On 2014-02-21, at 16:23, Neil Duffee wrote:
>
> Another thought... you wouldn't be lucky enough to use the same number of
> levels in your test naming scheme ie. HLQ.LV1.LV2.LV3.LLQ[.DATA|.INDEX]? If
> so, removing "Include Additional Qualifiers" would drop all the .DATA &
> .INDEX entries.
>
By all reason, it should work that way: If I remove "Include Additional
Qualifiers", I'd expect it not to include additional qualifiers. Alas,
it doesn't work that way. The .DATA and .INDEX entries are always listed.
I mentioned this, with overt displeasure, on one of these lists a few
years ago. The consensus of replies was that it was considered better
that way. I still don't understand why.
But I hacked at my UNIX/Rexx/TSO approach:
# ###############################################################
user@HOST: head -99 `whence qwikdel rexx`
==> /u/user/bin/qwikdel <==
#! /bin/sh
# Doc: Experiment with parallel data set delete.
Del1() { # Delete one data set.
read DSN || exit # Quit if no more.
rexx "trace Err; address TSO 'DELETE ''$DSN'''" &
return $?; }
cp "//'*redacted*'" /dev/fd/1 |
sed -n 's/^ *DELETE *\([^ ]*\) .*/\1/p' | # Extract DSNAMES from IDCAMS
JCL
( while :; do
Del1; Del1; Del1; Del1; Del1
# Wait a bit; five dozen TSO address spaces makes the Iceweasel cower
in fear.
#
sleep 1
jobs; done )
==> /u/user/bin/rexx <==
/* REXX Useful! */ interpret arg(1)
user@HOST
# ###############################################################
Works well enough. The chief disadvantage is I don't have time
to go for coffee. I wouldn't inflict it on a customer.
-- gil
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
