Forgot your password?
typodupeerror
The Courts GNU is Not Unix Open Source Linux Your Rights Online

Court Case To Test GNU GPL 371

Posted by Soulskill
from the doesn't-play-nice-with-others dept.
ciaran_o_riordan writes "Tomorrow, a German court will hear the case of AVM, a distributor of Linux-based routers, which seeks to block Cybits from distributing software that modifies the routers' software to add content-filtering functionality. Free Software Foundation Europe explains: 'AVM justified its position using three arguments. First, they stated that their whole product software must be regarded as an entity under AVM copyright, and that this entity must not be modified. The position Mr. Welte [founder of gpl-violations.org and copyright holder of several parts of the Linux kernel] took was that the whole product software would in that case be a derivative work according to the GPL, and thus the whole product software should be licensed under the GNU GPL. AVM then switched to a second argument: that the software embedded on its DSL terminals consisted of several parts. According to Mr. Welte, AVM could then not prohibit anyone from modifying or distributing the GPL licensed software parts. The final argument by AVM was that the software on their DSL terminals is a composition of several different programs, which, due to the creative process, would be a protected compilation and thus under the copyright of AVM and not affected by the copyleft of the GPL.'"
This discussion has been archived. No new comments can be posted.

Court Case To Test GNU GPL

Comments Filter:
  • by pieterh (196118) on Monday June 20, 2011 @02:12PM (#36504398) Homepage

    I've been making my software GPLd since 1991. If someone uses it, for free, and makes a derived work, and then tries to stop others from sharing that work, they are ethically and intellectually challenged.

    This is not a test case for the GPL, it's a straight-forward copyright violation case, where AVM is taking the work of tens of thousands of people, using it under a license that permits remixing, and then attempting to ignore those license conditions.

    Take them to the cleaners, Harald!

    • by swillden (191260)

      This is not a test case for the GPL, it's a straight-forward copyright violation case

      The really clever bit about the GPL is that any test case is a straight-forward copyright infringement case. People who violate the GPL find themselves in the unenviable position of having to try to use the GPL to justify their actions, because it's the only thing that gives them any shred of permission to use the code. All it takes, then, is for the court to evaluate the situation and realize that the GPL didn't actually give them permission because they didn't meet its requirements.

      • by Tetsujin (103070)

        This is not a test case for the GPL, it's a straight-forward copyright violation case

        The really clever bit about the GPL is that any test case is a straight-forward copyright infringement case. People who violate the GPL find themselves in the unenviable position of having to try to use the GPL to justify their actions, because it's the only thing that gives them any shred of permission to use the code. All it takes, then, is for the court to evaluate the situation and realize that the GPL didn't actually give them permission because they didn't meet its requirements.

        That's not necessarily true. Bear in mind that, if put to a legal test, parts of the license could be found unenforceable, in which case someone could escape the terms of the GPL. Another possible loophole is the definition of a "derivative work". Generally I think we assume that if a project pulls in some code that is identifiable as having been part of a GPL product, then that project (according to the GPL) must also be licensed by the GPL. But that assertion, too, is something that would have to be t

        • by afidel (530433)
          The problem with all of those arguments is that if you don't agree to the GPL you have *zero* legal right to use the code, in whole or in part (unless it doesn't meet some copyright test like API headers with no creative value). It's a well written legal contract and a LOT of companies with large legal teams (like Cisco for example) have looked at it and ultimately caved because they couldn't find enough of a loophole to drive their case through, I doubt a company that has needed to change their legal strat
          • by JAlexoi (1085785) on Monday June 20, 2011 @04:28PM (#36506036) Homepage
            Egh... GPLv2 specifies that a derivative work is defined by the relevant copyright law. In this case, it's the German Author's rights law.
          • by jd (1658)

            There's a difference between agreeing to the GPL and agrring on what it says. It is a little ambiguous on just what constitutes a derivative project, at least in v2, but the current case doesn't seem like it's an ambiguity issue. The fact that the defense keeps changing legal basis and the definition of their own work indicates that the defense knew in advance that they were violating any interpretation of the GPL.

