Xref: utzoo comp.unix.questions:12095 comp.unix.wizards:14977
Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!eecae!tank!mimsy!chris
From: chris@mimsy.UUCP (Chris Torek)
Newsgroups: comp.unix.questions,comp.unix.wizards
Subject: Re: Future at Berzerkeley
Message-ID: <16273@mimsy.UUCP>
Date: 8 Mar 89 07:12:53 GMT
References: <15184@cup.portal.com> <15407@cup.portal.com> <16230@mimsy.UUCP> <13324@steinmetz.ge.com>
Organization: U of Maryland, Dept. of Computer Science, Coll. Pk., MD 20742
Lines: 70

In article <13324@steinmetz.ge.com> davidsen@steinmetz.ge.com
(William E. Davidsen Jr) writes:
>  Is there a future for BSD? Ignoring the issue of when new releases
>will be available, I get the impression that virtually all of the
>hardware vendors have joined OSF or UNIX International. Since both of
>these systems will be SysV based, will there be a demand for BSD in
>three years? Five?

BSD is a research vehicle.  What does `vendor demand' have to do with
its existence?  Remember, *no* release of 2BSD has been officially
supported by *anyone*.  (2BSD may be dead---the 2.10 BSDers have been
trying to kill it for years---since PDP-11s are just Too Small these
days.  But there is still research interest in 4BSD.)  3BSD (the first
VAX system) was done at Berkeley for Berkeley.  4.0BSD and 4.1BSD ran
on VAXen when DEC's position was `VAX == VMS'.  4.2BSD primarily got
out of hand because the U.S. Federal government demanded both Unix
*and* TCP/IP.  Vendors found they had the choice of either adding
TCP/IP to SysV or using 4.2BSD---or losing the government market.

Government, of course, *is* *the* Big Business.  (The only one that can
afford to show a loss year after year, too.  Hmm....)  Vendors
scrambled to port Unix to their hardware.  Many of them quite sensibly
used 4.2BSD---it was easier to add any demanded SysV features to BSD
than it was to add networking to SysV---and, alas, many of them simply
ported the kernel, then stopped.  Since 4.2BSD was released before it
was finished, it was naturally quite buggy.  I cannot guess how much
that effect contributed to the move toward SysV, but I think it was
rather small in the end.  Making SysV standard is attractive to AT&T
for the obvious reason; it is good for IBM as well, since one of their
major government market competitors (DEC) has made a commitment to
BSD.  So there is plenty of money behind the `consider SysV standard'
campaign (which I naturally buck with `consider it substandard' :-) ).

Okay, so what does this imply for future government funding for BSD?
Suppose the Feds buy the `standard' line.  The bureaucrats grind out
new Official Regulations mandating that new buys meet the SVID.  But
say DoD wants feature X (ISO, Mach VM, pick your favourite).  They can:
    - get an exception so that they need not meet SVID
    - fund someone to put feature X in SysV
    - fund someone to make BSD SVID-compliant and put X in BSD
The second and third cost options about the same, and are cheaper
than the first.  Pouf!: instant research money.  (`Hey, Joe, run
off another 250,000 $20s'... :-> )  Who gets it?  Pouf!: instant
politics.  Obviously CSRG gets a shot.

Is option 2 cheaper?  I.e., would CSRG start doing SysV-based systems?
Maybe.  But the Feds are not going SVID: they are going POSIX, and
CSRG are going that way too.  The research money is sure to go to
some university, and universities prefer BSD.

Implication: future government funding is likely to be around.
CSRG may or may not get some.

Suppose the Feds pull out.  (Say George B. catches on to the fact that
half the Pentagon's paper-pushers are actually just moving the papers
round and round the building.)  The deficit starts shrinking and the
depression shows up full force.  Money gets scarce (as Joe in the print
shop stops the presses).  Only the Really Big universities are going to
be able to keep things going.  Berkeley is one.  Would CSRG survive?
No telling.  But they have a better chance than many.

All in all, the only thing sure to kill BSD is lack of interest (pun
required).

>  Information or rational comments desired.

Oh.  Too late.
-- 
In-Real-Life: Chris Torek, Univ of MD Comp Sci Dept (+1 301 454 7163)
Domain:	chris@mimsy.umd.edu	Path:	uunet!mimsy!chris