Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!husc6!cmcl2!rutgers!sunybcs!boulder!hao!oddjob!gargoyle!ihnp4!alberta!auvax!rwa
From: rwa@auvax.UUCP (Ross Alexander)
Newsgroups: comp.misc
Subject: Re: DAEMONs -- what does GCOS mean anyway ?
Message-ID: <350@auvax.UUCP>
Date: Wed, 14-Oct-87 20:00:24 EDT
Article-I.D.: auvax.350
Posted: Wed Oct 14 20:00:24 1987
Date-Received: Fri, 16-Oct-87 07:15:50 EDT
References: <1131@nrcvax.UUCP> <3550002@hplsdla.HP.COM> <392@xios.XIOS.UUCP> <3360@eagle.ukc.ac.uk>
Organization: Athabasca U., Alberta, Canada
Lines: 22
Summary: roff under GCOS

<void where prohibited by law>

Someone (I didn't bother to include the article, as you may have noticed
by now ;-) mentioned that `roff' was originally written in pdp11 assembly.
That would be Joe Ossana's code, I suppose.  But that drove the CAT photo-
typsetter, and would thus be the first version of `troff'.

I remember compiling `roff' in early 1975, on the UofWaterloo's H6050 under
GCOS.  It was written in B, the predecessor of C, and I am very sure of that
since I was at the time writing a (not very sophisticated) run-time profiling
package and the canonical test case was to compile roff with the profiling
turned on and then check that the resulting code ran correctly (which it
indeed did, eventually).

I shudder to think of that package (the profiler, I mean; roff was ok) - it
modified the object code on the fly, inserting instructions & generally
violated all the rules of good programming. I suppose it's a dusty deck now,
assuming it exists at all.

--
Ross Alexander @ Athabasca University,
<the known world>!ihnp4!alberta!auvax!rwa