Subject: Re: Lisp or Scheme?
From: rpw3@rigden.engr.sgi.com (Rob Warnock)
Date: 13 Jan 2001 13:25:40 GMT
Newsgroups: comp.lang.lisp
Message-ID: <93pl0k$nlllm$1@fido.engr.sgi.com>
Marco Antoniotti  <marcoxa@cs.nyu.edu> wrote:
+---------------
| Eli Barzilay <eli@barzilay.org> writes:
| >   (define (foo) (sleep 10) (printf "Hey!~%"))
|                    ^^^^^      ^^^^^^
| 		Not in R^nRS! Ditto!
+---------------

As I said in a previous message, "Scheme people" don't *care*!!!
It's only Common Lisp people that seem to get so bent out of shape
by the fact that the porduction implementations of Scheme are chock
full of stuff that's not in R5RS...

And where in CLHS, pray tell, are sockets, CLX/CLIM, threads, fork/exec,
FFI, and all the other stuff you need to add to CL to make *it* useful
in a production systems environment?

So enough already with this "not in R5RS" strawman, o.k.? The Scheme
standard is very, very small. The Common Lisp standard is very, very big.
We all know that. But neither standard is big enough to cover everything
most of us need for "real work". We all know that, too.

The solution in *both* cases is to pick an implementation that has
the things you need, and learn to live in/with that implementation
(and try *somewhat*, though don't be silly about it, to structure
your abtractions to if you *have* to change implementations it won't
be *too* painful). Since more of any "useful CL" is "standard", that
will almost always mean that the pain of moving between CL implementations
is less than the pain of moving between Schemes. So what? Big > little.

So please, let's stay away from the strawmen & cheap shots, o.k.?
There are probably plenty of serious issues we could talk about.
E.g., how about things where neither Scheme nor CL gets it quite right?


-Rob

-----
Rob Warnock, 31-2-510		rpw3@sgi.com
SGI Network Engineering		http://reality.sgi.com/rpw3/
1600 Amphitheatre Pkwy.		Phone: 650-933-1673
Mountain View, CA  94043	PP-ASEL-IA