In the last few months I've been drifting toward the idea that parametric polymorphism is in fact enough for everything, and subtype polymorphism isn't needed a

In the last few months I've been drifting toward the idea that parametric polymorphism is in fact enough for everything, and subtype polymorphism isn't needed at all. Also relevant reading in this context: Haskell Antipattern: Existential Typeclass http://lukepalmer.wordpress.com/2010/01/24/haskell-antipattern-existential-typeclass/ http://lukepalmer.wordpress.com/2010/01/24/haskell-antipatte...