[plt-scheme] Compiler optimizes Typed Scheme?!?

Of course, this can be viewed as an interoperability problem, and
Jacob's work is relevant there (POPL 2006).

Robby

On 2/26/07, Matthias Felleisen <matthias at ccs.neu.edu> wrote:
>
> On Feb 26, 2007, at 4:45 PM, Sam TH wrote:
>
> >>
> >> Humm, maybe I understand. You can't just mix typed and untyped scheme
> >> in the same module, can you?
> >
> > Right.  Typed or untyped is a module-level decision.
>
>
> Contrary to popular rumors it isn't clear yet what mixing typed and
> untyped code without boundaries means. (This is becoming known as
> "gradual typing" though I'm conjecturing that "complete confusion"
> would be much better. "-)
>
> As you may know and conjecture, we really try to do things here on
> very firm foundations. This foundation requires a boundary between
> the two kinds of code that allows the injection of contracts. Kathy
> Gray's work on Dynamic in Java is the closest in the direction of GT
> that I conjecture is truly sound.
>
> See Sam's Dynamic Languages Symposium/OOPSLA 2006 paper. -- Matthias
>
> _________________________________________________
>   For list-related administrative tasks:
>   http://list.cs.brown.edu/mailman/listinfo/plt-scheme
>