Bruce Perens: Statement on Busybox Lawsuits

Posted Dec 15, 2009 20:12 UTC (Tue) by JoeBuck (subscriber, #2330)
In reply to: Bruce Perens: Statement on Busybox Lawsuits by bangert
Parent article: Bruce Perens: Statement on Busybox Lawsuits

Bruce, if you're reading this discussion, could you point out where this detailed analysis goes wrong? What part of the current BusyBox do you claim authorship of? If you're arguing that you still have ownership because of they way they transformed away your code, leaving nothing left, wouldn't that put you on the USL side of the BSD lawsuits of the early 1990s?

Bruce Perens: Statement on Busybox Lawsuits

Posted Dec 15, 2009 20:50 UTC (Tue) by joey (guest, #328) [Link] (1 responses)

That analysis ends with a "still todo" section listing 50+ files, and the
current release of busybox still has 8 C files with Bruce's copyright on
them.

Bruce Perens: Statement on Busybox Lawsuits

Posted Dec 16, 2009 9:41 UTC (Wed) by niner (subscriber, #26151) [Link]

That analysis is also more than three years old. I'd guess things went on
since then.

Bruce Perens: Statement on Busybox Lawsuits

Posted Dec 15, 2009 21:30 UTC (Tue) by BrucePerens (guest, #2510) [Link] (8 responses)

Rob is mis-interpreting Judge Walker's means for determining the presence of non-literal copying. There is literal copying present. There is also a compilation copyright present.

Bruce Perens: Statement on Busybox Lawsuits

Posted Dec 17, 2009 21:57 UTC (Thu) by landley (guest, #6789) [Link] (7 responses)

There is literal copying but not _from_bruce_.

The idea of "one binary many functions dependent on name" came from _gzip_, which acted as gzip or gzunzip based on the name it was called as, and was one of the applets in the original debian bootloader.

The compilation copyright would be in the selection of which code to include. Ten years later we've thrown out several of the old applets completely, swapped out completely different implementations of a bunch, and added literally hundreds more.

By all means, if he'd like to make a compilation copyright argument in court, I'd happily be a witness _against_ him if asked. I made some of those changes and the "compiliation" was based on what the public susv3 spec listed plus the man pages of Red Hat, Ubuntu, SuSE, and Slackware. What the current code was doing wasn't particularly relevant, and that was a half-dozen years after he abandoned his bootloader aid and moved on. (Remember, Bruce never addressed the embedded space at all. That was Erik Anderson launching a _new_ BusyBox project.)

Bruce Perens: Statement on Busybox Lawsuits

Posted Dec 17, 2009 22:14 UTC (Thu) by BrucePerens (guest, #2510) [Link] (3 responses)

Rob the idea did not come from gzip and no person who has any idea about early Unix would say so. Using argv[0] to bring up different programs, with more than one program name linked to the same executable, was present in Version 6 unix. This came about in the 1970's.

I never claimed to have invented that feature, Rob. And in any case, it would be subject to a patent if I did, not copyright.

Your "forensic study" was nonsense, as far as I can tell. First, you used a method for identifying non-literal copying from CAI v. Altai when literal copying and a straight derivative of past work was present. Second, you ignored the presence of both code and copyright statements in multiple files that still exist today. Thus, I'm forced to conclude that you wrote that without ever looking.

We just made a compilation copyright stand in JMRI. The court ruled for summary judgement in Jacobsen's favor last week. I'm one of the experts on the prosecution's side.

Bruce Perens: Statement on Busybox Lawsuits

Posted Dec 18, 2009 0:23 UTC (Fri) by paulj (subscriber, #341) [Link] (2 responses)

This is not meant in a disrespectful manner, but I have to ask:

Is there not a conflict of interest if you are testifying as an expert witness in one case involving compilation copyright, when you are involved in a separate dispute that hinges on validity of such copyright? I.e. would you have the opportunity to influence case law in the one case favourably towards your interests in the other case in which you have a direct stake?

