The original analogy didn't talk about the Everest either. I guess Hadlock added it for effect. But just for the fun of it, let's bring it all to Everest sizes so... 15nm turns into 8848m, 590e9 bigger. 0.049 inches (head length) turns into 28,9e9 inches, some 73e9cm. Let's say 7.3e8m or 730000km. Never heard about a plane that big, did you?;)
I think you missed his point - successful desktop OS are successful because they just work - Linux is not there yet
Funnily enough, I think he didn't miss the point. The "Linux is not there yet" poster compares later in the thread the whole Windows software package with just the Linux kernel (because Gnome 2/3 aren't OSes by themselves). Never mind that thousands of public administration and education computers in Spain run just a modified Debian (Linex -which is possibly going to disappear-, Guadalinex, Molinux) with no real problems.
And the OEM argument is flawed too: Dell offers laptops with Linux (not all over the world, mind you) and so do many other OEMs. Is that enough "it just works" for you?
According to Upgrading and Repairing PCs it would currently be like having 4 Sears Towers floating on their sides, side by side, 5mm over the ground moving at almos 7800 kilometres per second while reading 2 centimetres long bits on tracks 30 centimetres away from each other.
Copyright and the market mix economics and authorship - I'm just bringing it up.
Oh, but the market didn't create copyright. The British Crown did, to keep printers checked. That it taxed the right to copy was a nice side-effect that quickly outgrew its mandate and turned into a lobby that managed to keep it going when the Crown found no more use for it.
As has already been said, Germany had more works created without copyright than Britain with it, in the same period. The breakaway North American colonies hurried to abolish copyright, thus making all continental production a free-for-all. That they later reinstated it says more about their printers' lobby power than their respect for it.
Contracts, as you say, keep being leonine, so little has changed with copyright. The entities persecuting copying aren't author guilds but industry associations, which clearly outlines who are the ones really not winning money over this whole deal. Of course there are weird chimeras like Spain's SGAE which include both editors and authors, but certainly you will see the senselessness of a union that includes both employers and employees: it really is a union of employers with employees thrown just to hide the fact.
But you are worried about stealing the authorship of a work, so I will show you an example of how it works: say you are a japanese manga artist who writes a successful manga series. In fact, it is so successful that it spawns anime recreations and sequels. And, suddenly, someone produces a a movie that is suspiciously like yours, which isn't too surprising since they learned coloring techniques from you when you were creating the manga. Do you get royalties or even acknowledgement? No, you don't, and if you dare speak you will probably get hammered with lawyers. So much for copyright protection.
Remember, authors usually have contracts that pay them per copy sold by the authorized publisher. As long as the publisher can keep making sales, the author can keep getting money. When an unauthorized printer sells copies of a book, it is extraordinarily rare that the author gets any money.
Remember that those contracts leave the author indebted unless he is a best-seller. So forget being paid per copy: most authors don't even get the chance to achieve that.
Samuel Clemens depended upon his writings to continue to sell, for his livelihood.
Which is not the same as depending on his copyright royalties, is it? Mind you, Samuel Clemens (better known as Mark Twain) was mainly a journalist. He got paid for his articles once (such is the fate of work for hire, isn't it?). He also was an orator, and he got paid for doing that too... once per show. He also got the patronage of a very rich man of the time. As you can see, it doesn't look like royalties were a problem until he took a stake in the intermediation business.
Would you have a problem with people taking games from the Humble bundle and selling them as their own work? How about someone's self published book that you didn't like?
Nice, you start mixing economics and authorship. I consider authorship something to be preserved, but authorship doesn't bring money to your table. Doing a good work does: when I develop software for an organization i get paid for the software itself, not for being its author. Of course you may argue that both are conjoined, but that won't happen if I'm hired to maintain someone else's software... and I will still get paid for my work. But I get paid once, not a thousand, a million times.
Even more: a company I worked for keeps selling the software I developed, making a profit on it and my authorship isn't recognized. As authorship is already not being recognized and already being profited from, what's your point about it?
How do you feel about a self-published author's copyright?
If his work is good enough I won't need to worry about that author's copyright: I will buy it. Actually, I already have: I've spend over 50 euros this Christmas buying copyrighted items. Of course having those items reasonably priced has helped a lot (see the $2.37M or so that the last Humble Indie Bundle has collected).
Excuse me, intermediaries aren't paying most authors anything at all. You know, 5% (if lucky) of profits and then they take all and sundry costs away from that 5%. Usually the author ends up owing money to the intermediaries.
Besides, if your reasoning was correct, then intermediaries would have never paid authors before copyright existed. You may be surprised, but they actually did. So the author getting money from his work is not dependent on the existence of copyright. It never was, and it never will be. What's even more: an author can now eschew all those intermediaries and reach the public directly. And people actually pay. It may not be a great amount, but it helps people make do if they are good enough.
Of course, there is the historical comment above mine to take into account. It's funny when facts contradict your theories, isn't it?
As many times as you need to hear that a download doesn't automatically equal a sale.
Like it or not, protection of a work is needed to keep the creative process going.
