On Sun, 6 Apr 2003, Chris Pine wrote:
> Then does it make any sense for UnboundMethod#to_proc to be defined?
> (This is another reason why it feels like Method should subclass
> UnboundMethod and not the other way around...)

If they have diverging interfaces such that the contracts conflict with
each other, then maybe they should two subclasses with a common
superclass, possibly Object or possibly a new class representing whatever
they might have in common.

________________________________________________________________
Mathieu Bouchard                       http://artengine.ca/matju