Re: Firefox/Thunderbird trademarks: a proposal
* Alexander Sack ([EMAIL PROTECTED]) wrote: [snip] > Mike appears to be subscribed to this list ... Eric, will you jump in if > you have any objections, etc.? Alright, jumping in... I'm not subscribed to the list and didn't really have any idea this was generating so much traffic. I went back and read all the threads (I think) and I hope I have a good understanding of what's going on. Unfortunately, I have a number of objections. Before I get to them, one of the interesting things pointed out in one of the threads is that the Trademark License might be more onerous then what trademark law (at least in the US) allows. Now, they're your trademarks, and I have every intention of respecting your wishes when it comes to using them. It may not paint the Mozilla Foundation in a good light if you are indeed trying to impose more restrictions than trademarks allow. Now then, I personally will not accept any deal that is Debian specific. Whether or not this is actually against DFSG #8 or not is beside the point, I don't want to play if it's only because we're the popular kid. This problem goes beyond Debian. Other distributions are not going to be able to use the trademark license as written. Are you going to cut deals with Fedora, Slackware and Gentoo as well? They're probably more likely to want to deal than us, but does that mean the little distros don't get to use the trademarks because they're not worth striking a deal with? The reason trademarks like the TeX one are more palatable is that there's a set of technical requirements and as long as you meet them you get to use the mark. You've already sort of done that with the Community Edition licensing, but you haven't gone far enough for what a distro is going to need: * Applying security fixes * Back porting bug fixes * Making distro specific integration changes and possibly a few others, that's just off the top of my head. I'm sure you'll have some more of your own. You're the ones who get to make the rules and if they're not too much of a burden then we'll play and won't have to go with iceweasel. But anyone who wants to follow the rules should be allowed to play, not just the big kids (aka Debian). Ok, I've just about exhausted the schoolyard metaphor, but I hope I made my point. As it stands now I'm not going to take any action until: * The Mozilla Foundation tells me to stop using the marks or, * The Trademark License is finalized (right now it still says version 0.7 draft, so I don't consider it in any way binding). PS please CC me on any replies, I'm not subscribed to debian-legal. -- Eric Dorland <[EMAIL PROTECTED]> ICQ: #61138586, Jabber: [EMAIL PROTECTED] 1024D/16D970C6 097C 4861 9934 27A0 8E1C 2B0A 61E9 8ECF 16D9 70C6 -BEGIN GEEK CODE BLOCK- Version: 3.12 GCS d- s++: a-- C+++ UL+++ P++ L++ E++ W++ N+ o K- w+ O? M++ V-- PS+ PE Y+ PGP++ t++ 5++ X+ R tv++ b+++ DI+ D+ G e h! r- y+ --END GEEK CODE BLOCK-- signature.asc Description: Digital signature
Re: Illustrating JVM bindings
Michael K. Edwards wrote: > On Fri, 14 Jan 2005 22:12:56 +, Andrew Suffield > <[EMAIL PROTECTED]> wrote: > [snip] > >>Some of those python scripts may be derivatives of GNU readline. Most >>are probably not. Those that are must be licensed under the GPL. The >>rest do not have to be. All this interpreter crud in between is >>*irrelevant*. If the same program written in C would be a derivative >>then it's still a derivative even when you insert an interpreter in >>the middle. > > I agree completely with Andrew here. I also think that "linking crud" > is every bit as irrelevant as "interpreter crud". I agree with this statement as well, but not in the way you have argued. "Linking crud" is irrelevant in that it does not in and of itself define whether something is derivative or not (leaving aside any issues of inlines and other code included in a compiled binary). Many cases of "A links to B" have A as a derivative work of B, while others do not, and in neither case is the mere fact that A links against B a relevant factor in the decision; it is simply a quick shortcut to guess that something may be a derived work. For example, it is quite arguable that a program written today against the readline API is not a derived work of either readline or editline, because of the existence of both, assuming it was intended to be usable on both. Similarly, a program written in ANSI C is certainly not a derivative of glibc or any other C library. On the other hand, a program written againt a unique GPLed library, with no other implementation, is almost certainly a derivative work of that library: you are combining two expressive and copyrightable works into a new whole which is greater than either. - Josh Triplett signature.asc Description: OpenPGP digital signature
Re: Eclipse 3.0 Running ILLEGALY on Kaffe
Brian Thomas Sniffen wrote: > Michael Poole <[EMAIL PROTECTED]> writes: >>Brian Thomas Sniffen writes: >>>But why do you think RMS is so keen to have a working, FSF-owned Hurd? >> >>NIH syndrome. What is your explanation? > > I'm sure he'd like to make a system with guaranteed only free > programs. As would many Free Software users and developers, but I seriously doubt that goal has anything to do with the HURD. Apart from the idea that that would run counter to the logic proposed in the decision to make glibc LGPLed, there is the issue of whether a program is a derived work or not. Linus has stated on several occasions that the statement he has made regarding the user/kernel boundary and the GPL was simply a clarification regarding "derived works": a program written to standard UNIX interfaces is clearly not a derivative of Linux, HURD, or any other particular UNIX system. - Josh Triplett signature.asc Description: OpenPGP digital signature
Re: Eclipse 3.0 Running ILLEGALY on Kaffe
Walter Landry writes: > What if there was a package wget++ that communicated with openssl > entirely through system() or exec() calls? It would construct > appropriate input and parse openssl's output. Would that constitute > linking? It ends up using all of the same code as the directly linked > version. > > If it is not linking, why couldn't you do this with all GPL'd > libraries? You could write a GPL'd wrapper around a library, and just > use the wrapper with exec(). > > In essence, why does using exec() suddenly break the chain, while a > linker or classloader does not? exec() does not break the chain of derivation. My understanding is that the Objective C front-end for gcc was released as free software because the (corporate) author's attempted exec()-based abstraction around the gcc back-end did not break the chain. Rather than fight it out in court, that company decided to make it free. >From the law's perspective, the intent and result (in terms of copyrightable expression) are far more important than the mechanism through which software works. That is why rules about static linking versus dynamic linking versus I/O streams versus other IPC can only be rules of thumb. They can hint at the type of relationship, but do not determine it. Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Eclipse 3.0 Running ILLEGALY on Kaffe
On Wed, Jan 19, 2005 at 10:30:30PM -0500, Walter Landry wrote: > > If so, what is the difference is between Y=Kaffe and Y=Linux? Linux > > exempts syscall-using clients from being directly covered by the GPL, > > That is the difference. Linux has an exemption and Kaffe does not. ... as far as I can tell, Linux has no special exemptions, because nobody has any authority to grant them. I agree with Raul: historically, the belief has been that the GPL allows this, though I'm not sure I could tie that belief to the GPL. Unless there really is an exemption--how?--the FSF seems to agree: http://www.gnu.org/licenses/gpl-faq.html#TOCPortProgramToGL unless they really failed to consider the kernel at all when writing that FAQ. -- Glenn Maynard -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Eclipse 3.0 Running ILLEGALY on Kaffe