That is false, was false and will keep being false. The only reason we have the second part of "El Quijote" was that someone wrote an apocriphal second part, for example. "Protection of a work" as you call it is needed to keep the intermediaries accumulating money, not the authors: it never was an author protection tool, it isn't now and it will probably never be.
The lost sale doctrine is fancy talk at the best, by the way. I can't help thinking about poor home to home ice sellers, losing sales because people started buying fridges.
R is under license A, S is under license B. If licenses A & B are incompatible and the authors of R & S don't give permission, R & S can't be conjoined. Really, I have been saying that the whole time. Perhaps I'm explaining myself wrongly but I'm running out of ways to make you understand that.
Your claim for example was that A submission of GPLv2 code constitutes a violation of the kernel's license, programs don't have legal rights corporations and individuals do.
Sorry, perhaps you don't understand that "kernel's license" means the license that goes along with the kernel. Most people do.
The kernel can't have a license if all the individual copyright holders have full control.
Of course it can if every part added to it is added under the same license. Which is the case because you can't legally add code under any license but GPLv2 only without the permission of the kernel developers.
You are begging the question. F says no contract existed and he never accepted the contract. That's counter evidence.
Ok, good luck trying that defense against infringing any other license or contract. "No, sir, I didn't understand that the terms of the license mentioned all over the place applied to me. By the way, I don't understand why the traffic regulations apply to me either".
Maybe, maybe not. I think it might be a lot easier to take parts of the kernel and conjoin them with GPLv3 code. Taking the entire kernel is likely to grant Linus standing, despite his belief he doesn't have standing. In other words I think Linus is making two mistakes.
Blablablah. References on "We have kernel developers on record saying that's not their understanding". You state that, you prove it. Anything else is trying to distract me and it won't work, at least this time.
Nottage v. Jackson the author is the person most closely responsible for the work being produced.
Very nice. That actually supports my stand that you can't happily relicense something without the author's permission.
Feist -- facts are not copyrightable but compilations of facts are
As the kernel is neither facts nor a compilation of facts, this doesn't apply.
Wheaton v Peters -- A conjoined work has its own license
Wheaton vs. Peters is a case of failure in applying for copyright: defective in publication is the term used. AFAIK this hasn't happened with the Linux kernel and even if it had, current law may have different provisions than XIX century law. Please, try to provide references not earlier to the advent of software. Copyright law on software has some specifics that may make relying on previous sentences unsafe.
Yes. That's exactly what it takes. There need to be provable events which constitute signing the license. Your entire argument is predicated on the belief that everyone who contributed to the kernel explicitly agreed to the kernel license including the no upgrade policy collectively and thus individually they cannot do anything more than relicense their own contributions.
Oh dear, you really don't get it. If you create half a book and I create the other half we both have the right to restrict the publication of that book... and keep the right to publish our own parts. What's even more, in the case of the Linux kernel, everything added to it is based on earlier versions of the kernel which, going back, return you to the 0.12 version in which Linus Torvalds (AFAIK, the only developer then) chose GPLv2-only as license. So yes, there was an original entity (Linus Torvalds) who held the copyright. And everything else derives from it and is forced to follow the original copyright restrictions. If what you say was correct, just reading the ROM BIOS on IBM's first PC wouldn't have been enough to disallow any developer to create a clone. And yet Compaq and others had to go through the "clean room implementation" hassle in order not to infringe on IBM's copyright. No matter
No that's false. That's called an implicit contract and the law does not accord them very much respect at all.
Implicit? It is so implicit that it is all over the kernel source files. You can only not see it if you willingly close your eyes. By the way, if that's implicit and invalid, then the rest of software licenses (both open and closed) are invalid. Of course, then all this discussion is futile, because they are implicit.
E can assert that F doing action K creates a contract, F can K and no contract is in place.
F accepts a contract which restricts F's actions (on how to distribute in this case). The contract exists before F does any other action.
We have kernel developers on record saying that's not their understanding. They believe that the kernel copyright is the standard GPLv2, but that the kernel collectively is not exercising the GPLv2 -> v3 transition. So again we know that's not true.
References, please. If what you say is true, there is no need to collectively transition to GPLv3. All it takes is someone taking the kernel and changing the license. They don't do it because they can't without permission of all the people that has provided code to the kernel. And that's because it is GPLv2 only.
That's again one of the myths of the open source community. Courts have held that entering into a conjoined work creates new licenses.
Excuse me? You talk about a license as being implicit and now dare to talk about new licenses being created? References, please.
In general publishers want unambiguous copyright but publishers have pushed ahead in conjoined works where individual copyright holders for a conjoined work didn't agree with relicensing. A good example of this is collections of articles originally published in magazines and then republished in a book.
I can't help but wonder what would have happened if the authors decided to sue... or why they didn't.
You are trying to have it both ways. There is no "kernel license" without a corporate entity that holds that license. Either there is a bunch of individual licenses or there is an implicit corporate entity.
What's an implicit corporate entity? "We, the kernel developers, being of caffeinated mind and tired body..."?
