[Simh] DECUS C

[Villy Madsen]

 I did some playing around with a pascal compiler from National Bureau of
Standards (it was an intern's summer project) in the late 70's.  We ran it under
RSTS/E.  It was a two stage compiler - the first pass generating p-code & the
2nd stage being platform/os dependent.  I was trying to replace a Bplus app that
we used for copying distribution tapes (we were a DEC VAR).  The BP app would
take hours to build a tape.

That first time that I ran it, I was sure that there was a problem - it looked
like the tape drive was just searching for the end of the tape!!

We looked at some of the machine code generate and couldn't believe the code
that was generated - our resident guru couldn't have done any better....

Villy

-----Original Message-----
From: simh-bounces at trailing-edge.com [mailto:simh-bounces at trailing-edge.com] On
Behalf Of brian
Sent: Sunday, August 15, 2010 22:27
To: simh at trailing-edge.com
Subject: Re: [Simh] DECUS C

Ethan Dicks wrote:
<...>
> 
> I don't think there was anything remarkable about Whitesmith's C 
> except that it was available to buy when others were not,

<...>

Whitesmiths. Ugh! There's a name which brings back some memories. I spent a
couple of years developing software on a PDP-11/44 under RSX-11M and the first
"compiler" we had was Whitesmiths Pascal.
"Compiler" because what it actually did in those days (early 80s) was to
translate the Pascal into C and then compile and link the C program. Since we
were debugging from the generated Macro-11 code by patching breakpoints into the
executable and then using ODT, we got to see how truly awful the generated code
was. Fortunately we then found out about Oregon Pascal, which I believe became
RSX Pascal, and the generated code improved beyond all recognition.


Brian.


_______________________________________________
Simh mailing list
Simh at trailing-edge.com
http://mailman.trailing-edge.com/mailman/listinfo/simh