On Wed, Feb 25, 2015 at 9:12 AM, Jonathan S. Shapiro <[email protected]> wrote:
> On Wed, Feb 25, 2015 at 12:26 AM, Matt Oliveri <[email protected]> wrote:
>>
>> I don't know if you think that issue has been resolved, but in my
>> mind, it hasn't. You still haven't given a correct account of how
>> upcast1_1to2 gets implemented.
>>
>> (You may have noticed that I also challenged Shap to implement it.)
>
> I think shap agreed yesterday that it can't be done using the primitives we
> have on hand.

I missed it. But I think Keean's proposal would allow that. Certainly
subtyping would, using Keean's order as the basis of the subtyping
relation. This is why I am surprised you think subtyping and
constraints would amount to the same thing.
_______________________________________________
bitc-dev mailing list
[email protected]
http://www.coyotos.org/mailman/listinfo/bitc-dev

Reply via email to