Further there is no obligation on any contract to enforce its terms. A3 is not infringing by indicating his understanding of the contract is more lenient.
Oh, that's true. You can ignore a contract... and you can ignore a law. What happens after that fact (losing your distribution license) doesn't change because you "misunderstood". There is no good (for some values of good) will eximent.
Illegal from whom? Who in your opinion has standing?
Illegal for A3, who is not the owner of the code he wants to relicense (if he doesn't have the permission of the authors). Of course, the ones entitled to do the relicensing are the authors and the people they collectively allow to relicense.
No, it is as simple as everybody knowing that's the kernel's license. It's there for everyone to read before they contribute and anyone who has spent the minimum time providing code already knows that Linux kernel's license is not the standard GPLv2 but a GPLv2 only one.
Massive it is, conjoined it may be too, but the licensing isn't ambiguous at all, at least on the part that specifically voids the possibility of turning it into GPLv3 or later.
You are just assuming that Linus' has authority to speak for the whole product but Linus does not have, nor even claim to have that right.
When did Linus enter this thread? I stated that Ax may sue about Ax's code on the cases I put forward. I stated that B may sue about B's code on the cases I put forward too. Why is it so difficult for you to understand that if you infringe the license of the code I developed I can take you to task about that code only? It doesn't matter if there is a thousand more developers. I'm responsible of taking legal care of the code I wrote, and that's it. Now, the developers can choose to present a united front in order to save themselves the hassle of proceeding with individual suits.
If A3 asserts that his code, in the kernel, does not have the no upgrade clause that's binding.
And A3's code won't have entered the kernel, because it is GPLv2 only. A3 will be infringing the kernel license by trying to insert standard GPLv2 code in it. Since that is illegal, A3 has no standing about trying to change the kernel's license.
Your whole argument turns around the point of sneaking non GPLv2-only code in the kernel. It doesn't work like that: if you do it you are infringing and cannot distribute. So please stop beating that argument to death, it's already buried under a kilometre thick of sedimentary rock. You can't lawfully insert standard GPLv2 code into the GPLv2-only Linux kernel. That's it, no two ways about it.
Just in case, here's another possible interpretation:
A writes some GPLv2 code call it X and puts it in the kernel.
Which makes the copy of X in the kernel GPLv2 with no option to change to GPLv3 or it can't enter the kernel.
B takes X takes a subset of it X'. X' is under GPLv2.
This can only happen if B is taking X from the standard GPLv2 source, not the Linux kernel.
B takes X' and mixes it with some code Y for which he has copyright, creating X'+Y. He releases the the combined work under GPLv3.
Correct as long as X' isn't taken from the X version in the Linux kernel.
C takes X'+Y and violates the GPLv3 but not GPLv2.
D takes X'+Y and violates both the GPLv3 and GPLv2.
1) Is B's release legal? Is GPLv2 compatible with GPLv3 or does the upgrade ban B from releasing X'+Y under GPLv3? Has B created a license which is unenforcible. If so, how is the new product licensed or is it just a violation of A's code and nothing more? Remember, Linus' claim is meaningless if it just applies to X', the real question is whether the can claim copyright authority over X'+Y on the basis of some sort of leadership over X'.
B's release is legal as long as X' doesn't derive from the GPLv2-only Linux kernel, which may happen if A publishes X as GPLv2. Since standard GPLv2 is forward compatible with GPLv3 you may relicense standard GPLv2 code as standard GPLv3 code. But you can't take GPLv2 only code from the Linux kernel and relicense it as GPLv3.
As you can see it all hinges on whether X' derives from the Linux kernel GPLv2 only version of X or some other standard GPLv2 version. In the first case you can't relicense as GPLv3. In the second one you can.
2) Can A sue B and win? Make it worse assume that X involves code from A1, A2, A3 and A4 with A1 and A2 granting permission while A3 and A4 don't. Does the fact that A1-A4 wrote their code solely for a collective work create collective responsibility? The popular answer on/. is no A3 and A4 retain all rights to subsets. But AFAIK courts do not look at things that way.
Your question has already been answered. And the partial permission issue isn't GPL specific. Refer to your friendly lawyer.
3) Can B sue C? Remember if the answer is yes then B has just upgraded the kernel to GPLv3. If the the answer is no, then you agree that once any part of the code license becomes questionable the whole thing unenforcible. This is exactly the situation with the entire kernel.
No, assuming that B could license his code as GPLv3 (automatically precluding its coming from the GPLv2 only Linux kernel), B can sue C over B's code. B can't sue C over X' and if C rewrites Y from scratch and stop using B's code he stops infringing on his further code. Infractions about previous code based on B's GPLv3 Y code may remain, though. But, of course, all this means you still think you can put standard GPLv2 code in the Linux kernel and get away with it. You can't: the Linux kernel is GPLv2 only, and if you want to put code in it you will need to make it GPLv2 only and thus unrelicensable as GPLv3... as long as it derives from the Linux kernel.
4) Can B sue D? What about A can he sue D?
A can sue D on the infractions related to A's code. B can sue D on the infractions related to B's code.
