On Mon, 16 Jul 2018 02:09:20 +0000, Jesse 1 Robinson wrote: >Program did not need APF. Shared production library contained other programs >that did need it. Until my bad JCL, it had never been a problem. I guess times >were simpler then... > I believe that until a fairly recent OS release LNKLST did not support a mixture of APF and non-APF libraries. So if you wanted your program accessible without STEPLIB it was (almost) necessary to put it in an APF library.
>-----Original Message----- >From: Charles Mills >Sent: Sunday, July 15, 2018 7:05 PM > >Putting on my security preacher hat, I might argue that programs that do not >need APF (i.e., test successfully without it) should not be in an APF library. >Granted, your story is from simpler times (I would assume). > There's the dilemma of how faithfully the test environment should mimic the production environment. Should the program have been tested at oh-dark-thirty, from an authorized library, with live data ...? -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
