Here goes. erase test data o
Ready; T=0.01/0.01 17:55:31 pipe literal | pad 10 | > test data . f 10 coerce allowempty Ready; T=0.01/0.01 17:55:41 l test data * (iso FILENAME FILETYPE FM FORMAT LRECL RECS BLOCKS DATE TIME TEST DATA O1 F 10 1 1 2008-04-02 17:55:41 Ready; T=0.01/0.01 17:55:44 pipe hole | > test data . f 10 coerce allowempty Ready; T=0.01/0.01 17:56:03 l test data * (iso FILENAME FILETYPE FM FORMAT LRECL RECS BLOCKS DATE TIME TEST DATA O1 F 10 0 0 2008-04-02 17:56:03 Ready; T=0.01/0.01 17:56:05 Paul, this is a very workable circumvention. Thanks very much!! -----Original Message----- From: CMSTSO Pipelines Discussion List [mailto:[EMAIL PROTECTED] On Behalf Of Paul Gilmartin Sent: Wednesday, April 02, 2008 5:37 PM To: [email protected] Subject: Re: [CMS-PIPELINES] Writing an empty file to set RecFM and LRecL On Apr 2, 2008, at 15:21, SPITZ, HOBART CTR DFAS wrote: > It's been a long day. Does this look right? If so why? > No, but it's likely WAD. CMS is historically hostile to empty files; even more to empty records, and will do anything it can to frustrate the programmer who attempts to employ them. I suppose the rationale is that in the absence of any record to establish the record length it reports a default value. I remember a similar problem when I once FILEDEFed a file as V 137. Later, when I re-opened the file for append (MOD), it loaded in the DCB not my specified 137, but the length of the longest record previously written. Ugh. > Ready; T=0.01/0.01 17:15:57 > > pipe hole | > test data . f 10 allowempty > > Ready; T=0.01/0.01 17:16:49 > > l test data * (iso > > FILENAME FILETYPE FM FORMAT LRECL RECS BLOCKS DATE > TIME > TEST DATA O1 F 80 0 0 2008-04-02 > 17:16:49 > Ready; T=0.01/0.01 17:17:21 > What happens if you initialize by writing a 10-byte record, then overwrite with an empty file? --gil
