Ray Dillinger scripsit:

> The problem is that the two different models of non-local control
> flow (exceptions, continuations) don't really play nice with
> each other.

I continue to be baffled by this line of reasoning.  Raising an exception
in the R6RS/R7RS model (which is closely related to Common Lisp's)
does *not* in itself constitute non-local control: it merely invokes a
procedure which is the (current) value of a parameter.  If that procedure,
the exception handler, chooses to execute a non-local return, it must do
so by way of a captured continuation, there being no other way to do it.

(It's possible that under the covers the `guard` macro uses some other
mechanism such as one-shot and/or delimited continuations that are
available in the implementation.)

--
We call nothing profound                        cowan@ccil.org
that is not wittily expressed.                  John Cowan
        --Northrop Frye (improved)

_______________________________________________
Scheme-reports mailing list
Scheme-reports@scheme-reports.org
http://lists.scheme-reports.org/cgi-bin/mailman/listinfo/scheme-reports