John writes:

> OK, I went ahead and tested it for everyone's edification. The result was
>  somewhat surprising but, upon reflection, kudos to the architect
>  responsible.
>
>  It appears that if the spooler runs out of available numbers below MAX, it
>  just keeps going with the next available number above MAX. IMO, the best
>  outcome, though it would be nice if there was a console warning message.
>  However, it is certainly better than having processes hang or blow up or,
>  even worse, a system abort. So, again, kudos to the architect responsible
>  for the design.

While I share John's enthusiasm for the correctness of the design decision, I
don't share his surprise. Rather, it seems to me that the design choice taken
by the author in this case is the minimum that you should expect -- if you're
designing equipment and processes to be used by ordinary people. The resort
to a process-terminating error message should always be the very last option,
especially if you have alternatives to keeping the process and alive and
working, as you do in this instance.

Resiliency, robustness and reliability are watchwords that should stay
centered squarely in the front of your mind while you're building constructs
to be used by humans.

Wirt Atmar

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *