Andrew Cooke's
comment on the
subject:
One reason haskell "needs" macros is that it's an interesting area of design space. Macros / staged evaluation in a lazy, typed system sounds like an intersting problem. This is the "fly to the moon" argument - there may not be a simple practical case, but who knows what we might learn on the way?