On Sun, May 13, 2018 at 3:36 PM, Seymour J Metz <[email protected]> wrote:
> I'm comfortablewith using Unix files. I'm not comfortable with packaging
> for MVS components that seems done by people without a clue, but it's not
> fair to blame that on the use of Unix.
>
I'm trying to figure out why IBM uses "pax" files in a z/OS UNIX
filesystem for their distribution. I know that they need some way to
distribute z/OS UNIX files and "legacy" datasets in a single package. I
have a couple things on the CBT which do that. What I did was use pax to
write a pax archive as a member of an FB/80 PDS. I then XMIT'd the PDS to a
sequential dataset. Of course, the reason that I did this is because that
is how Sam Golob wants the things on the CBT to be packaged. I don't really
know which method is better. Mine is at least more traditional. If the
XMIT'd sequntial dataset is really large, it might be a good idea to
AMATERSE it. I think my boss would prefer this, because he wouldn't need to
directly interact with UNIX files. The processing of a package would be
something like:
1) if needed: use AMATERSE to UNPACK the compressed XMIT'd PDS
2) RECEIVE the XMIT file ( RECEIVE INDATASET(...))
3) Use ISPF edit to display the PDS directory
The restored PDS would perhaps have members such as $DOC, $UNLOAD, and
other members which might contain PAX'd UNIX files (one member per
directory?), XMIT'd PDS libraries and DSORG=PS datasets. The $UNLOAD job
would have steps to "expand" all of the distributed libraries, PS datasets,
and UNIX files. All the DSNs and other "customizable" names would be
referenced via SET statements at the start of the job (rather than example
"hard coded" DSNs which each need to be edited.)
Just my thoughts, they are likely worth what you paid for them {grin}.
>
>
> --
> Shmuel (Seymour J.) Metz
> http://mason.gmu.edu/~smetz3
>
>
--
We all have skeletons in our closet.
Mine are so old, they have osteoporosis.
Maranatha! <><
John McKown
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
