Newsgroups: comp.lang.lisp
Path: cantaloupe.srv.cs.cmu.edu!bb3.andrew.cmu.edu!newsfeed.pitt.edu!godot.cc.duq.edu!news.duke.edu!news.mathworks.com!newsfeed.internetmci.com!in2.uu.net!world!wware
From: wware@world.std.com (Will Ware)
Subject: Re: Future of LISP for anything
Message-ID: <Dr0ts9.4ru@world.std.com>
Organization: The World Public Access UNIX, Brookline, MA
X-Newsreader: TIN [version 1.2 PL2]
References: <4lnrou$h06@crchh327.rich.bnr.ca> <JIM.96May2151830@vapor.Franz.COM> <JIM.96May3102529@vapor.Franz.COM> <4mh807$de8@dfw-ixnews1.ix.netcom.com> <831306063snz@wildcard.demon.co.uk> <4mkq7n$ooq@goanna.cs.rmit.edu.au> <831406871snz@wildcard.demon.co.uk>
Date: Tue, 7 May 1996 05:55:21 GMT
Lines: 52

I've been watching this thread progress for a couple of weeks now. My
background is that I studied Lisp for a short time in college, then
started working as a hardware engineer so I didn't see it again for about
15 years. In the last ten years, I've been doing increasing amounts of C
programming. Now I'm rediscovering Lisp thanks to CLISP running on my
Linux box. I'm delighted with the progress that Lisp has made in that
time, and I'd love to see Lisp gain popularity. There's probably a lot of
great software that isn't being written because people don't know Lisp,
and think they should avoid it for one reason or another. 

It would certainly help if Lisp was getting the kind of financial backing
that Java and C++ are enjoying, but that doesn't seem to be in the cards.
Looking at the gradual ascension of Linux, I wonder if there isn't
something that could be done at a grassroots level. One place to start is
to look at what are widely (but not accurately) perceived as shortcomings
of Lisp.

(1) The idea that Lisp is hard or obscure or confusing. These days I write
C for work and Lisp for fun, and the Lisp is easier to read and write. I
haven't invested lots of hours, and there was no traumatic learning curve.
I had to learn some new stuff about Common Lisp, but no big deal. If Lisp
is obscure, it's only because it isn't widely known. 

(2) The idea that Lisp is memory-greedy and slow. I've used CLISP on a DOS
machine with 4 megs of RAM, and I didn't have any memory or performance
problems. I didn't try to consciously benchmark it against an equivalent C
program, but I wasn't bothered by slowness. I've read that if one does
invest a lot of time, one eventually learns to write Lisp code that runs
about as fast as C, and some day I'd like to do that.

(3) The idea that Lisp is incomplete as a language, and lacks features
required to get real work done. This might have been a valid complaint
once, but not with all the stuff that's new in Common Lisp, e.g. access to
the file system. 

Something that might help would be if there were more tutorial material
around for Lisp. Web pages would be particularly good for this. The web
pages I've seen on Lisp are mostly resource lists (some very good ones,
but not tutorials). I've found Paul Graham's books really useful myself,
and if there were web pages like that, it would be great. I also found
that CLISP came with a useful file called LISP-tutorial.txt.

Big splashy apps like Emacs and Autocad definitely have their place in
keeping Lisp alive, but to somebody new to Lisp, huge things like that are
probably intimidating. A large number of smaller, simpler apps might do
more to win peoples' affection. One simple project I did when I was
getting started with CLISP was to replace the spreadsheet I used to
compute my monthly finances with a Lisp program. 
-- 
-------------------------------------------------------------
Will Ware <wware@world.std.com> web <http://world.std.com/~wware/>
PGP fingerprint   45A8 722C D149 10CC   F0CF 48FB 93BF 7289
