Thanks. I had in mind situations where although you have final say on the use of something -- say, your private museum --, if you let the public in, even for free, then it must be under conditions subject to other laws (e.g. non-discrimination, for an extreme example).
No no no no no. They have to simply have to prove that SCO does not own derivate code. It does not matter you how you define derivative.
Well, I hope you're right. But "own" may have been too strong a word. SCO claim that a license gives them some say on what IBM can do with what they call "derived code". (Specifically they claim that its terms prohibit IBM from open-sourcing "derived code".)
You're saying that SCO have no claim under any definition of "derived", period. I hope you're right. But if not, then I don't see how a discussion over that definition could be avoided (it already happened, to some extent, in the AT&T "clarification letters" that Dennis Ritchie posted somewhere). My fear is that causing such a discussion is exactly what the SCO Gambit is about.
It does not matter. Copyright itself is discriminatory and free-enterprise unfriendly.
So, I take it, your answer to the original question is that a copyright holder can grant "free redistribution, but only to males", period. I'm not so sure...
First of all it really does not matter how you define "derivative". IBM simply has to prove that they themselves own the derivative work and not SCO. They don't have to redefine the word
Well, IBM are up against extravagant "we own all derived works" claims by SCO. So they will have to argue the meaning of "derived", and make it weak enough to at least obliterate SCO's claims. My hope is that the GPL's (vaguely similar) clauses don't lose teeth in the process.
The whole GPL thing is secondary. IN this case IBM is countersuing using the GPL.
I'm not talking about this. My fear is that the SCO Gambit is designed to generally weaken what licenses can forbid to do with derived works. That precedent would then be used later in another case. (E.g., some company deliberately setting itself up to be sued for linking GPL code.)
Finally let's say that you can define a derivative in such a way that it completely invalidates the GPL. What would prevent me from getting a hold of windows code somehow and them making a minor change and calling it my own?
You might as well ask what prevents you from redistributing Mickey Mouse movies (with or without modifications). The difference is that Disney does not grant you that right, whereas the GPL does. It does "under certain conditions", but what if courts were to find those discriminatory, free-enterprise-unfriendly, or whatever? That'd bring us back to stage 1.
The GPL has some very tight definitions of what constitutes a derived work.
I think the FSF's definitions are good and hope a court will uphold them. But note (emphasis mine):
RMS: We have no say in what is considered a derivative work. That is a matter of copyright law, decided by courts (...) we cannot tell the courts to treat a certain kind of work as if it were derivative, if the courts think it is not. (...) I think we have a pretty good argument that nontrivial dynamic linking creates a combined (i.e. derivative) work.
Interestingly enough, this definition DOES NOT APPLY to kernel extensions. IOW, some of the sorts of things that SCO is attempting to to claim ownership of (XFS) would not be considered a derivative work even under the GPL.
That's good news -- that means SCO couldn't arrange to lose on their own preposterous "derivative" claims and take down the FSF's definition in the process. I hope this is true, but my fear is that's what they set out to do.
What about the author. The people who hold the actual copyright released it under the GPL because that was the most attractive license to them. If they wanted to release it under BSD or public domain they would have done so.
Right, they licensed under the GPL because they would like derivative works to remain open source.
But now SCO arrive with their preposterous, pre-emptive claim that because of their license, any "derivative works of Unix" belong to them and hence can't be open-sourced by IBM.
So IBM need a definition of derivative that invalidates SCO's claim but not the GPL, and I hope this is not too thin ice for them to tread.
For example:
The FSF's preferred definition would call a program derivative if it links to another. (Thus the need for the special LGPL when you want to allow linking.)
So, what if SCO claimed some IBM programs as derivative because they link? I have no idea is they can, but if they did, might IBM not be tempted to argue that linking does not, by itself, make a program derivative?
And if IBM wins (as few of us doubt), might this not establish enough of a precedent against the linking definition to support, someday, in some other staged lawsuit, the claim that GPL = LGPL because "linking does not, by itself, make a program derivative"?
My fear is that sending SCO to intentionally lose this case (why else would they behave like the fools they do...) is just a gambit to take down the GPL.
(No, that wouldn't kill Linux. But relaxing the definition of "derivative" might bring it back into the fold of proprietary development, and thus kill the momentum of Free Sotware.)
If the unlikely "partial invalidation" described comes to pass, the remnants of the GPL would come out more like a BSD license
Yes (we seem to have written this simultaneously) -- or, it now occurs to me, maybe even LGPL-like: letting people redistribute without letting them link from other code couldn't be enforced. Wouldn't that alone weaken the GPL enough to please the dark side?
