> On May 10, 2016, at 10:15 AM, Mark Zelden <[email protected]> wrote: > > I never responded / commented on Andrew's post, but since you brought the > subject up > again I will. > > I always use DDDEFs unless some products install / maintenance procedures > don't include > adding DDDEFs and the vendor supplies an SMP/E PROC or JCL with all the DDs > in them > instead. Even then I may add the DDDEFs depending on how many target and > dlibs there > are. > > In my original post the example was from a product that contains JES2 exits > and is > of course very sensitive to the the JES2 level and /or IBM maintenance to > JES2. The > exits are installed in the product as a usermod and you re-apply the usermod > to > force reassembly after JES2 maintenance is applied. Since multiple levels of > the > OS are being maintained I "APPLY REDO" pointing to a copy of the SMP/E > controlled loadlib using the proper level of the JES2 macro library. It > makes little > sense (at least to me) to run one apply, change the dddef, then run another > apply for > each JES2 level. It's much easier to keep two copies of the JCL with the > SYSLIB > override in JCL pointing to the correct macro library where I can see it. > There is > no "correct" DDDEF when I'm maintaining two different OS / JES2 levels. > > I actually have several products that I have to deal with this way while > maintaining > multiple OS / JES2 levels. > > Regards, > > Mark > —
Mark: One comment here. When I do JCL I have 1 PROC (total) for Receive, Apply and Accept. The symbolics in JCL allow for overrides for APPLY & Accept volumes. There is no second guessing on what you are applying and accepting. I find with DDEF’s there can be too many incidents of OOPS I didn’t mean to do that. Also with DDEFS there can be too many fingers in the pie and that leads to oops. Ed ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
