*BSD News Article 17438


Return to BSD News archive

Path: sserve!newshost.anu.edu.au!munnari.oz.au!news.Hawaii.Edu!ames!agate!soda.berkeley.edu!wjolitz
From: wjolitz@soda.berkeley.edu (William F. Jolitz)
Newsgroups: comp.os.386bsd.misc
Subject: Re: Digiboards and BSDI/386
Date: 23 Jun 1993 20:09:45 GMT
Organization: Computer Science Undergrad Assoc., UC Berkeley
Lines: 136
Message-ID: <20ada9$5k5@agate.berkeley.edu>
References: <1993Jun4.094829.6159@spcvxb.spc.edu> <1993Jun11.080552.6234@spcvxb.spc.edu> <1993Jun11.181807.8884@fcom.cc.utah.edu> <C8ryBL.Jyn@percy.rain.com>
NNTP-Posting-Host: soda.berkeley.edu


Back in 1989 when I originally created 386BSD (see "386BSD: A Modest
Proposal", which Keith Bostic was good enough to circulate widely on
the net at the time), the goals I had for it was simply to offer an
alternative platform for Berkeley *VAX* UNIX, to keep it from dying
out.

I think this goal has been achieved.

At the end of 1990, I contributed the work to the university in the
belief that it would have a profound effect at promoting use of the
wealth of ideas in BSD, in which I've had a stake since the 70's.

It seems to have been noticed.

In January 1991, DDJ started publishing the first article of a 17-part
feature series, describing the particulars of this work, so that others
may start to participate in 386BSD.

It would seem that they have.

After this work was completed, and during the publishing of my series,
Mike Karels and UUNET's Rick Adams asked me to help them make the work
available in a unencumbered forum for wide use by research,
educational, and commercial purposes. BSDi was formed to be a clearing
house for services, similar to Cygnus (or so I thought).

NET/2 is released, BSDi is operational with a "product" by summer.

What followed is a comedy of errors scenario which are almost
slapstick, to wit:

Rick gets the "gold bug", and the Three Stooges follow along. BSDi
pursues an insane direction of becoming USL, down to the "teamster"
tactics that it still occasionally employs. UC splits into two groups
-- CSRG and UC.  CSRG echos BSDi to the letter, but denies involvement.
UC adopts the ostrich position, with an index finger pointing towards
CSRG.

I walk, and return to my original work. USL, after being well provoked,
launches a noodle-brained attack, intending to psychologically encumber
NET/2 while baring corporate fangs at the BSDi "yapping" little dog and
the mindless, sleeping giant UC (fumbling along with it's hands in
other pockets, and perplexed by the Naked Guy, who isn't wearing any).

BSDi/CSRG, sensing potential competition, clumsily attempt to clean my
clock, leaving me "forever" in their debt, using inappropriate tactics
to further this end.  Yes, they already thought they were as powerful
as USL, and could do what ever they wanted.  UC, sensing that the wave
is about to crash, attempts to batten down the hatches, and preserve
the status quo, come what may. After all, the taxpayers will pay for
any legal inconvenience.

Release 0.0 is thus inspired, as a way to deliver on my obligations to
about 100 people. UC was given the option to release instead via
controlled means, but decline because of the budding conflict of
interest problem that continues to this day. They also didn't see any
value in the work -- so much for academics and their brilliant powers
of reason. That's why they're sitting in tenured jobs instead of
running businesses.

Other key people go completely ballistic, for very different reasons.
It's hard to justify a kilobuck for something you can get for nothing.

Meanwhile, back at the ranch, sufficient encouragement and assistance
allow us to put out a beta release, 0.1.  Original mission is complete,
albeit with a few holes.

The "greed" battle for control, or "cold" war begins. 386bsd florishes,
as real people finally get a chance to participate with BSD, something
they have waited for ten years for.

Our favorite judge sees through the silliness that USL insists is its
case, yet is not so taken in to let UC/CSRG/BSDi entirely off the hook.
Bit of the "pox on both your houses" is present, as he correctly
assesses the nature of both plantiff and defendant. Note that
USL/Novell still has the opportunity to launch significant actions,
which the judge is also aware of.  Obviously, the man is a keen student of
human nature.

BSDi incorrectly chooses to interpret this as a "win". It starts to
"organize" a nascent 386bsd industry, with all the polish of a
protection racket (e.g.  "you'd better play along, or we'll deal you
out"). Netbsd comes into existance about the time we politely decline
the invitation. Other commercial efforts spring into being, smelling
opportunity. The "cold" war moves into more obtuse and covert
directions, justified over the weird passion to be "The One, True,
BSD", regardless of personal liability or ethics. They are encouraged
by the cynical Three Stooges, already tarred and visible in the fray.

The user community interprets (correctly) that they are no longer
targetted, however indirectly, and breaths a collective sigh of relief.
They are completely bewildered by the lack of unity in the 386BSD
world, and target numerous "villains" for this. The boundaries of
commercial/free worlds are intentionally blurred for obvious, ruthless
ends, since the ethical success of 386bsd is more certain than the
legal successes possible. Subsuming and confusing becomes the
strategy.

BSDi correctly decides to start releasing code (init -- done two years
ago) as better versions (Christoph Robitscko's) appear publicly, as an
attempt to appear magnanomous without really doing anything.  Through
"good guy, bad guy", they can work both sides of the field. I have a
feeling they will need to release quite a bit of code soon.

Yet, the key weakness in all this gamesmanship has been the lack of
genuine cooperation, leading the nimble minds here to correctly be very
circumspect in collaborations, and extremely careful of "new work", as
it is a battleground for the control of "The One, True, BSD".

Meanwhile, UNIX continues to retreat in front of the looming Microsoft
cloud, polishing their gold braid and boasting of conquering worlds
while losing market share.

It is just good enough to make a TV movie or maybe even a book (hint).

However, the real answer continues to sit out in plain view, if one is
not so entranced in deceptions that they cannot look beyond their
nose.  Such work will continue only if it really *is* genuine, and
abandons the almost 30-year old mechanisms that are obvious weaknesses
present in 4.4BSD.  Otherwise, the "litigate if you dare" attitude will
become permanently attached to those various "BSD's", and eventually
will curse them as the average user finds the continued ambiguity
distasteful.

I am certain that we have not heard of the last legal actions here. The
pity is that it was all avoidable, except for the deliberate, conscious
choices by individuals who knew better. 386BSD is on a course away from
the pending maelstrom, and will not be used to prop up other
cynical and pointless ambitions. However, the right road will cost us
in the short-term as we build a firm foundation for the long-term.

Then again, the whole problem all along has been taking the greedy
short-term view, hasn't it.

Bill.