Why is my work derivative of Stallman's if I do a CALL to his library, but not derivative of Intel's if I do a REP MOVSB, thus invoking their microcode?
Because of the difference in the copyright licenses.
Look, I get you don't like RMS's licenses, fine, that doesn't change the fact that for this particular issue, there is no fundamental difference between microcode and software when it comes to copyright. They are both copyrightable, and our CPUs' microcode is almost certainly copyrighted. The only difference is you like Intel's copyright more than RMS's. Fair enough.
I don't understand why GPL proponents can't just stick with the facts instead of equivocating between the original source and the code added to it.
Because our whole subthread started with Insanity Defense saying:
The purpose of the GPL is to keep the project as a whole "free", not just the original version.
"not just the original", so I haven't been equivocating, it never has been about "original" versus "added" code, its been about the "project"... at least until you just added that junk in to change the subject.
what I'm proposing (as thought problem, which I assumed you would understand)
Oh great, here it comes...
Well, whether or not *I* get it, my point was that a judge will not be amused by your "thought problem", which in turn, renders its likely-hood of actually happening to virtually nil.
Lesse, midnight on a boring middle-of-the-week Wednesday, just got through watching an old rerun of Clint Eastwood in A Fistful Of Dollars on the WGN Late-Nite-At-The-Westerns, but there's nothing good on now, and nothing else to watch on DVD, so what is kdawson's answer to this dilemma?
"Eureka! A flame-fest between the BSD Zealots and the GPL Fanatics, that ought to keep me entertained for the next 4 or 5 hours!"
[rummages through the inbox looking for good dry kindling, a match, some dynamite, and...]
Come on, Guys and Gals, this is a setup piece for a flame-war, if I've *ever* seen one, you've *all* been had...
Hey I actually wrote all of Linux in my basement, and all these geeks stole it! I emailed it to Linus and he stole it, I just never bothered to mention it until now. How can you possibly prove otherwise! Yes you can't prove anything, not beyond a shadow of a doubt.
RMS: If you don't change to using the GPL, then you'll have to stop using readline.
Not "switch to the GPL or else I'll call out the lawyers", but "if no GPL *then* stop using readline". The option to stop using the GPLed code was explicitly mentioned at the beginning. As the clisp author mentions, readline was not essential to his app, he *could* have easily ripped it out if he wanted to; clisp in no way depended on it.
The clisp author didn't even admit RMS was right; he essentially said, "whatever, it's easier this way," and GPL'd clisp.
Oh really?
RMS: I hate to have to play this role with a fellow hacker, but...
BH: I'm sorry too, as I am very indebted to the GNU project.
RMS: If you can get away with this then any company can get away with it. ... Is that what you want?
BH: No, of course!
RMS: If you do succeed in circumventing the GPL for readline, you would be blazing a path for every commercial company that wants to do it. Would you really like that result? Wouldn't it be best to eliminate this dispute by using the GPL for Common Lisp?
BH: Would be best for me, true.
RMS: Totally aside from readline, the Lisp system would be more useful that way.
BH: This and the following are convincing me:
* Up to 1991 the decision whether using GPL or not, was simply a matter of philosophy or copyright policy. The success of Linux, however, demonstrates how a speed and quality of development was achieved which would have been impossible without access for everyone to the source of everything.
* Finding co-developers for other Lisp packages or testers for other hardware/OS platforms might be easier if I release full source.
So be prepared to seeing CLISP's source before Christmas.
I've snipped out the technical argument over the linking. (Can anyone blame BH for misunderstanding or disagreeing with that? Hell, just look at all the heat in this thread!) Otherwise your description of Bruno's response does not at all match up to what he actually said, methinks. Bruno didn't "give up" in the end, in the end, he *voluntarily* agreed with RMS for several reasons, even if he still didn't agree with, or understand, the FSF's linking argument.
A hint for you, Mr. piojo: If you're going to try your hand at revising history, you'd better make damn sure that the *unrevised* history still isn't out there on the web for all to see...
In a perfect world, I would look at the degree to which I'm using a library (and several other questions I and others have mentioned), and make a decision as to whether my work is a derivative work.
In other words, allow those accused of an offense to decide their own guilt? That's not my idea of a perfect world, but whatever...
