Charles wrote: >I wrote a (successful!) product that in one very peripheral feature took an >operand that could represent a member name in a default PDS, a dataset name, >or a zFS file name. I differentiated among the three based on length and the >presence or absence of periods and/or slashes. No one ever complained that >they had a dataset or a zFS file named SHORTNAM and could not reference it. >I think the convenience and simplicity of being able to say simply >FILENAME(whatever) outweighed the perils of the ambiguity. Product design >involves tradeoffs.
Agreed, this isn't tricky. If it starts with a slash (or possibly a period), it's a USS file; otherwise it isn't. So a zFS file named SHORTNAM would be listed as /u/whatever/SHORTNAM or, at worse, ./SHORTNAM and it Just Works. >From the USS side, support DD:ddname as a filename and you're good (from C I'm not actually sure you can avoid supporting that). We have such a use case and have never had a problem with it. ...phsiii ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN