The whole scenario is too complex for an e-mail here, but I have a program that needs to keep track of the difference between NEW and MOD to determine whether it is okay to open the dataset more than once for output (which is advantageous to the program but not absolutely necessary if it would cause other problems).
The whole NEW/MOD behavior is incredibly complex when you factor in the different behaviors for sequential versus partitioned datasets. I fear this is going to start a whole "I hate JCL" digression, but MOD means very different things for PDS versus sequential. MOD is "permissive" for sequential datasets (it might exist and it might not -- whatever) whereas MOD for PDS members means "must be a unique member name." Charles -----Original Message----- From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf Of Walt Farrell Sent: Wednesday, May 29, 2013 3:50 PM To: [email protected] Subject: Re: Getting DD DISP On Wed, 29 May 2013 16:54:49 -0500, Kirk Wolf <[email protected]> wrote: >Is there an easy way for a program to get the DISP (NEW/MOD/SHR/OLD) of >a DD from the TIOT, or do you have to do a RDJFCB? I'm somewhat curious why you'd want to know, from a program, Kirk. But however you get it, I'll mention that NEW and MOD are not necessarily distinct. That is, they are certainly separate keyword values, but if the data set does not already exist then MOD can act like NEW. That's a large part of my wondering why it's significant to the program, since I can't think of a way to make a meaningful distinction between those two values of DISP. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
