we can now say:

data T

for an empty datatype what about:

data T deriving (Show) -- or Eq, Ord, ...

personally I think this should be legal. the actual semantics are irrelevant, as far as I can tell, since you won't ever have something of type T other than _|_ and so you could simply define the derived instance of equals to be something like:

instance Eq T where { (==) = undefined }

though

instance Eq T where { _ (==) _ = True }

seems also to be resonable. this isn't a big deal, but i'd like it fixed. ghc panics when you try to derive something. hugs gives you a syntax error and i can't get nhc to accept empty datatypes at all... - hal -- Hal Daume III "Computer science is no more about computers | hdaume@isi.edu than astronomy is about telescopes." -Dijkstra | www.isi.edu/~hdaume