Brett Parker <[EMAIL PROTECTED]> wrote: > On Wed, Jan 19, 2005 at 12:52:29PM -0500, Brian Thomas Sniffen wrote: > > Michael Poole <[EMAIL PROTECTED]> writes: > > > > > As has been settled on this list, Eclipse is not a derivative of Kaffe > > > and does not contain any copyright-protected portion of Kaffe. It is > > > possible to claim that "Eclipse+Kaffe" is a work based on Kaffe, but > > > by the same argument, "Debian" is a work based on Kaffe, and the > > > rational interpretation is that both cases are mere aggregation. > > > > It seems to me that "mere aggregation" must be the smallest idea that > > is still aggregation. For example, Emacs and Vim are merely > > aggregated in Debian. wget and openssl are not merely aggregated, > > because there's more going on there. It's not necessary to look in > > great detail at what *is* going on there -- it's enough to say that > > there is more there, so it's not merely aggregation. It's aggregation > > and something else. > > wget and openssl are linked, openssl is a build depend of wget, it is > very much required to compile it. So, yes, it is not mere aggregation. What if there was a package wget++ that communicated with openssl entirely through system() or exec() calls? It would construct appropriate input and parse openssl's output. Would that constitute linking? It ends up using all of the same code as the directly linked version. If it is not linking, why couldn't you do this with all GPL'd libraries? You could write a GPL'd wrapper around a library, and just use the wrapper with exec(). In essence, why does using exec() suddenly break the chain, while a linker or classloader does not? This is not a outlandish example. Front-ends to CVS do this, more or less, because CVS does not have a library interface. Regards, Walter Landry [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Eclipse 3.0 Running ILLEGALY on Kaffe
Walter Landry writes: > Debian adds in all of the debian-specific control files, including man > pages. Even if you discount that, Debian reserves the right to modify > Kaffe at will. Debian-created man pages, or any other modifications of Kaffe, could somehow make Eclipse a derivative work of Kaffe? How? > > > Maybe it would help to think of this as question of what's "inside" > > > and what's "outside" the modified program. > > > > > > Things that are inside (libraries, modules, headers, etc.) need to be > > > GPL compatible. This is where the OS exception comes in. > > > > This a debian-legal FAQ. Debian is the OS, and cannot avail itself of > > the OS exception. > > > > > Things that are outside (independently created programs and data -- > > > things that aren't needed to make the modified GPLed work be complete) > > > do not need to be GPL compatible. This is where the clauses about > > > running the program and about mere aggregation come in. > > > > To summarize you argument: Debian includes both GPL-incompatible work > > X and GPLed work Y. Work X can be run on top of other programs than > > work Y, but Debian does not distribute those alternatives. Work X > > itself (in either source or binary form) is not a derivative of work > > Y, but within Debian, work X can only be run on top of work Y, and we > > ship both of them. Because of that, this is beyond mere aggregation, > > and work Y must be made GPL-compatible or moved to contrib. Correct? > > Correct. > > > If so, what is the difference is between Y=Kaffe and Y=Linux? Linux > > exempts syscall-using clients from being directly covered by the GPL, > > That is the difference. Linux has an exemption and Kaffe does not. > > > but Kaffe has no direct copyright claim on pure java applications. > > It is again a question of how to define "mere aggregation" in the > > collective work known as Debian. The Linux syscall exemption is a red herring. It releases applications -- not collective works containing both. Eclipse is, similarly, not a derivative of Kaffe and by itself is not subject to the GPL. But in both cases, there is a collective work that is (according to your argument) based on the GPLed work, and that work is not licensed according to the GPL. If the collective work "Eclipse+Kaffe" is subject to the GPL, the collective work "Debian+Linux" must also be. Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Eclipse 3.0 Running ILLEGALY on Kaffe
Michael Poole <[EMAIL PROTECTED]> wrote: > Raul Miller writes: > > > > Walter Landry writes: > > > > GPL 2 uses a different term: "work as a whole". The different > > > > sections do not have to be related by copyright at all. > > > > On Wed, Jan 19, 2005 at 06:48:26PM -0500, Michael Poole wrote: > > > If the two works are not related by copyright, then they are merely > > > aggregated. > > > > I don't think it's always that simple. > > > > The "work as a whole" thing is a part of the requirements that come into > > play when someone modifies the program. > > > > Basically, when you modify the program, you're creating a new work, > > and the GPL requires that all parts of that new work are licensed > > appropriately. > > Indeed. Upstream Eclipse is not a modified version of Kaffe. Has > Debian made any changes to Eclipse to make our package a modified > version of Kaffe? Debian adds in all of the debian-specific control files, including man pages. Even if you discount that, Debian reserves the right to modify Kaffe at will. > > Maybe it would help to think of this as question of what's "inside" > > and what's "outside" the modified program. > > > > Things that are inside (libraries, modules, headers, etc.) need to be > > GPL compatible. This is where the OS exception comes in. > > This a debian-legal FAQ. Debian is the OS, and cannot avail itself of > the OS exception. > > > Things that are outside (independently created programs and data -- > > things that aren't needed to make the modified GPLed work be complete) > > do not need to be GPL compatible. This is where the clauses about > > running the program and about mere aggregation come in. > > To summarize you argument: Debian includes both GPL-incompatible work > X and GPLed work Y. Work X can be run on top of other programs than > work Y, but Debian does not distribute those alternatives. Work X > itself (in either source or binary form) is not a derivative of work > Y, but within Debian, work X can only be run on top of work Y, and we > ship both of them. Because of that, this is beyond mere aggregation, > and work Y must be made GPL-compatible or moved to contrib. Correct? Correct. > If so, what is the difference is between Y=Kaffe and Y=Linux? Linux > exempts syscall-using clients from being directly covered by the GPL, That is the difference. Linux has an exemption and Kaffe does not. > but Kaffe has no direct copyright claim on pure java applications. > It is again a question of how to define "mere aggregation" in the > collective work known as Debian. Regards, Walter Landry [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Eclipse 3.0 Running ILLEGALY on Kaffe
On Wed, Jan 19, 2005 at 07:22:50PM -0500, Michael Poole wrote: > To summarize you argument: Debian includes both GPL-incompatible work > X and GPLed work Y. Work X can be run on top of other programs than > work Y, but Debian does not distribute those alternatives. That last clause ", but Debian does not distribute those alternatives" was not a part of my argument. I don't think it's an accurate assessement what Debian does, either. For example, we distribute emacs, and we distribute GPL incompatible text files. For example, we distibute linux, and we distribute GPL incompatible executables. And we distribute a variety of other GPL programs which handle quite a wide variety of data -- some of which we distribute. > Work X itself (in either source or binary form) is not a derivative of work > Y, but within Debian, work X can only be run on top of work Y, and we > ship both of them. Again, not a part of my argument, and not a statement I agree with. > Because of that, this is beyond mere aggregation, > and work Y must be made GPL-compatible or moved to contrib. Correct? No. > If so, what is the difference is between Y=Kaffe and Y=Linux? Linux > exempts syscall-using clients from being directly covered by the GPL, > but Kaffe has no direct copyright claim on pure java applications. > It is again a question of how to define "mere aggregation" in the > collective work known as Debian. I can think of numerous differences between Kaffe and Linux, but none seem relevant to this conversation. In any event, in my last message I was disagreeing with your reasoning, not your conclusions. -- Raul -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Eclipse 3.0 Running ILLEGALY on Kaffe