Does anyone know if SCO plans to claim some works as derivatives by using the "linking" criterion (again, maybe just in order to lose and invalidate the criterion)?
What you are supposing, in essence is "what if the court made the GPL equivalent to releasing into the public domain"?
Not necessarily, maybe equivalent to something like BSD. (Copyright retained, redistribution is allowed but unconditionally.) What strikes me is that
SCO behaves as if they wanted to lose their case. (Everyone is grasping for an explanation, but I've not seen one that's convincing.)
Like the GPL, that case they seem to want to lose rests on an "all derived works belong to us" claim.
If (when) they lose, might this not create a precedent against any such clause? Surely there are people who'd like to restore a situation where, if you allow modification/redistribution, then it has to be a la BSD...
(Kind of like "first sale doctrine": once you're letting others redistribute, you can't dictate how, to whom, etc. I know this is different and the conclusion ought to be different. But the question is, will a court necessarily think that way?)
IBM already have a proprietary UNIX. What would turning Linux into another one do to help them?
Right, the number of IBM Unixes going from 1 to 2 is no big deal. What would be a big deal, and possibly in IBM's long-term interest, is the number of Free (GPL, non-closeable) Unixes going from 1, or a few, to zero.
SCO succeeding or failing has absolute nothing, zero, nada to do with the legal validity of the GPL. That's really just a red herring. In fact, an invalid GPL would make SCOs distribution of the Linux kernel illegal (as it would fall back to regular Berne convention rules)
What if a court ever found:
enforceable, the part of the GPL that says recipients can redistribute;
non-enforceable, (some of) the extra conditions imposed on the redistributor.
Surely, "free redistribution, but only to males" would not hold. But then the question is how it would fail. If the copyright holder sues someone for redistributing to females, will the outcome be to remove the restriction, or (as you claim) to disallow redistribution entirely?
Concept of 'derivative work' is not GPL-specific, nor it is
an invention of SCO. It is a standard term from the copyright law.
The
question is not whether the concept of 'derivative work' itself is
valid (present form of the copyright law asserts that it is,) but
rather, if SCO's claims, that Linux Kernel IS, in fact, a 'derivative
work' from SVR4 code, are valid.
This is not what I'm getting at.
Consider this: In all likelihood, a court will not find Linux derivative enough (of System V) that SCO could claim it its own, and enforce its licensing scheme. We know the arguments: The ideas were widely available, published in books, and in BSD, etc., yadda. Under such circumstances, a court will say, SCO's "all derivatives are belong to us" clauses can't be enforced.
But then, who's to say that stuff published under the GPL is not fair game as well? Again, we know the difference, but will a court? What if they were to rule that allowing redistribution automatically puts you in the BSD category...? That would take off all the GPL's teeth, effectively suppressing any intermediate regime between "closed" and "BSD". Isn't that Redmond's dream?
SCO's apparent attempt to pre-empt the GPL (by claiming they had a "derivatives belong to us" clause first) is playing to the gallery. It won't work, and they know it. In fact they seem to be doing all they can to lose this argument. Might the goal be to strike down, in the process, any kind of "derivatives belong to us" clause?
SCO is making this affair revolve around the concept of "derived work".
Their IBM suit asks the courts to regard all things Unix as "derived" and hence theirs.
OK, this is so preposterous that virtually no one (including themselves, probably) thinks it will work.
But now, the GPL rests precisely on something similar: a license that claims derived works for the community (if you accept it by redistributing the software).
Might someone hope that a setback to SCO's ridiculous claims would ipso facto provide a precedent against the GPL?
(Note: I know, the GPL and whatever contracts SCO had with IBM are very different beasts. But the question is whether a court could be fooled to conclude otherwise.)
But there are a lot of people that don't want their ten-year-olds exposed to explicit sex acts, especially of the degrading types which are prevalent in pornography.
Prevalent? How do you know?
Personally, I don't go looking for what I would consider degrading. Hence, I'm rarely ever exposed to it. Hence, I have no way of knowing how prevalent it is. And I guess this is true of most everyone (who watches porn).
So before everyone on/. immediately pounces on this as a ploy by the evil record companies/music
companies/religious right, maybe we should stop and address the perfectly valid concerns of children seeing sick porn.
