Re: Mandatory Job Log

On 06-Jan-2012 11:48 , rob@xxxxxxxxx wrote:

One thing nice about command exit points is that IBM could replace
the command with a new release and you don't have to remember to
CHGCMDDFT. That, and a default is only a default. You can set up the
exit point to ignore what they type in for LOG and use what you want
them to use.

I agree. My main intent and primary response was to inform of the difference between the LOG() as "[produce] Job log [or not]" and the LOG "[establish] Logging level [for filtering and level of details for messages placed in a joblog that is written]".

FWiW: I had no intention to suggest the changing of the command default with the secondary portion of my response. Obviously, a new default value does not make the LOG(*LIST) action "mandatory" when there remains a choice other than the default :-) allowing a user to explicitly SIGNOFF LOG(*NOLIST) thus overriding that new default. I was offering that _if_ using changed defaults, via CHGCMDDFT as others had already suggested, then consideration should be given for the often forgotten system change management associated with its use; something which nobody who had mentioned CHGCMDDFT had warned.

FWiW: In some old release(s) there had been many customers who used customized copies of SIGNOFF with changed defaults. The *CMD suddenly stopped working [causing *FC on job termination IIRC] when the new release was installed. That transpired because they had failed to implement proper change management, to delete the old duplicate and then making customization to the newly duplicated object. Those with the errors typically either had no system change management procedure to replicate the customization or had ignored "already exists" on their CRTDUPOBJ to implement the change and thus had merely re-applied the same customization to the existing down-level object which was no longer compatible with the newer level of the OS; i.e. the CPP for SIGNOFF, the *PGM object QWCCLFEC, had changed in a manner that was no longer compatible with the previous release SIGNOFF command. The copy of the *CMD object in the QSYS was replaced by the OS upgrade, but the customers experiencing problems had failed to similarly effect the same "upgrade" to their custom copies.

Regards, Chuck