On Sun, 9 Oct 2022 12:51:07 +0000, Seymour J Metz wrote: >The point that I am trying to make is that the behavior in question did not, >as claimed, go back half a century. > A half century ago, attributes specified to Allocation dominated attributes intrinsic to the data set, and were in turn dominated by attributes coded in the DCB. That behavior has been preserved in the transition from 3350 to 3390, and to UNIX files.
>________________________________________ >From: Paul Gilmartin >Sent: Sunday, October 9, 2022 8:40 AM > >On Sun, 9 Oct 2022 04:46:17 +0000, Seymour J Metz wrote: > >>Anachronism alert! You wrote "Attribute overrides have worked that way for a >>half century, longer than MVS UNIX has existed." Half a century ago there was >>no PATH and no SDB. In fact, there was no SYSEXEC, only SYSPROC. > >And there was no 3390, but BLKSIZE=3120 still works. > >What point are you trying to make? > >________________________________________ >From: f Paul Gilmartin >Sent: Saturday, October 8, 2022 9:10 PM >> >In REXX, long ago, prior to SDB. I used to ALLOCATE PATH(whatever) , > LRECL(199) RECFM(V,B) ... >and the REXX RTL would choose a reasonable BLKSIZE. -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
