Re: ACP/TPF

Paul Hinman <paul.hinman@[EMAIL PROTECTED]
> writes:
> I heard all kinds of stories about how fast it was to IPL and bring up
> the applications compared to bringing up MVT/SVS/MVS and then once the
> OS was up one could begin starting the applications.

re:
http://www.garlic.com/~lynn/2007o.html#57
ACP/TPF

story about how fast cp67 could fail, automatically dump, and
reboot/ipl (especially compared to multics)
http://www.multicians.org/thvv/360-67.html

this was one the features that aided in cp67 being used as commercial
time-sharing system. an issue was having the system up and running
24x7 for (commercial) dial-in access. a lot of time off-****ft
operation would have very little useage ... making it difficult to
recover costs thru useage charges. minimizing off-****ft costs
contributed to leaving the system up and running around the clock.
auto-dump and reboot contributed to allowing off****ft system
"lights-out" operation. misc. posts mentioning cp67 (& vm370)
commercial time-sharing operation
http://www.garlic.com/~lynn/subtopic.html#timeshare

besides auto reboot/restart .... cp67 also was using "prepare" CCW i/o
command on terminal/dial-in lines. large numbers of 360s were leased
and montly leased charges were based on processor useage "meter". The
useage "meter" would run whenever the processor was executing and/or
whenever there was active i/o. the "prepare" CCW would suspend
operation on the line (halting the "meter" if nothing else was
running) until bits were moving on line(s). "prepare" CCW allowed
system to be up and available, but the useage meter wouldn't be
running when there wasn't actually anything going on.