As the Atheros driver issue continues to simmer on the OpenBSD -misc mailing list and the Linux Kernel mailing list, with debate continuing over when the license of source code can be altered or added to, Eben Moglen made a statement for the Software Freedom Law Center. He began by defending their own actions, "it might be useful to recall the first stage of this process, when OpenBSD developers were accused of misappropriating Atheros code, and SFLC investigated and proved that no such misappropriation had occurred? Wild accusations about our motives are even more silly than they are false." He went on to acknowledge, "we understand that attribution issues are critically important to free software developers; we are accustomed to the strong feelings that are involved in such situations. In the fifteen years I have spent giving free legal help to developers throughout the community, attribution disputes have been, always, the most emotionally charged." He added that the SFLC would be making no further statements until their work on this matter was complete, noting:
"Also, and again for the last time, let me state that SFLC's instructions from its clients are to establish all the facts concerning the development of the current relevant code (which means the painstaking reconstruction of several independent and overlapping lines of development, including forensic reconstruction through line-by-line code reviews where version control system information is not available), as well as to resolve all outstanding legal issues, and to make policy recommendations, if possible, that would result in all projects, under both GPL and ISC, having full access to all code on their preferred terms, on an *ongoing* basis, with full respect for everyone's legal rights. We continue to believe those policy goals are achievable in this situation. The required work has been made more arduous because some people have chosen not to cooperate in good faith. But we will complete the work as soon as we can, and we will, as Mr Garvik says, follow the community's practice of complete publication, so everyone can see all the evidence."
From: Eben Moglen [email blocked] Reply-To: Eben Moglen [email blocked] Subject: Statement by SFLC (was Re: Wasting our Freedom) Date: Sun, 16 Sep 2007 09:17:41 -0400 VM 7.19 under Emacs 21.4.1 On Sunday, 16 September 2007, J.C. Roberts wrote: Let's say someone took the linux kernel source from the official repository, removed the GPL license and dedicated the work to public domain or put it under any other license, and for kicks back-dated the files so they are older than the originals. Now take it one step further. For the sake of example, let's assume all of this atheros driver nonsense went to a German court and the GNU/FSF/SFLC/Linux or whoever you want to call yourselves lost a criminal copyright infringement suit. You have now been legally proven to be guilty code theft. After such a ruling let's assume some jerk was to do the all the horrific stuff mentioned in the first paragraph above to the linux source tree, along with a little regex magic to call it something other than "linux" and seeded the Internet with countless copies. None of this has happened. What has happened is that people who do not have full possession of the facts and have no legal expertise-- people whom from the very beginning we have been trying to help--have made irresponsible charges and threatened lawsuits, thus slowing down our efforts to help them. It might be useful to recall the first stage of this process, when OpenBSD developers were accused of misappropriating Atheros code, and SFLC investigated and proved that no such misappropriation had occurred? Wild accusations about our motives are even more silly than they are false. We understand that attribution issues are critically important to free software developers; we are accustomed to the strong feelings that are involved in such situations. In the fifteen years I have spent giving free legal help to developers throughout the community, attribution disputes have been, always, the most emotionally charged. But making threats of litigation and throwing around words like "theft" and "malpractice" was a Really Bad Idea, because once some people started using that language--thus making adversaries rather than collaborators of themselves--I had no choice but to ask my clients and my colleagues to stop communicating with them. Let me therefore point out one last time that if the threats of litigation and bluster about crime and malpractice--none of which has the slightest basis in fact or law--were withdrawn, we would be able to resume detailed communication with everyone who has a stake in the outcome. Also, and again for the last time, let me state that SFLC's instructions from its clients are to establish all the facts concerning the development of the current relevant code (which means the painstaking reconstruction of several independent and overlapping lines of development, including forensic reconstruction through line-by-line code reviews where version control system information is not available), as well as to resolve all outstanding legal issues, and to make policy recommendations, if possible, that would result in all projects, under both GPL and ISC, having full access to all code on their preferred terms, on an *ongoing* basis, with full respect for everyone's legal rights. We continue to believe those policy goals are achievable in this situation. The required work has been made more arduous because some people have chosen not to cooperate in good faith. But we will complete the work as soon as we can, and we will, as Mr Garvik says, follow the community's practice of complete publication, so everyone can see all the evidence. We will make no more public statements until the work is complete, and we will be neither hurried nor intimidated by people who shout at us instead of helping. -- Eben Moglen v: 212-461-1901 Professor of Law, Columbia Law School f: 212-580-0898 moglen@ Founding Director, Software Freedom Law Center columbia.edu 1995 Broadway (68th Street), fl #17, NYC 10023 softwarefreedom.org
Why does he just not admit
Why does he just not admit and apologize for changing another persons license and thus violating his copyright in the first place??
