Well, it should be easy enough to test. I get your point about depnding on undocumented bahavior. It's something of a judgement call as to how likely you think the behavior to be stable. For this particular case, I'd presume that if it works now, it always will. It *should* be consistent with DSN refer-back and COND. If it is, then I'd count on it.
sas On Fri, Dec 11, 2020 at 3:10 PM Charles Mills <[email protected]> wrote: > I wondered about that, but how do I test that? It is not documented. If it > seems to work once, is that a good enough test? Twice? Five times? How do I > know whether it will work if other things in the PROC or JOB change? > > "Hopefully" is not a great testing philosophy. <g> > > Charles > > > -----Original Message----- > From: IBM Mainframe Discussion List [mailto:[email protected]] On > Behalf Of Steve Smith > Sent: Friday, December 11, 2020 11:49 AM > To: [email protected] > Subject: Re: How test correct procstep RC with JCL IF? > > You may be over-thinking it. I know refer-backs to a stepname stay inside > a particular PROC invocation, so hopefully, the same rule applies to IF. > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to [email protected] with the message: INFO IBM-MAIN > ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
