[TriLUG] More computer trivia... ( I'm an older fogy than you).

Rick DeNatale rick.denatale at gmail.com
Tue Dec 13 18:54:36 EST 2005


On 12/13/05, Wing D Lizard <wingedlizard at nc.rr.com> wrote:
> One of my first paid jobs was converting some surveying
> trig algorithms from hp 25 rpn to a trs80 ( basic -- level 2!!).

Well, Mr. Woodbury probably has us all beat, but...

My first computer experience was sometime late in the 1950s when I
visited my Dad who worked for IBM down at 590 Madison Avenue on
Manhattan.  I got a tour of the computer room. I can't remember what
the machine was, probably either a 7000 or 1400 series, but I got to
mount a magnetic tape.

Other early "computer" experiences were an attempt to build a computer
which would compose music for a science fair project in the early
'60s. This was using an educational computer which used rotary
switches made up of masonite disks on a masonite panel, with nuts and
bolts and wires, to make up a logic circuit.  I also remember another
plastic educational "computer" which consisted of a few mechanical
flip-flops which you programmed by putting little short plastic soda
straw pieces on pegs so that they interacted with metal rods the
"clock" cycled. The clock was a plastic slide which you moved back and
forth by hand.  This allowed you to make sophisticated things like a
counter with a small number of bits.

My real computer experience started when I went to college (UConn) in
1970. At the time, I was a big electronic music fan due to a program
in my high school, and majored in EE with the intent of becoming the
next Robert Moog.  All freshmen engineers at the time had to take an
introductory Civil Engineering class which was half a semester of
engineering graphics (drafting with real T-squares, triangles, and
french curves) and half a semester of Fortran II programming on an IBM
1620.  That got me hooked, and I started being one of the guys who
haunted the university computer center.

I somehow arranged to get an account on the IBM/360 Mod 65 and started
learning every programming language I could. I learned PL/I by myself,
and then in a class which used PL/C. On the 1620 some of us started to
design a new language called SCRUBOL (which IIRC stood for
Scientifically Compatible, Relatively Unusual, Basic Operating
Language) the details of which are lost to the mists of time.  I also
picked up Lisp 1.5, Snobol, Formac, APL/360 and probably others I've
since forgotten.  Most of the interaction with the computer center's
machines was batch jobs submitted on punched cards, but the S/360 did
run an early timesharing extension to OS/360 MVT called CPS (for
Conversational Programming System) which used IBM 2741 and IBM 2260
terminals.  For the youngsters who wouldn't know and the senile who've
forgotten, the 2741 was IBM's equivalent to a teletype but using a
selectric typewriter mechanism and talking EBCDIC instead of ASCII,
and the 2260 was an early CRT terminal which used delay line memory in
the central terminal controller.

UConn was one of the first universities with an accredited
undergraduate CompSci degree due to the efforts of Taylor Booth, who
was the CompSci chairman there at the time and very active in CompSci
accreditation committees. Most of the comp sci courses used DEC
machines owned by the EE department. So I learned PDP-5 and PDP-8
assembler language, and played around a bit with the PDP-11.  I also
got to play around with Algol-60 a bit with a friend who worked on a
Data General Nova for the Psych department in a research assistant
position.

Of course, having become a language lover, I took a compiler course,
but this was back in the days before yacc and lex (bison and flex to
the gnus) or even the Dragon book. We used the text Compiler
Construction by David Gries which was the state of the art textbook at
the time. We used a compiler-compiler called Jossle which had been the
PhD dissertation of the professor John White, who later became the
manager of Xerox Parc and former President and now CEO of the ACM. 
The LR and LALR parsing algorithms were beyond the state of the art
back then. But when I got to IBM one of my garage projects was to get
Wilf Lalonde's LALR parser generator working on VM/CMS.  I finally met
Wilf some years later when I became heavily involved in the Smalltalk
and OOPSLA communities.

During my college days, and later when I went to work for IBM in 1974,
there was a perceived and real gulf between the IBM and non-IBM
computing cultures.  Looking back on it now there were really as many
similarities than differences.  The same things were happening with
somewhat different technology.  VM/370 and CMS provided the
infrastructure within IBM for the same uses of Unix elsewhere.  We had
our own email system, news groups (we called them fora) etc.  The fora
were wide ranging albeit more focussed on technical subjects than
usenet ( I can't recall the equivalent of alt.pyrotechnics) and
blossomed when the IBM PC was introduced.  Gradually the gates to the
outside opened up. At first you needed special approval to get an
external internet email address which was via a gateway to the VM mail
system.

I particularly remember one internal forum which started when Byte
magazine published a special issue on the history of computers, this
led to a new forum called byteyymm where yymm was the year and month
of that issue which contained all kinds of reminiscences like the ones
on this and related threads.  Of course human nature being what it is,
such reminiscences tend to crop up in all kinds of discussions. We
used to have a saying that all fora ultimately turned into byteyymm. I
guess that this rule applies to the trilug mailing list as well! <G>

--
Rick DeNatale

Visit the Project Mercury Wiki Site
http://www.mercuryspacecraft.com/



More information about the TriLUG mailing list