>I wonder if I copy the "swift" executable to a PDSE if I could even run the
>compiler in the same way. I'll try that out at some point. I assume EXEC PGM
>is restricted to 8 bytes upper case, which is why I'd have to do this. Let me
>know if that's not true!
IMHO, at a high level, there is really only one difference between starting a
program out of an MVS load library (be it a PDS or a PDS/E) versus running it
out of some directory in the z/OS UNIX file system tree: The MVS loader does
not support loading from the file system. So in order to load from the file
system, you need some other program that will load the target program from the
file system thereafter.
Some such other programs are: BPXBATCH, any UNIX shell program, COZBATCH, JZOS,
BPXWUNIX() in REXX, etc.
So, basically any z/OS UNIX program can be copied to a PDS/E and then invoked
the known "MVS way", i.e. EXEC PGM=, or LOAD/LINK/ATTACH.
There are, however, minor differences which may or may not be important:
- Environment variables are easier to setup from the shell before calling the
target program. However, they can also be set using //CEEOPTS DD and
ENVAR('.....') LE runtime parameter.
- The PATH environment variable might need to be setup explicitly when not run
from the shell. It is usually setup during shell startup already.
- Program parameters are more cumbersome from JCL than from the shell.
- There are probably more that don't come to my mind right now.
--
Peter Hunkeler
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