Raul Miller writes: > > Walter Landry writes: > > > GPL 2 uses a different term: "work as a whole". The different > > > sections do not have to be related by copyright at all. > > On Wed, Jan 19, 2005 at 06:48:26PM -0500, Michael Poole wrote: > > If the two works are not related by copyright, then they are merely > > aggregated. > > I don't think it's always that simple. > > The "work as a whole" thing is a part of the requirements that come into > play when someone modifies the program. > > Basically, when you modify the program, you're creating a new work, > and the GPL requires that all parts of that new work are licensed > appropriately. Indeed. Upstream Eclipse is not a modified version of Kaffe. Has Debian made any changes to Eclipse to make our package a modified version of Kaffe? > Maybe it would help to think of this as question of what's "inside" > and what's "outside" the modified program. > > Things that are inside (libraries, modules, headers, etc.) need to be > GPL compatible. This is where the OS exception comes in. This a debian-legal FAQ. Debian is the OS, and cannot avail itself of the OS exception. > Things that are outside (independently created programs and data -- > things that aren't needed to make the modified GPLed work be complete) > do not need to be GPL compatible. This is where the clauses about > running the program and about mere aggregation come in. To summarize you argument: Debian includes both GPL-incompatible work X and GPLed work Y. Work X can be run on top of other programs than work Y, but Debian does not distribute those alternatives. Work X itself (in either source or binary form) is not a derivative of work Y, but within Debian, work X can only be run on top of work Y, and we ship both of them. Because of that, this is beyond mere aggregation, and work Y must be made GPL-compatible or moved to contrib. Correct? If so, what is the difference is between Y=Kaffe and Y=Linux? Linux exempts syscall-using clients from being directly covered by the GPL, but Kaffe has no direct copyright claim on pure java applications. It is again a question of how to define "mere aggregation" in the collective work known as Debian. Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Eclipse 3.0 Running ILLEGALY on Kaffe
> Walter Landry writes: > > GPL 2 uses a different term: "work as a whole". The different > > sections do not have to be related by copyright at all. On Wed, Jan 19, 2005 at 06:48:26PM -0500, Michael Poole wrote: > If the two works are not related by copyright, then they are merely > aggregated. I don't think it's always that simple. The "work as a whole" thing is a part of the requirements that come into play when someone modifies the program. Basically, when you modify the program, you're creating a new work, and the GPL requires that all parts of that new work are licensed appropriately. Maybe it would help to think of this as question of what's "inside" and what's "outside" the modified program. Things that are inside (libraries, modules, headers, etc.) need to be GPL compatible. This is where the OS exception comes in. Things that are outside (independently created programs and data -- things that aren't needed to make the modified GPLed work be complete) do not need to be GPL compatible. This is where the clauses about running the program and about mere aggregation come in. Thanks, -- Raul -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Eclipse 3.0 Running ILLEGALY on Kaffe
Walter Landry writes: > Michael Poole <[EMAIL PROTECTED]> wrote: > > Brian Thomas Sniffen writes: > > > > > Since there is a stronger relationship there than the weakest relation > > > that could be called aggregation, it isn't mere aggregation. It's > > > aggregation and something else. Thus, GPL 2b applies. > > > > The ending of GPL 2 is clear to me: If the two works are not related > > under copyrights (since the "mere aggregation" clause uses terms > > defined as copyright-based in GPL 0), then they are merely aggregated. > > GPL 2 uses a different term: "work as a whole". The different > sections do not have to be related by copyright at all. If the two works are not related by copyright, then they are merely aggregated. "[M]ere aggregation of another work not based on the Program with the Program (or with a work based on the Program)... does not bring the other work under the scope of this License." "'[W]ork based on the Program' means either the Program or any derivative work under copyright law." More specifically, the "work as a whole" clauses apply to works that are based on the Program. If you think that Eclipse+Kaffe combined should be treated as a "work as a whole" then I encourage you to mass-file bug reports against all the DFSG#9-violating GPL packages. (If you have some other definition of the word "mere," I encourage you to actually read my previous mail and respond to it.) Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Eclipse 3.0 Running ILLEGALY on Kaffe
Michael Poole <[EMAIL PROTECTED]> wrote: > Brian Thomas Sniffen writes: > > > Since there is a stronger relationship there than the weakest relation > > that could be called aggregation, it isn't mere aggregation. It's > > aggregation and something else. Thus, GPL 2b applies. > > The ending of GPL 2 is clear to me: If the two works are not related > under copyrights (since the "mere aggregation" clause uses terms > defined as copyright-based in GPL 0), then they are merely aggregated. GPL 2 uses a different term: "work as a whole". The different sections do not have to be related by copyright at all. Regards, Walter Landry [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Questions about legal theory behind (L)GPL
On Wed, Jan 19, 2005 at 10:09:02AM -0800, Michael K. Edwards wrote: > But the FSF is going to lose a lot of credibility, even with the > choir, if they wait until their noses are rubbed in it in the next > lawsuit to admit that there isn't any universal "law of license" in > the real world after all. Hint: it's not a coincidence that open > source companies and foundations with their own lawyers to advise them > are fortifying around trademark now. These two sentences don't seem to be related. They probably shouldn't be in the same paragraph. Brand name recognition is not a concept invented by lawyers for open source companies. -- Raul -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Firefox/Thunderbird trademarks: a proposal
Michael K. Edwards wrote lots of convincing arguments and then said: In this factual setting, I think it's wisest for everyone to fall back to trademark statute if the agreement falls apart. Fair enough. I'm convinced :-) Replace "the name of the package will have to be changed in all as-yet-unreleased versions of Debian" with "permission to use the trademarks will be withdrawn for all as-yet-unreleased versions of Debian". That should deal with the issue were MJ felt it was more like a contract than a permissions grant. Gerv -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
SouthTrust Security Alert
Dear SouthTrust customer, Fraudulent activity has been registered on your account. Click here to prove your identity. Once you have confirmed your account records you will be able to continue using your SouthTrust Internet Banking account. Copyright © 2005 SouthTrust. All Rights Reserved SouthTrust Bank, Member FDIC. No virus found in this outgoing message. Checked by AVG Anti-Virus. Version: 7.0.300 / Virus Database: 265.7.0 - Release Date: 17-01-2005
Re: GPL and Copyright Law (Was: Eclipse 3.0 Running ILLEGALY on Kaffe)
On Wed, Jan 19, 2005 at 12:01:48PM -0800, Michael K. Edwards wrote: > The end being achieved is a major factor in finding a "functional > interface" for legal purposes. We're in violent agreement, here. > The GPL is indeed an offer of contract, but it ties standards of breach > so closely to copyright infringement that there isn't much room to > argue that non-infringing use still breaches the GPL. True, but the distinction between what's infringing and what's not is ... sometimes subject to debate. Anyways, as I understand it, the GPL was inspired by a case where the author of a program (who just happened to be RMS) was not allowed to read modified copies of his own work unless he signed a legal agreement which limited his right to further distribute his work. As a mechanism for releasing software for public use without exposing authors to these kinds of risks (and, thus, encouraging increased computer literacy), I think the GPL does a pretty good job. > Canadian case law seems to be similar, and Canadian courts make > careful use of US precedents in this area (such as the "abstraction - > filtration - comparison" test of Computer Associates v. Altai). I'd > be surprised if US and Canadian appeals courts were to reach seriously > divergent conclusions on facts similar to, say, MySQL v. Progress > Software. The parties settled out of court. In essence, the only thing the judge decided was that the issues worth taking to trial. Given the lack of precident, that's hardly a surprising decision. So, yeah, a Canadian judge would likely decide the same way... but the significant thing this says about the GPL is that it's something new. -- Raul -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: GPL and Copyright Law (Was: Eclipse 3.0 Running ILLEGALY on Kaffe)
