>Tiny chance, tiny, tiny, of runtime/compiler problem if you are on V5+. Nice
>to know, just to know whether to discount it completely.
We're on the way to Cobol V5 but production is still V4.
>With the COBOL program and the record causing the failure, it would be
>possible to identify the issue. Of course, it is not always possible to supply
>these things.
Yes, this is what I have asked the developer to do. However, there are two
possible inhibitor: a) We're not given permission to move the record from prod
to test, and b) the problem does not occur in test with even that record,
because the data on the data base is different.
I have little hope so far they will succeed to reproduce the problem in test
before we know more about the cause (chicken or egg problem).
>For the first, using compiler option SSRANGE (with LE Runtime option CHECK(ON)
>if you are before V5) will help.
Yes, but again, easy in test, impossible in prod.
>I don't look at programs the way a Sysprog does, and that's probably true vice
>versa. What is easy for me to say and do, is probably as much like moon-dust
>to you as a lot on this list is to me :-)
Fortuntately, I've been doing application programming a lot in my career as
well, so I know both sides very well. What I'm pretty ignorant at is Cobol. I'm
mainly a PL/I, Assembler and REXX programmer. But I'm learning
--
Peter Hunkeler
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
