Jousma, David wrote:
I agree, taking version number out of dataset names for running datasets
is a good thing, as it reduces the need to "fix" JCL. With that said, I
HATE ALIAS's. We install to ver.rel datasets, then at implementation
time, do dataset copies with rename to remove ver.rel. That way NO ONE
has your maintenance libraries allocated either. I don't mind limited
use of alias's to get out of a pinch TEMPORARILY, but not as a long term
solution.
Dave
We have a product used by several different groups. They never all want
to upgrade at the same time. My solution was to leave release numbers
in the datasets, and modify the JCL to have symbols for the version
number in the DSN's. An INCLUDE member specifies the version for each
group. Whenever one group is ready to upgrade the version number in the
INCLUDE member is updated.
INCLUDE member VER
//GRP1PRD SET G1PRD=V413
//GRP1TST SET G1TST=V510
//GRP2PRD SET G2PRD=V413
//GRP2TST SET G2TST=V413
PROC used by group 1 for test
//G1TEST PROC
// INCLUDE MEMBER=VER
//STEP1 EXEC PGM=xxx
//STEPLIB DD DSN=hlq.&G1TST..LOAD,DISP=SHR
Once setup the only change needed is to the VER
member. Of course if the groups create their own JCL
they have to use the INCLUDE and symbols correctly
for this to work. This does create more PROC's, so
if JCL changes have to be made for other reasons there
are more places to do it. Other INCLUDE's might be used
for some of this.
--
Richard
----------------------------------------------------------------------
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
