Experimental Physics and
Industrial Control System

[email protected] (Gordon Uchenick) · Thu, 31 Aug 95 16:47:03 EDT

This is a subject that is guaranteed to start heated debate, if not
fist fights (at least on the East Coast). First of all, initialization
of outputs may be hardware specific because some types of I/O are affected
by a reset and other types are not. Not all types of I/O will allow you to
read back a value. The "suitable" thing to do may also depend on how long
the IOC has been down (a few seconds? a few minutes? a week? How do you
know for sure?).

IMHO, it make sense for a database designer to be able to specify whether
or not he wants a certain value to be output on restart. What do you do
when he hasn't? The general practice that I've seen is that, during
initialization, an output block should write ABSOLUTELY NOTHING to the
hardware unless it has been commanded by either an operator or an upstream 
block. Aside from safety's sake, we must realize that we are just restarting 
an IOC. We are not necessarily restarting the process!

Gordon

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System