On Mon, 21 Jul 2003, John Clark wrote:

> Linux clearly was at free from any taint, but the current
> claim as I read it was that USL code had been introduced into
> the package somehow some way.
>
> What controls are there for NetBSD here, or in general
> the BSDs to avoid re-introducing USL code into the distributions.

One way is to make sure that when developers import code (theirs or
others), that they clearly log where the code originated from (like
author, dates, license).

This may not stop the actual introduction of tainted code, but at least,
it makes it easier for other developers to double-check the origination of
the work.

This morning, I had the opportunity to participate in SCO's media
teleconference. SCO said that "IBM put their signature on it", in regards
to some of the Linux 2.4 code. It seems like they indicated the history or
changelogs of the code, but in layman's terms.

Also, they said some of this problem code was originally based on Dynix.

SCO said some direct line-by-line code copying was done, including
developer comments, errors and typos.

SCO's president clearly said "We are not talking about BSD code."

See http://www.bsdnewsletter.com/2003/07/News91.html.

On a related note, some Linux code from a couple years ago, reused some
BSD-licensed code without retaining the copyright/disclaimer. I wrote
about it "Redistributions of source code must retain the above copyright
notice ..." (but that news website is now dead).

   Jeremy C. Reed
   http://bsd.reedmedia.net/