Newsgroups: comp.lang.scheme
Path: cantaloupe.srv.cs.cmu.edu!das-news.harvard.edu!news2.near.net!MathWorks.Com!udel!news.sprintlink.net!uunet!sytex!smcl
From: smcl@sytex.com (Scott McLoughlin)
Subject: Re: Scheme _in_ Emacs?
Message-ID: <8myisc1w165w@sytex.com>
Sender: bbs@sytex.com
Organization: Sytex Access Ltd.
References: <LORD.94Sep11153553@x1.cygnus.com>
Date: Mon, 12 Sep 1994 05:33:54 GMT
Lines: 33

lord@x1.cygnus.com (Tom Lord) writes:
> Yuck.  One extra argument should not slow function calls down that
> much.  Enclosing code in LAMBDA should not make debugging awkward.
> This sounds like a clear case of programming style being distorted
> needlessly by deficiences in language implementation.
> 

Howdy,
        Hear, hear! Every time I pick up a PD language, I run
a version of the TAK benchmark.  Heap allocating argument lists
unecessarily _or_ not providing the GC machinery to deal with
this practice becomes readily apparent.
        PD implementations of "dynamic languages" (better term?)
on the PC are getting better. CLISP (Common Lisp) and CAML Light
are pretty nice. They don't support Windoze, but that'l come when
Chicago/Daytona are finally released.
        It's interesting to note that both of these come out of
Europe rather than the U.S.  The nifty implementations that I
know of in the U.S. (CMUCL, SML/NJ...) are stuck in the "Unix
ghetto" (actually, I like Unix - just not its dominance in some
areas to the exclusion of the 486/8meg box me and just about a
gadzillion other folks have at work and home).
        Anyway, I'd love to see a < $200 Scheme system that ran
under Windows 3.1, had a FFI, had a nice Windows based GUI
API already canned and some STRUCTURE and MODULE system and, oh
yeah, a finalization scheme for foreign data in Programmer's
Paradise or advertised in Dr. Dobb's Journal.
        Oh well, a man can dream....

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