I do not mean to impugn your reputation with this question. Rather, this question arises from what you've written, in the mind of at least one LWN reader, and I just wanted to highlight that for you so that you could clarify it (this theoretical conflict of interest may matter not a jot to a court with an adversial system - I guess, but I don't know).

--paulj

Bruce Perens: Statement on Busybox Lawsuits

Posted Dec 18, 2009 0:27 UTC (Fri) by BrucePerens (guest, #2510) [Link]

Is there not a conflict of interest if you are testifying as an expert witness in one case involving compilation copyright, when you are involved in a separate dispute that hinges on validity of such copyright? I.e. would you have the opportunity to influence case law in the one case favourably towards your interests in the other case in which you have a direct stake?

As it happens I probably won't have this particular conflict. But I would think it happens all of the time, in that a plaintiff can bring multiple cases regarding something of theirs which has been infringed, and can cite the previous cases in the pleadings of the later ones.

I think this is the way that the court intends it to work.

Thanks

Bruce

Bruce Perens: Statement on Busybox Lawsuits

Posted Dec 18, 2009 5:19 UTC (Fri) by landley (guest, #6789) [Link]

> Is there not a conflict of interest if you are testifying as an expert
> witness in one case involving compilation copyright, when you are
> involved in a separate dispute that hinges on validity of such copyright?

I'm not involved in the current case. (I posted a link to my December 15 blog entry explaining why.) All the SFLC cases I was involved in were resolved a year or more ago.

Also, Bruce is not a party to any current busybox-related case I am aware of. (That's what he's upset about, the SFLC's total lack of interest in him. They seem to think he's irrelevant too.) I'm just saying that if he was, I would be opposed to his participation, to the extent of testifying against him if asked.

That's not news, either: http://lwn.net/Articles/202106/

Bruce Perens: Statement on Busybox Lawsuits

Posted Dec 17, 2009 22:26 UTC (Thu) by BrucePerens (guest, #2510) [Link]

(Remember, Bruce never addressed the embedded space at all. That was Erik Anderson launching a _new_ BusyBox project.)

This is totally ignoring the role of Dave Cinege in the Linux Router project. He maintained Busybox for two years in that context.

Also, the one-floppy Debian installer had all of the qualifications of an embedded system - very limited resources, etc., - with one difference which was that it was ephemeral. Which as far as I can tell doesn't disqualify it.

For that project, I also made a program to cut down shared libraries to only the functions that were actually called for in a given set of executables. This allowed us to have an embedded-sized C library without new development.

Bruce Perens: Statement on Busybox Lawsuits

Posted Dec 23, 2009 2:21 UTC (Wed) by Baylink (guest, #755) [Link] (1 responses)

Bruce is right: "decide which function to perform by looking at argv[0]" predates gzip.

I have here in my hand Kernighan and Pike, _The_UNIX_Programming_Environment_, which suggests the idea on page 85, as "a novel use".

Publication date?

1984.

Anyone wanna go back any further than that?

Bruce Perens: Statement on Busybox Lawsuits

Posted Dec 23, 2009 22:51 UTC (Wed) by landley (guest, #6789) [Link]

My point was merely that gzip's use of the technique predated busybox.

Otherwise: dead thread, by the editor's request.

Bruce Perens: Statement on Busybox Lawsuits

Posted Dec 15, 2009 22:03 UTC (Tue) by ncm (guest, #165) [Link] (1 responses)

USL vs BSD was a very different kettle of fish, because USL was caught with dirty hands. That doesn't negate Joe's first question, though. I'd hate for historical confusion to interfere with getting a clear explanation from Bruce.

Bruce Perens: Statement on Busybox Lawsuits

Posted Dec 16, 2009 1:38 UTC (Wed) by BrucePerens (guest, #2510) [Link]

I think that USL did things to weaken their copyright that don't apply in my case, simply because of the date. Works are now copyrighted the moment they are set on paper, unlike back then.

It's hard for me to see how the program could not be derivative, given that it is an uninterrupted progression from my code base.