Wade Curry wrote:
>.. young age of 45</snip>
>> As I recall, the mainframe had 32K of 60-bit ram (core, or
>>..
> These days, mainframers call the RAM, "(main) storage". However,
Yes, I used the modern word. As I remember, it was just called "memory",
or sometimes "core". Now that I think of it, the word length was
probably less than 60 bits, since I remember that 60-bits was a big
bragging point of later/greater ControlData machines. I kinda-remember
that character-data was encoded in 6-bits, though. Can't remember any
details about integers and decimal/floats.
Neither do I remember the word "mainframe" being used in my 1962
exposure. Probably just "computer". I really wasn't an insider (of
computer-center operations), or even a computer-related student (don't
even know what/whether they were teaching -- the comp. center was run by
the (EE?) engineering department. I was just a user. I may have taken a
seminar-type course, read McCracken (?, I think), and helped a couple of
profs reduce some data, and tried a bit of homegrown programming for
classifying elementary particles.
Some years later (~1980), I ran into a reference to "tub memory". That
engendered quite a vision in my imagination. Does anyone know what that
might have referred to? Maybe it's a takeoff on (paging) drum. I never
met a drum device, either, but that must have been something to see the
innards of, itself. (..Carl?)
>..
> I am surprised that it didn't boot from tape. How many cards were
> required for that process. Booting is usually referred to as an
> IPL, now (initial program load). A warm restart of any system, OS
> or otherwise, is referred to as a "bounce".
To tell the truth, I don't recall the terminology -- it may indeed have
been IPL -- or maybe I just never ran into a proper discussion of what
was going on.
The card reader must have been memory-addressed, I suppose. I think the
initial deck was less than a few-hundred cards. There *were* multiple
instructions per card, but it still may not have been enough for a full
OS, so you may be right that there was a final load stage from tape.
Maybe not?
BTW, I have to correct my confusion re card columns in my post. The
cards were, of course, standard 80-column ("Hollerith") cards. For
Fortran, Cols 1-5 were for an (optional) statement number, column 6 was
the "continuation-column", cols 7-72 were for program statement text,
and cols 73-80 were for optional annotation use, typically card-deck
sequence.
Object decks used all(?) 80 columns for machine code. Maybe there was a
sort field, or maybe an explicit memory load address which would have
served the same purpose.
>..
>> entertainment.
> Card Jams... Sounds like an industrial, grunge band. Would they
> play at a KPLUG meeting?
Hey, is that domain name taken?
..j
--
KPLUG-List@kernel-panic.org
http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
