Newsgroups: comp.lang.dylan
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!newshost.marcam.com!charnel.ecst.csuchico.edu!olivea!news.hal.COM!decwrl!adobe!macb304.mv.us.adobe.com!user
From: mhamburg@mv.us.adobe.com (Mark Hamburg)
Subject: Re: message passing not-vs. multimethods
Message-ID: <mhamburg-070295094440@macb304.mv.us.adobe.com>
Followup-To: comp.lang.dylan
Sender: usenet@adobe.com (USENET NEWS)
Organization: Adobe Systems, Inc.
References: <9502041646.AA25511@speed.harvard.edu>
Date: Tue, 7 Feb 1995 17:43:55 GMT
Lines: 24

In article <9502041646.AA25511@speed.harvard.edu>, pg@das.harvard.edu
wrote:

>   From: hvandurm@vub.ac.be (Herman Van Durme)
> 
>   Why is Dylan using the multi-method paradigm instead of the more common
>   reciever/message-passing paradigm?
> 
> This is a question that seems to come up often.  It's a bit
> like asking, "why are you using polygons instead of squares?"
> Multimethods are a superset of message-passing.  If you only
> ever specialize the first argument of a method, then you have
> message-passing.  
> 
> Perhaps something about this should be added to the Dylan FAQ.
> 
> -- pg

I understand the benefits of multi-methods. I also understand the costs in
runtime and/or language implementation complexity with comparison to
allowing specialization only on one parameter. So the real question is, why
did the designers of Dylan feel that the benefits outweighed the costs?

Mark Hamburg
