
To do that I'd have to go back, take the time to re-read the GPLv3 *in* *depth*, think about each paragraph of each section individually...
Like I said, I just got a general impression that a lawyer would have a field-day with it.
Quoted 15 lines
Quoted 2 lines
What I was getting at, here, is that the GPLv3 isn't backwards compatible with GPLv2,
It couldn't possibly be. The whole point of upgrading the GPL is such that it complies better with its spirit of defending the freedoms, so as to keep free software free. This can only be accomplished with additional restrictions that stop practices that deny users' freedoms.
Relaxing the provisions, a necessary condition for compatibility, wouldn't make for better defenses.
Quoted 2 lines
because you aren't allowed to remove rights from the GPLv3. Remember, there are rights encoded in the GPLv3 that don't appear in v2.
Quoted 6 lines
I'm not sure what you mean by "rights" in the two sentences above. You know you can grant additional permissions, so I assume that's not what you mean, even more so because you *can* indeed take them out. Is it "conditions", "restrictions" or some such, that in turn translate into freedoms for downstream users, or is it about the granted rights per se?
Sorry, bad choice of words. There are "guarantees" encoded into every license. There are some encoded into the GPLv3 that aren't encoded into the GPLv2. You can't remove or restrict those guarantees without violating the license. And removing those guarantees would be the only way to make the GPLv3 fully compatible with the GPLv2.
Quoted 7 lines
Quoted 3 lines
In fact, if you want to use GPLv3 code in a GPLv2 project you have to use GPLv3. For some projects, like the Linux Kernel, the upgrade is impossible to accomplish.
Impossible is a bit too strong. I understand it would take a huge amount of work though, so I sympathize with "it wouldn't be worth it", even if, in my scale of moral values, I'd disagree.
In this case I wasn't speaking literally. I should have been a lot more specific there - it should say "practically impossible".
DRH
Signature · 1 line
Dialup is like pissing through a pipette. Slow and excruciatingly painful.
Quoted 9 lines · Alan Cox
Quoted 6 lines
Quoted 2 lines
I don't know any law that requires tivoization.
In the USSA it is arguable that wireless might need it (if done in software) for certain properties. (The argument being it must be tamperproof to random end consumers).
Quoted 5 lines
But this is not tivoization.
Tivoization is a manufacturer using technical measures to prevent the user from tampering (*) with the device, *while* keeping the ability to tamper with it changes itself.
You're splitting those hairs might finely. So when you ask whether there's any law that "requirse tivoization", you won't accept a law that creates a situation where the only practical solution is tivoization?
Quoted 11 lines
(*) tampering brings in negative connotations that I'd rather avoid, but since that was the term you used, and the term "modifying" might bring in legal-based technicalities such as that replacing isn't modification, I just went with it.
So, given a proper definition, do you know any law that requires tivoization?
Taking it further, do you know whether any such law requires *worldwide* tivoization, as in, applying the restrictions in the law even outside its own jurisdiction?
A law that requires certaint things be tamper-proof, where engineering realities requires that they be controlled by software and the software be upgradable (for security reasons and for support of future protocol revisions) isn't good enough for you?
DS
Quoted 8 lines · David Schwartz
Quoted 8 lines · Alan Cox
Quoted 6 lines
Quoted 2 lines
I don't know any law that requires tivoization.
In the USSA it is arguable that wireless might need it (if done in software) for certain properties. (The argument being it must be tamperproof to random end consumers).
Quoted 7 lines
Quoted 5 lines
But this is not tivoization.
Tivoization is a manufacturer using technical measures to prevent the user from tampering (*) with the device, *while* keeping the ability to tamper with it changes itself.
You're splitting those hairs might finely.
And I was wrong. Please see the "mea culpa on the meaning of tivoization" thread.
Quoted 3 lines
So when you ask whether there's any law that "requirse tivoization", you won't accept a law that creates a situation where the only practical solution is tivoization?
I guess it amounts to what you mean by "*only* practical solution".
"I can't fit the corresponding sources in this CD, so you won't get them." is no excuse to disrespect users' freedoms, why should this be different?
Quoted 8 lines
Quoted 3 lines
Taking it further, do you know whether any such law requires *worldwide* tivoization, as in, applying the restrictions in the law even outside its own jurisdiction?
A law that requires certaint things be tamper-proof, where engineering realities requires that they be controlled by software and the software be upgradable (for security reasons and for support of future protocol revisions) isn't good enough for you?
"engineering realities" is the weak point of your argument, see above. Is ROM still software? Is replaceable ROM still software?
Signature · 4 lines
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}Let's take a certain class of medical devices into account: ones that are absolutely definitely for medical treatment, but are not life threatening if they fail.
Say, a dental treatment device -- if the device produces a crown or bridge that doesn't fit properly, the dentist says "nope" and throws it away. No harm done.
Quoted 1 line · Alexandre Oliva
There may be business models that require the ability to make changes.
I'd say that its sensible for the manufacturer to attempt to retain this ability in every case. You never know what's going to go wrong, so it's a plus to have this option so that you can roll out some types of fixes without going bankrupt.
Now, for medical devices, this is tricky stuff: medical devices require all sorts of certifications, so modifying your product after you have certified it has it's complications. However, despite all the regulations it's realistic to be able to do this, and it does happen. Hell, windows-based devices in this field download new antivirus definitions and run windows update every few days.
Quoted 2 lines
Then it's fair to enable the user to make changes as well, such that they don't become dependent on the vendor
Now this is where the regulations get really heavy. If the user is offered the ability to modify the device, theres *no way* it would get certified. Your business is dead - you do not have a product you can sell. In such case, the license has completely excluded free software from the market and everyone is forced to use completely closed systems.
I realise that the latest GPLv3 draft would not pose restrictions here, as such devices would not be classified as consumer products. That said, talking purely in terms of business models and fairness: there ARE decent reasons for manufacturer lockdown in some industries.
Daniel
Quoted 7 lines · Daniel Drake
Let's take a certain class of medical devices into account: ones that are absolutely definitely for medical treatment, but are not life threatening if they fail.
Say, a dental treatment device -- if the device produces a crown or bridge that doesn't fit properly, the dentist says "nope" and throws it away. No harm done.
I've done quite a bit of research, I'm not nearly done.
These regulations (from what I can tell) seemed to follow suit with the National Electric Code (NEC) [latest] when dealing with mandatory isolated ground devices and special cabling methods when it comes into a device touching a patient. If that remains consistent, this won't be so bad.
If the patient never comes in contact with it, its not regulated as much and (from what I've seen) has no requirement for tamper proofing. I point out again, I am not _nearly_ done with my research.
I think of nothing else, anyone with an interest should closely monitor how these devices are being regulated by the FDA as more of them begin to look like penguins.
I won't argue one way or another as to the presence of benevolent intent in those laws-to-come, I'm simply pointing out the questionable technical competency of those who will be writing them and their need for guidance when doing so.
Best,
--Tim
Quoted 3 lines · Daniel Drake
I realise that the latest GPLv3 draft would not pose restrictions here, as such devices would not be classified as consumer products.
And even if they were, there's always ROM.
I don't know whether hardware seals that state "once you break this seal, law prohibits the use of this device with human patients".
Then the restriction is not being imposed by the manufacturer, only by law, and this does make lot of a difference as far as software freedom is concerned.
But then, law might not find this to be enough. Software patents are not the only stupid law that harms Free Software :-(
Signature · 4 lines
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}Quoted 10 lines · Alexandre Oliva
Quoted 3 lines · Daniel Drake
I realise that the latest GPLv3 draft would not pose restrictions here, as such devices would not be classified as consumer products.
And even if they were, there's always ROM.
I don't know whether hardware seals that state "once you break this seal, law prohibits the use of this device with human patients".
once you break the seal the device is no longer certified. an uncertified device cannot be used.
this is very common (in some areas it's widely ignored, in others it isn't)
this is just like the 'you void the warranty if you disrupt this sticker' stickers that you see on just about any hardware you buy today. some vendors are stickers for this, others don't really care.
David Lang
Quoted 8 lines
Then the restriction is not being imposed by the manufacturer, only by law, and this does make lot of a difference as far as software freedom is concerned.
But then, law might not find this to be enough. Software patents are not the only stupid law that harms Free Software :-(
Quoted 7 lines · On Jun 19, 2007, [email protected] wrote:
Quoted 6 lines · Alexandre Oliva
Quoted 3 lines · Daniel Drake
I realise that the latest GPLv3 draft would not pose restrictions here, as such devices would not be classified as consumer products.