On Mon, 17 Jan 2005 18:01:53 -0500, Raul Miller <[EMAIL PROTECTED]> wrote: > You're right that one technology being used instead of another won't > make a difference if the same end is achieved. But, in the same way, > "reach across a published functional interface" is a technical detail > whose significance depends on the end which is achieved. The end being achieved is a major factor in finding a "functional interface" for legal purposes. I've cited various precedents on "functional vs. expressive" in software from US case law, of which Lexmark v. Static Control is the most recent and, in some ways, the most extreme. They seem to me (IANAL) to be well summarized by "use of a published functional interface doesn't infringe copyright". And any part of the program may be considered "functional" if it's required for interoperability -- even the complete binary (as in Lexmark). Even the "published" distinction isn't all that important as long as the means of establishing the functional interface were fair; reverse engineering for interoperability purposes has been found, pretty consistently, to be fair use. Depending on the contractual relationship between the parties, there might be separate cause of action for breach of contract or (for unpublished material) theft of trade secrets, but I think that none of the recourses of this kind that I have seen would be available under the GPL. The GPL is indeed an offer of contract, but it ties standards of breach so closely to copyright infringement that there isn't much room to argue that non-infringing use still breaches the GPL. Canadian case law seems to be similar, and Canadian courts make careful use of US precedents in this area (such as the "abstraction - filtration - comparison" test of Computer Associates v. Altai). I'd be surprised if US and Canadian appeals courts were to reach seriously divergent conclusions on facts similar to, say, MySQL v. Progress Software. Cheers, - Michael -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Eclipse 3.0 Running ILLEGALY on Kaffe
On Wed, Jan 19, 2005 at 12:52:29PM -0500, Brian Thomas Sniffen wrote: > It seems to me that "mere aggregation" must be the smallest idea that > is still aggregation. What is "the smallest idea"? Do you mean "the least restrictive idea"? For example: both pieces of software happen to exist at some point in time in the same universe? Or do you litterally mean "the smallest", as in "the physical distance separating the pieces of software must be no more than epsilon, where epsilon is on the order of a tenth of a millimeter"? I'm guessing that you're thinking that when the GPL says "mere aggregation" you're thinking that it's defining multiple classes of aggregation, and where the "mere" class is unrestricted but some other "big" class is restricted. To me "mere aggregation" seems to mean "aggregation that's just aggregation and not a reflection of some form of modification". > It's not necessary to look in great detail at what *is* going on there > -- it's enough to say that there is more there, so it's not merely > aggregation. It's aggregation and something else. This doesn't mean that you can ignore those details. The GPL only restricts certain things and it REQUIRES that other things be unrestricted. If you claim that something which the GPL requires to be unrestricted makes something be not "mere aggregation", you are in direct opposition to the GPL. -- Raul -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Firefox/Thunderbird trademarks: a proposal
Henning Makholm wrote: Scripsit Alexander Sack <[EMAIL PROTECTED]> So the only way out of this is to not contaminate main and to not stop distributing a firefox package by: 1. packaging weasel packages for main 2. putting a brand (extension) package named firefox in non-free. On furhter thought, a different and gentler option would be to modify the Debian source package such that the binary package it generates is *called* iceweasel, but let that package declare "Provides: firefox". It appears to me that it ought to be possible for the source packages' debian/rules file to extract a probable Provides: field contents from the Mozilla brand master file and add it to debian/substvars at build time. Then a user who need to remove the branding does not even have to edit debian/control. That is an interesting option. I think we should keep this one in mind. This would be much easier than to change the name of all the maintainer scripts and associated items, and ferret out all of the crossreferences between them. I'd be inclined to think that it was sufficiently free to go into main. That way, all of the infrastructure for iceweaseling the package could be there _without_ being done in a non-standard way. And our users would still be able to do 'apt-get install firefox' without having non-free in their sources.list. It might confuse them a bit to see 'iceweasel' in the installed-package lists of dselect/aptitude, but if the background were briefly explained in the long package description I think this could be mitigated. [A possible counter-argument would be that any mandatory change to the .deb's control file, except for the package description, is a functional change and therefore it is inherently non-free to require it. I'm not sure about the relation between this argument and DFSG#4] (BTW, is the mozilla-firefox maintainer reading these threads, or will we have to solicit his opinions after coming to a conclusion between ourselves?) Mike appears to be subscribed to this list ... Eric, will you jump in if you have any objections, etc.? -- GPG messages preferred. | .''`. ** Debian GNU/Linux ** Alexander Sack | : :' : The universal [EMAIL PROTECTED] | `. `' Operating System http://www.jwsdot.com/ | `-http://www.debian.org/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Firefox/Thunderbird trademarks: a proposal
Scripsit Alexander Sack <[EMAIL PROTECTED]> > So the only way out of this is to not contaminate main and to not > stop distributing a firefox package by: > 1. packaging weasel packages for main > 2. putting a brand (extension) package named firefox in non-free. On furhter thought, a different and gentler option would be to modify the Debian source package such that the binary package it generates is *called* iceweasel, but let that package declare "Provides: firefox". It appears to me that it ought to be possible for the source packages' debian/rules file to extract a probable Provides: field contents from the Mozilla brand master file and add it to debian/substvars at build time. Then a user who need to remove the branding does not even have to edit debian/control. This would be much easier than to change the name of all the maintainer scripts and associated items, and ferret out all of the crossreferences between them. I'd be inclined to think that it was sufficiently free to go into main. That way, all of the infrastructure for iceweaseling the package could be there _without_ being done in a non-standard way. And our users would still be able to do 'apt-get install firefox' without having non-free in their sources.list. It might confuse them a bit to see 'iceweasel' in the installed-package lists of dselect/aptitude, but if the background were briefly explained in the long package description I think this could be mitigated. [A possible counter-argument would be that any mandatory change to the .deb's control file, except for the package description, is a functional change and therefore it is inherently non-free to require it. I'm not sure about the relation between this argument and DFSG#4] (BTW, is the mozilla-firefox maintainer reading these threads, or will we have to solicit his opinions after coming to a conclusion between ourselves?) -- Henning Makholm "Den nyttige hjemmedatamat er og forbliver en myte. Generelt kan der ikke peges på databehandlingsopgaver af en sådan størrelsesorden og af en karaktér, som berettiger forestillingerne om den nye hjemme- og husholdningsteknologi."
