>>Further, I interpreset your first Message as explaioned right now. So, by doing something wrongly, all forms have gone, but VFP.EXE is in the process list, and you have to kill it in order to start VFP again. Right ?
>>Your "sometimes" IMO is about doing something not neath, and all has gone except from VFP itself. And now I wonder : supposed W2K would kill VFP.EXE (as you state), is that a pro ? I'd rather see that I did something wrong.
>>I just don't like these kind of narrow escapes too much ...
>>
>>But maybe that's not what you meant.
>
>What I meant is that in normal circumstances, without any crash on VFP or without any manually need to kill the task, even if no VFP was in memory, showed in the task bar, I had instances in the task manager. That is something I have simulated several times. From boot, I started VFP, worked in it for about two hours, closed it. Bingo, no VFP showed in the task bar. But, it's in memory in the task list. Unsolved mysteries! :)

Not sure if we talk about the same, but my experience is similar;
I think this already can happen if you leave VFP with a form without Read Events. Thus, it fires something, but can't deal with it formally, and keeps on waiting for it.
My experience is on (something like) assiging an object to a variable (or array), and not releasing that variable. It remains with .NULL., thus the object isn't there anymore, but VFP remembers it. Just about cleaning up properly.
There is something about On Error in combination with On Shutdown too.
all (IMO) is related to the situations that you just can't quit in any way, unless certain (??) objects are re-created again, in order to release them normally.

Anyway, if you say that these situations won't appear in W2K anymore that is convenient, but on second thought it won't point us on doing things wrong. Because, in all these situations I found myself doing something wrong (like an object-variable containing .NULL. which just aint right; it should be Released).