Quoted 7 lines
Quoted 4 lines
And even if they were, there's always ROM.
I don't know whether hardware seals that state "once you break this seal, law prohibits the use of this device with human patients".
once you break the seal the device is no longer certified. an uncertified device cannot be used.
Yup. That's the law.
At which point it's not the hardware vendor imposing the restriction, so this use is perfectly acceptable.
One could presumably implement similar seals in software. Nothing wrong with a signature used to indicate that the device has been tampered with. Even a led somewhere that reflects this status.
None of this prevents the user from enjoying the freedoms he's entitled to, according to the laws of the place where he lives.
Signature · 4 lines
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}Quoted 2 lines · Alexandre Oliva
If the bug is in the non-GPLed BIOS, not in the GPLed code, too bad. One more reason to dislike non-Free Software.
Maybe the Tivo only loading signed kernels is a bug in their bios. :)
Quoted 7 lines
The freedom the GPL defends is not the freedom to modify and debug the system, but rather the covered software.
Now, if you find evidence that the "bug" is actually intentionally put there to stop you from doing what you wanted with the software, then there's clearly a violation of the spirit of the license, and you might even have a case of copyright infringement, but IANAL.
There are many interesting bugs out there. Who is to say what was intensional?
--
Len Sorensen
Quoted 5 lines · On Jun 20, 2007, [email protected] (Lennart Sorensen) wrote:
Quoted 2 lines · Alexandre Oliva
If the bug is in the non-GPLed BIOS, not in the GPLed code, too bad. One more reason to dislike non-Free Software.
Maybe the Tivo only loading signed kernels is a bug in their bios. :)
That might be so. And in the US, a court might end up finding that piece of code was taken at random, for no particular reason, from a sample of garbage produced by code monkeys (in a very literal sense) typing at random on computer keyboards over a large period of time ;-)
Signature · 4 lines
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}Quoted 5 lines · On Wednesday 13 June 2007 21:04:42 Alexandre Oliva wrote:
Quoted 1 line · Daniel Hazelton
Now stop parroting the FSF's worn and tired tripe.
Are you playing Linus' sheeple and parroting his lines just to make a point, or are you like that all the time? ;-)
Read the hover text on http://xkcd.org/c202.html
I'm wondering if it's time for a "Munroe's law" concerning that word...
Rob
Quoted 3 lines
Quoted 1 line
Only courts of law can do that.
Wrong! Anyone with half a brain can make the distinction. What TiVO did is
Maybe half a brain can, but anyone with a whole brain can assure you its a bit more complex and you are wrong..
Quoted 2 lines
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* -
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 see fit about their software and anything derivative of it.
Quoted 2 lines
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 -
You can't patent algorithms either
Alan
Quoted 8 lines · On Wednesday 13 June 2007 21:16:19 Alan Cox wrote:
Quoted 4 lines
Quoted 1 line
Only courts of law can do that.
Wrong! Anyone with half a brain can make the distinction. What TiVO did is
Maybe half a brain can, but anyone with a whole brain can assure you its a bit more complex and you are wrong..
Quoted 8 lines
Quoted 3 lines
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* -
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.
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 other Linux kernel)
Quoted 7 lines
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 see fit about their software and anything derivative of it.
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 should be")
Quoted 5 lines
Quoted 3 lines
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 -
You can't patent algorithms either
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.
DRH
Quoted 1 line
Alan
Signature · 1 line
Dialup is like pissing through a pipette. Slow and excruciatingly painful.
Quoted 2 lines · Daniel Hazelton
(and, in the case of a TiVO, the signing
keys are part of the installation, not the running or building.
Is installation not a precondition for running?
Signature · 4 lines
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}Quoted 6 lines · Alexandre Oliva
Quoted 2 lines · Daniel Hazelton
(and, in the case of a TiVO, the signing
keys are part of the installation, not the running or building.Is installation not a precondition for running?
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?
Quoted 4 lines · On Jun 14, 2007, [email protected] wrote:
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).
Signature · 4 lines
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}Quoted 10 lines · Alexandre Oliva
Quoted 4 lines · On Jun 14, 2007, [email protected] wrote:
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).
Do they have to provide a ROM burner if the ROM is socketed rather than soldered into place?
Quoted 8 lines · On Jun 14, 2007, [email protected] wrote:
Quoted 7 lines · Alexandre Oliva
Quoted 4 lines · On Jun 14, 2007, [email protected] wrote:
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?
Quoted 6 lines
Quoted 3 lines
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).
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.
Signature · 4 lines
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}Quoted 5 lines · Alexandre Oliva
Quoted 2 lines
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.
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
Quoted 2 lines
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
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.
Quoted 5 lines · Daniel Hazelton
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 should be")
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.
Signature · 1 line
dwmw2
Quoted 11 lines · On Thursday 14 June 2007 18:24:55 David Woodhouse wrote:
Quoted 5 lines · Daniel Hazelton
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 should be")
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.
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
Signature · 1 line
Dialup is like pissing through a pipette. Slow and excruciatingly painful.
Quoted 5 lines · Daniel Hazelton
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.
Yes.
Quoted 3 lines
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".
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.
Signature · 1 line
dwmw2
Quoted 15 lines · On Friday 15 June 2007 04:25:24 David Woodhouse wrote:
Quoted 5 lines · Daniel Hazelton
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.
Yes.
Quoted 3 lines
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".
Not necessarily. I'm not entirely sure where you got that idea from.
Quoted 8 lines
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.
-ELOGIC
Quoted 5 lines
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.
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
Signature · 1 line
Dialup is like pissing through a pipette. Slow and excruciatingly painful.
Quoted 10 lines · Daniel Hazelton
Quoted 8 lines
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.
-ELOGIC
What's logic got to do with it? It was fairly much a direct quote from the licence. You have _read_ the licence, haven't you?
Quoted 6 lines
Quoted 5 lines
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.
Quoted 10 lines
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.
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 combines it with other code of your own. But if you do, you're bound by the licence. And whether your code is a 'derived work' has nothing to do with it.
Yes, there are exceptions for mere aggregation onto a storage medium -- if the kernel and your own work are next to each other on a backup tape or your laptop's hard drive, or even both burned to a 'Gratis Software' CD as _separate_ works, then that doesn't count. But we're talking about a product which has a Linux kernel, a module built specifically for that kernel, and cannot function unless both of them are present. That ain't "mere aggregation on a storage medium".
Signature · 1 line
dwmw2
Quoted 11 lines · On Friday 15 June 2007 05:17:44 David Woodhouse wrote:
Quoted 10 lines · Daniel Hazelton
Quoted 9 lines
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.
Quoted 4 lines
Quoted 1 line
-ELOGICWhat's logic got to do with it? It was fairly much a direct quote from the licence. You have _read_ the licence, haven't you?
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 by the GPL.
Quoted 17 lines
Quoted 6 lines
Quoted 5 lines
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.
Quoted 10 lines
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.
Quoted 1 line
You're confused.
Nope. Not confused at all.
Quoted 4 lines
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.
But you obviously are. After all, what does this have to do with whether the GPLv2 can "magically" change the law?
Quoted 11 lines
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.
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 causes that module to become covered by the GPL?
Quoted 10 lines
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'.
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 this License.
In other words, even though I've built a program that isn't GPL'd, I can still put it on the same "volume of storage or distribution medium" as a GPL'd work and not have to put it under the GPL. Imagine that.
And again, I have to point out that, no matter what RMS and the rest of the people that wrote the GPLv2 may believe, it can't change the law on which it is based. If the law says "mechanical translation or processes cannot produce a new work", then no matter what the GPLv2 says, I can write a parser skeleton in the "YACC" language, run it through Bison and *NOT* need the "exception" that the FSF "thoughtfully" provides. Why? Because the *skeleton* I wrote carries the copyright - not the output of Bison, despite what the GPL *might* say on this in Section 0. A license *cannot* change the law, because it gets all its power from the law.
Quoted 5 lines
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.
Hrm... I see, you've included the section unmolested here - but you still seem unable to read it correctly. Perhaps I'm wrong though.
Quoted 4 lines
It's your choice -- you're not _forced_ to use the kernel, and you're not _forced_ to distribute a product which combines it with other code of your own. But if you do, you're bound by the licence. And whether your code is a 'derived work' has nothing to do with it.
And is this what the process of making a module does? Because that *IS* what I had mentioned. If the mail I'm responding to was a response to that mail then you are sadly confused as to what I was talking about.
