[bitc-dev] bitc and call/cc

[Jonathan S. Shapiro]

On Wed, 2005-08-10 at 23:57 -0400, Swaroop Sridhar wrote:
> Should Bitc consider reifying continuations?

The short answer: no.
The long answer: hell no.

The problem with call/cc is that your example:

> (begin
>    (call/cc (lambda k (set! f1 k)))
>    (f1 ())
> )

is completely impossible to analyze. In any case, one does not need to
do so -- with care, the assembly sequence that you wrote can be done in
strictly first-order BitC.

shap