If it is, my work needs to be GPL.
Or you just stop using the GPLed work in your work. *Your* work being GPLed is not the only way to resolve the dispute. There may be some financial damages awarded regardless, but ultimately whether *your* work gets GPLed or not is up to you, not the plaintiff. Stop using the other code, and your code can stay your own.
Um, no. There would have to be criminal intent. Without intent, it's strictly a civil issue.
FYI, "willful copyright infringement" (willful == intent), does allow for incarceration as a possible punishment, although the kind/amount of punishment is up to the judge, not the plaintiff. I think thats what the GP was trying to refer to.
If established case law (and I'm not up to date on this) says that linking to libraries (without distributing those libraries) doesn't constitute a derived work, how can a copyright license (which then wouldn't apply) change that definition?
Maybe because your "established case law" doesn't actually exist?
Why is everyone in this thread assuming the FSF's interpretation is somehow "obviously invalid"? The GPL has been around for 20 years, if it was clearly wrong, like, for example, it contradicted "established case law", don't you think we'd know that by now? Do you really think the FSF would keep sticking their neck out on this point, practically begging someone to take a swing, for 2 decades now, if they *knew* it contradicted "established case law"?
Reality check: there won't ever *be* any "established case law" on this issue until someone actually decides to challenge the FSF's interpretation in court... 20 years later... still waiting...
That's _reference_, not _derivation_. That (IMO, IANAL) no more makes the application derivative...
IANAL either, but see swillden's response to my post just above yours:
In the case of the IBM attorneys I worked with, there is simply no question: They're pretty sure the court would side with the FSF, and in any case don't want to expose IBM to the risk that the court might side with the FSF,...
If my work is not a duplicate or derivative of yours (regardless of what nasty hack I used to avoid making it so),
If we assume that a judge *agrees* that your "nasty hack" does indeed avoid infringement, then I'll agree with you (but not necessarily with the judge).:)
I was going by the GP's tone and assuming that the hack itself was legally questionable, thus proving bad intent almost by itself.
Thus, despite the fact that Apple has a copyright to Mac OS X, they don't get to control distribution of programs written for it.
Well, technically, they could *try* to make that claim (if this were absolutely impossible via copyright, no OS would need to explicitly permit it - and yet many/most(/all?) do just that in their licenses), though obviously no OS maker would do this since then no one would use their OS.
Despite the fact that Intel has a copyright on the microcode in its processors, they don't get to control distribution of programs written for them.
Same reason as above: no one would use their chip if they tried this...
Though I'm less sure about the possible legality of their even trying - this may already be handled by legal precedent. I dunno, I've never *seen* a copyright license for a CPU's microcode...:)
The whole point of this sub-thread is comparing the GPL to the BSD. I'm just saying that "keeping the project as a whole "free"" isn't a logical reason to use the GPL instead of the BSD
And my point was that a BSD project can have a proprietary fork, but *not* a GPLed project, thus there *is* a difference: a *proprietary* fork *might* end up meaning no reciprocation, no give-back, no collaboration on a common base... it all depends on the "good will" of the company doing the fork.
Right or wrong, some people, *and* other companies!, don't put much faith in seeing "good will" from a profit-motivated corporation, especially ones, like, say, MS...
Take the AARP. It charges its members a membership fee and it's clearly an organization just as much as the FSF or IBM.
Is the AARP distributing modified GPL apps to their members without the source code? If not, then your point is?
My point wasn't very clear: That "fee" will show the "willful" intent of your action, which is to, in effect, "make a buck off somebody else's work". That is not by itself illegal, but when you add that to the copyright infringement, because distribution is happening without providing the source code, the judge will more than likely put them together and come up with "this ain't no accident, they're doing all this *knowingly* and *purposefully*". Intent does matter (to a judge as well as a jury). As for all your hand-waving about distribution (from GPL3):
To "propagate" a work means to do anything with it that, without permission, would make you directly or secondarily liable for infringement under applicable copyright law, except executing it on a computer or modifying a private copy. Propagation includes copying, distribution (with or without modification), making available to the public, and in some countries other activities as well.