Because it's nonsense, if
Because it's nonsense, if you don't know how CVS (work in process for developers) works. And furthermore, after the hint about it, they changed it. But you cannot see this behaviour in Linux community, they just keep ignoring it. So it's rather nice to see all the Linux zealots crying "... but Theo", it's crap, get it. Linux defines itself through the hate on Windows and it sees itself as primus inter pares in opensource. This is the whole motivation, axe every *other* free operating system out there. There is no real community possible with most of the Linux crowd.
Theos big mistake, don't criticize Linux, do it and the crowd will kill you instantly without any reasoning involved. Only Linus can do so, badmouthing people and axe other peoples projects and the crowd will applaud him.
troll much?
Sheesh... your post appears to be a mere rant peppered with wild accusations.
What!? Why did you even
What!? Why did you even reply to my post above if it was rant about Linux Theo and other nonsense
CVS is [in most cases] a
CVS is [in most cases] a public repository anyone who runs -CURRENT or wishes to check out a CVS server runs it. Licensing should be adequite. Yes, errors may appear, but one still should be very careful. If in doubt, run a private CVS.
A proposed patch on a mailing list is something totally different. Its a [in most cases] public post, yes. But it isn't real yet. Its a proposal. Go look that word up in your dictionary.
uhm, sure. I'll just go and
uhm, sure. I'll just go and stand over there and mingle a bit with the *sane* BSD and linux folks. I'm sure there's a wall somewhere that is fascinated by your theories.
Jesus Christ. Linux is a
Jesus Christ.
Linux is a project trying to write software.
OpenBSD is a project trying to write software.
Pull your head out of your arse and do something useful. Nobody cares.
why not you stfu
A BSD derivative work can be licensed as GPL. The point of view of some people is "that is what happened, stop trolling"
Theo de Raadt may be wrong, don't follow him so blindly.
Re:Why does he just not admit
because he's a lawyer and not a coder and didn't do anything with the code perhaps???
a lawyer and not a coder
If you read Eben's short bio, you'll see he coded for mainframe systems once upon a time. But you're right that he had nothing to do with this current code wars.
Admit what?
Eben Moglen had absolutely nothing to do with the current Atheros driver wars. From his message, he's trying to evaluate two things:
1. Was code misappropriated?
2. How can GPL and ISC live together?
You can expect that Eben will be looking at this from a legal point of view and not only from US law. He also has a day job and can't work on the Atheros problem 100% of the time. If he'd make an immediate statement and judgement just because of Theo's ranting, I wouldn't trust his legal abilities. If you want an immediate "Oh yeah. that's wrong - let's sue!" then you should be talking to Darl McBride and ask him who to hire...
Clean room implementation
Well, lets just do a clean room implementation instead.
* http://en.wikipedia.org/wiki/Clean_room_design
We get one team to extract specifications from the BSD Atheros code. Then we get another team todo the implementation of the specification, that way there are no copy of code or copyright or anything.
It is a shame it has to come to this. Just over some code in a CVS repository that was never released, it turned out into this flamewar, and the BSD and Linux community seem to have forgotten that we essentially are the same people, geeks, with pretty much the same ideals of freedom and a free operating system.
The Linux kernel can legally
The Linux kernel can legally use the BSD's Atheros code because the ISC license does not set any additional restrictions to the GPL license — it is said to be "GPL-compatible". This is not even under debate, so there is no reason to rewrite anything.
The current issue is whether the ISC license permits changing the license of an ISC-licensed code to GPL, and whether the SFLC was giving correct advice by stating that this is legally possible.
