*BSD News Article 35045


Return to BSD News archive

Newsgroups: comp.os.386bsd.development
Path: sserve!newshost.anu.edu.au!harbinger.cc.monash.edu.au!msuinfo!agate!library.ucla.edu!csulb.edu!nic-nac.CSU.net!charnel.ecst.csuchico.edu!yeshua.marcam.com!zip.eecs.umich.edu!newsxfer.itd.umich.edu!uunet!brunix!mhw
From: mhw@cs.brown.edu (Mark Weaver)
Subject: 386BSD Remarkable? (was Re: Call for 386BSD Rel.1.0 SIG (Special Interest Group))
Message-ID: <1994Aug27.175807.20821@cs.brown.edu>
Followup-To: comp.os.386bsd.misc
Sender: news@cs.brown.edu
Organization: Brown University Department of Computer Science
References: <FOO-MAN.94Aug23191512@raven.raven.csrv.uidaho.edu> <jmonroyCv2Iw2.AD9@netcom.com> <1994Aug25.074246.4082@cs.brown.edu> <Cv5u8o.CEn@calcite.rhyolite.com>
Date: Sat, 27 Aug 1994 17:58:07 GMT
Lines: 58

In article <Cv5u8o.CEn@calcite.rhyolite.com>,
Vernon Schryver <vjs@calcite.rhyolite.com> wrote:
>In article <1994Aug25.074246.4082@cs.brown.edu> mhw@cs.brown.edu (Mark Weaver) writes:
>> ...
>>I'm not saying that 386BSD 0.1 wasn't a remarkable accomplishment.
>>It *WAS*.  If it wasn't for his work, we'd probably all be using
>>Linux right now. ....
>
> [...]
>
>The first sentence is an overstatement at least by implication.  386BSD
>0.1 was remarkable in the literal sense, as in "worthy of notice," but
>technically it was at best unremarkable.  Hundreds of other people had
>ported UNIX to more difficult platforms over the previous decades, and
>done better jobs of it.  386BSD 0.1 was remarkable mostly because it
>was free.  

Actually, I think you're overlooking something here.  386BSD was
not merely a port, because it was based on Net/2, which was not a
complete operating system.  Trying to get a port up when you only
have to write a few simple routines and drivers is one thing.  But
I think 386BSD was a bit more than that.  For instance, the VM
system was replaced with Mach's VM.

I may be showing my ignorance here, because I really don't know
for sure how much is missing from Net/2.

>Before you praise 386BSD 0.1, boot it.  Remember how easy it was to
>crash the tinybsd floppy?  Remember how much of a mess it was to install
>on a hard disk?  It was perhaps more than you'd expect from the first
>3-12 weeks of a porting job, which would be fine except that that was
>all there ever was.  The "dog (disk crash) ate my homework" story was
>ok for a few weeks, but smelled bad after 6 months.

Since the hardware on the PC architecture is so varying and
unreliable, it seems to be very difficult to write robust drivers
that work for everyone.  I think you're being a little unfair.

>You could argue that 386BSD 0.1 was remarkable as a very bad thing, that
>it caused the fragmentation into NetBSD/FreeBSD/Linux.  Imagine what
>might have happened if 386BSd 0.1 had not appeared.  What if we had only
>commercial source products like BSDI and free Linux?  Linux's NFS and
>network code might be far stronger, and benefiting the efforts of people
>who know the BSD source.  That would have been a better world for many 
>reasons.

Actually, this point I might agree with.  Linux has been getting
better and better, and for those who haven't looked a Linux in a
few months, you'd be surprised.  It's extremely stable and fast,
and the networking is ALMOST there.  My only beef with it is that
the C library is primitive compared to NetBSD.

Followups to comp.os.386bsd.misc.

	Mark
--------------------------------------------------------------------
Email: Mark_Weaver@brown.edu           | Brown University
PGP Key: finger mhw@cs.brown.edu       | Dept of Computer Science