"making available to the public" is the key. Judges are savvy enough to know that an organization defined as "members of the public which pay a fee" is, in fact, "the public", and the only reason for said organization's existence is to make you a profit off the act of copyright infringement (by "making available" without the source code). Your lawyer trying to argue otherwise is going to sound like a Nigerian scam artist making his pitch, and will just piss the judge off. Most judges aren't idiots: they've seen plenty of scams and scam artists in their career.
At this point, you have "willful copyright infringement", which can go beyond just monetary damages and right into jail-time, if the judge so chooses. This is why you don't want to annoy a judge with an obvious subterfuge. They see a lot of that anyway, so usually have a very low tolerance for BS.
In other words: you aren't going to fool anyone with this gag.
Yep, you're right, it goes into more detail (I didn't check it - just looked at the GPL), but you'd almost *expect* the LGPL to go into more detail, it *has* to, since this is why it exists (explicitly allowing more interaction with other non-GPL code).
I'm still not happy about "linking", though: can you provide a waterproof definition of linking in the context of Java programs?
If you mean "waterproof" in a *legal* sense, I don't believe thats possible for *any* language.:)
Thats what I was referring to at the end about "vagueness of the copyright law itself".
As far as I'm concerned a licence should be as short as it's possible to make it while communicating clearly what permissions it does and doesn't grant. Trying to make a licence too general results in it becoming longer and harder to understand. Using specific terms with agreed meaning is simply good communication.
Ok, I agree with all that, however the point I was trying to make is that the lack of "specific terms with agreed meaning" is inherent in the copyright law itself. Copyright licenses get their meaning and value from copyright law, they do not "stand alone".
Simple (probably bad) analogy: The copyright licenses are the "frame of your house", and the copyright law is the "foundation that you build your house on". No matter how strong you make the frame, if the foundation has a weakness, you're still in danger of losing your house, and sometimes, you have to *try* to overcome a possible weakness in the foundation by reinforcing the frame, which would *not* be necessary if the foundation was known to be solid.
The reason licenses aren't clear and easy to grok is because the underlying law isn't *helping* to *make* them clear and easy to grok.
The absolute worst offender, when it comes to unhelpfulness, is copyright law's concept of "derivative work". The meaning of this changes almost every bloody time the Supreme Court, or an Appeals Court, makes a decision concerning it! All the arguments elsewhere in this thread about linking are basically arguments over *technicalities*, and *none* of them can be clearly answered until we, or at least our lawyers, have a clear and precise legal definition of what a "derivative work" is. Alas, technology is moving much faster than the law, and the courts aren't able to update the law fast enough to keep up with reality. This is our current dilemma.
Attention Moderators: The Troll moderation does not mean "I disagree". If thats what it *meant* then thats what it'd be *called*.
you might consider that people claiming to be open WRT the GPL may imply open and can never be closed
This is a perfectly polite, and perfectly VALID *opinion*, whether you agree with it or not. If the parent is a Troll, then so is the grand-parent, the great-grand-parent, and virtually every freakin' post in this entire discussion thread!
Not until the EU can explain why they "allow" them, just don't "enforce" them. That doesn't make any sense: why allow software patent application if you never intend to enforce it... unless the "enforce" part is just an implementation detail left for v2.0 that is on its way?
"The purpose of the GPL is to keep the project as a whole "free" not just the original version."
A group that forms the core development for a project maintains control of that project even if somebody forks it to create their own proprietary project.
Of course, with a GPLed project, it *can't* be forked to a *proprietary* project, which is precisely why some prefer it. Another GPLed fork is no big deal: its still under the GPL, its code remains "free", *and* modifications in one can still be used in the other...
That last bit about modifications is probably why GPL forks don't happen so often.
start a GPL-Ripoff organization that you can join for a fee
Last I heard "employees" don't pay their "employers". What you're describing here is not a "company doing in-house work" by anyone's common-sense understanding of the phrase.
We take GPL'd code, modify it and allow you to use it
FOR A FEE...
The rest of your fanciful theory is irrelevant because as soon as a judge hears about the fee, you'll be *toast*.
What, you think judges aren't annoyed by deliberate and *obvious* subterfuge? You'd better look up "willful copyright infringement" in the copyright law before you go to the trouble of setting up "GPL Ripoff Inc."...
This issue will probably be tested in court some day.
