Re: Binding Directories

["Phil Hall" <hallp@xxxxxxxx>]

> >For the CRTPGM to make the choice for you, it would
> >need to 'know' more about your code. It's just protecting you from the
very
> >real horrors of binding to the incorrect one.
>
>   I disagree.  The /400 is designed in all other aspects to use the
> *LIBL.  I don't care for this design.  If you really want a different
> order, the CRTPGM supports explicit lib/module if you want to override the
> *LIBL.

True, the *LIBL concept is used widely on the AS/400, but in this case, it
can be more of a headache (my view) than a help. If I have a *MODULE and a
*SRVPGM that both export PROC_A [but the *MODULE one takes a pointer as the
parm, the *SRVPGM one takes a PKD as the parm] in my *LIBL which one should
CRTPGM 'choose' to use for me ? Hitting this problem actually displays the
poor design of the whole BNDDIR concept - that you can't have nested
BNDDIR's, instead we either accept that we'll get binding errors (that can
be overridden, in my view dangerously, with OPTION(*DUPPROC)) or we create
lots of custom BNDDIRs and use them on the BNDDIR parm to CRTPGM to filter
out the duplicates before they get to CRTPGM.

As for the explicit lib/module parameter of the CRTPGM cmd, sadly, CRTPGM
binds all the modules listed in the list, even if they are never referenced,
leading to bloated objects containing orphaned *MODULEs.

--phil

+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
| To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---