More corrections...

> When NetBSD broke off it was considered militant. The was partly
> because Chris Demitrious did not get along with people. Other people
> were upset becuase their patches (submissions to the patchkit effort)
> were not accept. There was much ill feelings. Chris is now a different
> person, I think he learned things. Those other peoples are now
> the core team at NetBSD.

There were some patches released in patchkit format (reverse
engineered) that ignored the need to serialize operations.  I
think that after the people who did this had "Makefile"s
explained to them, and were offered the real patchkit tools,
the conflict problems went away.  NetBSD was (per my other
posting) mostly people who were tired of waiting, and thought
progress was too slow, and weren't willing to leave it in what
they percieved as "too slow" hands.  Let's be honest: it's
still "too slow" for some of us... no matter what camp we are
currently in.


> AT&T also added pressure at the time by claiming Unix was a
> National (treasure??) and therefore should be consider un-exportable.

AT&T (USL) tried to claim trade secret status for UNIX; BSD
Net/2 contained the components they were complaining about,
but following disclosure, they had no trade secret status.
UCB was not accountable anyway, since the code derived from
code licensed without non-disclosure clauses, under the old
Western Electric license.  UCB EECS in fact did not renew
their UNIX license when the Western Electric license changed
to prohibit disclosure, so the cat was already out of the bag,
even if one ignores the Lions book, published by the University
of New South Wales, under the same Western Electric license,
lacking a non-disclosure requirement.

AT&T was, in fact, under a consent decree based on an antitrust
action under the Sherman Antitrust act, preceeding the breakup
(the "Judge Green Decision"), prohibited from making money on
software, or of obtaining any intellectual property protection
for UNIX, whatsoever.

Technically, then, the "harm", even if provably real, wasn't
monetarily recoverable; a circuit court judge basically said
that, when he admonished USL about their attempt at a restraining
order.  The part that got the press, though was when he called
their claims frivolous on trade secret grounds.

I never heard the export issue, but I would think that if it
were an issue, it would be a National Security issue, not an
issue of national propriety, since UNIX was used in most of
the digital telephone switches manufactured, particularly
those from Northern Telecom and AT&T.  These switches were
sold outside the US at the time, anyway.

Mostly, USL continued the suit out of a risk-reward calculation
(IMO), and it escalated to places other than BSDI because of
early briefs filing for summary dismissal on the basis of
"failure to exercise due dilligence" (basically, some people
who I believe are long gone from BSDI "hid behind" UCB when the
bully came out to beat them up [an overreaction to the "yo mama"
of the "1-800-ITS-UNIX" phone number])


> AT&T sold Unix to Novell for $1 Billion dollars in the middle
> of this and in reallity it was Novell that settled. BTW, this
> $1B almost bankrupted(sp?) Novell.

Novell bought USL for $80M, which is only 8% of the figure
you quote.  This is the same price they charged Sun to get
out of royalty payments, and the later sale of USL to SCO was
nothing but gravy for them: very good ROI, in fact.

Novell settled because it was a P.R. nightmare, and many of
the "Novell USG - UNIX Systems Group" and now dissociated Bell
Labs people, including Dennis Ritchie, threatened to testify or
file Amicus Curie briefs on behalf of UCB, totally undermining
USL's legal case.


The $1B purchase made at around the same time was the purchase
of Word Perfect.  Along with AppWare (another company started
with finding from the Noorda Family Trust, and later purchased
by Novell), which provided Novell with "COM" and "DCOM"-like
technology, any the purchase of spreadsheet software from Borland,
this was Novell's entry into competition with Microsoft.  The
AppWare purchase triggered the others, since Novell found out
that third party companies would not voluntarily commoditize
their software into invisible non-logo'ed component-ware, unless
you bought them and forced them to do it (or you were able to
wield monopolistic power in the marketplace to force them to do
it, as Microsoft later did with COM).

The reason the purchase was such a bone-head move was that it was
the first time that a company had used the well-known Novell
company valuation benchmark (PPE - Profit Per Employee) to
inflate their apparent value to astronomical heights.  In order
to do this, Word Perfect cancelled free support, cancelled all
forward looking products, all pen-based products, all marginal
products (like post 4.2 for UNIX and VMS), etc., and then let
those employees go.  Cutting your workforce nearly in half at
the end of a corporate fiscal year does wonders for PPE (if
you can't change the numerator, change the denominator).  Don't
think that Novell didn't learn from this: big does not necessarily
equal stupid.

Novell was (and still is) a pretty shrewd company; any company
that sticks around for any real length of time, and outlives
the initial cult-of-personality that started it by a management
generation plus one, is probably in it for the long haul.  It's
definitely on my "hold" list, and I'm likely to upgrade it, if
the price continues to stay low with the current P/E ratio for
much longer (and I can get it for the price of a fractional long
term capital gains cash-out of some other investments ;-))...



					Terry Lambert
					terry@lambert.org
---
Any opinions in this posting are my own and not those of my present
or previous employers.