Eben Moglen would win this case in his sleep... Hell, Eben probably wouldn't even have to get up and say anything: the judge will more than likely blow his/her stack even before your lawyer finished with their fast-talking...
Yea, the FSF would just **love** for someone to try something this asinine.
But how can Bob prove that the work was ever licensed under the GPL if it is not included in a major distribution's repository? What happens if a judge rules that Alice had no right to distribute the work in the first place, such as if the GPL work infringed a third party's look-and-feel copyright?
Wait, s/GPL/AnyOtherLicense/g and these hypothetical problems could still happen. How are these examples GPL-specific?
If it had been licensed under the BSD license IBM could certainly have created a proprietary version without giving back, but to what end? They already had their own perfectly adequate UNIX system.
They *still* have their own "perfectly adequate" proprietary UNIX, so why are they contributing to Linux despite that? This question is especially pertinent here when you consider that much of what they're contributing to Linux actually comes from their own proprietary UNIX!
I doubt that Linux developers would just stop development because their grandfathers' Big Computer Company was competing with them.
Uhh, IBM *is* one of the "Linux developers" now. Try grepping the Maintainers file for "ibm.com". You might be surprised.
IBM isn't worried about competition from Linux, they're worried about competition from their corporate *rivals*, a worry that goes away, however, if they only release their stuff under the GPL, since then their *real* competitors can't use their own stuff against them.
Slashdot Mirror
as an end user of the CPU, you don't get one, nor need one.
*sigh*
Doesn't matter, I justed checked the net, and Intel's downloadable/modifiable microcode is copyrighted.
One example here: https://issues.rpath.com/browse/RPL-2546
Why is my work derivative of Stallman's if I do a CALL to his library, but not derivative of Intel's if I do a REP MOVSB, thus invoking their microcode?
Because of the difference in the copyright licenses.
Look, I get you don't like RMS's licenses, fine, that doesn't change the fact that for this particular issue, there is no fundamental difference between microcode and software when it comes to copyright. They are both copyrightable, and our CPUs' microcode is almost certainly copyrighted. The only difference is you like Intel's copyright more than RMS's. Fair enough.
I don't understand why GPL proponents can't just stick with the facts instead of equivocating between the original source and the code added to it.
Because our whole subthread started with Insanity Defense saying:
The purpose of the GPL is to keep the project as a whole "free", not just the original version.
"not just the original", so I haven't been equivocating, it never has been about "original" versus "added" code, its been about the "project"... at least until you just added that junk in to change the subject.
GPL allows distribution within an organization
What you've described is not an "organization"...
what I'm proposing (as thought problem, which I assumed you would understand)
Oh great, here it comes...
Well, whether or not *I* get it, my point was that a judge will not be amused by your "thought problem", which in turn, renders its likely-hood of actually happening to virtually nil.
intra-organizational distribution
Oh, is *that* what they call it now?
Fine, whatever...
Lesse, midnight on a boring middle-of-the-week Wednesday, just got through watching an old rerun of Clint Eastwood in A Fistful Of Dollars on the WGN Late-Nite-At-The-Westerns, but there's nothing good on now, and nothing else to watch on DVD, so what is kdawson's answer to this dilemma?
"Eureka! A flame-fest between the BSD Zealots and the GPL Fanatics, that ought to keep me entertained for the next 4 or 5 hours!"
[rummages through the inbox looking for good dry kindling, a match, some dynamite, and ...]
Come on, Guys and Gals, this is a setup piece for a flame-war, if I've *ever* seen one, you've *all* been had...
Hey I actually wrote all of Linux in my basement, and all these geeks stole it! I emailed it to Linus and he stole it, I just never bothered to mention it until now. How can you possibly prove otherwise! Yes you can't prove anything, not beyond a shadow of a doubt.
[Citation ne...
Ah hell, 4 digit UID... never mind.
it means that proponents of the GPL will use legal pressure to require you to do this.
Bullshit.
http://clisp.cvs.sourceforge.net/*checkout*/clisp/clisp/doc/Why-CLISP-is-under-GPL
RMS: If you don't change to using the GPL, then you'll have to stop using readline.
Not "switch to the GPL or else I'll call out the lawyers", but "if no GPL *then* stop using readline". The option to stop using the GPLed code was explicitly mentioned at the beginning. As the clisp author mentions, readline was not essential to his app, he *could* have easily ripped it out if he wanted to; clisp in no way depended on it.
