I haven't encountered a situation where I wanted to run two invocations of a 
utility concurrently, and I have no problem with using a ddname list should I 
ever need concurrent utilities. Changing Allocation to support multiple TIOTs 
within a jobstep would be a massive undertaking, and I can't conceive of a 
plausible business case.

SVC screening seems like overkill; I would expect a subsystem to be far less 
fragile.


--
Shmuel (Seymour J.) Metz
http://mason.gmu.edu/~smetz3


________________________________________
From: IBM Mainframe Discussion List <[email protected]> on behalf of 
Paul Gilmartin <[email protected]>
Sent: Wednesday, October 23, 2019 2:54 PM
To: [email protected]
Subject: Re: Is there a TSO equivalent to JCL DDNAME=?

On Wed, 23 Oct 2019 17:34:37 +0000, Seymour J Metz wrote:

>I agree that IBM should provide a means to dynamically assign an alias to a 
>ddname, but IMHO ATTACH is the wrong place to do it. DYNALLOC would seems to 
>be the obvious place to do it.
>
An imporrtant use of alternate DDNAMEs is avoiding DDNAME conflicts:
concurrently running two utilities both coded to use (e.g.) SYSUT1,
but with different data sets allocated, perhaps to generated names.
I don't see how this can be accomplished with DYNALLOC aliasing.  I'd
welcome an enlightening example.

The value is greater if the "aliasing" is invisible to the ATTACHed/LINKed
subprogram.

>In the meantime, it is a SMOP to write a subsystem to redirect the I/O. I 
>don't recall the file, but on the CBT there is a subsystem to allow utilities 
>that do not support Wylbur compressed data sets to access them transparently.
>
Long ago someone posted here that he had accomplished (but not entirely)
DDNAME aliasing with SVC screening.

>DDNAME= is implemented in the C/I; there should be no need for Initiator 
>involvement.
>
I'll stand corrected.  In any case, before the job step program gets control.

________________________________________
From: Paul Gilmartin
Sent: Wednesday, October 23, 2019 11:20 AM
    ...
It's regrettable that alternate DDNAME requires each program to
code for it rather than being a part of ATTACH and transparent to
the ATTACHed program.

-- gil

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

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

Reply via email to