>>On a side note on COM. I talked to Randy Brown of Microsoft and he said that >VFP6 will NOT provide non-blocking COM objects. In other words every request >to an instanstiated VFP COM object will not be processed if there is a >previous request already being processed by the object. I thought MS >Transaction Server was going to do away with this limitation but it will not. >This makes VFP COM objects completely unscalable and unusable as 3-tier >business objects!!! I cannot believe the Microsoft VFP team thinks this is >acceptable! ( Sorry just venting a bit.. ) Do you have any insite on this?
>
>
>Yes, this is a big problem. But, before you throw out the baby with the bath-water, I would suggest testing VFP COM-Servers in your scenario. As far as MTS goes, only 1 instance of the runtime gets loaded. This is where the scaleability issues are coming into play. Without MTS, an instance of the runtime is loaded for each class that is loaded. FWIW, with or without MTS, VB servers do not have this limitiation. IMHO, this is new ground for VFP- MTS that is. And, we need to give the team a chance to work things out. I can tell you this issue is at the top of the VFP team's to do list.

John,

According to Randy Brown the blocking is on a method-by-method basis. In other words, two clients can run different methods concurrently, but not the same method. I guess the effects of this depend on your class design. I haven't had a chance to test this. Anyone else tried it out?