Quoted 7 lines
Yes, there are exceptions for mere aggregation onto a storage medium -- if the kernel and your own work are next to each other on a backup tape or your laptop's hard drive, or even both burned to a 'Gratis Software' CD as _separate_ works, then that doesn't count. But we're talking about a product which has a Linux kernel, a module built specifically for that kernel, and cannot function unless both of them are present. That ain't "mere aggregation on a storage medium".
Yet it still doesn't make them a "combined work". If it did then the simple act of me installing a copy of the ATI or NVidia modules makes them GPL'd. But it doesn't - if it did I'm sure that somebody would have filed a lawsuit over this already.
DRH
Signature · 1 line
Dialup is like pissing through a pipette. Slow and excruciatingly painful.
Quoted 11 lines · Daniel Hazelton
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 by the GPL.
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_?
That's... an interesting interpretation.
Quoted 7 lines
Quoted 4 lines
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.
But you obviously are. After all, what does this have to do with whether the GPLv2 can "magically" change the law?
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 "magical" change to the law required.
Quoted 7 lines
Quoted 7 lines
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).
Quoted 8 lines
Quoted 4 lines
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.
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 causes that module to become covered by the GPL?
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 were only _put_ there mechanically by the software).
Quoted 11 lines
Quoted 4 lines
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'.
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 this License.
Quoted 3 lines
In other words, even though I've built a program that isn't GPL'd, I can still put it on the same "volume of storage or distribution medium" as a GPL'd work and not have to put it under the GPL. Imagine that.
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 separate work' or whether it's part of a larger, coherent whole.
Quoted 8 lines
Quoted 5 lines
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.
Hrm... I see, you've included the section unmolested here - but you still seem unable to read it correctly. Perhaps I'm wrong though.
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 interpreted in court. But I'm willing to place bets :)
Quoted 8 lines
Quoted 4 lines
It's your choice -- you're not _forced_ to use the kernel, and you're not _forced_ to distribute a product which combines it with other code of your own. But if you do, you're bound by the licence. And whether your code is a 'derived work' has nothing to do with it.
And is this what the process of making a module does? Because that *IS* what I had mentioned. If the mail I'm responding to was a response to that mail then you are sadly confused as to what I was talking about.
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 the GPL's requirements about collective works. What part of that confuses you? You seem to _keep_ going back to talking about derived works.
Quoted 8 lines
Quoted 7 lines
Yes, there are exceptions for mere aggregation onto a storage medium -- if the kernel and your own work are next to each other on a backup tape or your laptop's hard drive, or even both burned to a 'Gratis Software' CD as _separate_ works, then that doesn't count. But we're talking about a product which has a Linux kernel, a module built specifically for that kernel, and cannot function unless both of them are present. That ain't "mere aggregation on a storage medium".
Quoted 4 lines
Yet it still doesn't make them a "combined work". If it did then the simple act of me installing a copy of the ATI or NVidia modules makes them GPL'd. But it doesn't - if it did I'm sure that somebody would have filed a lawsuit over this already.
The GPL applies if you _distribute_ a work which combines such modules and the kernel into a larger whole. You can do what you like if you don't distribute it.
The 'mere aggregation' thing means you can even distribute them together if they just happen to be side-by-side on a storage medium as if by coincidence. But putting them together into a product which actually uses and requires both of them is not 'mere aggregation'. Some people _were_ bundling the ATI and nVidia modules together with the kernel in a 'product', and they stopped under threat of legal action.
The module on its own is questionable -- it is a matter of opinion whether that's a derived work or not. But when you ship a product which combines both kernel and module into a coherent whole, it doesn't _matter_ if the module is a derived work or not. You are not permitted to distribute the kernel unless your module is also licensed under the GPL. Not because it's a derived work, but because the licence of the kernel says that's the price you pay if you want to distribute the kernel.
Signature · 1 line
dwmw2
Quoted 13 lines · On Friday 15 June 2007 06:49:05 David Woodhouse wrote:
Quoted 11 lines · Daniel Hazelton
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 by the GPL.
Quoted 6 lines
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_?
That's... an interesting interpretation.
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 the linux kernel itself.
Quoted 31 lines
Quoted 7 lines
Quoted 4 lines
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.
But you obviously are. After all, what does this have to do with whether the GPLv2 can "magically" change the law?
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 "magical" change to the law required.
Quoted 12 lines
Quoted 11 lines
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.
Quoted 7 lines
Quoted 3 lines
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 causes that module to become covered by the GPL?
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 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 license to the GPL. If you think otherwise then you are sadly mistaken.
Quoted 4 lines
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 were only _put_ there mechanically by the software).
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 source file is a mechanical translation.
Quoted 11 lines
Quoted 11 lines
Quoted 4 lines
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'.
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 this License.
Quoted 8 lines
Quoted 3 lines
In other words, even though I've built a program that isn't GPL'd, I can still put it on the same "volume of storage or distribution medium" as a GPL'd work and not have to put it under the GPL. Imagine that.
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 separate work' or whether it's part of a larger, coherent whole.
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 GPL.
Quoted 6 lines
Quoted 6 lines
Quoted 5 lines
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.
Quoted 7 lines
Quoted 2 lines
Hrm... I see, you've included the section unmolested here - but you still seem unable to read it correctly. Perhaps I'm wrong though.
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 interpreted in court. But I'm willing to place bets :)
Nope. To mean "source code files" or "functions in the source code". "Linker Sections" is your idea.
Quoted 5 lines
Quoted 5 lines
Quoted 4 lines
It's your choice -- you're not _forced_ to use the kernel, and you're not _forced_ to distribute a product which combines it with other code of your own. But if you do, you're bound by the licence. And whether your code is a 'derived work' has nothing to do with it.
Quoted 7 lines
Quoted 3 lines
And is this what the process of making a module does? Because that *IS* what I had mentioned. If the mail I'm responding to was a response to that mail then you are sadly confused as to what I was talking about.
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?
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 memory *separately* from the kernel "combine GPL'd code with your own". If that's what you think then you are wrong - see my comments about "mechanical processes" and such.
Quoted 5 lines
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 the GPL's requirements about collective works. What part of that confuses you? You seem to _keep_ going back to talking about derived works.
Because the *only* relevant passages in the GPL, as applied to not-in-tree kernel modules is that about "derived works". *IF* distributing GPL'd binaries as part of an "Operating System" created a "collective work" then the FreeBSD people, who - last I looked - shipped with GCC as part of the OS, are in direct violation of the GPL. That means that either your argument is wrong or that every non-GPL'd OS that ships with GCC as an integral part is in violation of the GPL.
Quoted 24 lines
Quoted 8 lines
Quoted 7 lines
Yes, there are exceptions for mere aggregation onto a storage medium -- if the kernel and your own work are next to each other on a backup tape or your laptop's hard drive, or even both burned to a 'Gratis Software' CD as _separate_ works, then that doesn't count. But we're talking about a product which has a Linux kernel, a module built specifically for that kernel, and cannot function unless both of them are present. That ain't "mere aggregation on a storage medium".
Quoted 4 lines
Yet it still doesn't make them a "combined work". If it did then the simple act of me installing a copy of the ATI or NVidia modules makes them GPL'd. But it doesn't - if it did I'm sure that somebody would have filed a lawsuit over this already.
The GPL applies if you _distribute_ a work which combines such modules and the kernel into a larger whole. You can do what you like if you don't distribute it.
The 'mere aggregation' thing means you can even distribute them together if they just happen to be side-by-side on a storage medium as if by coincidence. But putting them together into a product which actually uses and requires both of them is not 'mere aggregation'. Some people _were_ bundling the ATI and nVidia modules together with the kernel in a 'product', and they stopped under threat of legal action.
Quoted 8 lines
The module on its own is questionable -- it is a matter of opinion whether that's a derived work or not. But when you ship a product which combines both kernel and module into a coherent whole, it doesn't _matter_ if the module is a derived work or not. You are not permitted to distribute the kernel unless your module is also licensed under the GPL. Not because it's a derived work, but because the licence of the kernel says that's the price you pay if you want to distribute the kernel.
Hrm... I see your point. When applied *solely* to the kernel and associated modules. But I can *still* see problems with that interpretation - very big ones that a good lawyer could exploit.
DRH
Signature · 1 line
Dialup is like pissing through a pipette. Slow and excruciatingly painful.
Quoted 11 lines · David Woodhouse
Quoted 7 lines · Daniel Hazelton
Quoted 5 lines
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.
-ELOGICWhat's logic got to do with it? It was fairly much a direct quote from the licence. You have _read_ the licence, haven't you?
