On Tue, 25 Sep 2012 17:24:32 -0500, Art Gutowski wrote:

>On Mon, 24 Sep 2012 09:18:09 -0400, Kurt Quackenbush wrote:
>>Presumably the MCS for each SYSMOD is in a unique data set?  If so, then
>>yes, the order in which the archives for those data sets appear in the
>>GIMPAF.XML file determines the order in which they will be processed
>>during the RECEIVE.
>
>Why does it matter whether each is in a unique dataset?  If FMIDs are selected 
>from an alphabetical list and dropped into a common MCS during build, wouldn't 
>this problem still manifest?  
>
I would assume so.

>With IBM products its usually Hccnnnn for base and Jccnnnn for dependent, so 
>GIMPAF order will be alpha order.  Apparently not everyone follows that 
>convention.
>
We don't.  And I even overlooked a note to myself telling me to
be careful of that.  Are there second-level dependent FMIDs?
Kccnnnn?

>Why does order matter?  Is it because the FMID list in the GZONE entry is not 
>updated until RECEIVE processing for that SYSMOD is complete, only after which 
>it moves on to the next?  Just curious and trying to contrast with ASSIGN 
>statements in RECEIVE processing where order is unimportant.
>
Perhaps even more relevant is the ordering among PREs and SUPs of PTFs,
where order of SYSMODs in SMPPTFIN under a single APPLY command doesn't
matter.

Thanks,
gil

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to