But either way, I absolutely agree that re-licensing code "just because you can" is a braindead thing to do for a free software project, and as far as I can tell, the changes were already rolled back in Linux repositories.
The changes were not 'rolled
The changes were not 'rolled back' because the changes were never made in the first place. There was a patch submitted to the mailing list, which was never applied.
THAT is the truly amazing point
>There was a patch submitted to the mailing list, which was never applied.
Read the words again and youll see how stupid this whole thing is.
Changes not committed to a
Changes not committed to a public repository? What's kernel.org then:
Right here:
http://git.kernel.org/?p=linux/kernel/git/linville/wireless-dev.git;a=bl...
Is reyks code with two other authors claming copyright over it having done nothing but
indentation, headers and variable renames.
When exactly the same thing happened in reverse it was immediately removed by OpenBSD.
This is still up on kernel.org. Why?
BSD allows derivs to be GPL; GPL does not allow derivs to be BSD
>> Is reyks code with two other authors claming copyright over it having done nothing but indentation, headers and variable renames.
*****
Section 101
-- A "compilation" is a work formed by the collection and assembling of preexisting materials or of data that are selected, coordinated, or arranged in such a way that the resulting work as a whole constitutes an original work of authorship. The term "compilation" includes collective works.
-- A "collective work" is a work, such as a periodical issue, anthology, or encyclopedia, in which a number of contributions, constituting separate and independent works in themselves, are assembled into a collective whole.
-- A "derivative work" is a work based upon one or more preexisting works, such as a translation, musical arrangement, dramatization, fictionalization, motion picture version, sound recording, art reproduction, abridgment, condensation, or any other form in which a work may be recast, transformed, or adapted. A work consisting of editorial revisions, annotations, elaborations, or other modifications which, as a whole, represent an original work of authorship, is a "derivative work".
Section 103(a)
The subject matter of copyright as specified by section 102 includes compilations and derivative works, but protection for a work employing preexisting material in which copyright subsists does not extend to any part of the work in which such material has been used unlawfully.
(b)
The copyright in a compilation or derivative work extends only to the material contributed by the author of such work, as distinguished from the preexisting material employed in the work, and does not imply any exclusive right in the preexisting material. The copyright in such work is independent of, and does not affect or enlarge the scope, duration, ownership, or subsistence of, any copyright protection in the preexisting material.
*****
You get copyright to the derived work but not to the portions that are already copyrighted (excluding fair use).
>> When exactly the same thing happened in reverse it was immediately removed by OpenBSD. This is still up on kernel.org. Why?
Here is the difference.
I have BSD, I make a derivative work (ie, I make a new work that uses old BSD stuff). The derivative work has copyrights (see above quote, esp. part (b)). I make it GPL. The stuff I borrowed from BSD is still BSD and the BSD license ALLOWS me to use that work in a deriv as long as I attribute, etc. The BSD license DOES NOT say that if I make the derive GPL I lose my BSD license to the original works. So copyright law allows me to make a deriv to a BSD work and make the deriv GPL. The BSD doesn't stand in the way.
I have GPL, I make a derivative work (ie, I make a new work that uses old GPL stuff). The derivative work has copyrights (see above quote, esp. part (b)). I make it BSD. The stuff I borrowed from GPL is still GPL but the GPL license DOES NOT ALLOW me to use that work in a deriv THAT IS NOT ALSO GPL. The GPL license DOES say that if I make the derive NONGPL, I DO lose my GPL license to the original works. So copyright law allows me to make a deriv to a GPL work and make the deriv BSD. The GPL stands in the way, however, in that I then lose licensing rights to the original code.
To recap:
Copyright law allows you to make the derivative anything you want.
The BSD allows extensions to be GPL without giving you headaches.
The GPL does not allow extensions to be BSD if you want to include the original GPL along with it.
In either case the new additions/modifications belong to the new author, thus the new material IS copyrighted by the new authors.
Note, that the headaches the GPL gives you are protection for the authors of the code. It prevents monopolists and other powerful groups from taking the work of the GPL authors, leverage their business savvy or monopoly power, and lock the original authors out of the market.
