Why is the class heirarchy set up such that 2^(-1) gives and error in hugs and ghc? I don't know if there is some deep issue that requires there to be some exponential function with this property but this seems like a pretty serious bug. The obvios solution: expon x y = if y<0 then 1 / x^y else x^y has type (Fractional a, Integral b) => a -> b -> a So I gues my real question should be: Why dose Num not require Ord? Jeff [see Section 6.3.3 of the report -moderator]
- Re: 2^(-1) Jeff P. Burdges (Weasel)
- Re: 2^(-1) Lennart Augustsson
- Re: 2^(-1) Jerzy Karczmarczuk