On Wed, 11 Mar 2020 08:03:54 -0500, Peter Relson wrote: >... >o The Ref. doesn't specify which dominates when both a JCL and a SYSTEM > symbol definition are available. ></snip> > >I suppose I should have double-checked before my previous response. >No one will ever say that it's necessarily easy to find what you're >looking for, but this is documented. > Thanks. I often raise an issue on this list before going to RCF because I may have overlooked that which is not "necessarily easy to find". You may be a resource more valuable to IBM than a pubs editor, so perhaps I should go to RCF first. People who hand me citations should expect me to read the entire chapter and pick additional nits.
> (from the JCL Reference >https://www.ibm.com/support/knowledgecenter/SSLTBW_2.4.0/com.ibm.zos.v2r4.ieab600/jdefine.htm >): The right place, but indeed not "necessarily easy to find" among 1492 words of mixed text and examples. If the semantics were more uniform, the Ref. could be thinner. For example: If you specify duplicate JCL symbols on an EXEC statement, the system uses the first substitution text as the default. but: If you define duplicate JCL symbols on a SET statement, the system assigns the last substitution text to the JCL symbol. Couldn't they have been consistent and used "last" (my preference) in all cases? (I know; Bad History.) The use of "nullify" throughout disagrees with my use of the word. To me, "nullify" means "remove the definition of". The examples seem to show that it means "redefine as the empty string". For example (RCF needed?) if I SET a JCL symbol that duplicates a system symbol, then later "nullify" the JCL symbol, does it revert to the system symbol's substitution text or remain the empty string? Can I make that happen? (Competent C programmers know the difference between NULL and "".) The verbiage about repeated apostrophes shows a desperate need for an analogue of HLASM's DOUBLE BIF. I prefer even more Rexx's simple convention that substituted text is never rescanned. If all occurrences of symbols were recognized and substituted rather than only those enumerated elsewhere, the Ref. could be thinner by that amount. (I know; Bad History. And even Rexx enumerates exceptions to symbol substitution.) -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