Why anyone would use BSD as a matter of course is beyond me.
[Note, by BSD, I mean the BSD license being used. Also, IANAL.]
Further discussion
I forgot to clarify in the "recap" that the original work stays under its existing license unless the license itself (or some other authority) said you could change the terms (ie, relicense).
Also, let me stress that any modifications belong to the new author; thus, it makes perfect sense (legally) to have the new author's name on the license. Unfortunately, this may incorrectly give the impression that the new authors made significant contributions to the work.
See also, http://www.linuxtoday.com/developer/2007091000726OSKNDV . A reply comment from a different poster appears to point to a portion of the GPL that is inconsistent with the claim I was making, but he made a mistake (make sure to read the postings from the 17th of September). The (I think correct) claim made boils down to saying that if you distribute a binary translation of this Atheros BSD file (as part of the GPL kernel), you only need to provide source for the minor changes made by the author of the GPL derivative work (ie, provide source only for the minor comment and various other irrelevant changes).. in addition to the source for the (majority of the) rest of the kernel, of course. You do not have to provide code for the majority of the Atheros file that still falls under the BSD license.
IANAL
Agreed
It was either stupid or a selfish change. Linux kernel doesn't need this kind of unneeded license mangling. As they reverted, it was probably a stupidity :)
What some linux hackers are afraid of (therefore they try to relicense some projects to GPL-only) is that vendor could take this driver and start releasing it's own proprietary updated version of it (it's what ATI basically did with DRI/DRM). However splitting a joint work undertaken by Linux and BSD's to get a damn driver isn't really worth it. It's not only about legality (btw. I think it's perfectly legal to relicense that piece to GPL-only, but IANAL), it's also about OSS ethics - having some respect for decisions of previous authors, as here their motive for dual-licensing a codebase in both kernels.
BSD does not allow changing terms, ie, relicensing.
>> (btw. I think it's perfectly legal to relicense that piece to GPL-only, but IANAL)
No. You can't relicense the portions that are licensed only as BSD since that license doesn't say you can change the terms.
However, any derivative work has copyrights which can be licensed as the author wants. The derivative can be GPL but only the *new* changes are covered under the GPL.
As another example. If you give me an essay that is BSD, I can rewrite a new essay that looks identical to yours except for an additional period at the end. I can make the new essay GPL, but I must still adhere to the BSD license for the portion of the essay that is identical with what you gave me (ie, for the majority of the essay). This portion that is (still) under BSD does not fall under the GPL of the new work I made. So my new work is GPL, but this GPL covers only the new portions over which I have copyright (the additional period at the end).
IANAL. I come up with these ideas from my interpretation of USC Title 17 Sections 101, 103, and 106 mostly.
Also, I am assuming we are talking about a file licensed only as BSD (ie, you have no other copy licensed otherwise or any other rights granted).
Both sides have legitimate grievences, IMO (IANAL)
If Theo threatened Eben Moglen with a malpractice suit, then I would not condone that. All that such threats do against lawyers is slow down help and cause the loss of allies.
However, I personally think that the SFLC is making a dangerous mistake in how its lawyers read the BSD License. I don't think that this license does allow sublicensing or relicensing of existing code, and my discussions with authors of works under such licenses have generally agreed. Instead they intend to offer a grant of permissions from the original author to anyone who obtains a copy of the source code to use it in accordance with the license. In the substantive legal issues over what the BSD License allows, I agree with Theo for reasons I have stated elsewhere.
One thing we *all* need to remember is this: if you treat everyone like criminals, you will find yourself without friends.
Best Wishes,
Chris Travers
Not a lawyer either, but there are obvious errors in your post.
"However, I personally think that the SFLC is making a dangerous mistake in how its lawyers read the BSD License. I don't think that this license does allow sub licensing or re-licensing of existing code, and my discussions with authors of works under such licenses have generally agreed."