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
Quoted 4 lines · Linus Torvalds
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.
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 into that category and what doesn't.
Quoted 7 lines
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.
Yep, those are two extremes which fall either side of the grey area and are relatively easy to agree on... perhaps.
Quoted 4 lines
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.
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 _that_ necessarily means anything -- anyone can make baseless threats).
Quoted 3 lines
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.
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 embedded device, for example a router -- and they distribute a 'blob' of firmware for it, containing both the kernel a binary-only network driver module. Again we have to ask ourselves "is this a work based on the kernel?". Obviously there isn't a 'right' answer outside a court of law, but personally I reckon it's a fairly safe bet that it _is_ going to be considered to be a work based on Linux. Especially as the thing would be totally useless if you took away either part. They're both fundamental to its operation.
Quoted 2 lines
Which is why I think you're both *potentially* right. Which one of you is *actually* right will depend on the exact circumstances ;)
True.
Signature · 1 line
dwmw2
Quoted 3 lines · David Woodhouse
Actually, I don't see where it explicitly states that it only covers derived work.
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 notion of "derived work" - as defined by copyright law, and NOT the GPLv2.
Quoted 7 lines
The case which interests me most is when someone makes an embedded device, for example a router -- and they distribute a 'blob' of firmware for it, containing both the kernel a binary-only network driver module. Again we have to ask ourselves "is this a work based on the kernel?". Obviously there isn't a 'right' answer outside a court of law, but personally I reckon it's a fairly safe bet that it _is_ going to be considered to be a work based on Linux.
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
Quoted 12 lines · Linus Torvalds
Quoted 3 lines · David Woodhouse
Actually, I don't see where it explicitly states that it only covers derived work.
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:
Quoted 6 lines
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 notion of "derived work" - as defined by copyright law, and NOT the GPLv2.
Yep. And §2 talks explicitly about independent and separate works when they are distributed _with_ the Program, as part of a larger work based on the Program.
Quoted 10 lines
Quoted 7 lines
The case which interests me most is when someone makes an embedded device, for example a router -- and they distribute a 'blob' of firmware for it, containing both the kernel a binary-only network driver module. Again we have to ask ourselves "is this a work based on the kernel?". Obviously there isn't a 'right' answer outside a court of law, but personally I reckon it's a fairly safe bet that it _is_ going to be considered to be a work based on Linux.
Hey, I kind of disagree.
Quoted 8 lines
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?
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 claim that the distribution is a 'coherent whole' too.
Quoted 3 lines
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.
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.
Signature · 1 line
dwmw2
Quoted 8 lines · David Woodhouse
Quoted 3 lines
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.
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.
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 LGPL/GPL.
Ingo
Quoted 6 lines · Ingo Molnar
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.
Not _necessarily_ a collective work. But not necessarily _not_ a collective work either.
Quoted 2 lines
For driver blobs that are shared between Windows and Linux it would be hard to argue that they are derived from the Linux kernel.
You're back to the 'derived work' thing again, which wasn't relevant.
Quoted 3 lines
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 LGPL/GPL.
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.
Signature · 1 line
dwmw2
Quoted 8 lines · David Woodhouse
Quoted 6 lines · Ingo Molnar
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.
Quoted 7 lines
Not _necessarily_ a collective work. But not necessarily _not_ a collective work either.
Quoted 2 lines
For driver blobs that are shared between Windows and Linux it would be hard to argue that they are derived from the Linux kernel.
You're back to the 'derived work' thing again, which wasn't relevant.
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 collective work) might or might not apply.
Quoted 8 lines
Quoted 3 lines
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 LGPL/GPL.
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.
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
Quoted 7 lines · David Woodhouse
Quoted 7 lines
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?
Quoted 7 lines
Quoted 2 lines
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?
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.
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 hopefully _independent_ of both parties, so when a judge makes an arbitrary decision, the "arbitrariness" is hopefully at least somewhat "fair".
Hmm?
Linus
Quoted 6 lines · Linus Torvalds
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.
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 _stop_ shipping those modules, when once they did.
Quoted 9 lines
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 hopefully _independent_ of both parties, so when a judge makes an arbitrary decision, the "arbitrariness" is hopefully at least somewhat "fair".
Indeed.
Signature · 1 line
dwmw2
Quoted 19 lines · On Friday 15 June 2007 15:49:15 David Woodhouse wrote:
Quoted 10 lines · Linus Torvalds
Quoted 2 lines · David Woodhouse
Actually, I don't see where it explicitly states that it only covers derived work.
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:
Quoted 7 lines
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 notion of "derived work" - as defined by copyright law, and NOT the GPLv2.
Quoted 5 lines
Yep. And §2 talks explicitly about independent and separate works when they are distributed _with_ the Program, as part of a larger work based on the Program.
From section 0 of the GPLv2:
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.
DRH
Quoted 38 lines
Quoted 13 lines
Quoted 7 lines
The case which interests me most is when someone makes an embedded device, for example a router -- and they distribute a 'blob' of firmware for it, containing both the kernel a binary-only network driver module. Again we have to ask ourselves "is this a work based on the kernel?". Obviously there isn't a 'right' answer outside a court of law, but personally I reckon it's a fairly safe bet that it _is_ going to be considered to be a work based on Linux.
Hey, I kind of disagree.
What is a DVD? It's just a "blob" of a UDF image, potentially containing the Linux kernel.
Quoted 6 lines
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?
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 claim that the distribution is a 'coherent whole' too.
Quoted 3 lines
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.
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.
Signature · 1 line
Dialup is like pissing through a pipette. Slow and excruciatingly painful.
Quoted 3 lines · David Woodhouse
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,
... unless the other parts of the whole fall under the mere aggregation exception, methinks, but IANAL.
Signature · 4 lines
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}Quoted 4 lines · David Woodhouse
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.
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
Quoted 4 lines · Ingo Molnar
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".
I was actually trying to avoid the question altogether. It's not that interesting, largely because the answer is indeed 'talk to your lawyer'.
Quoted 4 lines
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.
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 product, the GPL has something to say about that.
Quoted 2 lines
but lets note that this is irrelevant to the Tivo argument. Tivo is not using bin-only modules AFAIK,
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.
Signature · 1 line
dwmw2
Quoted 8 lines · David Woodhouse
Quoted 4 lines
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.
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 product, the GPL has something to say about that.
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
Quoted 3 lines · Ingo Molnar
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)
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.
Signature · 1 line
dwmw2
Quoted 11 lines · David Woodhouse
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.
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 ... this is a much greyer area than the key stuff.
Quoted 4 lines
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.
i'm not that cynical about US courts.
Ingo
Quoted 9 lines · Ingo Molnar
Quoted 7 lines · David Woodhouse
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.
Quoted 7 lines
Quoted 4 lines
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.
as long as it's not distributed in one collective work, where is the problem?
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.
Signature · 1 line
dwmw2
Quoted 8 lines
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?
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. Anything else is mere aggregation.
Quoted 3 lines
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?
These terms are synonymous. And neither of them can apply to something that is not a work.
Quoted 4 lines
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'?
The license is just clarifying copyright law. Even if it intended to do something else, it can't. The license cannot set its own scope.
Quoted 2 lines
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.
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 something.
DS
Quoted 14 lines · On Friday 15 June 2007 09:12:43 David Woodhouse wrote:
Quoted 13 lines · Ingo Molnar
Quoted 11 lines · David Woodhouse
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.
Quoted 5 lines
Quoted 2 lines
as long as it's not distributed in one collective work, where is the problem?
As long as it's not distributed "as part of a whole which is a work based on the Program", there's no problem.
Agreed.
Quoted 4 lines
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?
Yes, you are.
Quoted 6 lines
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'?
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".
DRH
Quoted 2 lines
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.
Signature · 1 line
Dialup is like pissing through a pipette. Slow and excruciatingly painful.
Quoted 3 lines · Linus Torvalds
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.
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 explicitly allowed but I think I can defend my case".
Quoted 1 line
Only lawyers want a black-and-white world.
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 is white in front of the judge depending on their customer's needs.
Quoted 8 lines
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. Whatever. It's all good!
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
Quoted 9 lines · Linus Torvalds
Quoted 5 lines · Alexandre Oliva
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?
No. The anti-DRM language is still there, and no, it was never a misunderstanding.