A writes some GPLv2 code call it X and puts it in the kernel.
And it becomes GPLv2 only or it infringes the kernel's license.
B takes X takes a subset of it X'. X' is under GPLv2.
No, it is under GPLv2 only.
B takes X' and mixes it with some code Y for which he has copyright, creating X'+Y. He releases the the combined work under GPLv3.
Which he can't because X' can't be GPLv3 as it is GPLv2 only. Do I need to go further? Your whole chain fails in the beginning because the Linux kernel is GPLv2 without the "or later" option so you can't take it and relicense it as GPLv3.
That's _exactly_ the issue. The Linux kernel _can't_ be automagically be relicensed as GPLv3 and thus fails your example because X' is GPLv2 only. Actually, when A wants to put his GPLv2 or newer code in the kernel he will have to relicense it to GPLv2 only or it just won't be accepted because the kernel is GPLv2 only. Userspace may be different, true, but userspace is not the kernel.
I must be confused. You are saying that if someone takes from GPL, modifies it, then distributes it, that they are *not* required to distribute the source, but that you are allowed to violate any license they attach (if any).
Perhaps I didn't explain it right. I wanted to say that you aren't force to distribute your product at all, and if you decide to distribute it you must do it in a specific way (making the source code available to those that the software is finally delivered).
GPL isn't reinstating them. GPL is forcing distributors to follow "extra" rules.
Either GPL is giving the end user extra rights or it is reinstating them, because you certainly can not modify the software according to current copyright law unless the author allows you to. The distributors can choose either follow the extra rule (no plural there) or not distribute... like they would have under the basic copyright law. But, of course, GPL wants to protect use, and distribution is not it.
It feels more like you are arguing that "here's how I feel and how I justify my personal opinion, and isn't a discussion about what happens or how it works in law or such."
Ad hominems aside, the facts are that basic copyright law doesn't allow you to use or distribute the software as you want, BSD doesn't guarantee your recipients' ability to use the software as they want and GPL does. In contrast, BSD allows you to distribute as you like (even denying the right to use and distribute to those that receive your version) and GPL doesn't in order to keep the freedom to use the software and, mostly, the freedom to distribute it. And I think giving the buyer/recipient the ability of using the software as he wants independently of the wishes of the seller/originator is quite something.
But the GPL *DOES* restrict other people. Its mythology that it doesn't.
Congratulations. Did I say at any time that GPL doesn't restrict people? Didn't think so: I have said several times it restricts distribution in order to keep allowing use. And this thread is all about the misconception of GPL restricting use when it doesn't.
The GPL is viral and ties to force me to make my code fall under GPL if I use GPL code with it. RMS has stated many, many times that this is his goal. People need to quit pretending otherwise.
No. Your code can be under any license you want _unless_ you distribute it. You are just required to distribute it under GPL.
There are many valid situations where I cannot give source back or make it open, and the GPL interferes with this all the time. It is not evil or bad to restrict source, there are times when this is a necessary and positive thing.
Not all the time, mind you. Just when you wish your right to ignore my conditions was more important than my requirements that you don't. By the way, I would like you to provide some examples of the positive side of restricting source code.
GPL is not freedom, it is giving up freedom in exchange for what some people think are more important freedoms. Its also incorrect to state that GPL is about the freedom of the code. It greatly impacts the freedom of the coder, you can't neatly separate that out like it doesn't matter.
There is not absolute freedom. GPL isn't absolutely free, BSD isn't absolutely free, Public Domain isn't absolutely free. And they aren't absolutely free because there is more than one person on the planet to use them, so some people won't be able to get what they want. GPL is free to use: you don't have to abide to any restriction as long as you just use it. And I, personally, consider GPL better than BSD because it will continue to be free: there is no option for anyone (besides the original authors, of course) to close the source down.
and not a grant of extra rights to the buyer/consumer.
No, of course not. The buyer/consumer is entitled to doing anything with the software, like making copies, modifying it, reverse engineering it... oops, copyright law requires the author to allow me to do that. If GPL isn't granting extra rights to the consumer then copyright law is taking them away (and GPL is reinstating them).
But if you distribute it, you are, required to distribute. The GPL exerts force on the sellers
You are not forced to distribute. You are forced to distribute in a specific way. But distribution is not use.
Your comment is interesting in two ways: what it answers (would you please reference Linus' assertion?) and what it doesn't (everything else). Excuse me for not trusting your expert opinion on software quality, by the way;)
Since the GPL specifically lets you make verbatim copies (as long as you don't distribute them), GPL doesn't restrict your use of the software when you run it (making a copy of it in memory, as you say).
How you are not "repeating" what others say when you sell the software they made, even if modified? GPL allows me, when I buy GPL software from you, to repeat (read publish/sell) what you have "said" even if you don't want me to, the same right you had when you received it. GPL doesn't force you to repeat anything: you can use your GPL software in-house and never tell anyone about it.
It doesn't AFAIK. At most you could consider the license termination clauses which forbide you from conveying the software to other users either in original or modified form unless you comply with the license. Since distribution isn't use, I consider it isn't covered. You can do whatever you want with the code as long as you don't turn copies of it to other people: it is then when you need to check compliance.
