[linux-elitists] [mds@hsmds2.prnewswire.com: SCO Suspends Distribution of Linux Pending Intellectual Property...]

Martin Pool mbp@sourcefrog.net
Thu May 15 16:58:27 PDT 2003

On 15 May 2003, Ben Woodard <woodard@redhat.com> wrote:
> Don,
> How can they prove legally that one piece of source code was taken from
> another? i.e. How can they prove that it isn't convergent evolution #1
> and that the code didn't migrate the other way?

I suppose you would get several experts to testify that it was
extremely unlikely that the code was coincidentally similar.  Having a
few algorithms that are the same, or consistent naaming patterns is
unsurprising, particularly if the structure is inspired by BSD or
Minix or something else well published.  Having whole files with the
same function names in the same order, and similarly structured code
inside them is pretty conclusive.

(Supposedly original proprietary SMB servers have contained the string
"Andrew Tridgell". :-/)

Of course once they nominate which files are supposed to infringe, it
should be trivial to resolve -- the original contributor can explain
how they developed it, and whether it was based on some public code.

I would guess the most likely place for this to have happened is in
the driver for some undocumented hardware, where a person who happened
to see the SCO code used it as inspiration for writing a Linux driver.
That seems like the place where the temptation is greatest.  Perhaps
not copy&paste, but arguably a violation of SCO's confidential
knowledge about how the hardware works.  But I don't think even that
is very likely.

> > <ianal>If Sontag is lying, the kernel developers have a good basis
> > for a libel suit.</ianal>

I (idly) wonder if there are grounds for a libel or defamation suit in
Australia, given that they are damaging the reputation of Linux
developers without substantiation.


More information about the linux-elitists mailing list