Hi Jeremy,
"...I've no idea if that is or was correct ..."
Have you ever heard of the CSI (Consolidated Software Inventory)?
This/these VSAM Dataset(s) has/have been part of SMP/e since the day it
came out in the '80s.
Regards,
David
On 2023-05-30 18:20, Jeremy Nicoll wrote:
On Tue, 30 May 2023, at 21:41, Seymour J Metz wrote:
SMP/E uses VSAM
I've no idea if that is or was correct
and doesn't have the wonky member names of SMP through SMP4.
You keep saying this ... and you're still wrong. I don't know why you're so
sure. Maybe it's changed since then (sometime between 1985 and 2000
or so, I expect).
I only ever used SMP/E, and wonky member names did exist in one (at least)
of the PDSes it used to store stuff. I seem to remember also that the PDS
I found them in had a vast number of members in it - can't recall just how
many that was though.
Last time we discussed this (around 09 APRIL this year) I guessed which PDS
it might have been - and was probably wrong. Someone else said they
thought it was perhaps the SMPSCDS.
I've just run a search of my PC copies of some old mainframe notes, JCL
etc, looking for "SMPSCDS" and I found a reference to this exact issue
inside the transcript of a discussion I had with IBM support (from, I think,
the year 2000), when we were using os390 2.6 ... though that's not the
os in which I first noticed the weird member names).
The discussion was on the wider issue of poor documentation of
just exactly which characters were valid in datasetnames. For
example I'd found it was psossible to create dsnames with dashes
/ minus signs / hyphens in their qualifiers.
IBM said it was impossible. It wasn't - I'd successfully created
MYHLQ.TEST--
via ispf option 3.2 and also
MYHLQ.TEST-
in JCL, without having single quotes around those dsnames.
Another character one could get into a qualifier was the left curly
bracket (of course those are in some SVC member names) and
perhaps the code that allowed those in ispf accidentally allowed
them in qualifiers as well. I can't remember.
The conversation then moved on to implications of not being
able to SMS-manage datasets whose names didn't meet the
tighter rules that SMS used.
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
