Ah, I see, very clever! It uses two virtual method calls, and takes advantage of the fact that once the first is started, the compiler knows the type of that first object and so can find the right overload for the second object's intersect. I'm not sure what the advantage would be over the RTTI solution, although it would be interesting to compare the performance.
in reply to Re^6: Runtime introspection: What good is it?
in thread Runtime introspection: What good is it?