  • I don't think so (Score:5, Informative)

    by Rene S. Hollan (1943) on Monday June 20, 2011 @02:21PM (#36504526)

    1. This violates the GPL over the GPL-licensed parts of the whole, and therefore the GPL-licensed parts have to be excluded when the whole is distributed. The same reasoning applies if the restriction is due to a compilation copyright over the whole or a license regarding use that prohibits modification.

    2. Irrelevant: the GPL parts continue to be licensed under the GPL or they can not be redistributed. The GPL permission for compilations does not weaken the GPL license for GPL components so compiled.

    3. All a compilation copyright means is that no one else can redistribute the compilation. However, it can not restrict redistribution of the GPL licensed parts.

    About the only case that can be made is that modified routers can't be sold, but routers could be sold along with the means to modify them. Or a router could be sold, and someone hired to modify it.

    Of course, I am not a lawyer, and would welcome one correcting any error I made above.

    • In addition, a compilation is a derived work and the author needs to secure the copyrights to the parts. Of course, often a compilation is made from parts that have lost their copyright status, or are to small to be considered a work, but that does not mean that creating a compilation magically waives the rights of the creators of the parts.

      IANAL, and also not yours.

      • by Xtifr (1323) on Monday June 20, 2011 @03:04PM (#36505092) Homepage

        In addition, a compilation is a derived work

        Not legally, no. To create a compilation, you do need rights to publish each part, but you do not need permission to modify or prepare derivative works. And even if that weren't the case, the GPL explicitly has an exception.

        From v3: "Inclusion of a covered work in an aggregate does not cause this License to apply to the other parts of the aggregate."

        From v2: "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."

        • Thanks, I didn't know that.
        • by Raenex (947668)

          To create a compilation, you do need rights to publish each part, but you do not need permission to modify or prepare derivative works.

          This is a rather strange thing to say. If a part of the compilation is under copyright, then the compilation itself is a derivative work. That's why you need permission to publish each part that is under copyright.

          And even if that weren't the case, the GPL explicitly has an exception.

          Quite true, but if the compilation is considered a whole work worthy of copyright in and of itself, then the GPL must apply to the whole work. So you need to be very careful on the nature of the compilation.

      • by mwvdlee (775178)

        I think that is the clearest explaination of what's going on here; making a compilation of copyrighted materials does not terminate the copyrights of those materials.
        Also, as far as I know the whole "copyrighted compilation" thing only goes when significant and non-trivial effort has gone into compilation, neither of which is the case here.

        • Its akin to someone publishing a book that includes some Shakespeare plays, the Harry Potter series, and their own novels.

          They'd be sued for including the Harry Potters without permission, despite having created a new compilation work.

    • Now, the argument could be made that AVM needs to control quality over its products, and unauthorized modification may compromise that quality -- you don't want to be on the hook for someone else's changes to your router.

      Therefore a license restriction to ensure continued support starts to make sense.

      However, what that is is a warranty issue: "Warranty void if operating firmware modified". There, how hard is that?

      • What's at issue is a third party adding functionality to AVM's product that likely competes with another AVM product sold at a higher price.

    • About the only case that can be made is that modified routers can't be sold, but routers could be sold along with the means to modify them.

      Not really, because if they're placing additional restrictions on the redistribution of someone else's GPLed code, they lose the right to distribute that code under the terms of the GPL.

      • No, you can still redistribute the modified GPL code, but not the non-GPL bits that make up the compilation. At least, not unless the doctrine of first sale applies, and that strikes me as (a) weak, and possibly whisked away by, say, a reseller-contract, and (b) likely not applicable in Germany, anyway.

        I can't take a Red Hat Linux CD, add stuff to it (assuming no difference between making a modified copy and modifying the physical original, which can't be done easily with CDs), and redistribute it unless (a

        • Of course, this does not appear to apply to the instant case, because apparently Cybits is not modifying routers purchased for resale, but rather distributing software to modify them by their owners.

