[MLton] HOL's Moscow ML implementation, and pushing MLton to emulate it

[Daniel C. Wang]

Stephen Weeks wrote:
{stuff deleted}
> Basically, I think the problem comes down to figuring out a way to
> represent all of the types known at compile time using the host
> representation, to embed these in the interpreter's universal type
> (the exception trick isn't really necessary since we've got the whole
> program), and to figure out how to make the interpreter always use the
> right representation of a type so that there are no coercions.  

Let me rewrite the above as..


 > Basically, I think the problem comes down to figuring out a way to
 > represent all of the types known at compile time using an *unboxed*
 > representation, to embed these into a universal *boxed* typed, and to
 > figure out how to make the *polymorphic function* always use the
 > right representation of a type so that there are no coercions.

See the below for a pretty good solution....
   http://flint.cs.yale.edu/flint/publications/flex.html