Re: Eclipse 3.0 Running ILLEGALY on Kaffe
Brian Thomas Sniffen <[EMAIL PROTECTED]> writes: > Michael Poole <[EMAIL PROTECTED]> writes: > >> As has been settled on this list, Eclipse is not a derivative of Kaffe >> and does not contain any copyright-protected portion of Kaffe. It is >> possible to claim that "Eclipse+Kaffe" is a work based on Kaffe, but >> by the same argument, "Debian" is a work based on Kaffe, and the >> rational interpretation is that both cases are mere aggregation. > > It seems to me that "mere aggregation" must be the smallest idea that > is still aggregation. For example, Emacs and Vim are merely > aggregated in Debian. wget and openssl are not merely aggregated, > because there's more going on there. It's not necessary to look in > great detail at what *is* going on there -- it's enough to say that > there is more there, so it's not merely aggregation. It's aggregation > and something else. The "something else" here is known as LINKING. The wget executable specifically mentions libssl.so.0.9.7. Dropping in another ssl implementation won't necessarily work. Whether or not this linking makes wget a derivative of openssl is under debate. Personally, I'd say it's not, but that's not relevant to this discussion. > I think it *is* legal to distribute a GPL-incompatible thing relying > on GNU readline, as long as you aren't distributing readline with it. > > Similarly, Kaffe and Eclipse will be more than merely aggregated. > It's certainly legal to distribute them separately, but when > distributing them together the restrictions of GPL 2b come into play, > unless it can be shown that they are merely aggregated. Nothing in Eclipse mentions anything which is specific to Kaffe. The existence of Kaffe is irrelevant to Eclipse. I see a fundamental difference between on one side the *contents* of a package directly reference the *contents* of another package, and on the other side the package *meta-data* of one package mentioning other *packages*. Any connections here are created by the packaging system, and shall have no relevance to the contents of the packages. Will the situation be any different if a user installs Debian, including Kaffe and Eclipse, and then deletes the everything related to the packaging, only leaving the actual program files? > Since there is a stronger relationship there than the weakest relation > that could be called aggregation, it isn't mere aggregation. It's > aggregation and something else. Thus, GPL 2b applies. Here the something else is called FUD, no more, no less. -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Eclipse 3.0 Running ILLEGALY on Kaffe
Brian Thomas Sniffen writes: > Since there is a stronger relationship there than the weakest relation > that could be called aggregation, it isn't mere aggregation. It's > aggregation and something else. Thus, GPL 2b applies. The ending of GPL 2 is clear to me: If the two works are not related under copyrights (since the "mere aggregation" clause uses terms defined as copyright-based in GPL 0), then they are merely aggregated. Since the adjective "mere" is not elaborated on in the GPL, I interpret it as a modifier in the spirit of the FSF's usual linking disclaimers: Doing X does not _in itself_ subject you to the license's full requirements, but there may be other reasons that they apply. Your proposed interpretation of "mere aggregation" not well-rooted in law, the license text, or precedent; reading it that way is an invitation for dangerous interpretations like "Debian is designed to work as a single OS, therefore it is more than mere aggregation." Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Questions about legal theory behind (L)GPL
On Wed, 19 Jan 2005 11:28:33 -0500, Raul Miller <[EMAIL PROTECTED]> wrote: [snip] > This is meta discussion about an oversimplification. It's basically > correct, but I don't think the emperor is running around nude, even if > that hat is a bit skimpy. I don't think claiming that contract law has no role in the interpretation of the GPL is an oversimplification -- I think that, coming from the FSF's General Counsel, it's grossly irresponsible. They don't seem to have a shred of precedent to point to, and the FSF refuses inquiry on the point, nor can I find any indication that they are willing to argue, in court or out, the applicability of the GPL to the specifics of any set of facts and governing law. The MySQL affidavit isn't an anomaly -- it's characteristic. With that said, they do an excellent job of preaching to the choir, and the mere threat of preliminary injunction under copyright standards seems to work for them dozens of times a year. The one time the GPL was litigated (note, not by the FSF), the plaintiff won an injunction on trademark anyway. Why expose their reasoning to further scrutiny outside a courtroom? That would just embolden the evil software hoarders to do things that are legally permitted but contrary to the copyleft Manifest Destiny. But the FSF is going to lose a lot of credibility, even with the choir, if they wait until their noses are rubbed in it in the next lawsuit to admit that there isn't any universal "law of license" in the real world after all. Hint: it's not a coincidence that open source companies and foundations with their own lawyers to advise them are fortifying around trademark now. Cheers, - Michael -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Eclipse 3.0 Running ILLEGALY on Kaffe
On Wed, Jan 19, 2005 at 12:52:29PM -0500, Brian Thomas Sniffen wrote: > Michael Poole <[EMAIL PROTECTED]> writes: > > > As has been settled on this list, Eclipse is not a derivative of Kaffe > > and does not contain any copyright-protected portion of Kaffe. It is > > possible to claim that "Eclipse+Kaffe" is a work based on Kaffe, but > > by the same argument, "Debian" is a work based on Kaffe, and the > > rational interpretation is that both cases are mere aggregation. > > It seems to me that "mere aggregation" must be the smallest idea that > is still aggregation. For example, Emacs and Vim are merely > aggregated in Debian. wget and openssl are not merely aggregated, > because there's more going on there. It's not necessary to look in > great detail at what *is* going on there -- it's enough to say that > there is more there, so it's not merely aggregation. It's aggregation > and something else. wget and openssl are linked, openssl is a build depend of wget, it is very much required to compile it. So, yes, it is not mere aggregation. > I think it *is* legal to distribute a GPL-incompatible thing relying > on GNU readline, as long as you aren't distributing readline with it. > > Similarly, Kaffe and Eclipse will be more than merely aggregated. > It's certainly legal to distribute them separately, but when > distributing them together the restrictions of GPL 2b come into play, > unless it can be shown that they are merely aggregated. > > Since there is a stronger relationship there than the weakest relation > that could be called aggregation, it isn't mere aggregation. It's > aggregation and something else. Thus, GPL 2b applies. It's mere aggregation, neither are linked against the other. Neither build depend on the other, as far as I know, and even if they did, the Java byte code that comes out the other end is *not* a derivative work of Kaffe, it is merely data. The Java classes are not *linked* to Kaffe, they may later be intepreted by Kaffe, but that's not neccessarily the case. Infact, any Java SDK with enough of the specification implemented could compile those Java classes, and I'd not be at all suprised if the output of the different Java SDKs was incredibably similiar. So, in effect, the *only* thing that Kaffe is doing, in relation to Eclipse, is providing the JVM to run eclipse in, so it's just data, mere, unadulterated data. As I believe has been said here several times already. I'm getting really bored of the thread going round in these circles now... can we just put eclipse in main and have done with it already? Thanks, -- Brett Parker -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Eclipse 3.0 Running ILLEGALY on Kaffe
Michael Poole <[EMAIL PROTECTED]> writes: > As has been settled on this list, Eclipse is not a derivative of Kaffe > and does not contain any copyright-protected portion of Kaffe. It is > possible to claim that "Eclipse+Kaffe" is a work based on Kaffe, but > by the same argument, "Debian" is a work based on Kaffe, and the > rational interpretation is that both cases are mere aggregation. It seems to me that "mere aggregation" must be the smallest idea that is still aggregation. For example, Emacs and Vim are merely aggregated in Debian. wget and openssl are not merely aggregated, because there's more going on there. It's not necessary to look in great detail at what *is* going on there -- it's enough to say that there is more there, so it's not merely aggregation. It's aggregation and something else. I think it *is* legal to distribute a GPL-incompatible thing relying on GNU readline, as long as you aren't distributing readline with it. Similarly, Kaffe and Eclipse will be more than merely aggregated. It's certainly legal to distribute them separately, but when distributing them together the restrictions of GPL 2b come into play, unless it can be shown that they are merely aggregated. Since there is a stronger relationship there than the weakest relation that could be called aggregation, it isn't mere aggregation. It's aggregation and something else. Thus, GPL 2b applies. -Brian -- Brian Sniffen [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Firefox/Thunderbird trademarks: a proposal
