I
joined the S/3
software effort early in 1968. At that point my group was developing
the SCP and RPG-II Compiler for the card version
and another group was doing the same for the disk version. I
was programmer number 19 on the entire effort.
We
had an S/3 assembler on the S/360 Mod 50. We had to take our
source programs down stairs and turn them in at a window…lucky
to get two runs a day. We did the source on IBM 029 key
punches.
We had engineering level hardware in model shops throughout the
development lab…no emulators. The hardware was quite unsafe,
no
covers and jumpers and wires all over the place. I would say that we
were debugging the hardware and software at the same time.
We thought we had it made when we got IBM 2260 terminals so we could do
our source on them and submit our jobs without having
to take them to the window. We still had to go and pick up our object
decks from IS.
I
think the S/3 disk operating system was tailored after the S/360 OS-PCP
system. Also the minimum storage on the disk system was 8K.
I am quite sure that was the reason only one // LOAD was permitted per
procedure.
The
work environment we had was the best one I had in my 30 years with IBM.
We were given the hardware and software specifications
for what we
were working on and the schedules. We were a small group and management
stayed out of our way unless we needed them.
I think about 75% of
us were new hires with IBM. Very few had computer science degrees…the
universities didn’t have that degree yet then.
The
RPG-II Card Compiler
took up almost 2 boxes of 96 column cards (almost 4,000) and used one
stacker of the MFCU.
Very few of us dropped the compiler more than once. We didn’t
have a native assembler for the card version.
When the S/3 Mod 6 came to be I wondered how you could ever do anything
without punched cards.
There are so many more great memories from that projects and its
people. I think that was when I learned that it was sometimes
easier to get forgiveness than permission.
Dave;