The clisp author didn't even admit RMS was right; he essentially said, "whatever, it's easier this way," and GPL'd clisp.
Oh really?
RMS: I hate to have to play this role with a fellow hacker, but...
BH: I'm sorry too, as I am very indebted to the GNU project.
RMS: If you can get away with this then any company can get away with it.
...
Is that what you want?
BH: No, of course!
RMS: If you do succeed in circumventing the GPL for readline, you would be blazing a path for every commercial company that wants to do it. Would you really like that result? Wouldn't it be best to eliminate this dispute by using the GPL for Common Lisp?
BH: Would be best for me, true.
RMS: Totally aside from readline, the Lisp system would be more useful that way.
BH: This and the following are convincing me:
* Up to 1991 the decision whether using GPL or not, was simply a matter of philosophy or copyright policy. The success of Linux, however, demonstrates how a speed and quality of development was achieved which would have been impossible without access for everyone to the source of everything.
* Finding co-developers for other Lisp packages or testers for other hardware/OS platforms might be easier if I release full source.
So be prepared to seeing CLISP's source before Christmas.
I've snipped out the technical argument over the linking. (Can anyone blame BH for misunderstanding or disagreeing with that? Hell, just look at all the heat in this thread!) Otherwise your description of Bruno's response does not at all match up to what he actually said, methinks. Bruno didn't "give up" in the end, in the end, he *voluntarily* agreed with RMS for several reasons, even if he still didn't agree with, or understand, the FSF's linking argument.
A hint for you, Mr. piojo: If you're going to try your hand at revising history, you'd better make damn sure that the *unrevised* history still isn't out there on the web for all to see...
Nor (and this is the disputed point) do they get to control the reproduction, distribution, and/or use of a work which merely references their work.
???
If there's a 'copyright infringement' hiding anywhere in that "merely references", then yea, this is *definitely* the disputed point.
In a perfect world, I would look at the degree to which I'm using a library (and several other questions I and others have mentioned), and make a decision as to whether my work is a derivative work.
In other words, allow those accused of an offense to decide their own guilt? That's not my idea of a perfect world, but whatever...
If it is, my work needs to be GPL.
Or you just stop using the GPLed work in your work. *Your* work being GPLed is not the only way to resolve the dispute. There may be some financial damages awarded regardless, but ultimately whether *your* work gets GPLed or not is up to you, not the plaintiff. Stop using the other code, and your code can stay your own.
Um, no. There would have to be criminal intent. Without intent, it's strictly a civil issue.
FYI, "willful copyright infringement" (willful == intent), does allow for incarceration as a possible punishment, although the kind/amount of punishment is up to the judge, not the plaintiff. I think thats what the GP was trying to refer to.
If established case law (and I'm not up to date on this) says that linking to libraries (without distributing those libraries) doesn't constitute a derived work, how can a copyright license (which then wouldn't apply) change that definition?
Maybe because your "established case law" doesn't actually exist?
Why is everyone in this thread assuming the FSF's interpretation is somehow "obviously invalid"? The GPL has been around for 20 years, if it was clearly wrong, like, for example, it contradicted "established case law", don't you think we'd know that by now? Do you really think the FSF would keep sticking their neck out on this point, practically begging someone to take a swing, for 2 decades now, if they *knew* it contradicted "established case law"?
Reality check: there won't ever *be* any "established case law" on this issue until someone actually decides to challenge the FSF's interpretation in court... 20 years later... still waiting...
That's _reference_, not _derivation_. That (IMO, IANAL) no more makes the application derivative ...
IANAL either, but see swillden's response to my post just above yours:
In the case of the IBM attorneys I worked with, there is simply no question: They're pretty sure the court would side with the FSF, and in any case don't want to expose IBM to the risk that the court might side with the FSF, ...
If my work is not a duplicate or derivative of yours (regardless of what nasty hack I used to avoid making it so),
If we assume that a judge *agrees* that your "nasty hack" does indeed avoid infringement, then I'll agree with you (but not necessarily with the judge). :)
I was going by the GP's tone and assuming that the hack itself was legally questionable, thus proving bad intent almost by itself.
