At Wed, 12 Sep 2012 11:14:49 +0700, Robert Elz <kre%munnari.OZ.AU@localhost>
wrote:
Subject: Re: 80386 support
>
> Date: Tue, 11 Sep 2012 19:14:42 -0700
> From: "Greg A. Woods" <woods%planix.ca@localhost>
> Message-ID: <m1TBcTS-004lScC%more.weird.com@localhost>
>
> | However I don't think I ever ran any kind of unix on any 32-bit machine
> | with just 2MB of main memory, or if I did it was just to see if it could
> | boot.
>
> I have no idea what this thread is actually about these days (the recent
> messages don't seem to contain any notable content) but don't any of you
> people remember the "definition" of a useful workstation ...
Indeed I remember Bill Joy giving a talk about the nearly available
1M/1M/1M workstation of the future! At the time a megapixel display was
still rather high-end graphics (at least in my estimation at the time)
as was a workstation with more than 256KB of RAM. I wouldn't have been
self-hosting a whole 32-bit unix on such devices either, and besides,
they were usually diskless. The Sun-1 at the time still only ran a V7
port, not a true 32-bit VM unix (more about my distinction below).
Anyway, my comment was mostly aimed at i386 systems of course. :-)
I ran NetBSD on a Sun 3/260 with only 16MB of RAM for a while, but I
would never have even dreamt of trying to build a kernel on it, or
anything else for that matter. I used it as a workstation for a while
but I tried not to run too many xterm processes on it -- just the
Xserver, an xclock, and maybe one or two local xterms with the rest
running on a bigger machine in the basement. The 3/60 that was my main
workstation at the time was full of SIMMs, so 24MB IIRC, and I didn't do
any compiles on it either.
I guess either could have self-hosted with a full NetBSD build, but what
was the point? With little or no disk on them there had to be a bigger
NFS server somewhere else anyway, so why not do the builds on the bigger
server directly -- even old SCSI disks were a heck of a lot faster on
the server where they were directly connected than via NFS over good old
"slow" 10Mbit/s Ethernet.
> The first 32 bit system I ran unix on had 384KB of RAM, that also ran
> unix, self hosted, and handled multiple users (no networking, networking
> for unix had yet to be invented, not even uucp in the early days). What
> it ran initially was a variant of 6th edition unix that was designed to
> be able to run on non split I/D PDP-11's (which meant total program size
> for any program, including the kernel, < 64KB (56KB for the kernel to allow
> memory mapped I/O controller access). Even with the natural expansioon
> from using 32 bits everywhere instead of 16, and a less effecient
> instruction set than the PDP-11's, 384KB was quite a lot of RAM
> (and expensive, and used a lot of physical space.)
So it wasn't really running a true 32-bit Unix, was it. :-) It probably
wasn't even running as much code per process on average as Unix 32V.
For all intents and purposes even 32V was just running a 16-bit unix in
32-bit mode.
I don't think it's fair to try to equate the code sizes and memory
requirements of a 16-bit system even if it was modified to run on 32-bit
hardware. The purpose of my original reply to Mouse was to point out
that a modern 32-bit system inherently requires more memory (and more
patience too as there's more code to run overall).
4BSD with full TCP/IP networking support would in my mind be the first
true 32-bit unix-flavour system that it would be fair to compare NetBSD
with in any way. IIRC a Sun-2, the first Sun to run a 4BSD variant, and
going by Wikipedia it was apparently typically configured with at least
2MB of memory, and probably 4MB if you wanted to do any significant
development work on it. It had a pretty primitive, and small, compiler
and toolchain too, even compared to later SunOS m68k compilers.
I used a VAX 11/780 running 32V for a year -- but it had 4MB of RAM
right from the first day, IIRC. The PDP-11/40 sitting beside it even
had 4MB of RAM, but it replaced an older 11/60 with somewhat less RAM.
Both 11s ran V7. The VAX ran 3BSD briefly at the beginning of the next
term and then 4BSD after that, but I seem to remember it needed more
memory after the 4BSD upgrade, even before any new terminals were added,
and the compiler ran noticeably slower after the upgrade as well. The
last time I used that VAX I think it had 16MB of RAM and 64 terminals,
but with all 64 in use it seemed slower than the old 11/60 was with 16
active users.
Now modern systems such as NetBSD have things like SSL libraries, locale
support, wide-char support, GCC!, etc., etc., all adding megabytes more
code.
--
Greg A. Woods
Planix, Inc.
<woods%planix.com@localhost> +1 250 762-7675 http://www.planix.com/
Attachment:
pgpXOOtMJXR0u.pgp
Description: PGP signature