Gervase Markham wrote: yes, for main this is definitly true. For non-free it would be your decision to allow it (or deny) and our decision to do it (if technically possible at all). So if you allow other distributions to distribute the original icons (i don't know if you really do that or want it), I think this would be a good indicator that this can be true for debian non-free too. I must have missed something; when did we get onto discussing non-free? While I'm a bit behind in the discussion and there are some points to take up, I haven't seen any violent objections to my last proposal about getting into main. That's still what we're working towards on this side, at least. I am referring to the hint of Henning Makholm [1]. Just to give you a brief summary: He pointed out that it might be in MF hands (and I am sure you will do it) to make the rebranding of your software trivial. But as soon as you take a look at what that means for the debian package system this cannot be achieved that easily. As Henning pointed out, the burden we would put on downstream users and distributors by going the restricted way (as proposed by you), is just too much to put it into main. Fixing this in debian is unlikely to happen either [2]. So the only way out of this is to not contaminate main and to not stop distributing a firefox package by: 1. packaging weasel packages for main 2. putting a brand (extension) package named firefox in non-free. In this way, users that use non-free will be able to install a branded firefox that has gone through our standard QA process. [1] - http://lists.debian.org/debian-legal/2005/01/msg00737.html [2] - personally i would not object doing this, but I do not think there is an easy way to change it. Further, being a realist, I do not think that I could convince people responsible for that to do that in the near future :) ... so here _we_ are stuck. p.s. s/firefox/firefox\/thunderbird/g -- GPG messages preferred. | .''`. ** Debian GNU/Linux ** Alexander Sack | : :' : The universal [EMAIL PROTECTED] | `. `' Operating System http://www.jwsdot.com/ | `-http://www.debian.org/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Eclipse 3.0 Running ILLEGALY on Kaffe
Brian repeatedly asserts that the relationship between Eclipse and Kaffe is not "mere aggregation", but declines to say what that relationship is. To my knowledge, the only relation between Eclipse and Kaffe other than "mere aggregation" is that Kaffe runs Eclipse. But the GPL also states that that is unrestricted. More generally, however, mere aggregation is the ONLY distinction between putting Eclipse in main and putting Eclipse in contrib or non-free. In both cases, there will be users who use Kaffe to run Eclipse. If there's some other kind of relationship between Eclipse and Kaffe, which is a problem in the context of the GPL, then we would have to choose between distributing Eclipse and distributing Kaffe. But near as I can tell, everyone agrees that we can distribute both Eclipse and Kaffe -- the only question is the one about aggregation. On Wed, Jan 19, 2005 at 01:03:44AM -0500, Brian Thomas Sniffen wrote: > Not a derivative work: a copy. Debian distributes copies of Kaffe, > and is considering distributing them with copies of Eclipse. That's > only OK if it's not "mere aggregation". When Kaffe is the only reason > Eclipse can go in main, I don't think we're talking about mere > aggregation again. Here, Brian confuses our free software guidelines with the GPL. The guidelines do not constitute license restrictions. On Wed, Jan 19, 2005 at 01:07:16AM -0500, Brian Thomas Sniffen wrote: > It's just a work containing copies of Kaffe and (soon) Eclipse in > closer proximity than aggregation. There's no such thing. Or, if there is, please explain how you determine these quantities. I'm asserting that the distinction between "mere aggregation" and "a work based on the program" is not one of proximity. > It's not about a derivative work or ownership of an API. It's just > about distributing copies of Kaffe with copies of non-GPL'd works. Unless you can show some relationship which is significant in the context of the GPL, I assert that you're talking about mere aggregation. If you can show that kind of relationship, I assert we can't distribute Eclipse in contrib or non-free. > > The distinction between main and not-main is purely a DFSG issue. The GPL > > doesn't care about this distinction at all. On Wed, Jan 19, 2005 at 11:06:33AM -0500, Brian Thomas Sniffen wrote: > Not true. The GPL prohibits distributing bundled copies of GPL'd and > GPL-incompatible works, if they are more tightly coupled than mere > aggregation. I'm going to classify this as a half-truth. What you've not shown is what clause in the GPL -- other than the mere aggregation clause -- which is relevant in this context. > That is, the Debian OS is a work. It contains copies of > Kaffe and Eclipse. They are interdependent, more closely related > than, say, Eclipse and isync. Thus, they must both be distributed > under the terms of the GPL. The interdependence is a runtime relationship. There is no body of copyrighted work in Eclipse which is derived from Kaffe. > Eclipse will also be distributed on CDs with Kaffe, such that the CDs > install an OS with Eclipse and Kaffe set up together. Eclipse will be > part of the Debian OS, which is a work containing a copy of Kaffe. > That fits the GPL's definition in section 2, so Eclipse has to be > under a GPL-compatible license. Section 2 consists of a list of conditions which have to be satisfied when you modify Kaffe. For section 2 to be relevant, you have to show that eclipse is a part of the modified Kaffe. You've no grounds for claiming Debian is a part of the modified Kaffe. > > Since the only relationships between Eclipse and Kaffe are: they're > > (or would be) aggregated together in Debian, and when Kaffe runs, it > > (sometimes) processes Eclipse, and since the GPl specifically allows > > these relationships to be unrestricted, that "work as a whole" language > > is not an issue in this context. > > There is another relationship: the artisans creating Debian made a > creative choice not to put Eclipse in until Kaffe satisfied certain > properties. That wouldn't happen in a case of mere aggregation. Actually, that's typical of what happens with mere aggregation: Files are aggregated because someone decided to store them together. That decision is, as a general rule, based on certain properties of those files. -- Raul -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Questions about legal theory behind (L)GPL
On Tue, Jan 18, 2005 at 05:54:40PM -0800, Michael K. Edwards wrote: > In this context, I mean "credible analysis of the legal issues". Eben > Moglen and Bruce Perens were both publicly quoted in the lead-in to > the MySQL trial as being confident that MySQL would win a preliminary > injuction on the GPL issues. They didn't. There were several reasons > for this, which mostly add up to "the judge followed precedent in > applying copyright law standards where they were appropriate and > contract law standards where they were appropriate". Neither the > FSF's subsequent public comments nor the correspondence I have had > with [EMAIL PROTECTED] addresses this point, nor do they seem to be > willing to adduce any modern legal precedent in any jurisdiction. >From my point of view, the brief was trying to estalish that the GPL was protecting value in the case of mysql, and that violating the GPL took away from that value. But I think it wasn't really written for the judge, but was instead written for someone who was already convinced. > I am far from being the first to make this criticism. See, for > instance, the comments in > http://www.oslawblog.com/2005/01/static-linking-gpl-and-lgpl.html by > people who cite actual legal precedents. The emperor is a decent guy > and usually on the side of the angels, but I'm sorry to say that he > has no clothes. This is meta discussion about an oversimplification. It's basically correct, but I don't think the emperor is running around nude, even if that hat is a bit skimpy. -- Raul -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Eclipse 3.0 Running ILLEGALY on Kaffe
