Yesterday, we reported that the Software Freedom Law Center had started a lawsuit against several companies who they claim violated the GPL. The subject of the violation was BusBox, and the SFLC claims it is operating on behalf of the authors of BusyBox. Original BusyBox author Bruce Perens, however, begs to differ.
Perens claims that this lawsuit is being undertaken without his consent, even though the version of BusyBox disputed in the lawsuit is mostly his work – in other words, he holds the copyright. “First, I’d like to point out that I’m not represented in these lawsuits, and that the parties and the Software Freedom Law Center have never attempted to contact me with regard to them,” Perens says, “As far as I am aware, and under advice of various attorneys, I still hold an interest in Busybox through both content and compilation copyrights.”
The issue here is that while Perens indeed originally wrote BusyBox, several other developers have contributed code as well. Later on, maintainership of BusyBox has been handed down to other people, including Erik Anderson. In a comment on Slashdot, Perens explains what happened.
“The current suit is brought in the name of Erik Andersen. Erik worked for an embedded Linux company, now defunct, for a few years and was paid to maintain Busybox during that time,” he states, “During that time the company’s name appeared in copyright statements, and mine mostly disappeared.”
Perens claims that Anderson, the company he worked for, and Robert Landley have altered license statements and have removed copyright statements from other developers. “[They] appear to have removed some of the copyright statements of other Busybox developers, and appear to have altered license statements, in apparent violation of various laws,” Perens claims, “Mr. Landley once claimed that all of my contribution had been completely removed from the Busybox program, using a misinterpretation of Judge Walker’s methods for identifying non-literal copying to justify his claim. As far as I’m aware, he was incorrect.”
It is unclear to me what Perens refers to when he speaks of the Walker methods, but I’m assuming it has to do with Computer Associates Int. Inc. v. Altai Inc. It would certainly be appreciated if someone could correct me on this – or offer confirmation, of course. Of note is that if Anderson and Landley indeed contributed code, they are being infringed upon just as much, and are free to start a lawsuit.
Perens also doesn’t have a lot of love for the SFLC. “SFLC, which is supposed to represent Free Software developers without charge and without prejudice, seems to have been selective in which of the Busybox developers it chooses to represent, and has in the past been either guarded or hostile in its correspondence when contacted by other developers of the Busybox program,” Perens claims.
It will be interesting to see where this ends. “Much as other Busybox developers wish to support the general cause of getting companies to comply with simple Free Software Licenses, some of the other developers and I are becoming annoyed with Mr. Andersen and Mr. Landley’s apparent violation of our own rights, and SFLC’s treatment of our interest,” Perens ends his statement, “We have held off, to date, to avoid confusing issues, but our patience is limited.”
Since I’m not entirely sure what the details are of this “Copyright replacement” scheme that Perens alludes to, I have to assume it goes something like this:
1) Find a module with an author we don’t want any more
2) Rewrite said module, changing some percentage of the code
3) Remove previous author’s name, replace with our own.
If this was done methodically, with the intention of removing the original author, that’s just plain nasty.
It’s one thing to rewrite a module to be cleaner code, more efficient, follow a better coding style, etc. – but to intentionally remove the previous author is of questionable ethic. The original author went through all the trouble of defining the behavior and logic, restating the same code with your own style/method doesn’t make the original author’s work any less important.
Note: I have no clue if this does/doesn’t stand up in court, nor whether this is what actually took place – it’s just my speculation.
Well, the new code will still be derived work of the old code, so the original author still retains the copyright.
Basically, Perens can appeal to common decency in this, but I don’t think he has legal basis for his complaint (not does he insist on having one). He’s just saying that the lawsuit in question is a dirty trick (but that’s what lawsuits usually are).
OTOH, if the offending distributors of Busybox failed to get something as simple as GPL compliance done right (which is basically about providing a tarball of the code they used to compile the thing), I’m surprised they managed to deliver a device in the first place.
I’m betting they thought that since it was embedded stuff, no one would bother to dig inside and check. Hopefully this will teach future companies that they will be found out, so just release the code, its painless enough if you do it right from the beginning.
No, Bruce is not claiming that the lawsuit is a dirty trick. His whole argument is that he thinks he is a copyright holder in the current BusyBox and the current developers say that he is not.
I didn’t see that in Bruce’s post.
Moreover, Bruce has been around the block for a while, and is well aware that disclaiming his copyright is impossible under the terms of GPL.
A lot of the details were discussed in the slashdot comments by Bruce himself – since he submitted his blog post there and it made the slashdot front page.
If derived code is extended to mean code that was originally based on something someone once wrote even though it doesn’t include any of the original code, then Wine, GNU, all the BSDs and even Linux are violating copyright.
Wine devs are in the best position as they can claim that they have not ever seen a line of Windows code(dubious as that would be) but their software is still obviously built after Windows and using windows as a reference.
As for others, all the original GNU and BSD devs had read and learnt from AT&T code as had Linus from Tanenbaum’s. They also used that code as a base for their original systems linking to parts of it either directly or indirectly. The FSF claims that separate binaries are infringing too if they can only work with help from GPL code, so it should work the other way as well.
Basically if that reasoning were to be upheld in Court, most open source software would be infringing on someone’s copyright.
dubious? why? AFAIK Wine developers are very careful that noone who has actually seen Windows code can make contributions. If code looks fishy then they don’t accept it. Exactly because it would get them in a lot of legal trouble. There was even the thing about ReactOS, with someone accusing them of having used or seen Windows code and they did a full audit to make sure that wasn’t the case.
Firstly quite a bit of code people looked at was/is public domain. Secondly a lot of the programming in Linux BSD … was actually done according to standards (POSIX anyone) so looking at the standard not the original code. Thirdly if you only take a small part of a program, especially if you even cite the original implementation, that probably falls under fair use.
IMO you’re incorrect. Thing about a book, if you took e.g. Harry Potter and rewrote the whole book changing all the formulations, but not the story you would quite certainly be infringing copyright and probably be sued.
I do believe that they try to write original code clean from any MS influence. And I think they are doing a good job.
However, are we to believe that no one has ever seen a function disassembled in a debugger even before working on Wine? That no one has ever used Windows dlls to fill gaps in unreleased versions while they develop a replacement?
I have nothing against any of the projects I named, but if copyright worked as busybox author thinks they do, they too would be infringing. Fair use isn’t even a legal figure in most countries.
Uhm no. Some of the alternative utilities were PD. UNIX code and more importantly the kernel and the original C libraries were not PD and all of them started from there.
By the way, the first POSIX version was released in 1989. GNU was started in 1983, and BSD in 1975.
That is called plagiarism. A close relative of trademarks. I don’t think it has ever been applied to software itself as it doesn’t make sense.
http://en.wikipedia.org/wiki/Plagiarism
Edited 2009-12-16 12:59 UTC