Joe Fasel argued for the inclusion of S or W in the prelude on the grounds that a complete combinator base would be "neat". But the majority of the Haskell committee didn't buy that. -- Lennart Peter G. Hancock wrote:
Lennart Augustsson wrote (on Mon, 14 Feb 2005 at 14:55):
> Any definition can be made point free if you have a > complete combinator base at your disposal, e.g., S and K.
> Haskell has K (called const), but lacks S. S could be > defined as > spread f g x = f x (g x)
> Given that large set of Haskell prelude functions I would > not be surprised if spread could already be defined point > free in Haskell. :)
It sometimes surprises me the prelude doesn't have
diag f x = f x x
(aka W. It already has B, C, K and I: (.), flip, const and id.)
Peter Hancock _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe