Hello Donna,

I think the problem is the file equation which lacks ",old".
For example:

FILE X=<whatever>,OLD;SEMI;GMULTI

Without that what you wind up with is that the first time

input _s_rec;wait=60 <*x >*x

executes it times out on a real empty message file but the
>*X creates a new temporary file of the same name. Each
additional time this command is executed in the same session
your reading the temporary copy and not the real message
file.

I should also add that what Paul Christidis said is correct. The
>*X will begin writing to the file providing data that satisfies the
read preventing the timeouts. I altered your script to loop 10 times
echoing the _S_REC and HPCIERR vars. I see -9003 twice.

Pretty annoying little problem and I killed a perfectly good 979-320
before I realized all it was was a tempfile!

Hope this helps,

Bill
HP/CSY


>hi again!
>
>i'm stumped....  i've got a bit a ci script that should work
>and actually sometimes works <sigh>
>
>to describe the environment -- i've got a message file (file
>x=...;semi;gmulti;acc=in) being watched by the following:
>while 1=1
>  setvar hpcierr 0
>  input _s_rec;wait=60 <*x >*x
>  if hpcierr = _s_time_out
>  ....
>
>_s_time_out is set to -9003 -- the number that hpcierr
>should be set to if the input function times out.  the first
>time thru the loop (allowing it to timeout), i get the
>expected result hpcierr = _s_time_out = -9003.  the next
>(and the next and the....) time thru the loop (still
>allowing it to timeout), hpcierr is 0 and obviously not
>equal to -9003 and things go sour.  shouldn't hpcierr always
>be set if the input's wait time is met?              - d
>
>--
>Donna Garverick     Sr. System Programmer
>925-210-6631        [log in to unmask]
>
>>>>MY opinions, not Longs Drug Stores'<<<
>