On Thu, 20 Jun 2013 21:24:05 +0000, Pew, Curtis G wrote:
>
>
>For any interested, the problem was that the CLIST I use to allocate ISPF
>libraries had a non-existent data set in the SYSEXEC allocation, so it was
>failing. This was on a test LPAR, not production.
>
In Rexx, not CLIST, I use BPXWDYN to allocate individual data sets for
each of my library concatenations, using RTDDN() for each catenand
except the first, and ignoring any failures. Then I use BPXWDYN( 'concat ...')
to concatenate the successes. Sporadically, I examine messages and resolve
the failures.
It works to include UNIX directories in SYSEXEC, except as the first catenand,
although IBM tells me it's not supported.
It's a PIT{A|Y} that CONCAT requires at least two operands -- I must count
and bypass when there's only one. That effort should be done by CONCAT,
not abdicated to the caller. Bad boundary condition design.
-- gil
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
