Re: [Haskell-cafe] "Natural" polymorphism for n*(n+1)/2