>In my experience, the client generally doesn't give you time to do nifty interfaces anyway....they want the job done....there are those in our profession who have garnered bad reps by being too slow to deliver when trying to perfect an interface.

I think that's a key point. Most of Cooper's suggestions mean more work for designers and programmers. On the other hand, the people who can spend the time to break away from old interface paradigms may end up making more customers happy, and make more money thereby.

>I guess I have the same aversion to blind adherence to an interface *theory* as I do to blind adherence to OOP theory....compromises and the purpose of the application *have* to be considered.

I haven't yet heard anyone suggest otherwise.