On Sun, Apr 5, 2009 at 11:18 AM, Edward Jaffe <[email protected]> wrote: <snip> > The approach I took, which I still regard as an excellent one from an upward > compatibility and interoperability standpoint, was to replace the CLISTs > one-by-one with REXX counterparts over several releases. That way, there was > no pressure to change/redesign everything at once. Also, because this was an > infrastructure within which customers often plugged-in their own "home > grown" CLISTs. I did not want to force customers to rewrite all of their > CLISTs in REXX. > > My approach was to write an external REXX function to perform IKJPARS > emulation. In this design, any CLIST in the calling hierarchy could be > transparently replaced by a REXX. Parameters are passed to/from the external > function via the data stack. Just takes a few QUEUE/PULL statements at the > top of the REXX to provide full CLIST compatibility. > > I'd still be interested to hear if others were able to emulate IKJPARS in > fewer lines. One thought I had, only after it was too late, was to make the > external REXX function an assembler language program. In that case, I could > have called IKJPARS directly. One advantage to what I've done is that it > does not need TSO/E. The same code should work on any REXX platform: z/VM, > Windows, OS/2, etc.
Is this code you can share? It would be interesting/fun to see it (let the style wars begin!). ---------------------------------------------------------------------- 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
