On Sat, 11 Oct 2008 07:11:12 -0500, Kenneth E Tomiak wrote:
>No, you do not need rexx in a comment. It depends. TSO will read them from
>SYSEXEC and run them without rexx in the comment. z/OS UNIX Services, in
>my limited testing, requires rexx to be lower case where any other
>environment/platform could care less about the case.
>
No -- I've tried a couple case variants and they work. Of course,
the case in the invoking shell command must match the case of the
UNIX path to the EXEC.
However, z/OS UNIX Services requires that the "/* Rexx ..." comment
begin in column 1, unlike TSO and CMS. If the comment is indented
I get:
[EMAIL PROTECTED]:119$ INDENT
BPXW0000I Exec not found
[EMAIL PROTECTED]:120$
... the code explanation is in
C.1 "z/OS V1R10.0 Using REXX and z/OS UNIX System Services"
__________________________________________________________________________________
C.1 BPXW0000I
___________________________________________________________________________________________
BPXW0000I Exec not found
Explanation: The REXX program could not be found.
System Action: The REXX program is not run.
User Response: Check the format of the REXX program ...
> As was pointed out, if you get
>an IRX message than TSO already figured out it is rexx and the comment is
>unnecessary.
>
Likewise in this case, it appears that exec() recognized the script
as Rexx, and then called the Rexx interpreter, which somehow rejected
it.
I have found no IBM documentation that requires the comment to begin
in column 1. Both TSO and CMS accept EXECs with the Rexx comment
indented. I reported this in a PMR to IBM, who claimed the problem
is impossible to fix. (I regard this as nonsense; IBM merely lacks
the determination to make the repair.)
-- gil
----------------------------------------------------------------------
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
