*BSD News Article 35060


Return to BSD News archive

Newsgroups: comp.os.386bsd.development
Path: sserve!newshost.anu.edu.au!harbinger.cc.monash.edu.au!bunyip.cc.uq.oz.au!munnari.oz.au!constellation!qns1.qns.com!news.sprintlink.net!redstone.interpath.net!ddsw1!panix!MathWorks.Com!yeshua.marcam.com!charnel.ecst.csuchico.edu!olivea!trib.apple.com!amd!netcomsv!calcite!vjs
From: vjs@calcite.rhyolite.com (Vernon Schryver)
Subject: Re: 386BSD is dead; long live 386BSD...
Message-ID: <Cv1oHM.24v@calcite.rhyolite.com>
Organization: Rhyolite Software
Date: Wed, 24 Aug 1994 15:03:22 GMT
References: <33b84i$482@spruce.cic.net> <CEB.94Aug24000713@netcom16.netcom.com>
Lines: 74

In article <CEB.94Aug24000713@netcom16.netcom.com> ceb@netcom16.netcom.com (Ch. Buckley) writes:
>In article <33b84i$482@spruce.cic.net> pauls@locust.cic.net (Paul Southworth) writes:

> ...
>Very well said.  But it's not quite fair to say that the Jolitz' code
>development will ever be irrelevant to the *BSD family of operating
>systems -- they were the first-movers, and it would not have happened
>as fast or at all if they hadn't put the bugs in people's ear.  

That is a good point.  Actually sitting down and doing a straight 4.3BSD
80386 port instead of just talking about it, and then dragging that
laptop around the country side was admirable and very influential.


>  ...
>It seems to me that a version of this dream (if I've guessed it
>correctly) would be in the interests of many who are active in this
>community, and to the academic community at large  ...


If the code in 386bsd 0.1 that was not a straight copy of 4.3BSD had
not been so bad, that idea might not be so distressing.  All we need is
more people that produce things of the bug-free, clean coding style of
386bsd, and are not forthcoming about which parts were novel and which
copied from existing, publically available source.

There is another, perhaps worse problem with that dream, although it's
painful to contemplate something worse than a generation of new-grads
cranking out code of 386bsd 0.1 quality and setting and meeting schedules
in the 386bsd mode.  Few active, productive communities need or will
tolerate non-participating czars, people who direct the work of the
community without writing any of it.  There are plenty of commercial
outfits and consortia with that style.  Those who like the environment
or the products it produces currently can choose System V Release 4,
COSE, or OSF.

As we all recall, CSRG, starting at least with 4.3BSD was different on
all of those counts.  They were careful to give credit to other people's
code.  Their own code was of good quality.  They put their code where
their mouths were, or just produced code and kept quiet.

The main problem with the 386bsd phenomenon is that personality cults
are bad.  It doesn't matter whether the object of the cult is as powerful
and effective as Hitler or Stalin, average like Peron, or just a Breznyev.
The results are never good and sometimes catastrophic.  You might say
that there are cults that follow D.Knuth or V.Jacobson, but because both
stay so far away from their admirers, those cults are no worse than the
Peronists of today.

The work in 386bsd does not justify a cult nor a chair at a major
university.  It is the same as the work of hundreds of people who worked
on UNIX ports during the previous decade.  There were many dozens of 1
to 10 person groups who ported 4.1aBSD through 4.3BSD-reno to platforms
that were mostly much less like a VAX than a 386 PC.  Some did cleaner
jobs.  Some did it faster.  Some did it worse and took more time.  Some
got very rich.  Most are unknown.  Most of the results looked less like,
and had more novel or "experimental" code than 386bsd.  Practically none
brag or even talk about the work except on their resumes.  There were
substantial commerical outfits whose main businesses included UNIX
porting for hire, and that had people who spent years grinding out one
port after another.  Doing a port without a paying customer or boss and
without paying attention to what other people wanted is both admirable,
and instant disqualification as *BSD czar or even guru.  (Well, there
was the pay from BSDI, but that's not a pretty story nor a story where
the parties agree about the facts, so forget it.)  Working on a free
port is also admirable, but there are now hundreds of other people with
that distinction, and even rms got only the one MacArthur grant.


In this business, no one who cares really what you did last year.  All
that matters is what you're working on now and what you released last
week.

Vernon Schryver    vjs@rhyolite.com