The Author's Edition states for BLOCK number NETDATA:
http://publibz.boulder.ibm.com/cgi-bin/bookmgr_OS390/BOOKS/dfsc4a00/4.4.35
...
! ... The output buffer is
! flushed after an \INMR6 record has been written. This ensures
! that any stacked file begins in a separate record.
With PIPINX086I CMS/TSO Pipelines, 5654-030/5655-A17 1.0112
(Version.Release/Mod) - Generated 6 Jun 2011 at 08:31:39
I have pretty good evidence that this isn't true. While the record
being built may be padded to "number" characters (I can't tell),
it appears that it isn't flushed until all inputs to BLOCK are
severed. What I'm trying to do is (abbreviated):
CP DEFINE PUNCH CUA
callpipe [stages] | faninany | block 80 netdata | console | punch CUA
[ write stuff]
'callpipe literal '20'x'INMR06| [netdata]'
[write null record to flush punch]
CP DETACH CUA
Unless I sever all stages having input to BLOCK, I get a failure
>>> "callpipe CP DETACH 0011 | console"
RDR FILE 9046 SENT FROM user PUN WAS 9046 RECS 0008 CPY 001 A NOHOLD NOKEEP
PUN 0011 DETACHED
pty line { {Empty line by <CR><LF> { One dot {.. {Last Line {. {QUIT \INMR06
PIPIOS159E Device 0011 no longer exists.
PIPMSG004I ... Issued from stage 4 of pipeline 1 name "P01".
PIPMSG001I ... Running "punch 0011".
Ready(00159); T=0.03/0.04 08:49:17
I assume this is when BLOCK attempts to PUNCH the last record when
the Rexx stage terminates.
Thanks,
gil
BTW, the doc refers sometimes to "INMR06", sometimes to "inmr6". This
makes string searches tricky.