Slashdot Mirror
The original analogy didn't talk about the Everest either. I guess Hadlock added it for effect. But just for the fun of it, let's bring it all to Everest sizes so... 15nm turns into 8848m, 590e9 bigger. 0.049 inches (head length) turns into 28,9e9 inches, some 73e9cm. Let's say 7.3e8m or 730000km. Never heard about a plane that big, did you? ;)
I think you missed his point - successful desktop OS are successful because they just work - Linux is not there yet
Funnily enough, I think he didn't miss the point. The "Linux is not there yet" poster compares later in the thread the whole Windows software package with just the Linux kernel (because Gnome 2/3 aren't OSes by themselves). Never mind that thousands of public administration and education computers in Spain run just a modified Debian (Linex -which is possibly going to disappear-, Guadalinex, Molinux) with no real problems.
And the OEM argument is flawed too: Dell offers laptops with Linux (not all over the world, mind you) and so do many other OEMs. Is that enough "it just works" for you?
According to Upgrading and Repairing PCs it would currently be like having 4 Sears Towers floating on their sides, side by side, 5mm over the ground moving at almos 7800 kilometres per second while reading 2 centimetres long bits on tracks 30 centimetres away from each other.
Copyright and the market mix economics and authorship - I'm just bringing it up.
Oh, but the market didn't create copyright. The British Crown did, to keep printers checked. That it taxed the right to copy was a nice side-effect that quickly outgrew its mandate and turned into a lobby that managed to keep it going when the Crown found no more use for it.
As has already been said, Germany had more works created without copyright than Britain with it, in the same period. The breakaway North American colonies hurried to abolish copyright, thus making all continental production a free-for-all. That they later reinstated it says more about their printers' lobby power than their respect for it.
Contracts, as you say, keep being leonine, so little has changed with copyright. The entities persecuting copying aren't author guilds but industry associations, which clearly outlines who are the ones really not winning money over this whole deal. Of course there are weird chimeras like Spain's SGAE which include both editors and authors, but certainly you will see the senselessness of a union that includes both employers and employees: it really is a union of employers with employees thrown just to hide the fact.
But you are worried about stealing the authorship of a work, so I will show you an example of how it works: say you are a japanese manga artist who writes a successful manga series. In fact, it is so successful that it spawns anime recreations and sequels. And, suddenly, someone produces a a movie that is suspiciously like yours, which isn't too surprising since they learned coloring techniques from you when you were creating the manga. Do you get royalties or even acknowledgement? No, you don't, and if you dare speak you will probably get hammered with lawyers. So much for copyright protection.
Remember, authors usually have contracts that pay them per copy sold by the authorized publisher. As long as the publisher can keep making sales, the author can keep getting money. When an unauthorized printer sells copies of a book, it is extraordinarily rare that the author gets any money.
Remember that those contracts leave the author indebted unless he is a best-seller. So forget being paid per copy: most authors don't even get the chance to achieve that.
Samuel Clemens depended upon his writings to continue to sell, for his livelihood.
Which is not the same as depending on his copyright royalties, is it? Mind you, Samuel Clemens (better known as Mark Twain) was mainly a journalist. He got paid for his articles once (such is the fate of work for hire, isn't it?). He also was an orator, and he got paid for doing that too... once per show. He also got the patronage of a very rich man of the time. As you can see, it doesn't look like royalties were a problem until he took a stake in the intermediation business.
Oh, and yes, I'm alright with copying self-published works too. But perhaps private copy being a right in my country colours my view of the issue.
Would you have a problem with people taking games from the Humble bundle and selling them as their own work? How about someone's self published book that you didn't like?
Nice, you start mixing economics and authorship. I consider authorship something to be preserved, but authorship doesn't bring money to your table. Doing a good work does: when I develop software for an organization i get paid for the software itself, not for being its author. Of course you may argue that both are conjoined, but that won't happen if I'm hired to maintain someone else's software... and I will still get paid for my work. But I get paid once, not a thousand, a million times.
Even more: a company I worked for keeps selling the software I developed, making a profit on it and my authorship isn't recognized. As authorship is already not being recognized and already being profited from, what's your point about it?
How do you feel about a self-published author's copyright?
If his work is good enough I won't need to worry about that author's copyright: I will buy it. Actually, I already have: I've spend over 50 euros this Christmas buying copyrighted items. Of course having those items reasonably priced has helped a lot (see the $2.37M or so that the last Humble Indie Bundle has collected).
Besides, if your reasoning was correct, then intermediaries would have never paid authors before copyright existed. You may be surprised, but they actually did. So the author getting money from his work is not dependent on the existence of copyright. It never was, and it never will be. What's even more: an author can now eschew all those intermediaries and reach the public directly. And people actually pay. It may not be a great amount, but it helps people make do if they are good enough.
Of course, there is the historical comment above mine to take into account. It's funny when facts contradict your theories, isn't it?
Like it or not, protection of a work is needed to keep the creative process going.
