Dean Herington: > I strongly prefer that that behavior _not_ be provision of any > normal Int value.
I agree. I think a operation to let you examine the (possibly) underlying 2s complement representation without raising an exception should be quite separate from the operation to convert between signed and unsigned values possibly raising an exception even though they have the same type. (Note that they'd have different types if Haskell followed Java's example and made exceptions part of the type.) -- Alastair Reid [EMAIL PROTECTED] http://www.cs.utah.edu/~reid/ _______________________________________________ Haskell mailing list [EMAIL PROTECTED] http://www.haskell.org/mailman/listinfo/haskell
