Sean,
I didn't do the actual conversion here myself, but I've looked at the
resulting code. Like you, we took the path of putting all function in an
'Exit 54' that is now called for Exit-4 as well. The JCTUSERn fields are
still available to keep track of processing across exit calls. Wouldn't
that satisfy your persistence requirement? It's how I've always coded JES2
exits in the past for inter-call communication and still works for us.
There's one new wrinkle you have to be careful about that I'm not sure is
documented. We had to change our strategy of inserting cards individually
because of a JES2 logic change in 1.7: each card inserted on a single call
must pass a new validity check. For many years before 1.7, we inserted a
series of cards like this:
// JCLLIB=(dsn1,
// dsn2,
// dsn3)
In our early conversion attempt, the first card above caused a JCL error
because of some new validation that never occurred before at that point.
Our solution was to build the entire sequence of cards in memory and then
insert them all at once via RJCB processing.
At SHARE in Seattle, we noted that early documentation of RJCB was
lacking. Tom Wasik was at the session and immediately stepped up to
improving the doc.
"Smith, Sean M" <[EMAIL PROTECTED]>
Sent by: IBM Mainframe Discussion List
06/16/2006 07:28 AM
Please respond to
IBM Mainframe Discussion List
To
IBM-MAIN@BAMA.UA.EDU
cc
Subject
JES RJCB Processing
Hello all,
I am converting some Exit-4 code to Exit-54 and have changed to using
the RJCB process to insert cards to the job stream. The problem that I
am having is that there is a flag in the $XPL to identify when cards are
to be inserted from JCTXWRK but there is not a flag to indicate when
cards have been inserted by a previous pass.
I realize that these are two different requirements. The first tells
the system to insert cards from the JCTXWRK and my requirement is a flag
to indicate that cards have been inserted by a previous pass through the
exit. In one of my exits I can just a flag byte to the persistent
storage I acquire. But in another there is no persistent storage. Any
tips on how to flag that the exit has already inserted cards? Could I
just check the X054RJCP to see if it has an address on subsequent
passes?
--
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