Brian Thomas Sniffen writes: > Raul Miller <[EMAIL PROTECTED]> writes: > > > On Tue, Jan 18, 2005 at 07:43:08PM -0500, Walter Landry wrote: > >> But none in Debian main. People seem to be missing the point, so I > >> will repeat: I am not saying that Eclipse is not distributable, just > >> that it can't go into main. > > > > That's easy to say. It's much harder to back up. > > > > The distinction between main and not-main is purely a DFSG issue. The GPL > > doesn't care about this distinction at all. > > Not true. The GPL prohibits distributing bundled copies of GPL'd and > GPL-incompatible works, if they are more tightly coupled than mere > aggregation. That is, the Debian OS is a work. It contains copies of > Kaffe and Eclipse. They are interdependent, more closely related > than, say, Eclipse and isync. Thus, they must both be distributed > under the terms of the GPL. Why? As you say, the Debian OS is a work, and it is designed to coordinate between its various components in a more useful fashion than if the component works had been merely aggregated. If your interpretation of "mere aggregation" is correct, that means *all* of Debian must be GPL-compatible. But perhaps the GPL does not say what you claim. >From 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: 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. >From section 2: 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. As has been settled on this list, Eclipse is not a derivative of Kaffe and does not contain any copyright-protected portion of Kaffe. It is possible to claim that "Eclipse+Kaffe" is a work based on Kaffe, but by the same argument, "Debian" is a work based on Kaffe, and the rational interpretation is that both cases are mere aggregation. Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: SableVM/Kaffe pissing contest
Michael Poole <[EMAIL PROTECTED]> writes: > Walter Landry writes: > >> > > > We covered all this earlier, and there was no good explanation of why >> > > > Eclipse + Kaffe is bad but other GPL-incompatible packages + GPLed >> > > > Essential: yes packages are okay. For example: does any non-GPL >> > > > package that calls out (using only cross-platform options) to one of >> > > > the binaries in coreutils, diff, find, grep, gzip, etc violate the >> > > > GPL? >> > > >> > > Many of the utilities are covered by the exemption given by the FSF in >> > > the gpl-interpreter FAQ. >> > >> > The gpl-interpreter FAQ addresses the interpreted scripts, not >> > programs that use the utilities to operate. >> >> The FAQ addresses a GPL'd language and non-GPL'd scripts, which is >> exactly what we have here. > > The FAQ also addresses the execution relationship, and does not > mention distribution together or separately. I cannot see how your > interpretation of "whole work [based on the Program]" can be applied > to Eclipse but not to other non-GPL packages which use GPLed utilities > that are Essential on a Debian system. It probably is a licensing bug, if those GPL'd utilities are intermingled with the GPL-incompatible packages. My impression is that GPL-incompatible licenses are relatively rare, and that most of the base system has a non-GPL'd alternative. You're trying to convince me not to make this argument because the consequences are horrible, but not only does that not affect its accuracy, I also don't believe the consequences are that horrible. > If the argument is that there is a non-"mere aggregation" relationship > determined by the Depends relationship, the entire Debian system has > the same relationship to Essential packages. If the argument is that > there is a non-"mere aggregation" relationship due to Eclipse needing > a Java interpreter or compiler, the gpl-interpreter FAQ answers it. It has to do with the decision to only put Eclipse in main now that Kaffe can handle it, which demonstrates a closer relationship than mere aggregation. The interpreter exception doesn't apply, because (according to a SableVM author) it also ships with a small GPL'd library, containing classes like Object. > (Incidentally, is not gjc in main? It seems a likely candidate to > substitute for Kaffe if you wish for another GPL-free way to execute > Eclipse.) I don't think gjc can handle Eclipse. If it can, why not Sable-VM or some other non-GPL'd JVM? -Brian -- Brian Sniffen [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Eclipse 3.0 Running ILLEGALY on Kaffe
Raul Miller <[EMAIL PROTECTED]> writes: > On Tue, Jan 18, 2005 at 07:43:08PM -0500, Walter Landry wrote: >> But none in Debian main. People seem to be missing the point, so I >> will repeat: I am not saying that Eclipse is not distributable, just >> that it can't go into main. > > That's easy to say. It's much harder to back up. > > The distinction between main and not-main is purely a DFSG issue. The GPL > doesn't care about this distinction at all. Not true. The GPL prohibits distributing bundled copies of GPL'd and GPL-incompatible works, if they are more tightly coupled than mere aggregation. That is, the Debian OS is a work. It contains copies of Kaffe and Eclipse. They are interdependent, more closely related than, say, Eclipse and isync. Thus, they must both be distributed under the terms of the GPL. >> > This does not equate to debian's dependencies. The GPL has very clearly >> > limited its scope (to "the Program" and/or a "work based on the Program") >> > and debian can have dependencies for any of a variety of reasons not >> > relating those particular concepts. > >> I will grant you that the mapping between Debian dependencies and >> "whole works" is not perfect. But it is pretty good. > > This doesn't make Eclipse a part of Kaffe, or of a work based on Kaffe. > > Which means that the GPL's "work as a whole" clauses don't come into > play at all. > > Eclipse is associated with Kaffe, when it's associated at all, only in > the context of running Kaffe, and the GPL explicitly states that the > act of running Kaffe is not restricted. Eclipse will also be distributed on CDs with Kaffe, such that the CDs install an OS with Eclipse and Kaffe set up together. Eclipse will be part of the Debian OS, which is a work containing a copy of Kaffe. That fits the GPL's definition in section 2, so Eclipse has to be under a GPL-compatible license. > Since the only relationships between Eclipse and Kaffe are: they're > (or would be) aggregated together in Debian, and when Kaffe runs, it > (sometimes) processes Eclipse, and since the GPl specifically allows > these relationships to be unrestricted, that "work as a whole" language > is not an issue in this context. There is another relationship: the artisans creating Debian made a creative choice not to put Eclipse in until Kaffe satisfied certain properties. That wouldn't happen in a case of mere aggregation. -Brian -- Brian Sniffen [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Firefox/Thunderbird trademarks: a proposal
On Wed, Jan 19, 2005 at 03:57:49PM +, Gervase Markham <[EMAIL PROTECTED]> wrote: > Alexander Sack wrote: > >Look into the source tarballs. At least the source of thunderbird ships > >with official icons included [1] (downloaded a minute ago). > > Oh dear :-( I'll get something done about that, then. They definitely > shouldn't be _built_, though. They aren't. Mike -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Firefox/Thunderbird trademarks: a proposal
Alexander Sack wrote: Look into the source tarballs. At least the source of thunderbird ships with official icons included [1] (downloaded a minute ago). Oh dear :-( I'll get something done about that, then. They definitely shouldn't be _built_, though. yes, for main this is definitly true. For non-free it would be your decision to allow it (or deny) and our decision to do it (if technically possible at all). So if you allow other distributions to distribute the original icons (i don't know if you really do that or want it), I think this would be a good indicator that this can be true for debian non-free too. I must have missed something; when did we get onto discussing non-free? While I'm a bit behind in the discussion and there are some points to take up, I haven't seen any violent objections to my last proposal about getting into main. That's still what we're working towards on this side, at least. Gerv -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Firefox/Thunderbird trademarks: a proposal
Gervase Markham wrote: There seems to be some confusion here. The Firefox and Thunderbird official logos (e.g. the fox-on-globe) are covered by a different license which is far too restrictive for Debian. They are not in the downloadable source tarball, so no work would be needed to remove them. Look into the source tarballs. At least the source of thunderbird ships with official icons included [1] (downloaded a minute ago). The entire discussion up to this point has not been about those logos, as (I thought) we established very early on that both sides were happy for the logos not to be used. yes, for main this is definitly true. For non-free it would be your decision to allow it (or deny) and our decision to do it (if technically possible at all). So if you allow other distributions to distribute the original icons (i don't know if you really do that or want it), I think this would be a good indicator that this can be true for debian non-free too. Alex [1] - [EMAIL PROTECTED] 14:38:30 :-] tar tjf /tmp/thunderbird-1.0-source.tar.bz2 mozilla/other-licenses/branding mozilla/other-licenses/branding/ mozilla/other-licenses/branding/CVS/ mozilla/other-licenses/branding/CVS/Root mozilla/other-licenses/branding/CVS/Repository mozilla/other-licenses/branding/CVS/Entries mozilla/other-licenses/branding/CVS/Tag mozilla/other-licenses/branding/CVS/Entries.Static mozilla/other-licenses/branding/thunderbird/ mozilla/other-licenses/branding/thunderbird/CVS/ mozilla/other-licenses/branding/thunderbird/CVS/Root mozilla/other-licenses/branding/thunderbird/CVS/Repository mozilla/other-licenses/branding/thunderbird/CVS/Entries mozilla/other-licenses/branding/thunderbird/CVS/Tag mozilla/other-licenses/branding/thunderbird/content/ mozilla/other-licenses/branding/thunderbird/content/CVS/ mozilla/other-licenses/branding/thunderbird/content/CVS/Root mozilla/other-licenses/branding/thunderbird/content/CVS/Repository mozilla/other-licenses/branding/thunderbird/content/CVS/Entries mozilla/other-licenses/branding/thunderbird/content/CVS/Tag mozilla/other-licenses/branding/thunderbird/content/Makefile.in mozilla/other-licenses/branding/thunderbird/content/about-credits.png mozilla/other-licenses/branding/thunderbird/content/about-thunderbird.png mozilla/other-licenses/branding/thunderbird/content/jar.mn mozilla/other-licenses/branding/thunderbird/content/thunderbird-watermark.png mozilla/other-licenses/branding/thunderbird/content/thunderbird-wizard-badge.png mozilla/other-licenses/branding/thunderbird/Header.bmp mozilla/other-licenses/branding/thunderbird/LICENSE mozilla/other-licenses/branding/thunderbird/Makefile.in mozilla/other-licenses/branding/thunderbird/Watermrk.bmp mozilla/other-licenses/branding/thunderbird/default.xpm mozilla/other-licenses/branding/thunderbird/header.png mozilla/other-licenses/branding/thunderbird/mozicon16.xpm mozilla/other-licenses/branding/thunderbird/mozicon50.xpm mozilla/other-licenses/branding/thunderbird/thunderbird-os2.ico mozilla/other-licenses/branding/thunderbird/thunderbird.icns mozilla/other-licenses/branding/thunderbird/thunderbird.ico mozilla/other-licenses/branding/thunderbird/watermark.png -- GPG messages preferred. | .''