GPL Gets Its Day in Court in Israel 232
MadFarmAnimalz writes "In what appears to be the first court test for the GPL in the Middle East, Alexander Maryanovsky, the author of the GPL licensed Jin Chess Client is taking IchessU to court for violations of the GPL license."
Re:Right.... bit of clarification (Score:2, Interesting)
Personally I don't see how the GPL can be violated here and yet we still have PC's shipping with GPL software on them.
Re:Right.... bit of clarification (Score:1, Interesting)
In a way, this is very lucky timing for GPLv3 (Score:3, Interesting)
It's great that we will get the benefit of this ruling during the year when GPLv3 [fsfeurope.org] is being written. This sort of thing provides great suggestions for what should be clearer or worded differently.
Re:Right.... bit of clarification (Score:3, Interesting)
What happens if you release a GPL'd program with a plug-in API? That's fine.
Now you release a closed-source plug-in. That's not fine, since the plug-in is a derived work of the GPL'd plug-in system; it won't work without it.
Now, what happens if you license the plug-in API as a separate module under the LGPL (for example). The closed-source plug-in only depends on the plug-in module, which is LGPL'd. It inherits the GPL when used with a GPL'd application, but you could argue that the plug-in does not require a GPL'd program.
Now, when you use the plug-in, you link it with a GPL'd application. This would be a GPL violation, except for the fact that the GPL only comes into effect when distributing (it is not a EULA).
Doesn't matter (Score:5, Interesting)
To be specific - I am pretty sure the drivers use either a UNIX socket or a named pipe.
Re:In a way, this is very lucky timing for GPLv3 (Score:5, Interesting)
There are very good reasons why this is impossible. Actually, what you are proposing is a very natural notion, that sadly turns out to be wrong. I say 'natural', because during the first third or so of the 20th century, philosophy (of language, in particular), was seeking exactly what you are driving at - a 'pure' language, free from ambiguity. This would have had benefits for legal matters, as well as philosophical ones, and even metamathematical implications. But this was shown to be a futile attempt (Wittgenstein being the major figure showing this). I'll briefly summarize why this is so.
First, when you refer to human-related things - as laws are, they mention e.g. 'assault', 'homicide', and so forth - there is no way to 'clean up' the language. It cannot be made unambiguous, because the underlying concepts are ambiguous. Try to define (as the famous example goes) 'game'. For any suggested definition, there are counterexamples (e.g. not all games are about winning or losing, not all games have scores, not all games are fun, etc. etc.). This is a simple consequence of the fact that life is complex - we use the word 'game' in many contexts, in many ways. Unlike in math, where we start with definitions, in the law we start with pre-existing human concepts and try to work with them. We therefore cannot arrive at unambiguous statements.
Second, and this is a more subtle issue, language is meaningless without a context of use. By this I mean, that if you see some scribbles on a page, they are worthless without someone to read them. A sentence + a reader are what is necessary for 'meaning' to exist. Thus, even if we write what we believe to be unambiguous text, we can never remove the element of the reader: for us, the statement is unambiguous, but for another person, with a slightly different mindset, it may not be so. You may claim that your interpretation is 'correct', but that will not avail you when a matter is put before the public, i.e. open to interpretation by many people, as the law must be.
Sorry to go on at length, but this is a fascinating topic for me.
Not pipes or sockets ... merely source (Score:3, Interesting)
I've been watching the kororaa [kororaa.org] project for a while, ever since we did one tandem session with XGL and OS X on two machines, watching XGL rule - especially in the video across cube faces demo. But a few weeks later, the developer announced that he's stopping Kororaa because of GPL issues with properietary drivers. And here's a reply by the FSF [kororaa.org].
Now, the point to note here is that GPL is redistribution license. The way the nVidia folks handle it is to give the user some code, a binary blob and effectively tell them to build it themselves. The code they distribute does not link to the Linux kernel *yet*, while the binary blob is the closed source bit. Now, what the user does is to link them all up and there you go - which is not the distributor's fault. And this works because they are not redistributing any code that is copyrighted by a Linux kernel author (for example laf0rge [gpl-violations.org]).
The whole model makes the user violate GPL in principle, while the distributor (i.e nVidia) is in the gray area of legality. This is of course, my understanding from following up all this (and then had an argument with a paralegal @ work about GPL).
But I could be wrong you know ...
Re:Doesn't matter (Score:3, Interesting)
Perhaps that was just an unfortunate choice of words, or a small language barrier if you're not a native English speaker. However, unless your law works very differently to most, you can't force anyone else to licence their code in any specific way, including under the GPL. To say "the other piece is also covered by the GPL" would therefore be wrong.
Of course, it may be a copyright infringement for them to use your (GPL'd) code in their product if their code is not released with a compatible licence, but that has very different implications. If they're infringing your copyright, a court might be able to issue an injunction banning the use of your code and/or award you compensation, for example. However, the code behind "the other piece" hasn't magically become licensed under the GPL, and no-one has any magical rights to see/reuse the source code.
(IANAL, don't get legal advice from Slashdot, etc. etc.)
Re:Not pipes or sockets ... merely source (Score:3, Interesting)
The whole model makes the user violate GPL in principle
Actually, since the end user doesn't distribute anything there is no violation there. So the end user is still in the white. Depending on the interpretation of the GPL Nvidia may be in a gray/black area.