+1 Unless you have a VERY slow or busy machine, does it really matter? The point of a timeout usually is to keep a transaction from taking "forever" and if both ECBs are posted, then you achieved that goal.
If it does matter, why couldn't either of the POST routines check the *other* ECB and if posted, set a flag "the other guy got here first." CM On Tue, 6 May 2025 16:00:46 -0700, Michael Stein <[email protected]> wrote: >Do you really care which ECB got posted first? > >I'm guessing that you want a timeout on the service being called. > >One approach is to have the mainline just check if the service completion >ECB is posted, if so the service completed and you don't need to deal >with the timeout case. If the completion is not posted (yet) then deal >with the timeout (considering that the service still might compilete at >any moment...). ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
