[MLton-user] MLton performance

[Stephen Weeks]

> Are you interested in cases where SML/NJ outperforms MLton?

Sure.

> I have been playing around with some projects of my own in an
> attempt to learn SML and found that, for one program, SML/NJ runs in
> about 70% of the time of MLton

This is not unheard of, but it certainly is unusual, and so may be
worth looking into.

> Is there some profiling I can do before I dump 800 lines of badly
> written ML into your laps?

Yes, MLton's profiling is described at

	http://mlton.org/Profiling

I'm not sure what is the state of profiling for SML/NJ.