Newsgroups: comp.lang.lisp
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!europa.eng.gtefsd.com!howland.reston.ans.net!math.ohio-state.edu!jussieu.fr!univ-lyon1.fr!swidir.switch.ch!newsfeed.ACO.net!Austria.EU.net!EU.net!uunet!sytex!smcl
From: smcl@sytex.com (Scott McLoughlin)
Subject: Re: Common Lisp' dual name space
Message-ID: <Xo0aVc1w165w@sytex.com>
Sender: bbs@sytex.com
Organization: Sytex Access Ltd.
References: <CyrvCG.8wH@cogsci.ed.ac.uk>
Date: Sat, 5 Nov 1994 09:32:32 GMT
Lines: 30

jeff@aiai.ed.ac.uk (Jeff Dalton) writes:

> I will, however, note that types are not 1st-class objects in CL.
> 
> If you write a CL interpreter (say), the "environment" has to deal
> with only 4 namespaces (unless X3J13's added another one while I
> wasn't looking).  There are, of course, some kinds of objects that
> have names (e.g. packages).  There's typically a mapping from the
> names to the objects.  But I regard that as essentially a separate
> issue.
> 
> -- jeff

Howdy,
        Yes. When I read Jeff Dalton's previous post on this
issue, I too thought of "when writing an interpreter/compiler
what environments do I need" situation.  You need 4 to handle
the environments where nested bindings can shadow one another.
        OTOH, I'm not at all comfortable with the idea that
lexical shadowing is a requirement of anything we might want
to call a namespace.
        Anyway, it seems we agree that CL is at least a
LISP-4, so my original point that LISP-2 is a misnomer
wasn't out of place. Recently someone on comp.lang.scheme
used the term "Lisp-N", and I was pleased ;-)

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