That's total nonsense. BSD object code is distributed under any number of licenses, most often without any of the source available to determine its authorship or any modifications made. With the elimination of the "advertising clause" there is no obligation on the part of the re-distributor to disclose anything or even credit the originator of the code. All that's necessary is the inclusion of the "Portions Copyright(c) , " disclaimer. After that, your code is theirs, if the resulting binary is licensed in some draconian fashion you've just made a contribution to the aggravation of the public. BSD is amenable to proprietary conversion, inclusion into a GPLed whole doesn't change that or change the underlying BSD licensed base. If no notice as above was included, then that should be corrected. After all, the ISC license just says:
"Copyright (c) 4-digit year, Company or Person's Name
Permission to use, copy, modify, and/or distribute this software for any
purpose with or without fee is hereby granted, provided that the above
copyright notice and this permission notice appear in all copies.
THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE."
The "this permission notice" clause is more restrictive than BSD, but meaningless in the case of object code.
No one is discussing object
No one is discussing object code.
reverse engineerable bytecode - use that binary portion of BSD
It is very simple, compile to reverse engineerable bytecode format, move copyright notices to the read-me file, decompile the bytecode, voila, no BSD.
Reverse engineering an
Reverse engineering an already free driver?
Sure, you could, if you could convince anyone it was really clean.
So if you're truly a GPL zealot, and can't stand a portion of the code being more free than the GPL, this is the way to go.
However that isn't what was done.
What was done was to run the code through a different indent and rename the variables. That's not reverse engineering. and that's not authorship of a derived work. That is adapting the code to run on Linux, just like translating a book. It doesn't give someone enough to claim authorship. No matter what license they are using.
"So if you're truly a GPL
"So if you're truly a GPL zealot, and can't stand a portion of the code being more free than the GPL"
That was a nice sane untainted piece of text there.
I think you are wrong, again IANAL.
"That's total nonsense. BSD object code is distributed under any number of licenses, most often without any of the source available to determine its authorship or any modifications made."
It is true that the BSD license does not require that the source code be distributed. This does not mean that there is an implied sublicensing right. Once again, IANAL.
The permission grant you quote is:
"Permission to use, copy, modify, and/or distribute this software for any
purpose with or without fee is hereby granted, provided that the above
copyright notice and this permission notice appear in all copies."
Ok, this provides permission to use, copy, modify, and distribute the software. It does not provide a right to sublicense it because this right is not implied in or required in order to fulfill any of the other rights.
Furthermore these rights are subject to the condition that the author's original permissions grant be reproduced on the code.
What this license does *not* do is prevent you from enforcing your *own* copyrights over your own original added copyight-worthy content in any way you see fit. This is entirely different from relicensing or sublicensing the original BSD-licensed code.
The software which hasn't
The software which hasn't been distributed in the absence of the copyright notice and the text you've quoted is a device driver, nothing more, and perhaps substantially less. It appears that the strictly ISC licensed portion is the hardware abstraction layer which shims the driver to an entire work.
In allowing modification for any purpose the "license" allows a compilation in machine language to be distributed under arbitrarily more restrictive terms. This is *not* entirely different from relicensing the original BSD-licensed code, the BSD license in that case has become irrelevant. It also doesn't (entertainingly enough) provide for very effective enforcement of the original author's copyright.
Don't assume what I'm arguing for me; I'm new to this particular tempest in a teapot and trust me that I disagree with any patch which if applied would have removed the BSD license text. Hyperventilating over the outrage against "true freedom" down the road that should have been traveled or some such nonsense, on the other hand, certainly doesn't get any issues resolved.
It appears on only cursory inspection that patches were generated by a single individual improperly, whether they were ever applied I don't know, whether any result was distributed, I don't know. Potential relief available would be restoration of the unpatched original. To the best of my knowledge this is already done. At that point rants on the LKML are simply entertaining flame-fests and threats of criminal copyright infringement in German courts are only icing on the cake for true fans of net.kookdom. The bonus of Theo holding up MicroSoft as a paragon of virtue in its respect for the BSD license is ice cream on the iced cake, with whipped cream and nuts.
Don't think so.
Note that the word "entirely" here is quite correct. However in my view this is not entirely the same as "relicensing" original BSD code either. The functional difference is that people can still use the BSD source code as they see fit. I.e. using BSD code in a GPL'd application doesn't make that code fragment effectively under a differnet license (now, once you add substantive portions you hold copyrights to that is a different matter."
