So you are launching a compiled app from a form of another compiled app? Dan Goldberg
-----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of Bob Thompson Sent: Monday, August 24, 2015 9:13 AM To: Dan Goldberg Subject: [RBASE-L] - Re: Button EEP versus Timer EEP I can try the memory flush, but I do not think that is it. This is a compiled app. When the CFA stops, it returns to the external form. I can then click a button eep that calls the same CFA and it runs fine. I would think a memory issue would linger without an exit or flush. As far as speed, I am very satisfied. We previously used a commercial third party MRP program. It would not do exactly what we needed and it took well over 4 hours to run. So now we have the exact capability we need and in 1/4 the time! Thanks, Bob Thompson LaPorte, IN 219-363-7441 Sent from my iPod On Aug 24, 2015, at 10:20 AM, Dan Goldberg <[email protected]> wrote: > You might try to flush the memory after it runs to see if that helps. > > PROPERTY APPLICATION FLUSHMEM 'TRUE' > > Just a tip, you might trying to declare your cursors before the while loops > to speed it up. > > Dan Goldberg > > -----Original Message----- > From: [email protected] [mailto:[email protected]] On Behalf Of Bob Thompson > Sent: Monday, August 24, 2015 6:32 AM > To: Dan Goldberg > Subject: [RBASE-L] - Re: Button EEP versus Timer EEP > > Thanks to Karen and Mike for the responses. > > The code stops on record 1083. > > There is approx. 400 lines of code that executes properly before the process > stops. I have bee unable to determine the exact code line that it stops on > as TRACE or running it outside the timer does not exhibit the problem. The > code does stop inside a structure of: > > Declare C1 > While > Delare C2 > While > While > While > Stopping inside the last while loop. > > I have checked for improper breaks, goto's, etc. (It runs fine outside the > timer. ) > > This is a MRP program running for 1000+ items with multiple nested component > levels for 14 inventory locations. It is a significant chunk of code logic. > It takes about an hour to run, which makes multiple test runs somewhat time > consuming.
