Newsgroups: comp.lang.dylan
Path: cantaloupe.srv.cs.cmu.edu!rochester!udel!gatech!news.sprintlink.net!noc.netcom.net!netcom.com!netcom8.netcom.com!haahr
From: haahr@netcom.com (Paul Haahr)
Subject: Re: Argument ordering conventions?
In-Reply-To: davis@ilog.fr's message of 12 Jul 1995 09:03:20 GMT
To: davis@ilog.fr (Harley Davis)
Message-ID: <HAAHR.95Jul13093320@netcom8.netcom.com>
Sender: haahr@netcom8.netcom.com
Organization: NETCOM On-line services
References: <DB9FuA.BrB@aplcenmp.apl.jhu.edu>
	<21295.9507091551@bessie.harlequin.co.uk>
	<3tpga5$upa@acs6.acs.ucalgary.ca> <3tpip0$rak@bdt.bdt.com>
	<DAVIS.95Jul12110320@passy.ilog.fr>
Date: Thu, 13 Jul 1995 08:33:14 GMT
Lines: 23

Harley Davis <davis@ilog.fr> wrote:
> By the way, has there been any work in implementations to statically
> resolve keywords in function calls?

Yes.

> If keywords are optional and unordered, it must be pretty inefficient
> to treat them completely at runtime.

It would be, wouldn't it.

> Is there a way to put keyword information in module
> interfaces for compile-time resolution?

I'm not sure what you're asking.  Module and library specifications are
just lists of imported and exported names.  But it's certainly possible
to include information about keywords in the library export information
that is dumped by a compiler;  what that data looks like is, of course,
not specified by the language.

As with most things in Dylan, sometimes you can't avoid fully dynamic
behavior, and that has run-time costs, but most of the time it is
possible to move a lot of the cost to compile-time.