In short the difference is what restrictions you can impose on what someone downstream can do with the code. My position is that you can only enforce your copyrights on your own work, not those on the works you merely used with permission.
IANAL, etc.
Two different issues
Regarding the first issue of distribution of dual licensed code
"how its lawyers read the BSD License. I don't think that this license does allow sublicensing or relicensing of existing code,"
That is NOT the BSD license they read, but the author statement that explicitly said that you can distribute this code using either the BSD or the GPL. The author gave the choice of license to each licensee, which therefore can choose to re-distribute using either or both licenses. Double-licensing is exactly the same as publishing it twice: once with a BSD license, the other with a GPL license. The practice to publish once with both license attached is merely a convenience to avoid publishing twice every time.
As an example: MySQL publish it's code both with an GPL like license and a Commercial license. In other words they dual-license, just like part of atheos is dual-licensed. The difference in the case of MySQL, is that no-one is claiming that the terms one license apply to the other and vice versa... That would be dumb since they are mutually exclusive.
Now regarding the other issue, which is mixed contributions under different licenses:
let's say there are a 1000 lines source under BSD. with copyright to A. let's say someone (B) do a patch of 100 line , replacing 50 lines, adding 50 others... , let's say he release it's patch under the GPL (the path itself, not the combined work). can the A apply B's patch on it's work and publish the resyult under BSD license ? NO B's contribution license doesn't permit it
Can B apply the path to A's work, and use it in a GPL environment: YES A's license allows it.
A's code - if he didn't dual licensed - cannot be changed, and the term of he BSD license still appears, but they do NOT apply to B's substantial contributions.
Wrong. B and A can both
Wrong. B and A can both apply the patch. The issue is when A distributes B's patch.
Publish!
The poster did say "and publish" which is a form of distribution once your webserver (ftp server, etc) sends the bytes over in response to the browser tcp/http request.
I think you are wrong on a number of counts
First, I dont think that all of the code was dual-licensed by the author. Some portions were BSD-only. IANAL, once again.
Now, having something which is dual-licensed poses a serious issue that I dont think has gotten the discussion that it needs: That of intent. Is the intent to allow a sublicense? In which case wouldn;t it be better to put a sublicensing notice on the top? Or is it to encourage derivatives of either license?
What is shmget wrong about?
I don't think it was stated that every piece of code in dispute was dual licensed.
And I don't see where relicensing/sublicensing comes in? Using the interpretation by shmget, you have two separate copies. You can develop either one as you see fit. Assuming this was the licensing option available, one copy is sitting in the attic, virgin, gathering dust while the other copy was just modified and thrown into the kernel purely and entirely as GPL (with 2+ GPL authors).
Yeah but theres 2 situations
Yeah but theres 2 situations one of which is BS. The first one is software dual-licensed under GPL and BSDL. The Linux developers, preferring GPL, can simply pick GPL whereas a BSD developer, preferring BSDL, can simply pick BSDL.
The other and only situation of importance is Reyk's. Don't relicense his code, but allow yourself to link seems to be the solution to the problem. But IANAL and we'll hear it from a lawyer sooner or later.
As for that slower working cause of legal threats. That can also be a revenge in itself, and Theo needs a few lessons. Not learned from DARPA, Theo?
So, you're saying the BSD is the GPL?
Before I go further let me make clear that I think it was rude for someone to propose changing the license on this stuff to GPL only, and would have been rude for the Linux kernel people to have actually done it, and so it's a good thing they didn't. But this is just strange:
"However, I personally think that the SFLC is making a dangerous mistake in how its lawyers read the BSD License. I don't think that this license does allow sublicensing or relicensing of existing code"
This is very weird. The fact that the GPL does not allow sublicensing or relicensing of existing code is *precisely* what people who favour the BSD have against the GPL. Allowing sublicensing and relicensing of existing code is, for BSD advocates, the BSD license's *core feature*. And when corporations selling closed source, proprietary software grab existing BSD code and relicense it closed, people nod sagely and say yes, this is what the BSD license is all about--maximum usability for everyone.
