On Mon, 22 Jul 2013 21:25:23 -0400, Robert A. Rosenberg wrote:
>
>This is an issue since the design of RESTORE is broken due to its
>poor handling of SUPS/PREs If I can Apply a SYSMOD I should be able
>to restore it without needing to also restore any SYSMODs it PREs.
>IOW: If it PREs SYSMOD1 which has Elements EL1 and EL2 and contains
>only a new EL1, then a restore should be able to just select EL1 from
>SYSMOD1 and not need to also restore SYSMOD1 (along with its PREs
>which then need to be Applied again).
>
Several releases ago, at OS/390 V2R5.0 SMP/E, SMP/E introduced the
"Improved load module build processing", which for building _new_
load modules during _APPLY_ processing supported fetching ++MOD
elements from the GLOBAL zone.  Alas, it does not support such
selection when updating a load module as in RESTORE, nor selecting
other element types such as ++MAC from the GLOBAL zone.  It's a
pity that SMP/E did not follow through to provide such function as
you (and I) want for greater flexibility of RESTORE.  If that were done,
ACCEPT would become a needless function.

IBM has stated that the design objective of RESTORE is to reset
objects to an ACCEPTed level; as such it performs its function
admirably and needs no enhancement.  You and I feel differently
from IBM here.

The corresponding VM facilities, VMFMERGE and VMSES/E have
no requirement for a function analogous to ACCEPT.  Accordingly,
they provide the flexibility we'd like to see in SMP/E RESTORE.

-- gil

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

Reply via email to