Care to explain exactly how it's a valid concern? Prepubescent children are just that: prepubescent. Which means in particular: mostly indifferent to adult sex or "sick porn". Just as they are mostly indifferent to fine wines, cuban cigars, or elegiac poetry. Those tastes comes later in life.
So either present compelling evidence that ten-year-olds seeing some underage and probably illegally compelled porn acress getting anally raped by twenty guys isn't damaging his/her attitudes towards themselves and the opposite sex.
Oh, the clichés... So any porn actress is "probably" compelled, underage, and raped. Care to name even one example?
The fact that you probably can't, might say something about supply and demand in porn. Don't know about you, but for the people I know, sex is interesting between people who are into it.
So in porn, I prefer an actress who looks into the camera and seems to say: "This is sex, I know what I'm doing, and I'm doing it." Not the sort of acting that a raped underage easily delivers.
Plus, we're talking about underage viewers, not underage actors. It would be nice if you weren't mixing everything.
And have that evidence be compelling enough to persuade some typical suburban parents. Or present a compelling solution to the problem, since censorware is so universally reviled, and generally
ineffective anyway.
Typical suburban parents are much more reasonable than you give them credit for. They might realize that if there is no solution, it could be that there is no problem. Kids will learn about what adults do in bed. So what?
And no, neither "watch your kids 24 hours a day" or "teach your children about sex and pornography at an early age" are compelling enough. 24-hour surveillance is never possible with kids
"Watching your kids 24 hours a day" = prohibition. As you note, it's impractical. It's also stupid. Let people (and teenagers are people) learn by themselves!
and no matter how much you teach them, their views on the world will still be influences by the world around them.
Thanks God, yes, they'll learn about the real world by being exposed to it. They might even grasp, as you don't seem to, that porn is fiction. Anyway, kids mostly don't learn about sex from their parents; they learn from their peers.
Or make the consequences of shutting them down be so horrible that it's worth having a nation of sex
perverts.
So if teenagers see some porn, we shall become a nation of sex perverts! Again: porn = fiction.
I suppose you also believe that allowing alcohol under 21 would make us a nation of alcoholics? Well, most of Europe doesn't have an age limit for alcohol. (There may be one in the books, but not in practice. Your kid can buy your beer at the grocery store.)
Result, everybody is relaxed about it, and they have none of the stories of underage freshmen dying in stupid drinking contests, which fill half our campus papers' headlines every fall.
(The other headlines are usually about abortion. Why they are not found in most of Europe either, is left as an exercise for the reader.)
Not to mention ordinary (voice) directory assistance,
for which phone operators are starting to charge more and more.
A disturbing trend...
In Switzerland, for instance, unlimited local directory assistance
used to be part of the basic phone subscription, and you could
get it from any pay phone for the price of a local call.
(Or even for free, I don't remember. Plus every booth had all
the phone books in it.) Nowadays they charge outrageous prices
- roughly $1
from home and $1.50 from a pay phone, per inquiry. Last I checked they also charged for using the online white pages, but on the bright side, this seems to have changed.
France still has free directory assistance through minitel, but IIRC, you gotta pay dearly if you also want an old-fashioned phone book.
Live near NYC, need a copy of the Manhattan Yellow pages? Bell Atlantic wants something like $50 for them. Better ask a friend to save last year's edition...
The kind folks at The Web Union used to provide free shells, and gratis ad-free web space for non-profit pages. As the link shows, they have discontinued this and plan "re-emerge at some point in the near future as a low-cost hosting provider".
Slashdot Mirror
Ah, sorry, this won't fly. You are forgetting that SCO have exclusive rights on good engineering !
OpenSSL: Fixes CAN-2003-0543, CAN-2003-0544, CAN-2003-0545 to address potential issues in certain ASN.1 structures and in certificate verification code. To deliver the update in a rapid and reliable manner, only the patches for the CVE IDs listed above were applied, and not the entire latest OpenSSL library. Thus, the OpenSSL version in Mac OS X 10.2.8, as obtained via the "openssl version" command, is: OpenSSL 0.9.6i Feb 19 2003
- enter your email adress HERE
- click OK!
This is the BEST, FOOLPROOF way to NOT GIVE YOUR ADDRESS AWAY!!"But I hope you're right in this case. If so (and if the courts uphold the FSF's definition of a program as derivative if it links), then the GPL seems safe from being dumbed down to an LGPL or BSD equivalent though this sort of strategy.
How is it a problem if half the code was leaked?
Tell us where to get the other half and then we'll start to worry.