Everyone except GPL users, apparently. If a GPL user tries it, suddenly half the BSD people ironically seem to utterly forget what the BSD is supposed to be and seem to imagine that it's intended not to allow relicensing--which is to say, imagine that it's a copyleft license just like the GPL, or at a minimum the LGPL. People, if what you really want is a copyleft license, use the LGPL or the GPL. If you *don't* want a copyleft license but instead want everyone to use your code and/or feel that restrictions in the name of later freedom are still unacceptable, use BSD. If you want that stuff but hate the GPL so much that you want it to be an exception, I'd say you need a new license; sort of a BSD license with an exception clause saying that people can do anything they want with the code *except* copyleft it. Which would be silly, and would still restrict freedom, but would be a license that genuinely allowed y'all to have it both ways.
If you modify the BSD
If you modify the BSD license to expressly state that the particular file cannot be relicensed under the GPL, then it would not qualify as an Open Source license. I think most of the developers who choose the BSD license would not want this.
I didn't recall reading any comment from BSD advocates that they want to allow relicensing, either. I do think that's the case, but could you please elaborate on this and paste some links to the comments themselves?
I have a feeling that the BSD and GPL licenses are just competitive -- the developers who choose GPL don't care what people say about them, as long as the source remains free (free as defined by the GPL); while the developers who choose the BSD license think that writing code under that license would make them famous and look more generous, thus more companies will offer them more jobs, or maybe those companies will contribute back their code. If what you say is true (that BSD allows relicensing, which I think is true but that I'm not sure about), then clearly the proponents of the BSD license feel threatened (as in having more competition) by the developers who choose GPL for their code. They don't feel threatened by the companies who take their source code and don't contribute back, because those companies are precisely the people they are trying to impress.
That's just my own theory -- feel free to disagree.
Where did the SFLC make a mistake?
>> However, I personally think that the SFLC is making a dangerous mistake in how its lawyers read the BSD License. I don't think that this license does allow sublicensing or relicensing of existing code
I though the SFLC hasn't come out with a position on this matter?
Read this comment: http://kerneltrap.org/Linux/SFLC_on_Atheros_Driver_Issue#comment-271380 .
A derivative is a new work and can be licensed GPL. What this means is that all the original items that make this a new work different from existing works fall under GPL. The portions that were BSD, remain BSD (meaning that source code to these BSD portions need not be distributed.. meaning that you can distribute only the few modifications that fall under GPL).
Will there be problems later in figuring out what portions fall under which license? Maybe. This is why I think there should be a link to the original BSD portions, if only in good faith. I think this is why most groups would rather code things themselves if they don't feel confident where any significant contributions may have come from.
This is a good controversy to be going through to help clear the air on misconceptions and hopefully help develop better practices going forward (if needed). Keep in mind, IANAL, and have been learning along with most everyone else. [Read the short sections of copyright law that are referenced from the link above.]
Let's not just remember what Plagiarism is kids.
Plagiarism is taking something someone else wrote and claiming you did, or changed it significantly.
When people are adding their name to reyk's code after variable renames and running it through linux indent - this is plagiarism. simple.
You can add your name to a copyrighted work, as a derivative work, when permitted only
when you make siginificant changes.
Translating something to run on linux and be formatted to fit the linux kernel's specs is not.
And yet code with those names on it is still published on kernel.org.
And yet it appears that in
And yet it appears that in spite of this, the SFLC is too busy staring at their navels rather than do anything about it. "Someone said something mean to us and called us a thief",
"so now we can't do anything about it because we got caught publicly."
Perhaps this is because they wish to condone plagiarism by different parameters of indent? I mean, after all, Eben wrote much of the gpl version 3, and yet Linus doesn't
like it in the kernel. Wouldn't it be convienent if we could take the linux kernel,
run it all through indent and rename the includes directory "gnulinux", call the project a fork, and slap a more restrictive license, like say, the GPL version 3 on it?
Would sure get around that inconvenience of Linus not liking the license the SFLC main people helped author, and constantly calling the OS linux"...
and slap a more restrictive
Sure it would be fine if the license allowed you to do it. Which it doesn't — most of the kernel is GPLv2-only. Which is precisely the point of licenses — a list of conditions specifying what you can and what you cannot do with the code.
