We need to develop a decision tree for when to use a SmartClient vs. a traditional Web Client.  Most architecture committees struggle with when and where to use a SmartClient and they want a simple matrix to follow.  Are you seeing the same thing?

It is easy to decide when you have local devices, constant use, low latency requirements and heavy data input...SmartClient all the way.  But when you make that decision, how do you partition your business logic?  General rule of thumb...if it affects data input...put it on the client, but there is more.  The general mindset of people looking at SmartClient find themselves tending to start thinking 2-tier, instead of making the leap to using them in 3-tier architectures (SmartClient calling a service facade).  This is a flaw, using a SmartClient as a thin composite solution that is working in a service oriented way provides a responsive (snappy) interface that makes users very happy.

It may seem obvious to us, but there is a real need for guidance...

This posting is provided "AS IS" with no warranties, and confers no rights.