[Prev][Next][Index][Thread]

Re: Dylan runtime checking vs C++



In article <953171385.368124@clint.waikato.ac.nz>, "Matthew Luckie"
<kluckie@ihug.co.nz> wrote:

> 0657.313
> Mark Utting
> 
> if you were there 84 he might have been studing with you ->
> 
> "In the early 1980's, I completed a BSc in Pure Mathematics and Computing
> and an MSc in Computing (thesis was on functional languages) here at Waikato
> University"

Yep.  Mark was a year ahead of me.  '84 was his 2nd year of MSc and my 1st
(I got a job and left without finishing).

Mark and his flatmate, Paul Brebner, designed and built (wire-wrapped) a
6809-based computer and I helped them with a few things -- I recall that I
designed a rather tricky clock circuit that could switch glitchlessly on a
cycle by cycle basis between 1 MHz, 2 MHz and a 1.33 MHz hybrid (500 nS
on, 1000 nS off) based on the address being accessed, because most parts
worked at 2 MHz but some components (ROM and I don't remember what else)
wouldn't.  I also recall helping out with the VAX-based BCPL to O-Code
compiler and O-Code to 6809 compiler, and no doubt other stuff I don't
remember any more.  I Mark did the PROLOG port by himself.  Paul and I
were the tutors in charge of the hardware/microprocessor lab in the K
block basement, so we had after-hours access to the stuff there :-) 
Against all the odds (and advice of the more experienced) we managed to
debug the thing with no more than a dual-trace oscilliscope and a listing
of the ROM.  Oh, the joys of continuously resetting the machine,
triggering the oscilliscope from the reset, showing the clock on one trace
and moving the other probe from address line to address line and then data
line to data line, figuring out what sequence of addresses it was
accessing and with what data...  As I recall it turned out there were only
two hardware bugs: the reset switch was wired push-to-open instead of
push-to-close, and the sequential-logic clock divider of our own design
(trying to get 9600 bps on the serial port to talk to a VT100) was wired
for divide-by-13 instead of divide-by-12 (or something like that).

Ah, we were real men in those days, and you tell it to the kids today and
they won't believe you... ;-)

Sorry to ramble off-topic.

-- Bruce



References: