It was Oct 1, 2000, 15:25, when Austin L. Denyer keyboarded:
>> Ha! I know tricks like that one too! I had a machine like that, and wrote
>> a simple but functioning accounting system in it :)
>
>One of my colleagues once tried to write a program to calculate the
>performance characteristics of large-bore oil hoses, and ran out of memory
>on a 16k machine. I then wrote the thing myself in ... wait for it ... 450
>BYTES!!!!! Needless to say, it was completely devoid of any eye candy, but
>it worked (well, I wrote it in an evening...)
>
>Out of interest, what machines were you using? How long ago?
Hi Ozz,
I was doing work on an IBM System/3 in those days, as well as on
an old Burrought B-7700. The Burroughs could only be fed through punched
cards (always racing to get the optical reader which was 40 times faster
than the mechanical one! ;-) for the average user, there were no keyboard
terminals attached. Except for the console, which had it's own printer for
system output. On that machine, programs could not be bigger than 64K
(equal to the IBM S/3). It had 1 meg of RAM, which was an indecent amount
already. It was for a technical highschool, had 200 users on it. (Try that
on the average machine these days, with 128Megs hahaha!)
System crashes would be saved on a special 64Meg removable diskpack, which
we would then debug by hand.
On the IBM we had been messing so much that not much of the executable
code had any bearing to its source code. Usually we'd patch the hex code
directly in memory and dump that back to disk. Using an 8 bit switch array
on the machine itself. Such fun!! ;)
Yeah, that's real programming. 450 bytes of code, assembler preferably,
that does what it has to do. FAST!
Paul
--
The Tao that is seen
Is not the true Tao,
Until you bring fresh toner.
http://nlpagan.net - ICQ 147208 - Registered Linux User 174403
-=PINE 4.21 on Linux Mandrake 7.1=-