It was claimed that GPLv3 would forbid implementations of DRM. That's just plain false. If you don't think so, please show what terms in the latest draft prohibit DRM (as opposed to merely making it ineffective, a necessary consequence of abiding by the spirit of all GNU GPLs)
Quoted 2 lines
It's offensive because Tivo never did anything wrong, and the FSF even acknowledged that.
Another misunderstanding. The FSF never said TiVo didn't do anything wrong. It only said it didn't think there was a license violation. I personally disagree with that assessment, but IANAL.
Anyhow, deciding whether it's right or wrong is not the same as deciding whether it's legal or illegal. Law doesn't define what's right or wrong. That's what morals and ethics do.
Quoted 3 lines
want to protect the integrity of that hardware.
The kernel license covers the *kernel*.
When they choose to include a copy of the kernel in their hardware that they can modify but others can't, they're failing to comply with the spirit of the license. For brevity, I won't repeat the quotes from the GPLv2 preamble, that I just included in the message I sent to Lennart Sorensen in this same thread. Can you justify how you came to the conclusion (if you did) that TiVo is abiding by the spirit of the license?
Quoted 6 lines
Quoted 4 lines
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.
No, it was not the former.
Wow, I didn't see that coming. Public admission of ill intentions? ;-) :-D :-P
Quoted 5 lines
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.
Let me see if I got this right. There was a section entitled "3. Digital Restrictions Management" in GPLv3dd1. Are you saying that, when people complained about the DRM clause, they actually meant the provisions in "1. Source Code", that established a requirement to include the source code corresponding to functional signatures, namely the signing keys, as part of the corresponding source code?
Quoted 1 line
And you're just parrotting their idiotic line.
Please watch your tone. If you find offense at the allegedly condescending tone in which the FSF says "misunderstanding", how do you expect me and the FSF to take this?
It is also odd that you claim the right to be entitled to your own opinion and reading about stuff, while denying myself the same right. Please don't do that. I have a mind of my own, and the fact that I reach similar conclusions doesn't make me a parrot. Even more so when I actually have some influence on those conclusions.
Quoted 13 lines
Quoted 3 lines
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.
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.
Quoted 4 lines
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 think so.
You are definitely confused. You're talking about the legal terms, while I'm talking about the spirit. The legal terms tried to reflect the spirit as best as they could, but they left some holes. Some people found them and started exploiting them.
Sure, if you want to leave those holes unplugged in your code, that's your decision. I don't doubt that the GPLv2 legal terms fit the bill for you. I think GPLv3 would do even better in this regard. But none of this is about the spirit of the GPL. Claiming GPLv3 changes the spirit is totally missing the point of what the spirit amounts to. The spirit is described in the preamble, it's not the legal terms.
Quoted 2 lines
The beauty of the GPLv2 is exactly that it's a "tit-for-tat" license,
Ok, let's explore this argument. In what sense is it tit-for-tat? What is tit-for-tat about it? What is the payback an author who releases software under the GPL can legitimately expect to get?
Quoted 5 lines
Quoted 2 lines
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.
That's a lot of bullshit. You are apparently the grand poobah, and can decide _which_ freedoms and for _what_ others' that matter.
The freedoms I'm talking about are very clearly described in the spirit (preamble) of the license you chose for your project. Go look at the preamble one more time, "grand poobah" ;-)
[...] 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.
Signature · 4 lines
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}Quoted 5 lines · Alexandre Oliva
Let me see if I got this right. There was a section entitled "3. Digital Restrictions Management" in GPLv3dd1. Are you saying that, when people complained about the DRM clause, they actually meant the provisions in "1. Source Code"
Yes. I said that multiple times. It was obvious. But people didn't listen.
It's now in Section 7, or whatever.
The section 3 never mattered.
Linus
Quoted 3 lines · Linus Torvalds
The beauty of the GPLv2 is exactly that it's a "tit-for-tat" license, and you can use it without having to drink the kool-aid.
One could even add that "tit-for-tat" appears to be the best strategy in game theory for continuous runs of the prisoners dilemma. At times I wonder why game theory isn't taught in schools yet - it might shorten discussions like these.
Jörn
Signature · 2 lines
All art is but imitation of nature. -- Lucius Annaeus Seneca
Quoted 10 lines · On Wednesday 13 June 2007 19:35:41 Jörn Engel wrote:
Quoted 2 lines · Linus Torvalds
The beauty of the GPLv2 is exactly that it's a "tit-for-tat" license, and you can use it without having to drink the kool-aid.
One could even add that "tit-for-tat" appears to be the best strategy in game theory for continuous runs of the prisoners dilemma. At times I wonder why game theory isn't taught in schools yet - it might shorten discussions like these.
Jörn
With the sheer amount of sheeple[1] in the world (and on this list), I doubt anything could make these discussions any shorter.
(While I hate thinking that sheeple are on this list, it is an unavoidable fact. (I had hoped I wouldn't find any sheeple here, as my favorite theory is that they are all "Fundamentalist Christians" like the "Creationist" fools))
DRH
1: Sheeple (n): People that act like sheep - ie: they cannot think or form opinions for themselves and always look to someone else for their thoughts and parrot the opinions of some "trusted" figure.
Signature · 1 line
Dialup is like pissing through a pipette. Slow and excruciatingly painful.
Quoted 3 lines · On Wednesday June 13, [email protected] wrote:
1: Sheeple (n): People that act like sheep - ie: they cannot think or form opinions for themselves and always look to someone else for their thoughts and parrot the opinions of some "trusted" figure.
I recommend that you avoid definitions like this. Using them simply makes you appear to have a very poor understanding of your fellow humans.
"cannot" and "always" are absolutes that never apply (well, hardly ever).
In my experience, most people do think for themselves and do form opinions about areas where they have an interest/ability, and tend to follow trusted others in areas where they have less interest or ability.
The problem that I think you see is particularly the "parrot the opinions" bit. When a person tries to argue a case based largely on the opinion of someone else with little personal understanding, they
are in very risky territory. This is akin to 'fundamentalism' that
you also mentioned in your post.
Accusing people of arguing opinions that are not their own may well be appropriate. Accusing them of not be able to think is not.
NeilBrown
Quoted 11 lines · On Wednesday 13 June 2007 20:14:47 Neil Brown wrote:
Quoted 3 lines · On Wednesday June 13, [email protected] wrote:
1: Sheeple (n): People that act like sheep - ie: they cannot think or form opinions for themselves and always look to someone else for their thoughts and parrot the opinions of some "trusted" figure.
I recommend that you avoid definitions like this. Using them simply makes you appear to have a very poor understanding of your fellow humans.
"cannot" and "always" are absolutes that never apply (well, hardly ever).
In this case I gave a very hasty definition of the term, though it does fit very well.
Quoted 4 lines
In my experience, most people do think for themselves and do form opinions about areas where they have an interest/ability, and tend to follow trusted others in areas where they have less interest or ability.
Agreed. But those *aren't* "Sheeple".
Quoted 8 lines
The problem that I think you see is particularly the "parrot the opinions" bit. When a person tries to argue a case based largely on the opinion of someone else with little personal understanding, they
are in very risky territory. This is akin to 'fundamentalism' that
you also mentioned in your post.Accusing people of arguing opinions that are not their own may well be appropriate. Accusing them of not be able to think is not.
I never accused any specific person of either. I was making an observation about the general nature of humanity as I've observed it. A better definition of "Sheeple" would be "Someone who continually parrots some "trusted" figures opinion even when presented with proof that that opinion is in error. They also, generally, do not "think deeply" on any topic."
And while I won't apologize to anyone that may have felt the "Sheeple" remark was directed at them, I will apologize for not taking more time to choose a better definition for the term originally.
DRH
Quoted 6 lines
NeilBrown
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Signature · 1 line
Dialup is like pissing through a pipette. Slow and excruciatingly painful.
Quoted 2 lines · Neil Brown
Accusing people of arguing opinions that are not their own may well be appropriate. Accusing them of not be able to think is not.
I agree with the latter, but the former is seldom appropriate. Any accusation ought to be provable, and it's nearly impossible to prove that an opinion held by someone is not his own. People quite often arrive at similar opinions independently.
Signature · 4 lines
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}Quoted 7 lines · Jörn Engel
Quoted 3 lines · Linus Torvalds
The beauty of the GPLv2 is exactly that it's a "tit-for-tat" license, and you can use it without having to drink the kool-aid.
One could even add that "tit-for-tat" appears to be the best strategy in game theory for continuous runs of the prisoners dilemma.
It is, indeed.
Now the remaining piece of the proof is to show that the GPLv2 is tit-for-tat.
Signature · 4 lines
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}[ Gmail did horrible things to the original post by giving it a base64 content transfer encoding, so majordomo@vger dropped it. It's just an off-topic digression, but I cared enough to resend anyway, fwiw. ]
Quoted 7 lines · Jörn Engel
Quoted 3 lines · Linus Torvalds
The beauty of the GPLv2 is exactly that it's a "tit-for-tat" license, and you can use it without having to drink the kool-aid.
One could even add that "tit-for-tat" appears to be the best strategy in game theory for continuous runs of the prisoners dilemma.
Tit-for-tat is the best *deterministic* strategy when playing iterated prisoner's dilemma. But note that "deterministic" and "rational" are not adjectives that go well with "humans", and most real-world (social) situations are noisy environments -- miscommunication and misunderstandings are the usual noise. A double-D noise perceived by any player would throw a tit-for-tat-playing couple into a perennial spiral of D's, for example, which is clearly not a Pareto-efficient solution for either.
Quoted 3 lines
At times I
wonder why game theory isn't taught in schools yet - it might shorten discussions like these.
Yes, and no.
Yes - for teaching game theory (and its social relevance) in schools; and add behavioral economics to this list :-)
No - it doesn't shorten discussions, however. And it shouldn't either. Human / social situations are complex, Jörn; tit-for-tat can win computer contests, for example, but it's not a behaviour one person would find as entirely agreeable in another.
Satyam
Quoted 4 lines · Tarkan Erimer
And maybe another questions should be : How long a copyright owner can hold the copyright, if died or lost for sometime ? if died, the copyright still should be valid or not ? If lost, what the law orders at this point for copyright holding ?
In most countries, copyright lasts for 70 years after the death of the author. Either way, it is most definitely valid.
-hpa
Hi Neil,
Quoted 8 lines · Neil Brown
Quoted 5 lines · On Saturday June 9, [email protected] wrote:
Hi,
As we know the forthcoming GPL V3 will be not compatible with the GPL V2 and Linux Kernel is GPL V2 only.
So, another point is, which is previously mentioned by Linus and others, that if it is decided to upgrade the Linux Kernel's License to GPL V3, it is needed the permission of all the maintainers permission who contributed to the Linux Kernel and there are a lot of lost or dead maintainers. Which makes it impossible to get all the maintainers' permission.
Quoted 8 lines
You don't need the permission of maintainers. You need the permission of copyright owners. The two groups overlap, but are not the same. Dead people cannot own anything, even copyright. Their estate probably can. I don't think it is theoretically impossible to get everyone's permission, though it may be quite close to practically impossible.
So, does it mean we can change the license of the dead people's code ?
Quoted 25 lines · On Sunday June 10, [email protected] wrote:
Hi Neil,
Quoted 14 lines · Neil Brown
Quoted 11 lines · On Saturday June 9, [email protected] wrote:
Hi,
As we know the forthcoming GPL V3 will be not compatible with the GPL V2 and Linux Kernel is GPL V2 only.
So, another point is, which is previously mentioned by Linus and others, that if it is decided to upgrade the Linux Kernel's License to GPL V3, it is needed the permission of all the maintainers permission who contributed to the Linux Kernel and there are a lot of lost or dead maintainers. Which makes it impossible to get all the maintainers' permission.Quoted 8 lines
You don't need the permission of maintainers. You need the permission of copyright owners. The two groups overlap, but are not the same. Dead people cannot own anything, even copyright. Their estate probably can. I don't think it is theoretically impossible to get everyone's permission, though it may be quite close to practically impossible.
Quoted 2 lines
So, does it mean we can change the license of the dead people's code ?
I presume the heirs of the dead people could change the license. And if they have no heir, then there is no-one to sue for breach of copyright, so I assume the copyright lapses.
And I wouldn't be surprised if there were some legal precedent that allowed for some process whereby we could make a "best effort" to contact copyright holders (including registered paper letters and entries in the "Public Notices" section of major newspapers) and if no-one stepped forward to claim copyright in a reasonable period of time we could assume that the copyright had lapsed. But you would need to ask a lawyer, and it would be different in different countries.
But I think this is largely academic. You only need a fairly small number of fairly significant contributors to say "no" and the rest of the process would be pointless. And at last count, the number of kernel people who were not keen on GPLv3 was fairly high. Of course no-one knows for certain yet what the final GPLv3 will be, and maybe lots of people would change their mind when it comes out.
There would certainly be value in a straw-pole once GPLv3 was out and had been discussed for a while - to see if a license change to GPLv3 would be accepted by a majority of current developers. Doing that would at least provide a clear statistic to point people at.
NeilBrown
Quoted 3 lines · Neil Brown
I presume the heirs of the dead people could change the license. And if they have no heir, then there is no-one to sue for breach of copyright, so I assume the copyright lapses.
In most of the law systems out there the copyright stays valid for 70 years (or so) after the holder's death.
Signature · 1 line
Jiri Kosina
Quoted 7 lines · On Sunday 10 June 2007 08:45:41 Jiri Kosina wrote:
Quoted 3 lines · Neil Brown
I presume the heirs of the dead people could change the license. And if they have no heir, then there is no-one to sue for breach of copyright, so I assume the copyright lapses.
In most of the law systems out there the copyright stays valid for 70 years (or so) after the holder's death.
I'm almost certain that it is the same in the US, not the death+90 previously stated. (I've read the copyright laws a number of times to deal with some involved conversations) In some of the writings tied the change that made it death+70 its stated that said change was made to "bring US laws in line with Europe and most of the rest of the world" (paraphrase - I didn't bother going and digging out the page again). It's a relatively common belief - and, IIRC, was even brought before the US Supreme Court - that the copyrights length was changed to give Disney longer protection, if just because a lot of Disney's copyrights were going to expire and the change was applied retroactively. (And, unsurprisingly, the suit was shot down - 70 years is still a "limited" period. However, IIRC, there was some noted concern by the Supreme Court that the US Congress would exploit the legal loophole and just keep extending the copyright period retroactively to make it, effectively, never-ending. That scenario, while not *technically* in violation of the language of the US Constitution (which grants the US Congress the power to set the length of copyrights) would violate the spirit of it)
DRH
Quoted 2 lines
In most of the law systems out there the copyright stays valid for 70 years (or so) after the holder's death.
[I don't want to appear picky but IMO it is important to be as precise as one could possibly be, therefor...]
That is not quite correct. At least for the german law system and presumably for most if not all similar systems the corrrect stmts is:
In most of the law systems out there the copyright stays valid for 70 years (or so) after initial publication of the copyrighted work. For work whose publication date can't be determined (*) (e.g. unpublished work) the creators death is assumed to be the publication date for the sake of determining the aforementioned 70 years protection period.
(*) "can't be determined" does not apply for published work with an insecure initial publication date. In those cases the earliest documented publication (i.e. prooveable in court) is considered the initial publication .date
Note that the original phrase could last indefinitely by simply moving the copyright from person to person.
Best,
Michael
Signature · 5 lines
Technosis GmbH, Geschäftsführer: Michael Gerdau, Tobias Dittmar Sitz Hamburg; HRB 89145 Amtsgericht Hamburg Vote against SPAM - see http://www.politik-digital.de/spam/ Michael Gerdau email: [email protected] GPG-keys available on request or at public keyserver
Quoted 6 lines · Michael Gerdau
Quoted 2 lines
In most of the law systems out there the copyright stays valid for 70 years (or so) after the holder's death.
[I don't want to appear picky but IMO it is important to be as precise as one could possibly be, therefor...]
Quoted 6 lines
That is not quite correct. At least for the german law system and presumably for most if not all similar systems the corrrect stmts is:
In most of the law systems out there the copyright stays valid for 70 years (or so) after initial publication of the copyrighted work.
...
Please read the German "Gesetz über Urheberrecht und verwandte Schutzrechte" before making completely false claims - in Germany it's 70 years after the death of the author.
Quoted 2 lines
Best,
Michael
cu
Adrian
Signature · 4 lines
"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 SeedQuoted 3 lines
Please read the German "Gesetz über Urheberrecht und verwandte Schutzrechte" before making completely false claims - in Germany it's 70 years after the death of the author.
Oups, I had been mixing up named published works and anonymously published works.
My previous claim is valid for anonymously published work only and thus holds no relevance for the situation discussed. I apologize for having raised a false claim.
The copyright for works with a known is valid for 70 years after the authors death.