That is false, was false and will keep being false. The only reason we have the second part of "El Quijote" was that someone wrote an apocriphal second part, for example. "Protection of a work" as you call it is needed to keep the intermediaries accumulating money, not the authors: it never was an author protection tool, it isn't now and it will probably never be.
The lost sale doctrine is fancy talk at the best, by the way. I can't help thinking about poor home to home ice sellers, losing sales because people started buying fridges.
See? If you had your kindle handy you would have known to put an h in there.
FTFY ;)
R is under license A, S is under license B. If licenses A & B are incompatible and the authors of R & S don't give permission, R & S can't be conjoined. Really, I have been saying that the whole time. Perhaps I'm explaining myself wrongly but I'm running out of ways to make you understand that.
Your claim for example was that A submission of GPLv2 code constitutes a violation of the kernel's license, programs don't have legal rights corporations and individuals do.
Sorry, perhaps you don't understand that "kernel's license" means the license that goes along with the kernel. Most people do.
The kernel can't have a license if all the individual copyright holders have full control.
Of course it can if every part added to it is added under the same license. Which is the case because you can't legally add code under any license but GPLv2 only without the permission of the kernel developers.
You are begging the question. F says no contract existed and he never accepted the contract. That's counter evidence.
Ok, good luck trying that defense against infringing any other license or contract. "No, sir, I didn't understand that the terms of the license mentioned all over the place applied to me. By the way, I don't understand why the traffic regulations apply to me either".
Maybe, maybe not. I think it might be a lot easier to take parts of the kernel and conjoin them with GPLv3 code. Taking the entire kernel is likely to grant Linus standing, despite his belief he doesn't have standing. In other words I think Linus is making two mistakes.
Blablablah. References on "We have kernel developers on record saying that's not their understanding". You state that, you prove it. Anything else is trying to distract me and it won't work, at least this time.
Nottage v. Jackson the author is the person most closely responsible for the work being produced.
Very nice. That actually supports my stand that you can't happily relicense something without the author's permission.
Feist -- facts are not copyrightable but compilations of facts are
As the kernel is neither facts nor a compilation of facts, this doesn't apply.
Wheaton v Peters -- A conjoined work has its own license
Wheaton vs. Peters is a case of failure in applying for copyright: defective in publication is the term used. AFAIK this hasn't happened with the Linux kernel and even if it had, current law may have different provisions than XIX century law. Please, try to provide references not earlier to the advent of software. Copyright law on software has some specifics that may make relying on previous sentences unsafe.
Yes. That's exactly what it takes. There need to be provable events which constitute signing the license. Your entire argument is predicated on the belief that everyone who contributed to the kernel explicitly agreed to the kernel license including the no upgrade policy collectively and thus individually they cannot do anything more than relicense their own contributions.
Oh dear, you really don't get it. If you create half a book and I create the other half we both have the right to restrict the publication of that book... and keep the right to publish our own parts. What's even more, in the case of the Linux kernel, everything added to it is based on earlier versions of the kernel which, going back, return you to the 0.12 version in which Linus Torvalds (AFAIK, the only developer then) chose GPLv2-only as license. So yes, there was an original entity (Linus Torvalds) who held the copyright. And everything else derives from it and is forced to follow the original copyright restrictions. If what you say was correct, just reading the ROM BIOS on IBM's first PC wouldn't have been enough to disallow any developer to create a clone. And yet Compaq and others had to go through the "clean room implementation" hassle in order not to infringe on IBM's copyright. No matter
No that's false. That's called an implicit contract and the law does not accord them very much respect at all.
Implicit? It is so implicit that it is all over the kernel source files. You can only not see it if you willingly close your eyes. By the way, if that's implicit and invalid, then the rest of software licenses (both open and closed) are invalid. Of course, then all this discussion is futile, because they are implicit.
E can assert that F doing action K creates a contract, F can K and no contract is in place.
F accepts a contract which restricts F's actions (on how to distribute in this case). The contract exists before F does any other action.
We have kernel developers on record saying that's not their understanding. They believe that the kernel copyright is the standard GPLv2, but that the kernel collectively is not exercising the GPLv2 -> v3 transition. So again we know that's not true.
References, please. If what you say is true, there is no need to collectively transition to GPLv3. All it takes is someone taking the kernel and changing the license. They don't do it because they can't without permission of all the people that has provided code to the kernel. And that's because it is GPLv2 only.
That's again one of the myths of the open source community. Courts have held that entering into a conjoined work creates new licenses.
Excuse me? You talk about a license as being implicit and now dare to talk about new licenses being created? References, please.
In general publishers want unambiguous copyright but publishers have pushed ahead in conjoined works where individual copyright holders for a conjoined work didn't agree with relicensing. A good example of this is collections of articles originally published in magazines and then republished in a book.
I can't help but wonder what would have happened if the authors decided to sue... or why they didn't.
You are trying to have it both ways. There is no "kernel license" without a corporate entity that holds that license. Either there is a bunch of individual licenses or there is an implicit corporate entity.
