Ulrich Krueger wrote:
You know, guys, as much as I'd like to see the JCL PARM expanded in length,
I'd hate to see the problems that will arise from it. Face it, how many of
your programs that process PARM values have a hard-coded 100-byte storage
area to receive the PARM string into? Just about all of them, I bet.
I move to leave PARM alone. If you were to mess with it, you'd probably get
all sorts of incompatibility issues, just like we had with Y2K dates.
How about something new instead: the "SET" JCL -statement is already
available.
e.g. //S1 SET P1=ABC,P2=123456 etc.
IIRC, you can have as many SET statements in your JCL as you need and you
can name each variable to just about anything you want (with a few
restrictions, I believe).
(And I'm just not sure about the length of each variable at the moment.)
So, all we now need would be a supported method to get at each SET variable.
How about something like a Language Environment callable service. You tell
it the variable name to retrieve (e.g., P1) and it returns the variable's
length and starting storage address (or a non-zero return code if not found
or length zero)?
With this information, your program can now process all the parameters it
needs to know about without having to be afraid of exceeding the PARM length
limitation.
And there'd be no impact on traditional PARM processing, either.
Regards,
Ulrich Krueger
-----Original Message-----
From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On Behalf
Of Rick Fochtman
Sent: Friday, January 04, 2008 10:28
To: [email protected]
Subject: Re: JCL parms
I'd be overjoyed if IBM would admit to the possibility of a parm string
up to 32,767 bytes, and design parm-driven software with that in mind.
And allow JCL to accept a parm of that length. Changes to code would be
minimal, since it already uses a halfword length. Changes to JCL
processing might be rather major, so I don't expect this to be done
overnight, as long as it's kep in mind.
Rick
----------------------------------------------------------------------
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
PL/I programs that use DCL EXEC_PARM CHAR(*)VAR can already handle a
length up to 32k and therefore should not need recompilation, especially
if it is not secondarily copied into something with a hard-coded length.
Walter Rue
----------------------------------------------------------------------
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
