Allan said:
"The major drawback is that each SYMBOLICRELATE alias must be updated
(manually) whenever the
underlying object changes."
I don't see why. In fact, that is exactly the reason we DO use them.
Here's a 3.4 of our TMS.PROD datasets, for example:
TMS.PROD.CTAPMENU *ALIAS
TMS.PROD.CTAPOPTN *ALIAS
TMS.PROD.CTAPPENU *ALIAS
TMS.PROD.CTAPSAMP *ALIAS
TMS.PROD.CTAPTENU *ALIAS
TMS.PROD.PPOPTION *ALIAS
Here's a listcat of TMS.PROD.CTAPMENU:
ALIAS --------- TMS.PROD.CTAPMENU
IN-CAT --- UCAT.VSYSXX1
HISTORY
RELEASE----------------2
ASSOCIATIONS,
SYMBOLIC-TMS.&TMSLEVL..CTAPMENU
RESOLVED-TMS.V126SP00.CTAPMENU
Here's my symdef for TMSLEVL:
SYMDEF(&TMSLEVL.='V126SP00') /* Level of TMS */
Here are my libdefs for TMS in ISPF:
"LIBDEF ISPMLIB DATASET ID('TMS.PROD.CTAPMENU')"
"LIBDEF ISPPLIB DATASET ID('TMS.PROD.CTAPPENU')"
"LIBDEF ISPTLIB DATASET ID('TMS.PROD.CTAPTENU')"
And here's my definition for
DEFINE ALIAS (NAME(TMS.PROD.CTAPOPTN) -
SYMBOLICRELATE(TMS.&TMSLEVL..CTAPOPTN))
I ran this a year or so ago when the low level qualifiers changed for CA1, but
other than that, these aliases never change. Perhaps you're thinking of a plain
old relate.
MA
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
