8 Aug
2004
8 Aug
'04
7:27 a.m.
On Wed, 2004-08-04 at 14:50, Paul Koning wrote:
Either would serve. But a software mechanism would be
really unlikely
in practice without something resembling a real OS, with an interrupt
mechanism or some other way to do things periodically (as on the CDC
6000 series).
I agree with you generally, but disagree only in that an "OS" is hardly
needed to take advantages of hardware facilities and side effects (in
fact you're better off without one, they usually get in the way of "fun"
:-)
As far as I can tell, the 650 had no such thing --
your application
was the only thing running. In theory you could keeep track of time
even though your code lives on a drum -- but in reality, how likely is
an application programmer, especially a business data processing
programmer, to want to deal with that kind of esoterica?
Agreed. I think it extremely unlikely that in a 650 an ordinary program
would jump through the hoops necessary to extract a timebase from this
particular machine.
It's possible, however unlikely, that there's a manufacturers test port
that counts drum revolutions or somesuch rot, but in hardware of that
vintage there tends to NOT be unexploited hardware, it was painful and
expensive enough as it is!
Short of a peripheral device in a catalog that provides clock-time or
real-time data, the answer is likely "no", 650's can't tell correct
time
(any more than you could tell it to
WRITE 100 H1,M1,S1
WRITE 110 D1,M1,Y1
100 FORMAT (I2,I2,I2)
110 FORMAT (I2,I2,I4)
I couldn't remember the details of the FORMAT statement, and none of my
books are new enough to have alphanumeric formats in them!)