Re: CHGOBJAUD of *CMD never gets logged; who did CHGWTR

Dennis,

Actually the help text for CHGCMDDFT discusses the use of the command when
changing the default for IBM provided commands, so it is considered a
supported interface.
The help text for CHGCMD, on the other hand, clearly states that one should
NOT change the CPP, VCP, or POP of an IBM provided command.

Bruce
Bruce Vining Services: Contract Programming
On Wed, Jan 27, 2010 at 11:45 AM, Dennis Lovelady <iseries@xxxxxxxxxxxx>wrote:

To avoid things like having our updated QSYS objects replaced during
PTF
application, there are best practices regarding how we handle
"replacing" an object in QSYS. Some replace an object directly in QSYS
and keep a copy in a "safe place" against those times when the QSYS
object will get updated by IBM. Sometimes, there is no alternative to
this approach.

Thanks, Mike. I was trying to say that best practices don't really include
"replacing" objects in QSYS (and similar) libraries. Where IBM has made it
possible to use exit points (and where those exit points satisfy whatever
need), the best practice is to utilize that method instead.

I also see issues with using VCP as something other than a validity checker
for a variety of reasons. One particularly compelling reason is that the
VCP will be called at times other than when the command is actually being
executed (for example, when entering the command into SEU for example).
Another is that the VCP does not have knowledge of the success or failure
of
the command itself. For example, I may do a RLSJOB 123456/MYUSER/MYJOB.
If
I have attached a VCP to RLSJOB command, the VCP will do its bidding. But
if
that job doesn't exist on the system, then whatever the VCP did should
probably be undone. But we don't have a method of doing that.

Of course there certain modifications to IBM-supplied objects that work to
our advantage. Most notable of these is to change the defaults for some
command to fit your site's purpose. Those I tend to do in QSYS, and never
manually. The changes will be made by a program named NEWRELEASE so that
we
know they can be reliably repeated when necessary.

I understand I can get beat up for saying that. After all, IBM doesn't say
that's OK any more than they say it's OK to add a VCP to their code. My
answer (to come full circle before exiting) is that IBM provides an
approved
method for accomplishing the one, but no method of changing default other
than direct object alteration.

Dennis Lovelady
http://www.linkedin.com/in/dennislovelady
--
Computers are not intelligent. They only think they are.

--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.