On Tue, 30 Oct 2007 07:20:12 -0700, Steve Comstock wrote: >John Eells wrote: >> >> "RECFM=U verification is designed to provide the same protection against >> writing programs into non-program PDS libraries as is provided for PDSE > >I agree that's a great safety valve. (What instructor hasn't seen >his or her students wipe out some libraries?) It didn't make my >list because I considered it a long-overdue feature that won't >change the way applications programmers work. But it is nice to >have. Finally. > Indeed, but too little too late. The OS (not utilities) simply should _never_ change the attributes of an existing nonempty data set. (Possible exceptions: increasing BLKSIZE, and for RECFM=V increasing LRECL.) If need be, the programmer can allocate a new data set and copy. The only time I've used the facility is to recover the mess when someone has misused it. And then I generally wind up with a PDS containing one member with RECFM=F among others with RECFM=V. The OS (not whatever utility) should prevented that happening.
-- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