Well, I hope you're right. But "own" may have been too strong a word. SCO claim that a license gives them some say on what IBM can do with what they call "derived code". (Specifically they claim that its terms prohibit IBM from open-sourcing "derived code".)
You're saying that SCO have no claim under any definition of "derived", period. I hope you're right. But if not, then I don't see how a discussion over that definition could be avoided (it already happened, to some extent, in the AT&T "clarification letters" that Dennis Ritchie posted somewhere). My fear is that causing such a discussion is exactly what the SCO Gambit is about.
So, I take it, your answer to the original question is that a copyright holder can grant "free redistribution, but only to males", period. I'm not so sure...
Well, IBM are up against extravagant "we own all derived works" claims by SCO. So they will have to argue the meaning of "derived", and make it weak enough to at least obliterate SCO's claims. My hope is that the GPL's (vaguely similar) clauses don't lose teeth in the process.
I'm not talking about this. My fear is that the SCO Gambit is designed to generally weaken what licenses can forbid to do with derived works. That precedent would then be used later in another case. (E.g., some company deliberately setting itself up to be sued for linking GPL code.)
You might as well ask what prevents you from redistributing Mickey Mouse movies (with or without modifications). The difference is that Disney does not grant you that right, whereas the GPL does. It does "under certain conditions", but what if courts were to find those discriminatory, free-enterprise-unfriendly, or whatever? That'd bring us back to stage 1.
I think the FSF's definitions are good and hope a court will uphold them. But note (emphasis mine):
Interestingly enough, this definition DOES NOT APPLY to kernel extensions. IOW, some of the sorts of things that SCO is attempting to to claim ownership of (XFS) would not be considered a derivative work even under the GPL.That's good news -- that means SCO couldn't arrange to lose on their own preposterous "derivative" claims and take down the FSF's definition in the process. I hope this is true, but my fear is that's what they set out to do.
Right, they licensed under the GPL because they would like derivative works to remain open source.
But now SCO arrive with their preposterous, pre-emptive claim that because of their license, any "derivative works of Unix" belong to them and hence can't be open-sourced by IBM.
So IBM need a definition of derivative that invalidates SCO's claim but not the GPL, and I hope this is not too thin ice for them to tread.
For example:
The FSF's preferred definition would call a program derivative if it links to another. (Thus the need for the special LGPL when you want to allow linking.)
So, what if SCO claimed some IBM programs as derivative because they link? I have no idea is they can, but if they did, might IBM not be tempted to argue that linking does not, by itself, make a program derivative?
And if IBM wins (as few of us doubt), might this not establish enough of a precedent against the linking definition to support, someday, in some other staged lawsuit, the claim that GPL = LGPL because "linking does not, by itself, make a program derivative"?
My fear is that sending SCO to intentionally lose this case (why else would they behave like the fools they do...) is just a gambit to take down the GPL.
(No, that wouldn't kill Linux. But relaxing the definition of "derivative" might bring it back into the fold of proprietary development, and thus kill the momentum of Free Sotware.)
Yes (we seem to have written this simultaneously) -- or, it now occurs to me, maybe even LGPL-like: letting people redistribute without letting them link from other code couldn't be enforced. Wouldn't that alone weaken the GPL enough to please the dark side?
Does anyone know if SCO plans to claim some works as derivatives by using the "linking" criterion (again, maybe just in order to lose and invalidate the criterion)?
Not necessarily, maybe equivalent to something like BSD. (Copyright retained, redistribution is allowed but unconditionally.) What strikes me is that
- SCO behaves as if they wanted to lose their case. (Everyone is grasping for an explanation, but I've not seen one that's convincing.)
- Like the GPL, that case they seem to want to lose rests on an "all derived works belong to us" claim.
If (when) they lose, might this not create a precedent against any such clause? Surely there are people who'd like to restore a situation where, if you allow modification/redistribution, then it has to be a la BSD...(Kind of like "first sale doctrine": once you're letting others redistribute, you can't dictate how, to whom, etc. I know this is different and the conclusion ought to be different. But the question is, will a court necessarily think that way?)
- enforceable, the part of the GPL that says recipients can redistribute;
- non-enforceable, (some of) the extra conditions imposed on the redistributor.
Surely, "free redistribution, but only to males" would not hold. But then the question is how it would fail. If the copyright holder sues someone for redistributing to females, will the outcome be to remove the restriction, or (as you claim) to disallow redistribution entirely?Consider this: In all likelihood, a court will not find Linux derivative enough (of System V) that SCO could claim it its own, and enforce its licensing scheme. We know the arguments: The ideas were widely available, published in books, and in BSD, etc., yadda. Under such circumstances, a court will say, SCO's "all derivatives are belong to us" clauses can't be enforced.
