[Haskell-cafe] Aren't type system extensions fun?

[Derek Elkins]

On Tue, 2008-05-27 at 00:19 +0100, Eric Stansifer wrote:

> Say, wouldn't a syntax like "(forall a => a -> a)" or "(a => a -> a)"
> or something similar to that be more consistent with syntax for
> contexts, e.g. "(Ord a => a -> a)"?

It's not remotely the same thing as a class constraint so why would we
want it to be "consistent" with it?