I would argue that cat should not document it, any more than the COBOL compiler doc should cover all the possible permutations of //SYSPUNCH DD ... They should reference the C library documentation and say "any of those formats."
I agree on the merging to DCB attributes. Charles -----Original Message----- From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On Behalf Of Paul Gilmartin Sent: Wednesday, September 1, 2021 10:02 AM To: IBM-MAIN@LISTSERV.UA.EDU Subject: Re: Copy RECFM=VBS/LRECL=X to zFS? On Wed, 1 Sep 2021 09:34:07 -0700, Charles Mills wrote: >The DD:ddname format for a "legacy" dataset is a supported and documented >feature of the C runtime library (even if not specifically for cat). > The "cat" developers are understandably unwilling to document and support that behavior of "cat". At very least they'd need to specify the behavior when allocation options conflict with options on the "cat" command. (I suppose "unpredictable" would suffice.) Bottom line: if it breaks, either immediately or at the next release, you get to keep both pieces. Related: I once allocated a RECFM=VB data set, overriding to RECFM=U and fetched it from DD:SYSUTx with FTP BINARY. I expected to see BDWs/RDWs. Didn't. In effect, FTP used the attributes from the DSCB and ignored the allocated attributes. "cat" might have similarly unexpected behavior. -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