          At this point, the argument becomes one of "Well, if a reseller can't sell modified routers because it violates their license regarding the copyright non-GPL bits", and Cybits facilitates such modification, Cybits is an "enabler" of copyright infringement.

          As much as I try to play Devil's Advocate for the plainti

    • by Xtifr (1323)

      IANAL either, but I think you basically got it right. The one detail you might have overlooked, which may be relevant, is that the GPL explicitly forbids sublicensing! So they can't apply a separate license (even the null license) to the compilation; only to its individual parts. This is why there is no overall license for, e.g., Red Hat or Debian. Red Hat has a license for its trademarks, but if you strip those (as with CentOS), everything else is freely distributable and modifiable.

  • by mark-t (151149) <markt@ l y n x.bc.ca> on Monday June 20, 2011 @02:23PM (#36504554) Journal
    *ANY* copyrighted work derivative must comply with the expectations and terms of the original copyright holder. How, exactly, is using the GPL any different from this?
    • by Xtifr (1323)

      Yup, not a GPL test at all. Simply a copyright case.

  • by doperative (1958782) on Monday June 20, 2011 @02:23PM (#36504556)

    "Ironically, by preventing others from enacting the rights granted by the GNU GPL, AVM itself is in violation of the license terms. Therefore they have no right to distribute the software" says Till Jaeger.

    • by Bengie (1121981)

      Get one of those injunction things that blocks them from selling those units until the case is straitened out.

    • by hey! (33014)

      That raises in interesting possibility I hadn't thought about. Just because GPL requires you to give certain rights to any derivative works does it mean that recipients of such works *actually receive those rights*? I'm thinking probably not.

      Party A licenses software a under GPL.

      Party B incorporates software a into proprietary licensed product b, thus violating GPL.

      Party C creates derivative work c from b.

      So, is it possible that both C and B are violating software licenses here? It'd work like this. Party

  • This doesn't even make sense without the GPL. Even if AVM owned the copyright to all code, it doesn't by itself prevent users from adding new code in place. It's as if the author of a book sued to prevent purchasers from the writing in the margins of their own copy.

    Unless German copyright law is very different from US law, there must be a clause in the license that TFA fails to mention. This clause presumably disallows consumers from making modifications to the software. Since the GPL is involved, this

  • by iplayfast (166447) on Monday June 20, 2011 @02:40PM (#36504752)

    It amazes me that companies feel justified in not paying for their software. The GPL's payment is in openness and code changes. So many companies try this (and get away with it). It would be similar to companies using cracked warze and then trying to justify it in court as their own works.

  • Of all these, I think the compilation argument is an interesting one and merits watching. It could have interesting results on other things based on open licenses such as Linux distributions and other devices based on Linux.

    I don't think I will roll out a complete comment on this just now. I will instead ponder the notion of thinking of all these computer programs as components and calling this a compilation which is "protected."

    There are many things that cause this to fall apart such as the individual li

  • by roguegramma (982660) on Monday June 20, 2011 @02:45PM (#36504808) Journal

    I think the case hinges on the right to sue of the parties.

    A contributor to the GPL'ed software would have the right to sue.

    The defendant, Mr. Welte, might not have the right to sue AVM to put their code under the GPL, but luckily he is the defendant.

    AVM is the plaintiff. They have admitted that their released work is a mix of their original work and GPL'ed work. In my opinion they still may assert their copyright for their contributions, but do not actually have the right to distribute their work.

    That is why I think the defendants primary cause, although I'm sympathetic to the defendant, is lost.

    However as a countermove, (I'm assuming the defendant owns an AVM router), I suggest he countersues AVM for damages, since they set him up with a router that cannot be operated legally since they don't own the rights to the software, and for wasting his time he spent coding in good faith that AVM actually intends to secure the rights to the software, which is hard to do without GPL'ing the software.

    IANAL, and not yours.

    • A contributor to the GPL'ed software would have the right to sue.

      Which Mr. Welte is. He wrote part of iptables, if I'm not mistaken.

      The defendant, Mr. Welte, might not have the right to sue AVM to put their code under the GPL, but luckily he is the defendant.

