Closed isovector closed 4 years ago
I don't think that Semigroup
for a :-> b
is implementable. One can make unlawful instance by only applying Monoid
to the a -> b
parts of Fun
, but then one could first applyFun
and then <>
. Can you motivate the need?
I think my motivation is misinformed here. I wanted to use Fun
as a denotational model, but realized this should be an implementation detail of my property tests, not of my model. Cheers!
It would be great to have
Semigroup
andMonoid
instances forFun
.