Thus, despite the fact that Apple has a copyright to Mac OS X, they don't get to control distribution of programs written for it.
Well, technically, they could *try* to make that claim (if this were absolutely impossible via copyright, no OS would need to explicitly permit it - and yet many/most(/all?) do just that in their licenses), though obviously no OS maker would do this since then no one would use their OS.
Despite the fact that Intel has a copyright on the microcode in its processors, they don't get to control distribution of programs written for them.
Same reason as above: no one would use their chip if they tried this...
Though I'm less sure about the possible legality of their even trying - this may already be handled by legal precedent. I dunno, I've never *seen* a copyright license for a CPU's microcode... :)
However, Larry has had a run-in with a fanfic writer
Ahh, a "fan"!
[insert clever reference to the movie 'Misery' here...]
That explains it. :)
The whole point of this sub-thread is comparing the GPL to the BSD. I'm just saying that "keeping the project as a whole "free"" isn't a logical reason to use the GPL instead of the BSD
And my point was that a BSD project can have a proprietary fork, but *not* a GPLed project, thus there *is* a difference: a *proprietary* fork *might* end up meaning no reciprocation, no give-back, no collaboration on a common base... it all depends on the "good will" of the company doing the fork.
Right or wrong, some people, *and* other companies!, don't put much faith in seeing "good will" from a profit-motivated corporation, especially ones, like, say, MS...
All very good points... and happily, none of them contradict mine.
Take the AARP. It charges its members a membership fee and it's clearly an organization just as much as the FSF or IBM.
Is the AARP distributing modified GPL apps to their members without the source code? If not, then your point is?
My point wasn't very clear: That "fee" will show the "willful" intent of your action, which is to, in effect, "make a buck off somebody else's work". That is not by itself illegal, but when you add that to the copyright infringement, because distribution is happening without providing the source code, the judge will more than likely put them together and come up with "this ain't no accident, they're doing all this *knowingly* and *purposefully*". Intent does matter (to a judge as well as a jury). As for all your hand-waving about distribution (from GPL3):
To "propagate" a work means to do anything with it that, without
permission, would make you directly or secondarily liable for
infringement under applicable copyright law, except executing it on a
computer or modifying a private copy. Propagation includes copying,
distribution (with or without modification), making available to the
public, and in some countries other activities as well.
"making available to the public" is the key. Judges are savvy enough to know that an organization defined as "members of the public which pay a fee" is, in fact, "the public", and the only reason for said organization's existence is to make you a profit off the act of copyright infringement (by "making available" without the source code). Your lawyer trying to argue otherwise is going to sound like a Nigerian scam artist making his pitch, and will just piss the judge off. Most judges aren't idiots: they've seen plenty of scams and scam artists in their career.
At this point, you have "willful copyright infringement", which can go beyond just monetary damages and right into jail-time, if the judge so chooses. This is why you don't want to annoy a judge with an obvious subterfuge. They see a lot of that anyway, so usually have a very low tolerance for BS.
In other words: you aren't going to fool anyone with this gag.
LGPL v3 has the following:
Yep, you're right, it goes into more detail (I didn't check it - just looked at the GPL), but you'd almost *expect* the LGPL to go into more detail, it *has* to, since this is why it exists (explicitly allowing more interaction with other non-GPL code).
I'm still not happy about "linking", though: can you provide a waterproof definition of linking in the context of Java programs?
If you mean "waterproof" in a *legal* sense, I don't believe thats possible for *any* language. :)
Thats what I was referring to at the end about "vagueness of the copyright law itself".
As far as I'm concerned a licence should be as short as it's possible to make it while communicating clearly what permissions it does and doesn't grant. Trying to make a licence too general results in it becoming longer and harder to understand. Using specific terms with agreed meaning is simply good communication.
Ok, I agree with all that, however the point I was trying to make is that the lack of "specific terms with agreed meaning" is inherent in the copyright law itself. Copyright licenses get their meaning and value from copyright law, they do not "stand alone".
Simple (probably bad) analogy: The copyright licenses are the "frame of your house", and the copyright law is the "foundation that you build your house on". No matter how strong you make the frame, if the foundation has a weakness, you're still in danger of losing your house, and sometimes, you have to *try* to overcome a possible weakness in the foundation by reinforcing the frame, which would *not* be necessary if the foundation was known to be solid.
