Subject: Re: a system for WWW applications to use heavyweight Scheme
From: rpw3@rigden.engr.sgi.com (Rob Warnock)
Date: 1997/06/12
Newsgroups: comp.lang.scheme
Message-ID: <5nni0i$33vb@tokyo.engr.sgi.com>

Bruce R. Lewis <brlewis@mit.edu> wrote:
+---------------
| > GJC, caught in the past as usual...
| 
| Feel free to stay in the past; R4RS was written in 1991.  With all the
| money I'm sure you're making from commercial WWW programming, you surely
| can afford to take a few days and bring SIOD into compliance.
+---------------

I've been looking at that, in my (non)copious spare time. What you would end
up with would almost certainly be substantially larger and slower than SIOD
is today. One of the main reasons is also one of your main complaints --
SIOD doesn't provide *any* of the R4RS numeric tower except reals, and even
its characters are just numbers (reals), too:

  > (mapcar putc '(72.561 101.9 108.35 108 111 44 32 87 111 114 108 100 33 10))
  Hello, World!
  > 

This is both good and bad. Good, because it cuts down interpretation overhead
(SIOD is a "pure" interpreter) and code size (for the interpreter); bad because
you don't get bignums and all the R4RS standard conversion routines (nor many
of the control structures, either, but you can fake most of them with macros)

When I need a "real" Scheme I use MzScheme or SCM or one of the several
available compilers. When I need a tiny Scheme script that runs in a
fraction of a second of elapsed time, I use SIOD.


-Rob

-----
Rob Warnock, 7L-551		rpw3@sgi.com   http://reality.sgi.com/rpw3/
Silicon Graphics, Inc.		Phone: 415-933-1673  FAX: 415-933-4392
2011 N. Shoreline Blvd.		[after 8/2/1997 ==> 650-933-xxxx]
Mountain View, CA  94043	PP-ASEL-IA