Re:

> > the timeout occurs the entire 'input' command along with the 'text'
> > corresponding to 'hpcierr' 9003 gets written to the message

The "input foo; wait=60 <*x >*x" will (when a record was found)
delete the record from the message file, and then try to append it
to the end of the file.  So, one has to ask: is this desirable/acceptable
to your application?  (Remember: there could have been multiple records
in the file, and now they're out of chronological order).

Setting HPMSGFENCE to 2 prior to the "input" will result in one
blank line being added to the message file in the event of a timeout.

So...there are two solutions I see:

   1)...why put it back?  Why not pass it to the application somehow?

      If you can avoid trying to put it back, you've solved the problem.

      I.e.: remove the ">*x" and do something like:

           run application;info ="!foo"
or

   2) put it back conditionally (instead of on the "input" command):

      setvar hp..  0
      input foo; wait=60 <*x
      if hp...  = -9003 then
         comment we timed out
      else
         comment we didn't time out...replace the record (albeit at the end
         comment of the file) and run the application...
         echo !foo >*x
         ...run application...
         endif



Stan Sieler                                           [log in to unmask]
www.allegro.com/sieler/wanted/index.html          www.allegro.com/sieler