They were introduced as part of the System Fc rewrite.
The Fc approach has the benefit of unifying a lot of the work on GADTs, functional dependencies, type and data families, etc. all behind the scenes.
Every once in a while, (~) constraints can leak into the surface language and it can be useful to be able to talk about them in the surface language of Haskell, because otherwise it isn't clear how to talk about F a ~ G b style constraints, which arise in practice when you work with type families.
On Sun, Jan 11, 2015 at 6:04 PM, Dominic Steinitz <[hidden email]> wrote: