On Wed, 25 Feb 2026 21:16:23 +0000, Seymour J Metz wrote: >Maybe it's only used to control translation during *SAM I/O. > ... *SAM IO is OMVS file savvy, so that should be no restriction.
Looking for documentation (my bad habit) I found: <https://www.ibm.com/docs/en/zos/3.2.0?topic=tapes-ccsid-decision-tables> AYFK!? I infer that over a half-century ago, a designer made a dreadful decision, and users have been cursed with needless complexity ever since/ >________________________________________ >From: Paul Gilmartin >Sent: Wednesday, February 25, 2026 2:00 PM > >On Wed, 25 Feb 2026 08:54:43 +0000, Colin Paice > >> I have >>//STDERR DD PATH='&PATH/stderr.D&YYMMDD..T&HHMMSS', >>// PATHOPTS=(OWRONLY,OCREAT,OTRUNC), >>// PATHMODE=SIRWXU >> >>and would like to be able to tag it so it gets processed properly when >>using OMVS programs >> ... >With respect to FILEDATA, in [articular, I found, many years ago >that the JCL Ref. stated that if FILEDATA did not appear on the >DD statement it was taken to be BINARY. However I found that >if an existing file had a contrary tag the access method respected >it (but see Hyrum's Law.) I believe, but my memory is less clear, >that FILEDATA in JCL was preserved when creating a new file. >I did not experiment with CCSID or with overriding tags. > >The Ref. seems to suggest that CCSID applies primarily (only?) >for tape. Is there hardware support in the tape control unit? > >In days of yore the Ref. contained a matrix of mutually exclusive >options. That has vanished, perhaps because it outgrew the >page size, replaced by scattered "Relation to Other Options" >sections. -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
