On 05/06/2015 5:36 PM, Matt Oliveri wrote:
Keean, the members of type classes are types, I believe, and triangles aren't types. You're on the right track, in that something just like type class resolution could propagate the rightness of a triangle. But it wouldn't actually be a type class because a triangle isn't a type.
I believe he's using using a generalization of type classes that serve as type functions, the way Habit uses them to prevent division by zero [1], ie. div :: (NonZero t t’) => t -> t’ -> t.
Sandro [1] http://hasp.cs.pdx.edu/habit-report-Nov2010.pdf _______________________________________________ bitc-dev mailing list [email protected] http://www.coyotos.org/mailman/listinfo/bitc-dev
