Dear Folks,
The LASTCLPA TSO command from CBT Tape File 133 all of a sudden (it
turned out it was on z/OS 2.2) started giving the wrong date. This
command has now been fixed, and the corrected version is on File 133 of
the Updates page of www.cbttape.org.
The reason is as follows: Before z/OS 1.2 (HBB7705) the "last
CLPA" time was found in the first record of the PLPA page dataset. At
z/OS 1.2 and afterward, the correct information for "last CLPA" time was
changed to the second record of the PLPA dataset. The test for the
proper release level was a CLC instruction against the string '7705'.
That's fine until z/OS 2.1, which is HBB7790. But z/OS 2.2, which is
HBB77A0, has a problem, in that '77A0' sorts "lower" than '7750' in
EBCDIC, and the wrong PLPA record was being pulled to extract the time
stamp.
I fixed the code, to also test for 'A' thru 'F' in the third
character of '77A0'. And we have to be aware of the future release
levels of z/OS, if this method will be used in the future. Maybe it's
better to just use the OSLVL bits in the CVT to make the judgment. But
I didn't do this (yet).
We try to notice this kind of stuff in programs that we use. If
any of you ever notices a breakdown in how one of the CBT Tape programs
works, please inform me as promptly as possible, so we can fix it, and
it will still remain useful and available for all the people to use.
Thanks again for your support.
Sincerely, Sam Golob
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