      He is not. He is not working or representing Cybits, the company sued by AVM.
      But he entered the dispute as someone who owns (part) of the copyright of the product in question.

      • Thanks, I didn't figure that out from the summary.

        I still think he cannot force them to release their code under the GPL. They can just say "no", and then he has the right to sue them and will win, making them pay damages and making them stop distributing their product.

        Of course, the threat of this might convince them, as it usually does in GPL cases.

        IANAL, and not yours.

        • by anyGould (1295481)

          I still think he cannot force them to release their code under the GPL. They can just say "no", and then he has the right to sue them and will win, making them pay damages and making them stop distributing their product.

          I think the logic is thusly:

          The gadget is either using GPLed code under a valid license (which means they can't prevent others from modifying it, thus they lose), or they are not using it under a valid license (which means that they're vulnerable to infringement claims themselves, and depending on exactly which bits were being modified, they might not have standing to protest in the first place).

        • I still think he cannot force them to release their code under the GPL. They can just say "no", and then he has the right to sue them and will win, making them pay damages and making them stop distributing their product.

          It is an interesting situation. Mr. Welte cannot force AVM to do something, but he can quite clearly tell them "if you do what you intend to do then you are committing copyright infringement" with the consequence that AVM might be in for some serious consequences.

          Actually, I think at this point AVM has already been distributing the software in violation of the GPL terms. So the GPL license doesn't apply here anymore. So the fact is: AVM took some copyrighted code that they have no right to distribute, mo

  • Could we stop saying "test the GPL"? If the GPL does not hold, then copyrights and contracts don't hold. You can't kill the GPL without killing more onerous EULAs, and as we all know, that will never happen.
    • by ArsonSmith (13997)

      You seem to hold the antiquated ideal of rule of law. Laws aren't made as inflexible as you seem to portray any more. There are many rules that can be flexed and others that can be out right broken without consequence. You just have to know the right people and have the right friends in government.

      • by mark-t (151149)
        Perhaps it may have escaped your attention that the only thing that is granting any privileges to copy GPL'd works at all is the text of the GPL itself. Even if the GPL is invalid, standard copyright law is still applicable, and people still need explicit permission of the copyright holder (which is what the GPL is supposed to do).
        • by ArsonSmith (13997)

          you seem to speak with things like "any privileges" or "explicit permission" where, if it was for the good of the people these things could be flexed a bit.

          • by mark-t (151149)

            This implies that you advocate the abolition of the current concept of copyright law (literally, the "right to copy")... which, if not exclusive, is no longer valuable.

            Of course, copyright itself is no longer that valuable anyways, considering the epidemic numbers of people that disregard it simply because they can.

            • You keep failing to get the GP's point, he is not advocating anything. Judiciary decisions aren't always logical, or clear, even when the law is logical and clear. Sometimes the decisions favor a contrived interpretation that lets the people with more money act freely, while the ones with less money are restricted. Other times the oposite happens, but then, it is just a matter of the part with more money not having money enough...

      • by JAlexoi (1085785)
        This being Germany, in case if AVM wins I would urge all original authors of the software AVM uses to assert their moral rights to attribution. And file a lawsuit to force AVM to send out all equipment owners attribution notices.
  • This is not a test of the GPL! It's simply a test of copyright law, and, in particular, the boundaries of the definition of "derivative work".

    If I were an author, and gave a publisher permission to publish a story I wrote, unmodified, they would be free to publish it in a collection with other stories by other authors, but not free to modify huge chunks of it. But, what if they got another author to write stuff that was inserted between key paragraphs of my story? My story is unmodified, in the sense tha

  • A more clear summary than the long-winded summary up top is this:

    Company A makes a widget based on GPL code.
    Company B makes code that you can buy and run on the widget, changing it's function
    Company A is suing company B to stop them claiming copyright violation

    There are multiple weird things about this case that I am very unclear on.

    - Does company B's distributed code even contain any code from Company A? If it doesn't, and it is literally just code that gets injected into the router, then it is not a copyr

  • Yes, they have copyrights on the compilation. But they have to respect the rights of the Linux copyright holders, and those rights say "you can only distribute the software under these conditions".

