cook@pinocchio.encore.com (Dale C. Cook) (07/17/89)
I've been enjoying the thread on "my first real machine" and thought
of a few more points to throw in. Hit 'n' now if you couldn't care
less about such things.
1) The profusion of no-ops mentioned *were* actually put to use by
we in the diagnostic department. By careful selection one could
have just the right delay in your no-op. :-) BTW, the whole time
I worked with 10's, the overblown instruction set was always under
debate. It was mostly defended on the grounds of it's appealing
symmetry. :-)
2) The 20 was a pioneer in funky (and mostly useless instructions).
My own favorite was the JFFO (Jump on First one FOund). We *always*
tried to work a few into each diagnostic as a matter of pride and
to keep zelots from removing ("cleaning up") the instruction set.
3) When I joined the diagnostic group there was already a great body
of support code. For example, we used an I/O macro package that
absolutely made Macro-10 as easy to use as any high level language.
(It used to be said that Macro-10 *was* the highest level language. :-)
All of the function calls were made through UUO (User Unimplemented
Opcodes) traps, as someone mentioned.
4) TOPS-20 broke some new ground of it's own. It used a program
called EXEC as an interface between the user and the O/S (called
The Monitor in those days.) At first, we 10 hackers thought it
was positively bizarre that every user on the system had to run
a copy of EXEC. Why not just put the damn commands into the O/S
where they belong, we said. Then someone pointed out that we
could "roll our own" commands.... Anyone for a shell?
5) I have to agree with the poster who commented that the 10's and
20's may not have been the best machine ever made, but damn they
were fun to program!
Anyone for a thread on SNOBOL?
- Dale (N1US) Encore Computer Corporation, Marlborough, Mass.
INTERNET: cook@encore.com
UUCP: {buita || talcott || husc6 || bellcore} !encore!cook