Sorry,
Michael
Signature · 5 lines
Technosis GmbH, Geschäftsführer: Michael Gerdau, Tobias Dittmar Sitz Hamburg; HRB 89145 Amtsgericht Hamburg Vote against SPAM - see http://www.politik-digital.de/spam/ Michael Gerdau email: [email protected] GPG-keys available on request or at public keyserver
Quoted 6 lines
But I think this is largely academic. You only need a fairly small number of fairly significant contributors to say "no" and the rest of the process would be pointless. And at last count, the number of kernel people who were not keen on GPLv3 was fairly high. Of course no-one knows for certain yet what the final GPLv3 will be, and maybe lots of people would change their mind when it comes out.
You can take a fair bet someone will say no, or much more likely they or whoever inherited their copyright will say $50,000
Quoted 9 lines · On Sunday 10 June 2007 09:40:23 Alan Cox wrote:
Quoted 6 lines
But I think this is largely academic. You only need a fairly small number of fairly significant contributors to say "no" and the rest of the process would be pointless. And at last count, the number of kernel people who were not keen on GPLv3 was fairly high. Of course no-one knows for certain yet what the final GPLv3 will be, and maybe lots of people would change their mind when it comes out.
You can take a fair bet someone will say no, or much more likely they or whoever inherited their copyright will say $50,000
I seeds shades of Merkey there :P
Seriously, though, this was all settled a long time ago. Linus said "While individual parts of the kernel *could* be licensed [with another license] the kernel as a whole is strictly GPLv2" (I've tried to get it right, but my memory isn't as good as it used to be when it comes to useful quotes like that)
DRH
Quoted 10 lines · Neil Brown
...
And I wouldn't be surprised if there were some legal precedent that allowed for some process whereby we could make a "best effort" to contact copyright holders (including registered paper letters and entries in the "Public Notices" section of major newspapers) and if no-one stepped forward to claim copyright in a reasonable period of time we could assume that the copyright had lapsed. But you would need to ask a lawyer, and it would be different in different countries.
...
A legal precedent valid in all jurisdictions?
Harald suceessfully takes legal actions against people violating his copyright on the Linux kernel under the terms of the GPL in Germany at German courts based on German laws.
If someone finds any legal precedent in Finland or the USA or Russia that some copyright would have lapsed for some reason, would this have any legal effect in Germany?
Quoted 1 line
NeilBrown
cu
Adrian
Signature · 4 lines
"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[legal precedence to force waiving copyright>
Quoted 9 lines
A legal precedent valid in all jurisdictions?
Harald suceessfully takes legal actions against people violating his copyright on the Linux kernel under the terms of the GPL in Germany at German courts based on German laws.
If someone finds any legal precedent in Finland or the USA or Russia that some copyright would have lapsed for some reason, would this have any legal effect in Germany?
That is difficult but AFAIK it work in the "increase protection" direction but not the other way around, at least in germany.
For example even though the copyright on Micky Mouse is no longer valid by german law it is still enforcible under german law because the german jurisdiction does acknowledge the (recently extended) longer such period in the US.
Apart from that I'm sure you'd find some state "legally" waiving a copyright on whatever. If that would start an avalanche of similar terminations everywhere else I'd be utterly surprised.
Best,
Michael
Signature · 5 lines
Technosis GmbH, Geschäftsführer: Michael Gerdau, Tobias Dittmar Sitz Hamburg; HRB 89145 Amtsgericht Hamburg Vote against SPAM - see http://www.politik-digital.de/spam/ Michael Gerdau email: [email protected] GPG-keys available on request or at public keyserver
Quoted 4 lines · Neil Brown
I presume the heirs of the dead people could change the license. And if they have no heir, then there is no-one to sue for breach of copyright, so I assume the copyright lapses.
Quoted 8 lines
And I wouldn't be surprised if there were some legal precedent that allowed for some process whereby we could make a "best effort" to contact copyright holders (including registered paper letters and entries in the "Public Notices" section of major newspapers) and if no-one stepped forward to claim copyright in a reasonable period of time we could assume that the copyright had lapsed. But you would need to ask a lawyer, and it would be different in different countries.
I've done some research on this and from what I can tell you are correct. There is some sort of "Due Diligence" law that you have to satisfy that is slightly different from country to country.
Quoted 1 line
From what I can tell the process is similar to changing your name, an ad
in the paper would be o.k. in most places. Since the intent of the copyright holder to make their work free is clear, its pretty clear cut.
It was hard to find specifics, because not many people worry about it if they know the copyright holder to be existentially challenged. I didn't find many cases of people even bothering with the formality.
I say existentially challenged because you don't need to be dead to vanish. I also read it as ".. a person of questionable existence." as well as "A null (or moot) party".
Quoted 6 lines
But I think this is largely academic. You only need a fairly small number of fairly significant contributors to say "no" and the rest of the process would be pointless. And at last count, the number of kernel people who were not keen on GPLv3 was fairly high. Of course no-one knows for certain yet what the final GPLv3 will be, and maybe lots of people would change their mind when it comes out.
I think its good that people have their own viewpoints instead of just watching the Linus - RMS tennis match as the points of the license get debated and assuming the views of whoever 'wins' in their eyes. Neither person is going to be around forever.
Quoted 4 lines
There would certainly be value in a straw-pole once GPLv3 was out and had been discussed for a while - to see if a license change to GPLv3 would be accepted by a majority of current developers. Doing that would at least provide a clear statistic to point people at.
Unless you are distributing the Linux kernel in a whole or parts of it in some way you don't even HAVE to accept the terms of the GPL2, or GPL3 for that matter. This doesn't really concern me as an end user, but it should interest me (and does).
I'm more worried about someone finding a way to leverage existing or future patents against users of the Linux kernel, or people who provide the use of a computer using the Linux kernel as a service.
I'm not *overly* concerned about that. Anyone trying to get people to pay for the right to use the Linux kernel would put themselves in clear and certain danger of becoming ' a person of questionable existence ' themselves.
Best,
--Tim
Quoted 3 lines · Tarkan Erimer
So, does it mean we can change the license of the dead people's code ?
Please realize that one doesn't need to be dead to become uncommunicative incapacitated or vanish. The only need to be somewhere other than where they were without updating anyone.
Here is a very humorous, but sort of scary theoretical :
"Linus was so disturbed by the code in a submitted patch that he had a nervous breakdown and spent the next 30 years in a padded room. Unfortunately, no provisions were left to determine what happens to his copyrights should he become incapacitated."
That is a LOT different than
"Linus passed away at age 397 today, and left an appropriate will so his work remains useful for those enjoying it."
If you are a substantial contributor, address this in a will should you die and limited power of attorney to someone that you trust should you fall off the face of the planet. Just like you would see to it that your assets were properly dispersed if something happened to you.
If your unfortunate enough to not have anyone in your life that you can trust, you can grant power of attorney to any reputable charitable organization that you feel mirrors your beliefs and ideals.
Prevention is so much better than a cure and quite cheap.
Best,
--Tim
Hi!
Quoted 7 lines
Quoted 2 lines
So, does it mean we can change the license of the dead people's code ?
Please realize that one doesn't need to be dead to become uncommunicative incapacitated or vanish. The only need to be somewhere other than where they were without updating anyone.
Quoted 6 lines
Here is a very humorous, but sort of scary theoretical :
"Linus was so disturbed by the code in a submitted patch that he had a nervous breakdown and spent the next 30 years in a padded room. Unfortunately, no provisions were left to determine what happens to his copyrights should he become incapacitated."
License is only promise not to sue. If Linus is safely in a padded room, I'll happily relicense Linux under GPLv17, knowing he is not going to sue me :-).
(And actually, if he's in a padded room, there's probably someone who can act in his name. At least czech law works like that.)
Pavel
Signature · 2 lines
(english) http://www.livejournal.com/~pavelmachek (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
Quoted 5 lines
But; if the Linux kernel should Dual-Licensed (GPL V2 and GPL V3), it will allow us the both worlds' fruits like code exchanging from other Open Source Projects (OpenSolaris etc.) that is compatible with GPL V3 and not with GPL V2 and of course the opposite is applicable,too.
That is a misleading claim. While being dual-licensed would make it either for other projects to adopt Linux code, it would have three downsides:
1) If Linux code were adopted into other projects that were not dual-licensed, changes could not be imported back into Linux unless the changes were dual-licensed which is not likely when the contributions are made to a project that's not dual-licensed.
2) Linux could no longer take code from other projects that are GPL v2 licensed unless it could obtain them under a dual license.
And, last and probably most serious:
3) Linux derivatives could be available with just a GPL v3 license and no GPL v2. license if the derivers wanted things that way.
