On Wed, 8 Mar 2017 11:32:48 -0600, John McKown wrote: > >> [Default] On 8 Mar 2017 07:17:56 -0800, in bit.listserv.ibm-main >> ajn...@ufl.edu (Nims,Alva John , Al) wrote: >> >> >My lousy $0.02 on this: >> >1. 71? Column 71 is the old "Continuation" column in JCL. >> >2. How would you code a 71 character DLM='-----71 Char----' when you >> already have to code "//name DD DATA,DLM='..." in the JCL, I HATE having to >> continue quoted strings, I never get it right! >> Dammi! I envisioned 71 (or 50), or whatever as a maximum. You should be free to use 13, or 42, or what you choose. Is the 18 in z/OS 2.3 a minimum, or 2 still be supported? Or intermediate values?
>My take on all this is that JCL should have a way to be "free format" and >not restricted to FB/80 card images any more. I don't keep my JCL in PO or >PO-E library, I keep it in a UNIX file. I would love to be able to do a >SUBMIT where the JCL images are as long as I need them. It would really >help when I need something in the PATH= on a DD where the PATH= value is >real long, like >/usr/lpp/internet/server_root/HealthMarkets/UserAdmin/somefile.cnf (66 chars >long). > Sometimes I break it up with SETs, often ignoring the warning that SET symbols should not be defined in terms of othe SET symbols. Is it possible to code a SYSIN DD with white space and superfluous parameters and SET symbols that evaluate to nothing so the DLM is split from column 71 to the next column 16? On Wed, 8 Mar 2017 13:16:54 +0000, Vernooij, Kees (ITOPT1) - KLM wrote: > > But as long as we have to distinguish all our production jobnames and all > our production sysinlibrary names with 8 characters, I think 18 DLM characters > for the 3 DLM strings we use yearly is very wealthy. ... Eerily reminiscent of "No one should ever need more than 640kB" I'd just like the phobic programmer to be as comfortable as possible using a random character string as DLM that the likelihood of a collision is negligible without scanning the SYSIN to verify that it doesn't occur as data. (I've done that in a script, retrying if there's a collision. But 'yy' always worked on the first try.) Would a modern cryptographer be comfortagle with 18? -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
