>Erik,
>
>Your approach is a Decorator pattern (according to Gamma and Helms "also known as wrapper" :-)). A direct quote from Gamma and Helms is "Decorators provide a flexible alternative to subclassing for extending functionality."

Thanks for the clarification.

>So I guess, on the semantics, I stand on John's side.

Actually, we probably all stand on the same side as far as far as nomenclature. I know it's not really subclassing- the following quotes are from my original post:

'allows me to effectively 'subclass' any COM object'
'analogous to real subclassing'

Throughout the post (except the title), the word subclass is used only in quotes.

I was just object to calling it a simple or classic wrapper, and was trying to defend against JVP's assertion's that:

'this is nothing new'
'the Access method just complicates the issue- what's the point'
'you can't override behavior'


Again, thanks for helping to clear things up.