contification paper

[Matthew Fluet]

> Any opinion on whether I should compile kit.sml and mlton.sml with
> -no-polyvariance?

I've always done mlton with -no-polyvariance.  But I think I had kit
working without it before,

Why don't you do them both the same, whatever you decide.  I'm saying that
those two benchmarks compiled and executed on a different machine
anyways, it makes sense to have them the same, even if we use
-no-polyvariance for them.