On Fri, Jul 04, 2008 at 02:29:24PM +0100, Ian Lynagh wrote:
Also, currently it derives Data.Typeable, which is unportable, but we can easily work around that. The only extensions that I don't think that we can do without are ExistentialQuantification and Rank2Types. DeriveDataTypeable makes the implementation easier, and DeriveDataTypeable and PatternSignatures make using it easier.
If Typeable is going to be in the language, then I think 'deriving Typeable' should not only be included, but be the _only_ way to get Typeable instances. There is a lot of room for compiler specific representations of TypeRep's and a lot of codes invarients can be broken by nasty Typeable instances. Though, this is more a haskell' issue. But I have no qualms about using DeriveDataTypeable in code if I am going to be using Typeable at all. John -- John Meacham - ⑆repetae.net⑆john⑈