[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Burrough's Algol Compiler

As I recall, there were only 2 Burroughs machines in academic settings, UD
(I think the computer center director, John Falcone?, used to work for Burroughs,
and of course the CS department loved a machine that actually used
(Extended) ALGOL) and the other was somewhere in California (or maybe
Australia?).  [The EE department was none too thrilled by a machine that
didn't do FORTRAN as well, not to mention how the computer center
was run, like counting as uptime the time when the CPU was up but the
I/O Processor was down, or giving resources to Data Processing instead of
academic work when resources were limited -- that sharing of resources
between DP & academic work was a big item of general philosophy in
the EE department.]

They had quite a few security problems because they hadn't been tested much
in such an environment.  For example: The compiler checked the user's privileges
and, if the user wasn't allowed to do privileged operations, it wouldn't set the
executable bit for the file.  (As opposed to checking the user's privileges at
run-time.)  But it still gave you the object code.  They also trusted the executable
bit that came off of a magtape.  So Greg Hostler wrote a program to do things
like create accounts, compiled it, put it onto magtape, took it to a certain PDP-11
running an operating system that had very flexible magtape handling,
copied the tape, setting the executable bit, copied the program back to disk
on the B-6700, and *viola!*.  When they disabled his account, he borrowed someone
else's and reenabled it.  I guess Burroughs tried hard to hire him, but he went to work
for DEC.  (I had heard that he left DEC after his fiance dumped him, but I don't
know what happened to him after that.)  He also ported an Extended ALGOL
compiler to the PDP-11.  (He wrote an interpreter for the stack machine code,
I think.) DECUS had it.

And speaking of power problems, when the B-7700 was installed, the diagnostic
processor was plugged into a mis-wired socket and some boards became fuses,
but not before some of the current got back into something else, the I/O processor
probably.  Dan said to not laugh, he'd seen DEC boards do the same thing.
It was amazing to see the etch literally blown off of a PC card.

- Aron

>From: "Gary L. Luckenbaugh" <garyll@ibm.net>
>To: <deltoids@mcws.net>
>Subject: Burrough's Algol Compiler
>Date: Sun, 28 Jun 1998 16:43:57 -0400
>X-MSMail-Priority: Normal
>X-Mailer: Microsoft Outlook Express 4.72.2106.4
>X-Mimeole: Produced By Microsoft MimeOLE V4.72.2106.4
>Sender: owner-deltoids@mcws.net
>Reply-To: deltoids@mcws.net
>Clark Baker writes in his memoirs, "George and I did a reasonable amount of
>hacking in ALGOL on the B5500 and later on the B6700. The systems
>programming on these computers was done in a dialect of ALGOL and we wanted
>to learn all about that. The computer center folks weren't too happy with
>our interest. "
>By a strange quirk of fate, I ran into the Product Manager for the old
>Burrough's Algol compiler a couple of months ago.  I was at a Lockheed
>Martin meeting in Valley Forge, and had dinner with a guy who was speaking
>at the same meeting as me.  When I asked him about his "heritage" (a
>Lockheed Martin term for, "Where did you work before Lockheed Martin
>acquired you?") he told me he had started at Burroughs, merged with Univac,
>was sold to Loral, and merged with Lockheed Martin.  When I asked him what
>he did at Burrough's he said he was the Product Manager of the B5500/B6700
>Algol compiler years ago.  He was very familiar with the UofD and said they
>were his best customer.