On Fri, 31 May 2019 08:26:50 -0400, Billy Ashton wrote: >Hi John, I am getting ready to go away for a few days, and when I get back >I will give you a mockup of my workflow so you can see what's what. Thanks >for helping! > Thoughts on this:
Might your dangling ENQ (EDC5061I) occur because you have _BPX_SHAREAS=MAYBE? Would _BPX_SHAREAS=NO cp mitigiate this? I believe ssh, scp, and sftp are all wrappers around a comon core, their functions and entitlements are largely interchanbable, and: o sftp performs pure binary transfers o scp is text-oriented (possibly performing CRLF conversion?) o z/OS (only?) ssh performs ISO8859-1<->IBM-1047 conversion. I've been able to circumvent this by piping through iconv on the z/OS side. (Do I hear Kirk snickering?) o Co:Z may provide options to suppress some of these misbehaviors. o I hate EBCDIC! AMATERSE UNPACK performs an inane check on (DEVTYPE of the first catenand of?) SYSUT1, prohibiting UNIX files. I can circumvent this by supplying as SYSUT1 a concatenation of: o (first) an empty temp PS CKD dataset o a UNIX file (I don't recall trying this with a pipe). Rexx may be more suitable than sh. z/OS likes Rexx. Rexx can invoke AMATERSE with ADDRESS LINKMVS AMATERSE. This passes allocated DDNAMES (with their ENQs) to AMATERSE. If Rexx is invoked from sh, it can invoke shell commands with ADDRESS SH. If Rexx is invoked from any environment it can invoke POSIX functions with ADDRESS SYSCALL (cumbersome) or shell commands with BPXWUNIX (simpler). There's a prevalent misconception that BPXWUNIX stdin, stdout, and stderr can only be compound stems or STACK. In fact, any of those streams can be: o stem, o STACK, o pre-allocated DDNAME, or o pre-existing descriptor 0, 1, or 2 (e.g. a pipe). -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