    So, if they want to distribute, they have to allow modification. If they don't allow modification, they can enjoy their own copyright in the privacy of their own home, but they can't distribute the compilation without violating someone else's copyright.

  • The final argument by AVM was that the software on their DSL terminals is a composition of several different programs, which, due to the creative process, would be a protected compilation and thus under the copyright of AVM and not affected by the copyleft of the GPL.

    Which means they're back to a derivative work ;-)

  • They seem to be confusing the GPL with the BSD license or the public domain. If they ignore the GPL, nothing else lets them use the GPL-licensed code, so they have to take it out. Schnell!

    • by mark-t (151149)
      Well, they can always *USE* it... they just can't legally copy it.
      • by idontgno (624372)

        Well, they can always *USE* it... they just can't legally copy it.

        And by "copy it", we really mean "distribute it". Yes, AVM can continue to use the Linux kernel on every little widget it uses inside its own facilities and processes. But it can't convey one copy in a product sold to a customer, because redistribution in a product is a form of "copying", and copying triggers the obligation to release source and permit modifications.

        Interesting. This is a legal attack on the practical usability of Free Sof

        • by mark-t (151149)

          "Although others own the copyright on some of the components we use, we assert definitive control on the overall package embodied in our device"

          Which is incompatible with the GPL and thus by their actions, they are openly admitting that they do not have permission from the copyright holder to copy the work. Commercially distributing a copyrighted work that you don't have permission to copy in the first place is generally seen as a Bad Thing.

          The only way one can take AVM's side in this is if they advoca

        • by mark-t (151149)

          Also...

          And by "copy it", we really mean "distribute it"

          Maybe you mean that... it's not what copyright says though, and not what I meant either. By "copy" I meant "copy". The catch is really for what purpose the copy is being made... distribution is largely irrelevant, except to the extent that it can be a purpose one has for making a copy and is generally deemed as sufficient evidence to convict for copyright infringement if one does not have permission from the copyright holder to have made copies for

  • ... if you want the damn code proprietary, then write the damn thing yourself. Why is this so hard to understand? People that whine about "Well, I wrote part X as a modification of GPL project Y, so why am I forced to use GPL on X!" The answer is simple, because you based your work on GPL! If you don't want to use GPL, then avoid GPL!

    Yes, the issue can be made much more complex. But at the end of the day, this "problem" hasn't changed, and it's not that complicated. Microsoft figured it out and they al

  • With open source code as pervasive as it is it's all to easy for a coder at a company to cheat and plagiarize. Maybe even sabotage his own company. Courts would need to look at the amount of code involved and determine if the company in question had reasonable knowledge of infringement. Damage? The answer is we need a clear decision as far as disclosure from companies when something like this happens. More or less depending upon the amount of code in question they must be forced for reply to a cease and

    • "I actually really want to see it go before SCOTUS and have lawyers try to explain pointers and addresses when they don't even understand it themselves. Then define execution."

      They were just recently discussing turing-completeness there (because of software patents), right untill they dismissed the entire concept as not a real thing...

  • by feenberg (201582) on Monday June 20, 2011 @04:56PM (#36506326)

    According to the article, the defendent is not distributing code containing GPL code. Rather, they are distributing a program that reads from a DSL router and modifies the (perfectly legal) GPL code on the router, reinstalling the modified code. The defendent doesn't think this is a violation, since he does not distribute any GPL code to users, only the binary "diffs". The modified code is never "distributed", only installed on the individuals own router. Since the GPL limits distribution, but doesn't affect "internal" use, there is an argument that the GPL is not violated. However, there is a further section in the GPL that takes up just this point, which is quite orthogonal to any of the arguments posted here. Even if this section of the GPL was not enforced in Germany, it wouldn't be the end of the GPL, as this is an extremely inconvinient way to distribute software, and the liklihood that the "diffs" didn't include GPL code is very small.

Those who do not understand Unix are condemned to reinvent it, poorly. - Henry Spencer, University of Toronto Unix hack

Working...