29 Jul
2004
29 Jul
'04
7:50 a.m.
On Thu, 2004-07-29 at 04:49, Vintage Computer Festival wrote:
It's come up in some litigation. The actual issue
at hand is that someone
was able to overturn a patent by claiming the IBM 650 had a real-time
clock built in. They "proved" this by submitting as evidence a printout
that had the date printed on it(!) I've checked the IBM 650 Manual of
Operation and it makes no mention whatsoever of a real-time clock. I
pretty much figured it wouldn't but I of course had to do due diligence.
Oh this makes it more interesting! I know that I have a book that
discusses the development of this, in practical terms, it would be
mid/late 1950's. It would take some research to find it.
Certainly there were time-periodic interrupts for isochronous purposes
by 1960, whether the 650 had one or not.
"Keep time when the power is off" -- umm kids, for giant old machines,
turning the power off was like a massive emergency/annual maintenance
event/likely disaster/call-the-boss-up-at-3am, retaining the time would
be looooow on the list of things-to-worry about.
1970's DG RTOS and RDOS, for example, supported time-of-day, maintained
by the RTC option on the CPU board -- which option was some TTL chip and
a crystal. It generated a periodic interrupt. Upon shutdown/reboot (it
was a small mini, no catastrophe :-) you'd enter the date & time.