What's an implicit corporate entity? "We, the kernel developers, being of caffeinated mind and tired body..."?
Further there is no obligation on any contract to enforce its terms. A3 is not infringing by indicating his understanding of the contract is more lenient.
Oh, that's true. You can ignore a contract... and you can ignore a law. What happens after that fact (losing your distribution license) doesn't change because you "misunderstood". There is no good (for some values of good) will eximent.
Illegal from whom? Who in your opinion has standing?
Illegal for A3, who is not the owner of the code he wants to relicense (if he doesn't have the permission of the authors). Of course, the ones entitled to do the relicensing are the authors and the people they collectively allow to relicense.
Massive it is, conjoined it may be too, but the licensing isn't ambiguous at all, at least on the part that specifically voids the possibility of turning it into GPLv3 or later.
You are just assuming that Linus' has authority to speak for the whole product but Linus does not have, nor even claim to have that right.
When did Linus enter this thread? I stated that Ax may sue about Ax's code on the cases I put forward. I stated that B may sue about B's code on the cases I put forward too. Why is it so difficult for you to understand that if you infringe the license of the code I developed I can take you to task about that code only? It doesn't matter if there is a thousand more developers. I'm responsible of taking legal care of the code I wrote, and that's it. Now, the developers can choose to present a united front in order to save themselves the hassle of proceeding with individual suits.
If A3 asserts that his code, in the kernel, does not have the no upgrade clause that's binding.
And A3's code won't have entered the kernel, because it is GPLv2 only. A3 will be infringing the kernel license by trying to insert standard GPLv2 code in it. Since that is illegal, A3 has no standing about trying to change the kernel's license.
Your whole argument turns around the point of sneaking non GPLv2-only code in the kernel. It doesn't work like that: if you do it you are infringing and cannot distribute. So please stop beating that argument to death, it's already buried under a kilometre thick of sedimentary rock. You can't lawfully insert standard GPLv2 code into the GPLv2-only Linux kernel. That's it, no two ways about it.
A writes some GPLv2 code call it X and puts it in the kernel.
Which makes the copy of X in the kernel GPLv2 with no option to change to GPLv3 or it can't enter the kernel.
B takes X takes a subset of it X'. X' is under GPLv2.
This can only happen if B is taking X from the standard GPLv2 source, not the Linux kernel.
B takes X' and mixes it with some code Y for which he has copyright, creating X'+Y. He releases the the combined work under GPLv3.
Correct as long as X' isn't taken from the X version in the Linux kernel.
C takes X'+Y and violates the GPLv3 but not GPLv2.
D takes X'+Y and violates both the GPLv3 and GPLv2.
1) Is B's release legal? Is GPLv2 compatible with GPLv3 or does the upgrade ban B from releasing X'+Y under GPLv3? Has B created a license which is unenforcible. If so, how is the new product licensed or is it just a violation of A's code and nothing more? Remember, Linus' claim is meaningless if it just applies to X', the real question is whether the can claim copyright authority over X'+Y on the basis of some sort of leadership over X'.
B's release is legal as long as X' doesn't derive from the GPLv2-only Linux kernel, which may happen if A publishes X as GPLv2. Since standard GPLv2 is forward compatible with GPLv3 you may relicense standard GPLv2 code as standard GPLv3 code. But you can't take GPLv2 only code from the Linux kernel and relicense it as GPLv3.
As you can see it all hinges on whether X' derives from the Linux kernel GPLv2 only version of X or some other standard GPLv2 version. In the first case you can't relicense as GPLv3. In the second one you can.
2) Can A sue B and win? Make it worse assume that X involves code from A1, A2, A3 and A4 with A1 and A2 granting permission while A3 and A4 don't. Does the fact that A1-A4 wrote their code solely for a collective work create collective responsibility? The popular answer on /. is no A3 and A4 retain all rights to subsets. But AFAIK courts do not look at things that way.
Your question has already been answered. And the partial permission issue isn't GPL specific. Refer to your friendly lawyer.
3) Can B sue C? Remember if the answer is yes then B has just upgraded the kernel to GPLv3. If the the answer is no, then you agree that once any part of the code license becomes questionable the whole thing unenforcible. This is exactly the situation with the entire kernel.
No, assuming that B could license his code as GPLv3 (automatically precluding its coming from the GPLv2 only Linux kernel), B can sue C over B's code. B can't sue C over X' and if C rewrites Y from scratch and stop using B's code he stops infringing on his further code. Infractions about previous code based on B's GPLv3 Y code may remain, though. But, of course, all this means you still think you can put standard GPLv2 code in the Linux kernel and get away with it. You can't: the Linux kernel is GPLv2 only, and if you want to put code in it you will need to make it GPLv2 only and thus unrelicensable as GPLv3... as long as it derives from the Linux kernel.
4) Can B sue D? What about A can he sue D?
A can sue D on the infractions related to A's code. B can sue D on the infractions related to B's code.
It can be convoluted, but not really difficult.
A writes some GPLv2 code call it X and puts it in the kernel.
And it becomes GPLv2 only or it infringes the kernel's license.
