On Thu, 26 Apr 2007 18:07:02 +0200, Jon Renton wrote: >Hello, > >we are experimenting with getting control of all ADRDSSU calls (the IBM DSS >user interface). To achieve this, we are thinking about renaming ADRDSSU and >supplying our own ADRDSSU frontend which will later call the real IBM >module. > >Now my question - has anyone ever renamed modul ADRDSSU? Does this have any >effect on other systems (such as HSM) which use DSS to move their data? In my experience one of the bigger issues with rename/front-end solutions is that they don't (can't) adequately cope with the strong possibility that the new 'secret' name for the old program will become known, which leaves you back where you started (but with extra baggage to maintain). My experience with this solution mechanism only goes back to 1974 though, so maybe your experience trumps mine. What works 'better' is to zap in an intercept within the existing module (or better still, use an exit if there is an exit point defined by IBM at that appropriate point) to gain control "early & always" to implement the controls that you are looking for. Sans exit, I'd strongly consider zapping an early STM or B to jump to a patch area to make the external call (essentially building an exit call into the beginning of IBM's module). All of this should, of course, be done as an SMP/E USERMOD. On the other hand, if you insist on going with the rename/front-end route then you should at least use XCTL (or maybe CSRL16J) to transfer from your front- end into ADRDSSU. That would keep the environment closer to what DSS expects at execution time. -- Tom Schmidt Madison, WI
---------------------------------------------------------------------- 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
