>
You have confirmed my point. That is that an MTDLL is NOT multi threaded and calling it multithreaded is technically incorrect. It causes confusion as people think it really is.
>

I think your point was that Claude was confusing the two concepts. I did nto see the basis or validity of your point.

Claude was right - what you call it is nothing more than semantics. The fact is, for what Fox developers need - what the MTDLL does - regardless of what you want to call it - fits that need.

>
In a multithreaded app you can spin off a new thread and continue processing while the new thread also does its processing.
>

Craig - nobody - least of all me - is arguing that Fox is multi-threaded. The fact is - when people say they need multi-threading - they really don't because there are safer - more appropriate alternatives out there. My response to your flawed question was a good example of that.

>
An MTDLL is in fact apartment model threaded. You can call it from multiple clients, but those clients will stop and wait for the MTDLL to finish processing. They will not continue to process while waiting for the MTDLL to do its processing. You can also call it from a single client, but it too will stop and wait for the MTDLL to finish processing.
>

Your talking about synch vs. async processing - which is a different concept. For example, eventhough VFP - as a UI client is single-threaded - I can make use of async calls in ADO. It does not do a lot of good unless you are taking advantage of events - since that is the only way you will know when something has completed.

>
John, you of all people, should know when to use the proper terminology. You've jumped on enough people in the past for not being technically correct.
>

Well, when I did it - my ducks were in a row. The same cannot be said of your argument here. The fact is - from the point of view of the MTDLL - it can process multiple requests at the same time.

The MTDLL was created primarily for web-apps. Basically - it is a masterful kludge that was created. Somehow, Calvin managed to weave MT simulation into something that does not support multi-threading.

Claude's basic point was correct.