DS
Hi David,
Quoted 14 lines · David Schwartz
Quoted 5 lines
But; if the Linux kernel should Dual-Licensed (GPL V2 and GPL V3), it will allow us the both worlds' fruits like code exchanging from other Open Source Projects (OpenSolaris etc.) that is compatible with GPL V3 and not with GPL V2 and of course the opposite is applicable,too.
That is a misleading claim. While being dual-licensed would make it either for other projects to adopt Linux code, it would have three downsides:
1) If Linux code were adopted into other projects that were not dual-licensed, changes could not be imported back into Linux unless the changes were dual-licensed which is not likely when the contributions are made to a project that's not dual-licensed.
Quoted 9 lines
2) Linux could no longer take code from other projects that are GPL v2 licensed unless it could obtain them under a dual license.
And, last and probably most serious:
3) Linux derivatives could be available with just a GPL v3 license and no GPL v2. license if the derivers wanted things that way.
Thanks for the corrections ;-) The whole picture is more clear now for me :-)
BTW,I found a really interesting blog entry about which code in Linux Kernel is using which version of GPL :
http://6thsenseless.blogspot.com/2007/02/how-much-linux-kernel-code-is-gpl-2.html
The work done on a Linux 2.6.20. The result is quite interesting. Because almost half (Around %60 of the code licensed under "GPLv2 Only" and the rest is "GPLv2 or above","GPL-Version not specified,others that have not stated which and what version of License has been used) of the code is "GPLv2 or above" licensed. And also stated in the article that some of the codes should be "Dual Licensed" not the whole Linux kernel needed to be "Dual Licensed". So,if it is really like this, maybe we can make,for example: "File system related Codes", "Dual Licensed" and it will allow us to port ZFS from OpenSolaris requested by a lot of people or other things maybe ?
Quoted 6 lines · Tarkan Erimer
Thanks for the corrections ;-) The whole picture is more clear now for me :-) BTW,I found a really interesting blog entry about which code in Linux Kernel is using which version of GPL :
http://6thsenseless.blogspot.com/2007/02/how-much-linux-kernel-code-is-gpl-2.html
You've got to take MODULE_LICENSE() into account. There is
MODULE_LICENSE("GPL");
MODULE_LICENSE("GPL v2");
MODULE_LICENSE("GPL and additional rights");
MODULE_LICENSE("Dual BSD/GPL");
MODULE_LICENSE("Dual MIT/GPL");
MODULE_LICENSE("Dual MPL/GPL");I think it's time to set things right, making
* MODULE_LICENSE the authoritative place for the license (also makes it easier to grep for)
* sync up license into MODULE_LICENSE
Jan
Quoted 6 lines · Tarkan Erimer
BTW,I found a really interesting blog entry about which code in Linux Kernel is using which version of GPL :
http://6thsenseless.blogspot.com/2007/02/how-much-linux-kernel-code-is-gpl-2.html
Quoted 10 lines
The work done on a Linux 2.6.20. The result is quite interesting. Because almost half (Around %60 of the code licensed under "GPLv2 Only" and the rest is "GPLv2 or above","GPL-Version not specified,others that have not stated which and what version of License has been used) of the code is "GPLv2 or above" licensed. And also stated in the article that some of the codes should be "Dual Licensed" not the whole Linux kernel needed to be "Dual Licensed". So,if it is really like this, maybe we can make,for example: "File system related Codes", "Dual Licensed" and it will allow us to port ZFS from OpenSolaris requested by a lot of people or other things maybe ?
Once code obtained under the GPLv2 only licence that the kernel is released under is modified and submitted back to Linus for inclusion, that code would become GPLv2 only - only the original would be BSD/LGPL/GPLv2+ and only separate changes to the original could continue to be available under dual licence. Since most files will have been modified at various stages in Linux's development when major internal changes occur, surely practically everything is now GPLv2 only?
Quoted 1 line
So, does it mean we can change the license of the dead people's code ?
If you can contact whoever currently owns the copyright, they can release it under another licence... however this is no good because the derivative work would be GPLv2 only. Perhaps if you got *everyone* at all stages of development (including code that has been removed if existing code is a derivative work of it) to agree - then it could work.
It only takes one person's code, uncooperative or not contactable, to prevent a change to the licence, so there's not much point in trying unless you intend to start replacing such code.
Quoted 8 lines · Jan Engelhardt
You've got to take
MODULE_LICENSE()into account. There isMODULE_LICENSE("GPL"); MODULE_LICENSE("GPL v2"); MODULE_LICENSE("GPL and additional rights"); MODULE_LICENSE("Dual BSD/GPL"); MODULE_LICENSE("Dual MIT/GPL"); MODULE_LICENSE("Dual MPL/GPL");
Surely that doesn't work since the entire Linux kernel is (and can only be) released as GPLv2? Wouldn't anyone making changes to those files need to obtain a copy under the other licence and explicitly release it under both licenses in order to maintain that?
Signature · 1 line
Simon Arlott
Quoted 15 lines · Simon Arlott
Quoted 5 lines · Tarkan Erimer
BTW,I found a really interesting blog entry about which code in Linux Kernel is using which version of GPL :
http://6thsenseless.blogspot.com/2007/02/how-much-linux-kernel-code-is-gpl-2.html
Quoted 8 lines
You've got to take
MODULE_LICENSE()into account. There isMODULE_LICENSE("GPL"); MODULE_LICENSE("GPL v2"); MODULE_LICENSE("GPL and additional rights"); MODULE_LICENSE("Dual BSD/GPL"); MODULE_LICENSE("Dual MIT/GPL"); MODULE_LICENSE("Dual MPL/GPL");
Quoted 4 lines
Surely that doesn't work since the entire Linux kernel is (and can only be) released as GPLv2? Wouldn't anyone making changes to those files need to obtain a copy under the other licence and explicitly release it under both licenses in order to maintain that?
http://www.eweek.com/article2/0,1759,1915720,00.asp
has the answer. Quoting Linus:
"If you want to license a program under any later version of the GPL, you have to state so explicitly. Linux never did."
Hence, unless there's a "GPL 2 or later", all the "unspecified GPL" files are GPL2 only.
Jan
Quoted 10 lines
http://www.eweek.com/article2/0,1759,1915720,00.asp
has the answer. Quoting Linus:"If you want to license a program under any later version of the GPL, you have
to state so explicitly. Linux never did."Hence, unless there's a "GPL 2 or later", all the "unspecified GPL" files are GPL2 only.
The GPL states the default position:
"If the Program does not specify a version number of this License, you may choose any version ever published by the Free Software Foundation."
Leaving the question of whether Linus's comment at the top of the license changes the default:
"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."
So we have dueling defaults. The GPL says the default is any version. Linus' statement at the top of the GPL says the default is v2 only. It's not clear, at least to me, that there is any clear reason why one should win out over the other.
DS
Quoted 18 lines · On Sunday 10 June 2007 16:42:35 David Schwartz wrote:
Quoted 9 lines
http://www.eweek.com/article2/0,1759,1915720,00.asp
has the answer. Quoting Linus:"If you want to license a program under any later version of the GPL, you have
to state so explicitly. Linux never did."Hence, unless there's a "GPL 2 or later", all the "unspecified GPL" files are GPL2 only.
The GPL states the default position:
"If the Program does not specify a version number of this License, you may choose any version ever published by the Free Software Foundation."
Leaving the question of whether Linus's comment at the top of the license changes the default:
Quoted 8 lines
"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."
So we have dueling defaults. The GPL says the default is any version. Linus' statement at the top of the GPL says the default is v2 only. It's not clear, at least to me, that there is any clear reason why one should win out over the other.
Except that Linux included a copy of the license it was distributed under in the tarball, which was GPLv2. There's a case to be made that this DOES count as "selection of the license version". It's hard to get more explicit than including the complete license text.
On top of that, Linus clarified his position back in 2000:
http://www.uwsg.iu.edu/hypermail/linux/kernel/0009.1/0096.html
This means that not only are his on contributions so licensed, but there's a strong case to be made that everything he merged since then is explicitly GPLv2 only when sourced from the Linux kernel. As maintainer there's a good argument that he has a compliation copyright on the kernels he releases. He was certainly accepting patches under GPLv2 only, and merging them into a GPLv2 only work. Sourcing ANY code from that work and declaring it non-GPLv2 is really fishy; much better to find a clean upstream source, such as the original authors.
Of course the FSF would very, very much like this not to have been the case, and they've been trying to wish it away ever since...
Rob