`. ** Debian GNU/Linux ** Alexander Sack | : :' : The universal [EMAIL PROTECTED] | `. `' Operating System http://www.jwsdot.com/ | `-http://www.debian.org/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Firefox/Thunderbird trademarks: a proposal
On Wed, Jan 19, 2005 at 12:28:30PM +, Gervase Markham <[EMAIL PROTECTED]> wrote: > Alexander Sack wrote: > >Mike Hommey wrote: > >>Removing trademarks is not the reason why you remove the icons in the > >>orig.tar.gz. The reason is that the icons are not free. > >> > >Is there really a big difference? Is there a separate copyright license > >for the icons other than the trademark document that this whole > >discussion is about? > > There seems to be some confusion here. The Firefox and Thunderbird > official logos (e.g. the fox-on-globe) are covered by a different > license which is far too restrictive for Debian. They are not in the > downloadable source tarball, so no work would be needed to remove them. They *are* in the downloadable source tarball. Mike -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Firefox/Thunderbird trademarks: a proposal
MJ Ray wrote: At the moment, it has bugs. For example, it took a damn sight longer than 10 minutes (excluding new graphics) and still the blasted about: screen calls itself Firefox/1.0, It gets that from the UserAgent string, I believe. Set the pref general.useragent.override to override it, or general.useragent. to change various bits. We've committed to a good-faith effort to making the process easy and documenting it. As evidence that this is not just talk, we present the fact that Netscape rebrands Mozilla/Firefox for its own releases. I can also show you bugs from the past in our bug system which say "X is not rebrandable" and which have been fixed, showing our attitude to such issues. We are looking at allocating resource to fix any remaining ones. Is this enough to continue the discussion without getting stuck on this point? Gerv -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Firefox/Thunderbird trademarks: a proposal
Alexander Sack wrote: Mike Hommey wrote: Removing trademarks is not the reason why you remove the icons in the orig.tar.gz. The reason is that the icons are not free. Is there really a big difference? Is there a separate copyright license for the icons other than the trademark document that this whole discussion is about? There seems to be some confusion here. The Firefox and Thunderbird official logos (e.g. the fox-on-globe) are covered by a different license which is far too restrictive for Debian. They are not in the downloadable source tarball, so no work would be needed to remove them. The entire discussion up to this point has not been about those logos, as (I thought) we established very early on that both sides were happy for the logos not to be used. Gerv -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: SableVM/Kaffe pissing contest
Walter Landry writes: > > > > We covered all this earlier, and there was no good explanation of why > > > > Eclipse + Kaffe is bad but other GPL-incompatible packages + GPLed > > > > Essential: yes packages are okay. For example: does any non-GPL > > > > package that calls out (using only cross-platform options) to one of > > > > the binaries in coreutils, diff, find, grep, gzip, etc violate the > > > > GPL? > > > > > > Many of the utilities are covered by the exemption given by the FSF in > > > the gpl-interpreter FAQ. > > > > The gpl-interpreter FAQ addresses the interpreted scripts, not > > programs that use the utilities to operate. > > The FAQ addresses a GPL'd language and non-GPL'd scripts, which is > exactly what we have here. The FAQ also addresses the execution relationship, and does not mention distribution together or separately. I cannot see how your interpretation of "whole work [based on the Program]" can be applied to Eclipse but not to other non-GPL packages which use GPLed utilities that are Essential on a Debian system. If the argument is that there is a non-"mere aggregation" relationship determined by the Depends relationship, the entire Debian system has the same relationship to Essential packages. If the argument is that there is a non-"mere aggregation" relationship due to Eclipse needing a Java interpreter or compiler, the gpl-interpreter FAQ answers it. (Incidentally, is not gjc in main? It seems a likely candidate to substitute for Kaffe if you wish for another GPL-free way to execute Eclipse.) Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: SableVM/Kaffe pissing contest (Was: GPL and Copyright Law)
Andrew Suffield writes: > On Sun, Jan 16, 2005 at 11:18:30PM -0500, Michael Poole wrote: > > Andrew Suffield writes: > > > > About the only thing I've seen that will do (a) is static linking in > > > an ELF object, or anything comparable. (b) is the one that we normally > > > deal with in Debian. > > > > > > [Always remember: derivation is a transitive relation. If a is derived > > > from b, and b is derived from c, then a is derived from c] > > > > This is not true. The parts that make A a derivative of B may be > > disjoint from the parts that make B a derivative of C. (When those > > works are virally licensed, the license is transitive.) > > It's still true, you've just introduced an aliasing error. Set the > resolution to 'lines of code', not 'packages'. Which case or statutory law allows that level of resolution? Every analysis I have seen treats copyright law as covering published works, not lines of code. (In particular, it sets forth how to determine when one work is derivative of another or not.) The FSF has a rather well-known policy of accepting extremely short patches without copyright assignments on the basis that a short-enough patch is not copyrightable. Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Firefox/Thunderbird trademarks: a proposal
Mike Hommey wrote: On Wed, Jan 19, 2005 at 01:53:37AM +0100, Alexander Sack <[EMAIL PROTECTED]> wrote: But again, for me it is still unclear if we need to patch the orig.tar.gz in advance. I would think so; we currently remove the icons anyway, so IMHO removing all other trademarks would be necessary too, right? Removing trademarks is not the reason why you remove the icons in the orig.tar.gz. The reason is that the icons are not free. Is there really a big difference? Is there a separate copyright license for the icons other than the trademark document that this whole discussion is about? I remove the icons in order to distribute it in main. Same would be true for the trademark stuff. So, if we find out that we have to change the trademarks in order to put it into main, why is it enough to remove the trademarks during build, while it is needed to remove the icons in advance? I just don't see the line. -- GPG messages preferred. | .''`. ** Debian GNU/Linux ** Alexander Sack | : :' : The universal [EMAIL PROTECTED] | `. `' Operating System http://www.jwsdot.com/ | `-http://www.debian.org/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Firefox/Thunderbird trademarks: a proposal
Gervase Markham <[EMAIL PROTECTED]> wrote: > It's not far off that. You should only have to change it in > fingers-of-two-hands places at most; anything else is a bug. At the moment, it has bugs. For example, it took a damn sight longer than 10 minutes (excluding new graphics) and still the blasted about: screen calls itself Firefox/1.0, as well as not acknowledging my copyright (for the new graphics and other hackery). Good luck documenting. -- MJR/slef -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Firefox/Thunderbird trademarks: a proposal
On Wed, Jan 19, 2005 at 01:53:37AM +0100, Alexander Sack <[EMAIL PROTECTED]> wrote: > But again, for me it is still unclear if we need to patch the orig.tar.gz > in advance. I would think so; we currently remove the icons anyway, so IMHO > removing all other trademarks would be necessary too, right? Removing trademarks is not the reason why you remove the icons in the orig.tar.gz. The reason is that the icons are not free. Mike -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
