*BSD News Article 4428


Return to BSD News archive

Xref: sserve gnu.misc.discuss:6100 comp.os.mach:2040 comp.unix.bsd:4476
Path: sserve!manuel!munnari.oz.au!network.ucsd.edu!usc!wupost!uunet!mcsun!uknet!gdt!aber!aberfa!pcg
From: pcg@aber.ac.uk (Piercarlo Grandi)
Newsgroups: gnu.misc.discuss,comp.os.mach,comp.unix.bsd
Subject: AT&T USL vs. BSDI/UCB, Mach3, OSF/1, GNU HURD, Linux
Message-ID: <PCG.92Aug30225902@aberdb.aber.ac.uk>
Date: 30 Aug 92 22:59:02 GMT
Sender: news@aber.ac.uk (USENET news service)
Reply-To: pcg@aber.ac.uk (Piercarlo Grandi)
Organization: Prifysgol Cymru, Aberystwyth
Lines: 71
Nntp-Posting-Host: aberdb


The USL suit has direr implications that most people realize. The FSF
has decided to abandon the BNR2SS server as it is derived from the
disputed NET2 sources, and CMU have withdrawn the BNR2SS sources
from public distribution. The FSF is considering building a Mach3 server
based on Linux.

Unfortunately, I reckon this is pointless. Let's look at USL claims
in the BSDI/UCB suit:

1) NET2 and thus BSDI/386 contain USL copyrighted text, or its
   derivative.

2) NET2 and thus BSDI/386 embody USL trade secrets licensed to
   UCB, and by UCB disclosed to BSDI or its employees.

The defense by UCB/BSDI is that:

1) all text copyrighted by USL has been removed from NET2 and thus
   from BSDI/386, and substituted with original text.

2) no USL trade secret was employed in writing the replacement text,
   but only original or publicly known techniques.

The interesting aspect of this matter is that the above points, both
USL's and UCB/BSDI's, apply *exactly* to Mach3 itself.

The lineage of Mach3 starts from 4.1BSD+Accent IPC, and by a process of
various releases all USL copyrighted text has been removed, and new text
written in its stead, until Mach3 was obtained, just like NET2 was.

Not only that, all the authors of Mach3 have been exposed to the very
same USL trade secrets to which the UCB/BSDI people have been, as both
groups of researchers have worked for (over) ten years on various
releases of USL source text.

It can actually be argued that if USL wins the UCB/BSDI case, then
*all* staff and students of *any* institution or company that has an
USL Unix source license must be presumed to have been exposed to the
trade secrets contained therein, unless it can proven otherwise.

  Doing so requires proving that the sources have been kept secure and all
  people who have accessed them have been logged -- once the license has
  been granted, all staff and students of the licensee are bound by it,
  and the burden of proving that they never had access to the secret text
  is on them, and even if they can I have doubts that it matters).

So, if USL wins the UCB/BSDI suit by having the court accepting its
contentions, then Mach3 most probably falls too as the next step, and
(less probably) so any and all Unix like code developed by staff or
students of any organization with a Unix source license.

It would astonish me if, having set a precedent w.r.t. UCB/NET2/BSDI,
USL were not to use it immediately w.r.t. CMU/Mach3/FSF. By doing so
they would effectively prevent the OSF, with their current staff, from
producing an OSF/n that does not require an USL Unix source license.

So, giving up on BNR2SS buys essentially nothing, if one continues to
use Mach3 as the substratum

The sure way for the FSF to have an unenncumbered OS would be to adopt
Linux, if it can be proven that Linus Benedict never worked or studied
at an organization with a Unix source license, or if he did, that he
never was exposed to it.

Or one could continue to use BNR2SS with Mach3, because they are bound
to stand or fall together.
--
Piercarlo Grandi                   | JNET: pcg@uk.ac.aber
Dept of CS, University of Wales    | UUCP: ...!mcsun!ukc!aber-cs!pcg
Penglais, Aberystwyth SY23 3BZ, UK | INET: pcg@aber.ac.uk