The reason licenses aren't clear and easy to grok is because the underlying law isn't *helping* to *make* them clear and easy to grok.
The absolute worst offender, when it comes to unhelpfulness, is copyright law's concept of "derivative work". The meaning of this changes almost every bloody time the Supreme Court, or an Appeals Court, makes a decision concerning it! All the arguments elsewhere in this thread about linking are basically arguments over *technicalities*, and *none* of them can be clearly answered until we, or at least our lawyers, have a clear and precise legal definition of what a "derivative work" is. Alas, technology is moving much faster than the law, and the courts aren't able to update the law fast enough to keep up with reality. This is our current dilemma.
Attention Moderators: The Troll moderation does not mean "I disagree". If thats what it *meant* then thats what it'd be *called*.
you might consider that people claiming to be open WRT the GPL may imply open and can never be closed
This is a perfectly polite, and perfectly VALID *opinion*, whether you agree with it or not. If the parent is a Troll, then so is the grand-parent, the great-grand-parent, and virtually every freakin' post in this entire discussion thread!
*sigh*
in the EU where soft-where patents don't count
Not until the EU can explain why they "allow" them, just don't "enforce" them. That doesn't make any sense: why allow software patent application if you never intend to enforce it... unless the "enforce" part is just an implementation detail left for v2.0 that is on its way?
"The purpose of the GPL is to keep the project as a whole "free" not just the original version."
A group that forms the core development for a project maintains control of that project even if somebody forks it to create their own proprietary project.
Of course, with a GPLed project, it *can't* be forked to a *proprietary* project, which is precisely why some prefer it. Another GPLed fork is no big deal: its still under the GPL, its code remains "free", *and* modifications in one can still be used in the other...
That last bit about modifications is probably why GPL forks don't happen so often.
and choose a different license then.
Uhmm, did you notice that the title being used for this subthread is "Re:GPL Fanatics"?
Honest mistake, sometimes I don't notice when the title has changed either...
start a GPL-Ripoff organization that you can join for a fee
Last I heard "employees" don't pay their "employers". What you're describing here is not a "company doing in-house work" by anyone's common-sense understanding of the phrase.
We take GPL'd code, modify it and allow you to use it
FOR A FEE...
The rest of your fanciful theory is irrelevant because as soon as a judge hears about the fee, you'll be *toast*.
What, you think judges aren't annoyed by deliberate and *obvious* subterfuge? You'd better look up "willful copyright infringement" in the copyright law before you go to the trouble of setting up "GPL Ripoff Inc."...
This issue will probably be tested in court some day.
Eben Moglen would win this case in his sleep... Hell, Eben probably wouldn't even have to get up and say anything: the judge will more than likely blow his/her stack even before your lawyer finished with their fast-talking...
Yea, the FSF would just **love** for someone to try something this asinine.
But how can Bob prove that the work was ever licensed under the GPL if it is not included in a major distribution's repository? What happens if a judge rules that Alice had no right to distribute the work in the first place, such as if the GPL work infringed a third party's look-and-feel copyright?
Wait, s/GPL/AnyOtherLicense/g and these hypothetical problems could still happen. How are these examples GPL-specific?
If it had been licensed under the BSD license IBM could certainly have created a proprietary version without giving back, but to what end? They already had their own perfectly adequate UNIX system.
They *still* have their own "perfectly adequate" proprietary UNIX, so why are they contributing to Linux despite that? This question is especially pertinent here when you consider that much of what they're contributing to Linux actually comes from their own proprietary UNIX!
I doubt that Linux developers would just stop development because their grandfathers' Big Computer Company was competing with them.
Uhh, IBM *is* one of the "Linux developers" now. Try grepping the Maintainers file for "ibm.com". You might be surprised.
IBM isn't worried about competition from Linux, they're worried about competition from their corporate *rivals*, a worry that goes away, however, if they only release their stuff under the GPL, since then their *real* competitors can't use their own stuff against them.
I think you've completely missed the GP's point.
Mad Max ... the most money or the biggest guns
No, no, no.
If you're going to use a Mad Max reference, the *only* proper analogy allowed here is: "the biggest V8".