B takes X takes a subset of it X'. X' is under GPLv2.
No, it is under GPLv2 only.
B takes X' and mixes it with some code Y for which he has copyright, creating X'+Y. He releases the the combined work under GPLv3.
Which he can't because X' can't be GPLv3 as it is GPLv2 only. Do I need to go further? Your whole chain fails in the beginning because the Linux kernel is GPLv2 without the "or later" option so you can't take it and relicense it as GPLv3.
That's _exactly_ the issue. The Linux kernel _can't_ be automagically be relicensed as GPLv3 and thus fails your example because X' is GPLv2 only. Actually, when A wants to put his GPLv2 or newer code in the kernel he will have to relicense it to GPLv2 only or it just won't be accepted because the kernel is GPLv2 only. Userspace may be different, true, but userspace is not the kernel.
I must be confused. You are saying that if someone takes from GPL, modifies it, then distributes it, that they are *not* required to distribute the source, but that you are allowed to violate any license they attach (if any).
Perhaps I didn't explain it right. I wanted to say that you aren't force to distribute your product at all, and if you decide to distribute it you must do it in a specific way (making the source code available to those that the software is finally delivered).
GPL isn't reinstating them. GPL is forcing distributors to follow "extra" rules.
Either GPL is giving the end user extra rights or it is reinstating them, because you certainly can not modify the software according to current copyright law unless the author allows you to. The distributors can choose either follow the extra rule (no plural there) or not distribute... like they would have under the basic copyright law. But, of course, GPL wants to protect use, and distribution is not it.
It feels more like you are arguing that "here's how I feel and how I justify my personal opinion, and isn't a discussion about what happens or how it works in law or such."
Ad hominems aside, the facts are that basic copyright law doesn't allow you to use or distribute the software as you want, BSD doesn't guarantee your recipients' ability to use the software as they want and GPL does. In contrast, BSD allows you to distribute as you like (even denying the right to use and distribute to those that receive your version) and GPL doesn't in order to keep the freedom to use the software and, mostly, the freedom to distribute it. And I think giving the buyer/recipient the ability of using the software as he wants independently of the wishes of the seller/originator is quite something.
But the GPL *DOES* restrict other people. Its mythology that it doesn't.
Congratulations. Did I say at any time that GPL doesn't restrict people? Didn't think so: I have said several times it restricts distribution in order to keep allowing use. And this thread is all about the misconception of GPL restricting use when it doesn't.
The GPL is viral and ties to force me to make my code fall under GPL if I use GPL code with it. RMS has stated many, many times that this is his goal. People need to quit pretending otherwise.
No. Your code can be under any license you want _unless_ you distribute it. You are just required to distribute it under GPL.
There are many valid situations where I cannot give source back or make it open, and the GPL interferes with this all the time. It is not evil or bad to restrict source, there are times when this is a necessary and positive thing.
Not all the time, mind you. Just when you wish your right to ignore my conditions was more important than my requirements that you don't. By the way, I would like you to provide some examples of the positive side of restricting source code.
GPL is not freedom, it is giving up freedom in exchange for what some people think are more important freedoms. Its also incorrect to state that GPL is about the freedom of the code. It greatly impacts the freedom of the coder, you can't neatly separate that out like it doesn't matter.
There is not absolute freedom. GPL isn't absolutely free, BSD isn't absolutely free, Public Domain isn't absolutely free. And they aren't absolutely free because there is more than one person on the planet to use them, so some people won't be able to get what they want. GPL is free to use: you don't have to abide to any restriction as long as you just use it. And I, personally, consider GPL better than BSD because it will continue to be free: there is no option for anyone (besides the original authors, of course) to close the source down.
and not a grant of extra rights to the buyer/consumer.
No, of course not. The buyer/consumer is entitled to doing anything with the software, like making copies, modifying it, reverse engineering it... oops, copyright law requires the author to allow me to do that. If GPL isn't granting extra rights to the consumer then copyright law is taking them away (and GPL is reinstating them).
But if you distribute it, you are, required to distribute. The GPL exerts force on the sellers
You are not forced to distribute. You are forced to distribute in a specific way. But distribution is not use.
Your comment is interesting in two ways: what it answers (would you please reference Linus' assertion?) and what it doesn't (everything else). Excuse me for not trusting your expert opinion on software quality, by the way ;)
Since the GPL specifically lets you make verbatim copies (as long as you don't distribute them), GPL doesn't restrict your use of the software when you run it (making a copy of it in memory, as you say).
How you are not "repeating" what others say when you sell the software they made, even if modified? GPL allows me, when I buy GPL software from you, to repeat (read publish/sell) what you have "said" even if you don't want me to, the same right you had when you received it. GPL doesn't force you to repeat anything: you can use your GPL software in-house and never tell anyone about it.
It doesn't AFAIK. At most you could consider the license termination clauses which forbide you from conveying the software to other users either in original or modified form unless you comply with the license. Since distribution isn't use, I consider it isn't covered. You can do whatever you want with the code as long as you don't turn copies of it to other people: it is then when you need to check compliance.