On 21 February 2015 at 21:39, Matt Oliveri <[email protected]> wrote:

>
> > Choose one gets the following type I think:
> >
> > fn (fn 'a 'b -> 'c) (fn 'a 'b -> 'c) (fn 'a 'b -> 'c) -> 'c
> >
> > Inferred from the application of 'f'.
>
> That can't be right. The third argument will be a bool. Did you even
> read the message I referred to? Rhetorical question. I'm fed up for

now.
>

Based on this:

On Mon, Feb 16, 2015 at 2:57 PM, William ML Leslie <
> [email protected]> wrote:
> I'm a bit uncomfortable with what happens in the case of let-bound
> native-arity-polymorphic case:
> let f = choose_one x y z -- different arities​
> in f n m
> ​​
> How do you choose a native arity for the callsite?


I don't see a bool argument? Then again I have no idea what the type of
choose_one is so I just guessed that it randomly picks on one x y and z
that each would have different arities inferred from their definitions.


> I'm done. We're not getting anywhere until you put more effort into
> making a self-contained explanation of your proposal that is not
> woefully incomplete.
>

It's a shame, I thought we were getting somewhere, and you made some very
useful contributions.

Maybe someone else would like to take a turn coaxing a real proposal
> out of Keean.


Well, its probably wasn't a very good idea in the first place. Still I
think I might implement it and see how it behaves.


Cheers,
Keean.
_______________________________________________
bitc-dev mailing list
[email protected]
http://www.coyotos.org/mailman/listinfo/bitc-dev

Reply via email to