On 2016-12-01 12:05, Farley, Peter x23353 wrote:
> Oh, and I forgot to say: For a DD * with symbol substitution:
>
> SYMBOLS=({JCLONLY|EXECSYS|CNVTSYS} [,logging-DDname])
>
> In anything *except* the PARMDD DD statement the " logging-DDname"
> subparameter will let you point to a logging DD (usually SYSOUT=*).
>
> In the JCL Reference under DD Statement / SYMBOLS parameter, I find the
> restrictions:
>
> Logging is not performed in the following cases:
>
> if logging-DDname is specified on the DD statement which describes an
> in-stream data set that is the target on the PARMDD keyword. See PARMDD
> parameter.
> if logging-DDname is specified on the SYSTSIN DD statement which describes
> input data for the TMP (Terminal Monitoring Program).
> if in-stream data set is opened by one task and then read by a different task.
>
I suspect the first two restrictions stated are consequences of the third.
> if data set specified by logging-DDname cannot be successfully opened.
>
That's almost a truism. I suppose it's likewise true if the sysin itself
can't be opened.
Bummer, bummer, bummer! So the solution (I tried it) is to preface a job
step that copies the substitutable sysin to a passed data set and use
that by referback as PARMDD.
I would hope (but hardly expect) that IBM would have the courtesy to
its customers to do that operation implicitly, transparently, behind
the scenes and spare customers the nuisance.
(RCF submitted on PARMDD, its homomorphism to PARM, 100-byte limit, and
APF entanglement.)
Thanks,
gil
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
