On Sat, Apr 4, 2015 at 4:40 AM, Keean Schupke <[email protected]> wrote:
> Here's a slightly modified version of the arity-inference algorithm, with
> the distinction that we now have a kind "fn" with possible values "call" and
> "arg", to make the interpretation of arrow types a bit clearer: Everything
> else is as before.

I like "call" and "arg" instead of "y" and "n" or "true" and "false".
But I don't think we should call the the kind "fn" (or "Fn") because
it's not a function kind. And I don't want "call" or "arg" in the
surface syntax; it's too long, and it looks like an identifier.
_______________________________________________
bitc-dev mailing list
[email protected]
http://www.coyotos.org/mailman/listinfo/bitc-dev

Reply via email to