[racket] the typed Values restriction

On Wed, Mar 18, 2015 at 5:11 PM Alexis King <lexi.lambda at gmail.com> wrote:

> It might be a good idea to have a special case for Values when producing
> error messages. Currently, even though the documentation mentions that
> Values is only valid as the return type of a function, the error messages
> can be a little confusing. For example, using Values as an expression
> produces the error “type name used out of context,” but it isn’t really a
> type.
>

Yes, this is a good idea.

Sam


>
> > On Mar 18, 2015, at 14:01, Sam Tobin-Hochstadt <samth at cs.indiana.edu>
> wrote:
> >
> > Type variables range only over types, and `(Values A B)` isn't a type.
> > If your program type checked, then the following program would too:
> >
> > (: my-force : (All (A) (-> A) -> A))
> > (define (my-force f)
> >  (let ([tmp : A (f)]) tmp))
> >
> > (my-force (lambda () (values (void) (void)))
> >
> > but this program has a runtime error with the wrong number of values.
> >
> > Sam
> >
> > On Wed, Mar 18, 2015 at 4:48 PM, Benjamin Greenman <blg59 at cornell.edu>
> wrote:
> >> Today I got a surprising type error that I think is worth sharing.
> >>
> >> (: my-force (All (A) (-> (-> A) A))
> >> (define (my-force x) (x))
> >>
> >> (my-force (lambda () (values (void) (void)))
> >> ;; ERROR! 'my-force' cannot be applied to argument.
> >> ;; expected "(-> A)", got "(-> (values Void Void))"
> >> ;; result type "A", expected result "AnyValues"
> >>
> >> (inst my-force (Values Void Void))
> >> ;; Parse Error in type: "Values" is unbound
> >>
> >> That is all. I see now that Values is only allowed in result positions,
> but
> >> I'd still expected this to work.
> >>
> >>
> >> ____________________
> >>  Racket Users list:
> >>  http://lists.racket-lang.org/users
> >>
> > ____________________
> >  Racket Users list:
> >  http://lists.racket-lang.org/users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20150318/80c5a1b9/attachment-0001.html>