Radoslaw > You used many words, citations, quotations, etc. but IMHO you should just say "I was wrong".
I don't think trying to make sense of what is written in the manual which is *supposed* to describe what the product does hoping that WAC corresponds to WAD can be considered wrong. The manual was guilty of some sort of disingenuity - sorry, words again but they are rather appropriate when writing! - in suggesting that DISP=NEW could involve an existing data set. I was going to say I was misled but it's more bemused. If you care to recall, I did say that in constructing the diagrams which that section of the manual first offered by Gerhard Adam cried out to be so represented, I tried initially to follow your contention - which is what I had assumed previously without even having given it all that much thought. Then I decided that first paragraph required what has turned out to be erroneous. If I'd placed more credence on what was said about DISP=OLD I would have been guided back to the path of righteousness. > I really don't care what's in the manual, ... Sadly, not having a "sandbox" to hand these days, I'm obliged to rely on whatever documentation I can find. But since you *do* have one or more "playpen" systems to hand, perhaps you'd like to verify my revised diagrams. It would, for example, be interesting to have plotted out what how the errors which have passed JCL reveal themselves. Looking back to your first post in this thread, I see you may have already identified "PDS with member name specified", "data set does exist and member does exist", "DISP=MOD" as *not* a JCL "fail" but abend SB14 whether "input" or "output". > I want the discussion to stay technical one ... What could be more technical than reporting live experiments! Chris Mason On Sun, 6 Mar 2011 20:02:56 +0100, R.S. <[email protected]> wrote: >W dniu 2011-03-06 17:19, Chris Mason pisze: >> Radoslaw >> >>> Yes, DISP=NEW fails if the PDS already exists. Even if nonexistent member is >> specified. >> >> I'm assuming that what you mean to say is the following >[...] >Chris, >You used many words, citations, quotations, etc. but IMHO you should >just say "I was wrong". >I really don't care what's in the manual, what somebody said, etc. DISP >in JCL is always for dataset, not the member, DISP=NEW always end with >JCL error when PDS(E) exist, and it it completely irrelevant whether you >specify member name or not or specified member exists or not. > >The only (not mentioned) case when DISP=NEW works (does not cause error) >for existing dataset is ... >(I think it wasn't mentioned here yet) >GDG. > >GDG dataset with "deferred roll in" (proper name can be slightly >different, wrote from my poor memory) is simply rolled it when DISP=NEW >is used. But it is very specific case, which precludes usage of PDSE and >membername of PDS. And was (no longer is) unsupported by IBM for PDS. > > > > >I want the discussion to stay technical one, not - let's say - >political, like USS meaning or Toronto airports. >Regards >-- >Radoslaw Skorupka ---------------------------------------------------------------------- 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
