---------------------------<snip>------------------------
Unlike most new threads that end up in ancient history and then have to
be killed, this one starts out with ancient history! What an opportunity
for us Jurassic-types. Since I was non-email-capable for two weeks, I
read all old posts before adding my 2 cents' worth, which embody
responses to many previous posts.
Don't know about MFT, but MVT was reclassified as Class C (meaning
frozen, no more new releases, no more fixes) in November, 1977. I
continued working with it and other OS/360 variants off and on until
late 1983.
PCP did indeed stand for Primary Control Program.
I used BPS, TOS, and DOS from mid-1966 to late 1971. The worst thing
that the tape-resident-SYSRES TOS had to do was to recover from an I/O
error on the SYSRES tape itself - backspace or forwardspace to fetch the
system module to do the recovery, whose logic said to retry the failing
I/O, so move the tape back to where the error was, re-read ten times,
then, if it still failed, move the tape way back to the beginning again
to locate the system module to do the ABEND process. Truly heinous and
egregious.
BPS had a 2-pass 8K card assembler available for hard-core warriors. You
put the standard BPS self-loading IPL deck (all of 6 cards) in the card
reader followed by the 8K card assembler phase 1 object deck followed by
your source deck and IPLed from the card reader. The first phase punched
one output card, containing intermediate Assembly data, for each input
source card. Then you put the 6-card BPS IPL deck in the card reader
followed by the 8K card assembler phase 2 object deck followed by all
the cards punched out in phase 1 and IPLed from the card reader again.
This second phase punched the final object deck. In order to run the
program thus assembled, you again put the 6-card self-loading IPL deck
in the card reader followed by this object deck and reIPLed from the
card reader. So you had an "operating system" with major limitations:
(1) only one program could run at a time, (2) you had to re-IPL whenever
you wanted to run a different program, and (3) the operator performed
the operating system's functions of running one job after another.
The first version of BPS did not support multiplexing on the multiplexor
channel. Regardless of how cleverly you tried to overlap I/O, when you
did the EXCP the supervisor would do a SIO to start the I/O and then a
TIO loop until the I/O completed before returning control to the
instruction just after the EXCP's SVC. More heinosity and egregiousness.
MVT was first virtualized in early 1974 as OS/VS2 Release 1, better
known as SVS (Single Virtual Storage). A fuller version, OS/VS2 Release
2, was available a year or so later, and it was quickly renamed MVS for
Multiple Virtual Storages. MFT evolved into VS1.
I heard about COS, for Compatibility Operating System, but I'm not sure
what was made compatible with what (maybe it was a 360/20 emulator
running on a 360/30?).
---------------------<unsnip>-------------------------
We should get together sometime, Bill. I'm in Woodridge.
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
