Old mini computer

Roger Ivie rivie at ridgenet.net
Tue Sep 25 23:26:58 CDT 2007


On Tue, 25 Sep 2007, Mr Ian Primus wrote:
> I am about 200 miles away. I looked up the area code
> too, and I emailed her - she sent me some pictures of
> the machine. It's definitely an interesting beast. I
> have no experiance with Burroughs machines, but still
> - I'm really tempted.

I have no experience with the B80, but used a B800 for a couple
of years in high school. I think the B80 is a modernized, snazzy
B800.

I was working for an accounting firm in Roosevelt, UT. The B800
had a CPU cabinet, a printing console terminal that had two
sections: a 17" wide section for interactive use and an 8.5"
wide section for interacting with the operating system. To talk to
the OS, you'd press SYS REQUEST and issue commands on the smaller
section of the console.

We had three disk drives; RK05 equivalent. Two removable, one
fixed. A small 9-track tape drive and four TD830 terminals. The
terminals are form-oriented beasts polled by the computer. You
load a form onto the terminal, enter data into the form, and punch
transmit to send the completed form back to the computer. Fields on
the terminal could be left or right justified, and you could put
read-only fields for doing things like identifying the form.

I started out doing strictly data entry, but managed to scare up a
compiler somewhere and starting programming the thing in COBOL.
I'm not a big COBOL fan, but this was a nice COBOL and I enjoyed it.
My previous programming experience was primarily BASIC on TRS-80s
and PETs and hand-assembled 6800 on my Heathkit microprocessor
trainer (no, not that one; the short-lived model with half the
memory in a cardboard box. Mine is serial number 8).

I managed to get a screen-oriented version of the CANDE editor from
a site in Duchesne, but I had to do some fancy jiggery-pokery with
the tape drive to load it in; the fellow who gave it to me had
written it out as a data file instead of an executable, so the
machine refused to read it as an executable. I convinced it to
load the program (from tape) by writing a second tape containing
a file of the same name that was executable, punching SYS REQUEST
after the system had read the label but before it read the file,
then swapping tapes (with the other tape pre-positioned after the
label) while it was waiting for me to type a command.

Screen-oriented CANDE worked by loading a page of text into the
terminal, which you edit locally and send back to the computer. It
would then load the next page into the terminal. During one summer,
I took a FORTRAN programming course over the space of a couple of
weekends at Utah State University in Logan, UT, which had a B6800.
It also used CANDE, but this was more interactive; the terminal the
instructor used was a keyboard wiht a built-in modem and a video
output. After lecturing for a while, the professor would leave and
everyong would head for the card punch room. Since I knew how to
use CANDE, I'd hang back in the classroom and dial in while the
riff-raff were using the card punches.

Our B800 had a chain printer with no lower case letters, which was fairly
annoying. The console was dot matrix and had European characters 
instead of lower case letters. The terminals did lower case just fine.

Booting the machine involved reading a small bootstrap paper tape
(which we kept in the carriage of the console terminal) into a
small paper tape reader on the front of the CPU. This gave it 
enough smarts to load the OS from the disk.

I think I scarfed the COBOL compiler from the same place I swiped
CANDE. They were running, IIRC, a B96 instead of a B800. I had to
play some games to get the compiler to run; it would only run if it
were a specific job number, so once I figured that out I wrote some
do-nothing programs that would just occupy job slots so the compiler
would be happy.

IIRC, the compiler had eight passes. It was very slow on the B800, but
(as I mentioned) was quite nice.

We also had a cassette drive, but I didn't do anything more than fiddle
with it a bit.

The accounting firm had an office in Salt Lake where they had a B80 and
a B1800. I did some data entry on the B1800 once, but that was pretty
much all I did for them in Salt Lake.

We had some manuals for a couple of other programming languages that 
were more system oriented (NCL?), but I never managed to scare up a
compiler for them, so I never used them.

The B80 system in Salt Lake had double-density 8" floppies on it, and
they treated the disks quite gingerly. Rumor was that they had to be
absolutely perfect for the machine to read them. Quite different than
my later experience with 8" SSSD CP/M machines.

This would all be about 1980 to 1981.

The Roosevelt office also had an L5000? machine in the corner that
was never used and a big long paper tape that claimed to be a BASIC
interpreter for it. I tried loading a few times, but it always reported
an error halfway through the tape.

I graduated from high school in 1981 and moved on up to college and
into the DEC sphere of influence (USU at the time was moving away from
the B6800 onto VAX-11/780s). Haven't used a Burroughs since, except once
or twice in college when the VAX was down.

The Burroughs 8" floppies were hard-sectored. The university designed
their own CP/M machine and a couple were built that could read Burroughs
hard-sector disks instead of the normal soft-sector ones. I never used
those particular machines; I gather that only a few were built.
-- 
roger ivie
rivie at ridgenet.net



More information about the cctech mailing list