Re: Saving to a save file

<snip>
Hmm. Not sure whether Rob was talking about modifying the call in the CL
program, REXX script, or whatever, or actually modifying the command
defaults, but "CLEAR(*REPLACE)" would be potentially useful in a CL
program that updates distribution save files, except that it's just as
easy to have such a program do a CLRSAVF first (which is what we do).
</snip>

Spot on! I am NOT talking about changing command defaults but used in
just the right application. And you're right, it could have just as
easily been done with CLRSAVF first, and not one person would have raised
an eyebrow over "arbitrarily clearing a savf".
The reason I don't? I am running this HA vendors script in "step mode".
Adding CLRSAVF would stop the operator during the switch for yet another
review of the screen to be sure they are processing it right and hitting
F23 to approve it.

I really do not see the big hairy panic in using CLEAR(*REPLACE) in just
the right place.

For example, recent versions of IBM i added the CONFIRM parameter to
PWRDWNSYS. So, would having a weekly batch job with a PWRDWNSYS ...
CONFIRM(*NO) in it be a big cause for panic? "OMG! You're IPLing the
system without stopping the user and asking them 'are you sure'?" It's
not like one is suggesting changing the QIBM_PWRDWNSYS_CONFIRM environment
variable. Or changing the command default on PWRDWNSYS.
Good grief!

And I remember the days when the operator thought answering a message with
C stood for Continue.

Rob Berendt