On Thu, Mar 24, 2005 at 12:39:40PM +0100, Benjamin Franksen wrote:
On Thursday 24 March 2005 04:14, Thomas Hallgren wrote:
Yes, I think this should be fixed, and perhaps it could be done in a backward compatible way? If classes were allowed to declare default methods for superclasses, then you could have
class Functor f where fmap :: ... class Functor m => Monad m where ...the usual stuff... fmap = liftM
Then declaring
instance Monad T where ...
for some T, would implicitly introduce an instance Functor T, if it is not defined explicitly...
Robert Will has written a fully specified proposal for this. He calls it "delayed method definition", see http://www.stud.tu-ilmenau.de/~robertw/dessy/fun/, sections 4.3.1 and 4.3.2.
I'd prefer something more modest: classes could contain default methods for their superclasses, but only for a superclass with exactly the same arguments, as in class Eq a => Ord a class Functor m => Monad m so there'd be no need for a special device to specify which class the default method is for. This would lose things like the ability to specify general equality on Collections in the class class (Eq a, Eq (coll a)) => Collection coll a but I think that's stretching subtyping a bit. It would also be simpler to keep the H98 rule for instances, namely that an instance of class C can contain only definitions of methods of class C.