On Thu, 8 Mar 2012 11:29:38 -0600, McKown, John wrote: >> -----Original Message----- >> [mailto:IBM-MAIN@bama.ua.edu] On Behalf Of Scott Ford >> Sent: Thursday, March 08, 2012 11:02 AM >> >> You saying the working directory on Z/os unix is different >> than the homes? >> I really, really hope that was intended as whimsy. > >Sure "current working directory" is not always the HOME subdirectory. It is >whatever directory was last referenced via the "cd" command or the C "chdir()" >function or via the callable services: BPX1CHD (24/31 bit) or BPX4CHD (64 bit). > Or with Rexx "address SYSCALL chdir ..." And this works even under TSO, and it stays changed after the EXEC or program exits.
Yes, I'd like to see DYNALLOC respect CWD. But the problem remains: should the binding be performed at time of allocation, or at time of OPEN? Even now, one can DYNALLOC a path; delete the referenced file, then OPEN, which fails. IBM has taken a RCF on this WAD behavior. Better: tilde substitution in the PATH. Where I work, user homes are not in /u. And some people mount HOME via NFS from a system with YA naming standard. -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: INFO IBM-MAIN
