Musing on PMD dispatch problems

[Paul Dufresne]

Frankly, it took me about 2 hours just to begin to understand your message.
And I think I am still a bit confused.

>so depending on which one it sees first, it can dispatch
>either "x" or "y" -- and in Bryan's case it was actually dispatching "x",
>which is not the most specific method!
To follows you, I have to suppose we had actually call C foo: C.

Suppose you call the first that fit, by going down in each column.
If I have a new case, independant from yours:
A foo: C [ A * C].
and now I call A foo: B. By following your algorithm, would I call [A * C]?
Of course you can't, but I do you avoid to call methods where not all 
arguments are defined?

_________________________________________________________________
MSN Messenger : discutez en direct avec vos amis !  
http://messenger.fr.msn.ca/