Re: Scheme & number crunching

[Matthias Felleisen <matthias@cs.rice.edu>]

If you're willing to do whole-program analysis, "Scheme" programs can beat
equivalent C number crunching programs. Both Gambit and Stalin (argh) have
proven this point over the years. The problem of infering decent types for
loop-ish programs is no harder than in ML. These Scheme implementations 
unbox, allocate in stack regions, call closures directly, etc. 

Having said that, I must admit that I am not sure Schemers should play this
game and I am sure that PLT won't play this game. I have watched this for
many years, and I just don't think this is going to give Scheme the break
it needs. 

We need to compete with the scripters and show that a good semantics
matters and that scripts with a good semantics can evolve into real
programs. 

And we need a killer app. 

-- Matthias