In a recent note, "Shmuel Metz (Seymour J.)" said:
> Date: Wed, 10 May 2006 00:23:16 -0300
>
> >JCL tailoring is a nightmare.
>
> Only if you didn't create your JCL with File Tailoring in mind.
>
They (not I) didn't. Which is why I regularized it.
But my base complaint is with the original and persisting
design of JCL, which was created without file tailoring in
mind. Yes, file tailoring existed in those days and should
have been a consideration, even in the rudimentary form of
inserting in or removing from a deck of punched cards one
card containing a single option without the need to add
or remove a comma from the preceding line.
This technique was well enough known to FORTRAN (at least)
programmers prior to the invention of JCL: "Always put
a label on a CONTINUE, not a nontrivial statement, lest
you later need to insert another statement before or after
that statement."
Improvement is still possible:
o Provide explicit assertion of default values for all
keyword options, such as "TYPRUN=NORMAL", or even
simply "TYPRUN=".
o Provide a dummy keyword option (e.g. "SKIP=") to use as
a placeholder. This would serve the same valuable purpose
as FORTRAN's "CONTINUE, or a "." alone on a line at the end
of a list of SMP/E options, or a blank line at the end of a
list of IDCAMS options each of which is followed by "-".
E.g.:
//NAME JOB programmer,account,
// TYPRUN=HOLD, /* This line may be deleted. */
// SKIP=
or, even:
//STEP EXEC PGM=IEFBR14,
// SKIP=
to allow inserting, perhaps, "COND=whatever". (But there
is an explicit default for COND, viz. "COND=(0,GT)".)
Ideally, such a dummy option should be distinguished by
being allowed to appear multiple times on a single statement.
-- gil
--
StorageTek
INFORMATION made POWERFUL
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
