Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!samsung!usc!rutgers!mit-eddie!uw-beaver!zephyr.ens.tek.com!tektronix!sequent!norsk
From: norsk@sequent.UUCP (Doug Thompson)
Newsgroups: comp.os.minix
Subject: Re: Once more. No more.
Message-ID: <36055@sequent.UUCP>
Date: 1 Jun 90 15:58:31 GMT
References: <20721@nigel.udel.EDU>
Reply-To: norsk@crg2.UUCP (Doug Thompson)
Organization: Sequent Computer Systems, Inc.
Lines: 33

In article <20721@nigel.udel.EDU> 702WFG%SCRVMSYS.BITNET@cornellc.cit.cornell.edu (bill gunshannon) writes:
>I&D = Instruction & Data
>The machine used 2 separate 64K blocks of memory one for data and one
>for executable junk.  Sounds kinda like INTEL small model when you get
>right down to it.
>It was because although Memory Management allowed up to 512K, the processor
>could still only work with 64K at a time.  There were worse.  I have an
>LSI 11/03.  It has 56K bytes of word addressable memory or the equivalent
>of 28K worth of address space.  Try cramming MINIX into that. :-)
>
>bill
>
>                                          bill gunshannon
>                                       702WFG@SCRVMSYS.BITNET

In 1980, at BYU, we had an old PDP-11/40 (No MMU). We were able to get a
copy of Mini-UNIX (v6) that supported 3 users. Like the 11/03, there
was only 28kwords of with core, the remaining 4kw  was IO space. The
kernel was 12k words, allowing only 16k words for program space. Time
sharing was like: swap user 1 in, run for a second, swap user 1 out, swap
user 2 in, run a second, swap user 2 out, swap user 3 in, run a sec, and
repeat. ...

The system compiled itself, and worked fairly good. I even did my compiler
course on it - the only one in the class who used a high level language -
all else used DEC asm. ugh!  Things were small then, but we got alot
done.


-- 
Douglas Thompson		UUCP: ..{tektronix,ogicse,uunet}!sequent!norsk
"The scientist builds to learn; the engineer learns in order to build."  
Fred Brooks