[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Multimethods



	From @udel.edu:carroll@cis.udel.edu Fri Mar 22 17:17:41 1991
 GMT
	To: self-interest@self.stanford.edu
	Subject: Multimethods
	From: "Mark C. Carroll" <carroll@udel.edu>
	Date: Fri, 22 Mar 91 20:09:55 -0500

	I find the idea of multimethods extremely interesting, and I've found
	multiple polymorphism to be extremely useful in languages like C++.
	But before something like that gets added to Self, I think some
	consideration should be given to what it does to the model of the
	language.

	Self currently has an extremely elegant model. It provides an
	extraordinarily simple, uniform way of doing things. The basic
	operation is the message pass, and if you understand the conceptual
	mechanics of message passing, you understand Self. (Well, it's a bit
	of an exageration, but, for the sake of argument, let it stand.)

	What happens to that model when we add multimethods? 

Great question! I think it's something like: "Ask a bunch of objects to do their thing"
instead of asking just one. But, it may be more profound than that.
We won't know till we try.

Dave