Newsgroups: comp.lang.lisp
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!udel!gatech!newsfeed.pitt.edu!uunet!sytex!smcl
From: smcl@sytex.com (Scott McLoughlin)
Subject: Re: Common Lisp' dual name space
Message-ID: <Ji47uc1w165w@sytex.com>
Sender: bbs@sytex.com
Organization: Sytex Access Ltd.
References: <39927h$e3q@tools.near.net>
Date: Thu, 3 Nov 1994 03:28:42 GMT
Lines: 30

barmar@nic.near.net (Barry Margolin) writes:

> In article <CynJ44.1A0@cogsci.ed.ac.uk> jeff@aiai.ed.ac.uk (Jeff Dalton) writ
> >Just because we have a name doesn't mean we have a namespace.
> 
> Sure it does.  A namespace is any mapping from names to bindings.
> 
> Not all of Common Lisp's namespaces permit there to be lexical nesting of
> namespaces, but that doesn't mean they aren't namespaces.  For instances,
> there are package and type namespaces, but just a single, global instance
> of each.

Howdy,
        I was about to thank Jeff Dalton for clarifying this
point, but I guess I'll wait ;-) (BTW, a buddy of mine thinks
the "seven namespaces" list is in Norvig, but I haven't found
the quote yet.)
        Right now, I am most concerned about DEFTYPE and what
macros and/or the compiler can assume about types so defined.
A function can be declared "inline". Should I think of DEFTYPE
as defining predicates that can be inlined or is this a 
no-no. My _guess_ is that I can, because types will presumably
be viewed as gospel by a compiler when used in DECLARE forms.
But this is only a guess.
        Anyway, any help would be appreciated.

=============================================
Scott McLoughlin
Conscious Computing
=============================================
