On Thu, Jul 15, 2010 at 12:42:41AM +0100, Thomas Schilling wrote: > > simplifications are possible. To make this efficient, the solver also > regularly canonicalises constraints. E.g., function symbols go to the > left and constructors to the right.
One minor correction: the canonicalisation of constraints actually has more to do with ensuring that the solving process terminates than it does with efficiency. -Brent _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
