>OK, you are right, but I see, that other application besides VFP still got
>systemmessages because they run in the background independently the user actions.
>My problem is, that I dont understand, why VFP stops NullEvent during some user actions.

It may seems like that but technically not. Again, this is because Windows is not a true multitasking system like you'll find in a Unix box for example. This is probably the most requested feature for the next version of Windows. If it look like that on another application, it may be because the technique being used is something like you mentioned or similar. It also depend on what kind of process is running. A SQL is really intensive on the CPU. I am doing that for the Universal Thread Navigator and sometimes, when a socket is receiving all the new messages, the intelligent socket which is responsible to send messages from the server to the client is not able to have its message delivered to the client because the intelligent socket's event at the client side never fired because the other process was running.