On Sun, 2004-08-08 at 10:51, Tore S Bekkedal wrote:
> Reading Scott Stevens' post regarding the FORTRAN decks, a bell suddenly
> rang inside my wee head.
> http://www.columbia.edu/acis/history/1965.html
> The 7090 had an accurate, mechanical TOD clock.
Cool find! Here's the cogent text from that page, though this vintage is
too old for me to have worked on, it brings back memories of weird
interfaces:
>From http://www.columbia.edu/acis/history/1965.html
"... The 407 [peripheral to 709x] also had the closest thing ... to a
time-of-day clock. It was a mechanical clock built into the 407 and
counted 100ths of a minute. The 709x would interrogate the clock at the
start and end of each job. Interrogating it was kind of interesting,
because it involved printing the clock value on the printer, terminating
the cycle before it actually made a mark on the paper, and then using
the "echo checking" feature of the printer to fetch the actual value.
Echo checking was a way for the program to make sure that the mechanical
printer had actually printed the correct digit on the paper. The side
effect of this is that if the program running interrogated the clock
(say, with a call from Fortran, or if the operating system did it
between steps of a job), the printer made a funny noise -- like
printing, but not quite."
Not unlike reading multiplexed seven-segment output from a watch chip,
no?!
Received on Mon Aug 09 2004 - 14:00:40 BST