Sheesh... You must be very
Sheesh...
You must be very upset that you post the same drivel for the umteenth time.
I think you forgot what your doctor told you: take your pills and relax.
(How old are you anyway, you sound like 12.)
Probably not plagiarism but attribution should be more clear
>> Plagiarism is taking something someone else wrote and claiming you did, or changed it significantly.
Then you said,
>> When people are adding their name to reyk's code after variable renames and running it through linux indent - this is plagiarism. simple.
The second doesn't necessarily follow from the first, but I agree that the author might be playing with fire.
Note, that many authors put their names on research when perhaps some of them did very minor work. The key to plagiarizing is to "claim you did" it or to imply it. Version control history may be sufficient attribution (besides that just putting your name doesn't imply any particular percentage) as long as this author never claims to have written parts not written. After all, you can only tell who wrote a part of an open source project by what the log says on version control. It would be a problem if version control was modified (well, this would be "forgery") or if the author made false claims.
I am not a lawyer, but I don't think this author will get into trouble legally just for putting the name behind two others while not claiming to have written portions of the code not written. I don't know what guidelines the courts use, so you may be correct that the actions taken constitute plagiarism.
I came across these links: http://www.plagiarism.org/learning_center/plagiarism_faq.html and http://en.wikipedia.org/wiki/Plagiarism .
I think a way forward is to provide a link to the original and then make the derivative GPL. If you don't put someone's name at the top (the person who made the changes and introduced the GPL licensing for modifications as allowed by Title 17 Section 103), then one could argue that the file was modified and is wrongly being attributed to the original authors. Do you see how not putting a name at the time of a modification can be a problem? Imagine if editors made changes to works but didn't clarify that they were publishing/distributing edited works (however minor the edits). They might be accused of trying to misrepresent the original authors. I would say that the kernel "editor" should make it clear they were an editor and not an original author, but again, with projects known to be version-controlled one would be remiss not to check the history yet make percentage assumptions about which author listed wrote how much.
I think what was done was close to typical procedure for version-controlled projects. Perhaps someone will come afterwards, study the history, make changes, and rightfully take this author off the list to add their own name. There should still perhaps be some sort of link or statement referencing the original file to help prevent misunderstandings later on.
The "documentation" could have been done better.
A sensible response
I've been following this story, including the never-ending debate on misc and this comment from Jose is one of the best responses I've seen.
Forget the licence and complaints that the BSD licences may or may not permit the 'rebranding' of code, what we have here is plagiarism plain and simple.
While the code remains under a single licence, it's easier to manage changes in the future. When people decide to make minor changes and slap a different licence on them, it's going to be a nightmare if the code needs a licence audit in the future.
Adding a second licence is not a good idea but if it must be done, at least make sure all the changes are clearly attributed in the code. i.e.
/* The function hello was added by Sean and is GPL licences, see the header of this file for details.
int hello()
{
printf("omg look at me, I write code!")
return 0;
}
I sincerely hope this whole situation is resolved soon because it's not reflecting well on the OpenBSD and Linux worlds.
Congress has never passed an
Congress has never passed an anti-plagiarism act. It never will.
Translation of anything is the creation of a derivative work, look it up:
"A “derivative work” is a work based upon one or more preexisting works, such
as a translation, musical arrangement, dramatization, fictionalization, motion picture
version, sound recording, art reproduction, abridgment, condensation, or
any other form in which a work may be recast, transformed, or adapted. A work
consisting of editorial revisions, annotations, elaborations, or other modifications,
which, as a whole, represent an original work of authorship, is a “derivative work”."
http://www.copyright.gov/title17/
Don't try to teach people things that you don't know, it's unproductive and leads to incorrect conclusions.
Thanks in advance for your cooperation.
US ! = World
Don't get confused about the US being the whole world -- it's International law that really matters.
Translation appears to not be enough to get a derivative work copyrighted. Read this example of the French highschool boy who was arrested and jailed for distributing his French translation of Harry Potter.
If translation was enough, people would translate works from English to another language, then from that other language back to English and they'd own the rights to the "derived" book in English. Obviously this is not the case.