Stephen,
Stephen Kellett wrote:
> In message <HyM8b.21296$dk4.662667@[EMAIL PROTECTED]
>, Randall R
> Schulz <rrschulz@[EMAIL PROTECTED]
> writes
>
>> I did learn PDP-11 assembly and wrote programs in it for one of
>> my undergraduate courses (among other things, this was the course
>> in which the concept of recursive subroutines were introduced in
>> the CS curriculum of that institution in that era). The
>> instruction set is so elegant and the opcode formats so regular
>> and predictable, that for years I could use the front panel
>> switches (!) to key in an interrupt driven routine to echo
>> keyboard input back to the TTY / CRT.
>
>
> You know, when I was a lad, I had to live in a hole in the road,
> and take a bath in the puddles when it rained. These youngsters
> today, they don't know how easy they have it!
Well, switches notwithstanding, MACRO-11 (the DEC-supplied assembler
for the PDP-11) was an amazing assembler (when that's one of your
primary programming tools, you make it good). I recall one assignment
where you could, given sufficient motivation and cleverness, get the
assembler to do all the computations needed and there was no need to
run any program--in fact, no instructions were emitted. I can't
remember what the program was supposed to do, unfortunately (I am,
as they say, older than dirt).
> Seriously Randall, thats a fabulous story you have there. I've
> never programmed anything via switches on the front. I think I
> should be relieved to say that. Either your college was
> underfunded, or you are older than I?
The University of Wisconsin certainly wasn't underfunded, but this was
the late 70s / early 80s. We had a PDP-11/20, a /40, a /45, a /34 and
eventually a VAX-11/780 and later a few /750s for distributed systems
research. The CS department also ran DEC's timesharing system (RSTS-E)
on a PDP-11/70. Back then, there was also a dual-processor Univac
mainframe and more than a few keypunches. I was back there a couple of
years ago, and the Univac machine room is now the campus computer
store and--are you ready for this?--there are no keypunches anywhere!
We used card input for our assembly programs on the 11/20, too, though
it did have a very low-res CRT and a keyboard. I guess there must have
been a line printer, too--maybe it spooled output through the Unix
system on the /45. The /20 had one of those head-per-track disk
drives, the /45 had two SMD drives (the ones that look like small
dishwashers or large trash compactors and shake like hell when the
system is driving them hard) for a grand total of 120 MB. Woo-Hoo!
I'm most proud of the time I fixed one of the Unibus serial cards (a
KL-11, maybe?) using a logic probe, a logic injector and the
schematics for the card. It was a bad 7400 series TTL chip (a hex
inverter, perhaps?).
Yeah, I guess we've come a long way. I guess... I miss all the
fla****ng lights. And nowadays you can't really fix much the way you
could back then. The PC board mounting technologies make it pretty
much impossible to replace components that aren't socketed.
If you want to see what crazy hardware is, check out DECtape:
<http://www.linuxguruz.com/foldoc/foldoc.php?query=DecTape>.
(We tried the swapping to DECtape hack they mention, by the way.)
Here's a page with pictures: <http://www.pdp8.net/tu56/tu56.shtml>.
> Stephen
Well, enough of this. The past is gone, may it rest in peace.
Randall
|
