Hi!

It looks like it is related to use of the common DLL libraries. I have this problem too on Windows NT SP6 - when I open anotehr instance of the VFP development environment, and run some APP in the first instance, I'm not able to do anything in the second instance until first instance finish code running and get to the waiting for input state. I always interested how to avoid it but did not seen anything about this yet.

Just a suggestion:
Try to put each VFP EXE in separate folder and put VFP run-time DLLs in each folder. This probably will cause using of the different instance of VFP run-time DLL for each EXE. I do not know if this can help, but if can, this probably will increase the amount of memory used by all VFP applications currently running.

Anther suggestion: In all your VFP applications keep _VFP.AutoYield property setting always .T. As far as I know, it allows processing of internal VFP events between code lines execution, that is probably also could be related to this problem.

HTH.

>I have a monitor that runs multiple VFP applications. These apps transfer and process data across remote servers. The monitor uses WSH’s oshell.run to launch the VFP app’s. My problem is when each VFP application terminates it takes the W2k server considerable time to release the application and free resources. I’ve noticed that when other VFP apps are executing certain tasks (such as returning data from remote source) it prevents the terminated apps from being released.
>
>I’m looking for any help on how I can force the release the terminated apps and free the resources.