Thanks for the tips, Isaac. I fixed my Ord decl. Too bad about the
run-time time & space overhead. I guess that's not fixable. Perhaps Simon
will decide to follow Pepe's suggestion of liberal deriving when
UndecidableInstances enabled.
Cheers, - Conal
On 8/21/07, Isaac Dupree <[EMAIL PROTECTED]> wrote:
>
> Conal Elliott wrote:
> > Thanks, Simon. The manual deriving is easier than I expected, by
> > boilerplate delegation of constraints & methods, as below. You may want
> to
> > give such an example when you document the change. Cheers, - Conal
> >
> > -- | Pairing of type constructors
> > newtype (f :*: g) a = Prod { unProd :: (f a, g a) }
> > -- deriving (Show, Eq, Ord)
> >
> > -- Deriving no longer works on types like :*:. So:
> > instance (Show (f a, g a)) => Show ((f :*: g) a) where
> > show (Prod p) = "Prod " ++ show p
> > instance (Eq (f a, g a)) => Eq ((f :*: g) a) where
> > Prod p == Prod q = p == q
> > instance (Ord (f a, g a)) => Ord ((f :*: g) a) where
> > Prod p < Prod q = p < q
>
> Hmm, the Eq and Ord instances won't have the efficiency of newtype
> deriving of sharing the Eq and Ord dictionaries with the (,) type to
> which the operations are delegated. In fact I think that Ord definition
> is not correct because it didn't happen to choose one of (<=) and
> (compare), Ord's minimal definition (and even if it did, it's still more
> slightly-runtime-inefficient not to implement all those methods,
> depending on (f a) and (g a)'s Ord instances)
>
> Isaac
>
_______________________________________________
Cvs-ghc mailing list
[email protected]
http://www.haskell.org/mailman/listinfo/cvs-ghc
