On Mon, Apr 9, 2012 at 12:10 AM, wren ng thornton <[email protected]> wrote:

> In the BitC context there's the
> additional constraint about making allocation explicit. I'm not sure
> what sorts of issues that'd introduce for the newtype approach. Does/did
> BitC even have newtypes as such? They're one of my favorite features of
> Haskell, but even Mark Jones dropped them from Habit.
>

BitC didn't have newtype, though it isn't clear that newtype really needs
to be deeply primitive aside from the desire for some semi-automated helper
functionality introduced by the front end. I did plan to add it, but it
didn't seem pressing.

I'm not aware that any new allocation issues are introduced by newtype.
_______________________________________________
bitc-dev mailing list
[email protected]
http://www.coyotos.org/mailman/listinfo/bitc-dev

Reply via email to