But then, who's to say that stuff published under the GPL is not fair game as well? Again, we know the difference, but will a court? What if they were to rule that allowing redistribution automatically puts you in the BSD category...? That would take off all the GPL's teeth, effectively suppressing any intermediate regime between "closed" and "BSD". Isn't that Redmond's dream?
SCO's apparent attempt to pre-empt the GPL (by claiming they had a "derivatives belong to us" clause first) is playing to the gallery. It won't work, and they know it. In fact they seem to be doing all they can to lose this argument. Might the goal be to strike down, in the process, any kind of "derivatives belong to us" clause?
SCO is making this affair revolve around the concept of "derived work".
Their IBM suit asks the courts to regard all things Unix as "derived" and hence theirs.
OK, this is so preposterous that virtually no one (including themselves, probably) thinks it will work.
But now, the GPL rests precisely on something similar: a license that claims derived works for the community (if you accept it by redistributing the software).
Might someone hope that a setback to SCO's ridiculous claims would ipso facto provide a precedent against the GPL?
(Note: I know, the GPL and whatever contracts SCO had with IBM are very different beasts. But the question is whether a court could be fooled to conclude otherwise.)
Prevalent? How do you know?
Personally, I don't go looking for what I would consider degrading.
Hence, I'm rarely ever exposed to it.
Hence, I have no way of knowing how prevalent it is.
And I guess this is true of most everyone (who watches porn).
Care to explain exactly how it's a valid concern?
Prepubescent children are just that: prepubescent.
Which means in particular: mostly indifferent to adult sex or "sick porn".
Just as they are mostly indifferent to fine wines, cuban cigars, or elegiac poetry.
Those tastes comes later in life.
Oh, the clichés...
So any porn actress is "probably" compelled, underage, and raped.
Care to name even one example?
The fact that you probably can't, might say something about supply and demand in porn.
Don't know about you, but for the people I know, sex is interesting between people who are into it.
So in porn, I prefer an actress who looks into the camera and seems to say:
"This is sex, I know what I'm doing, and I'm doing it."
Not the sort of acting that a raped underage easily delivers.
Plus, we're talking about underage viewers, not underage actors.
It would be nice if you weren't mixing everything.
Typical suburban parents are much more reasonable than you give them credit for.
They might realize that if there is no solution, it could be that there is no problem.
Kids will learn about what adults do in bed. So what?
"Watching your kids 24 hours a day" = prohibition.
As you note, it's impractical. It's also stupid.
Let people (and teenagers are people) learn by themselves!
Thanks God, yes, they'll learn about the real world by being exposed to it.
They might even grasp, as you don't seem to, that porn is fiction.
Anyway, kids mostly don't learn about sex from their parents; they learn from their peers.
So if teenagers see some porn, we shall become a nation of sex perverts!
Again: porn = fiction.
I suppose you also believe that allowing alcohol under 21 would make us a nation of alcoholics?
Well, most of Europe doesn't have an age limit for alcohol.
(There may be one in the books, but not in practice. Your kid can buy your beer at the grocery store.)
Result, everybody is relaxed about it, and they have none of the stories of underage freshmen dying in stupid drinking contests, which fill half our campus papers' headlines every fall.
(The other headlines are usually about abortion. Why they are not found in most of Europe either, is left as an exercise for the reader.)
Not to mention ordinary (voice) directory assistance, for which phone operators are starting to charge more and more.
A disturbing trend...
In Switzerland, for instance, unlimited local directory assistance used to be part of the basic phone subscription, and you could get it from any pay phone for the price of a local call. (Or even for free, I don't remember. Plus every booth had all the phone books in it.) Nowadays they charge outrageous prices - roughly $1 from home and $1.50 from a pay phone, per inquiry. Last I checked they also charged for using the online white pages, but on the bright side, this seems to have changed.
France still has free directory assistance through minitel, but IIRC, you gotta pay dearly if you also want an old-fashioned phone book.
Live near NYC, need a copy of the Manhattan Yellow pages? Bell Atlantic wants something like $50 for them. Better ask a friend to save last year's edition...
The kind folks at The Web Union used to provide free shells, and gratis ad-free web space for non-profit pages. As the link shows, they have discontinued this and plan "re-emerge at some point in the near future as a low-cost hosting provider".