RE: Never ending programs

>Having seen repeated recommendations to
>isolate the trigger program from
>the actual processing program, doesn't
>this therefore imply the use of a
>never-ending program ? ;)

Not at all, Evan.  My setup looks like this:

addpftrg ... pgm(generic)

pgm generic
*entry plist
   (trigger buffer definitions)

thisFile chain triggerControl
if triggerActive
  call thisFileTrigger
  parm (trigger buffer definitions)
endif

    pgm thisFileTrigger
    *entry plist
       (trigger buffer definitions)
    * perform actual trigger functionality here.

Make all the programs actgrp(*caller).

To replace the actual trigger program, just recompile thisFileTrigger.  You
can also turn the trigger off by setting the proper flag in the control
file.  Never having to actually do a rmvpftrg.
  --buck