>>=

I have an interest in something that would be like `TupleX` is to `Tuple2` if that's what you're thinking (for coproduct derivation). If it was flat, like tuple is, that...

btw, it would have the best perf if an `EitherX` could masquerade as an `Either(X + 1)`, associating from the right.

shapeless coproduct is just as inefficient as nested disjunctions... one object allocation for each element. Something like this ``` class CoproductN[A1, A2, A3 ...] { private[this] val: Any = ......

that's pretty much what I'm currently using (although I'm using actual tuples for the product case). It just feels like there is a huge perf sitting there in the coproduct...

@edmundnoble that's pretty much exactly what I'm suggesting ;-) but pregenerating for coproduct is a question of "when do we stop"? 22 is too low.

But dwarf isn't even my favourite doc! 

My eyes are acting up today so it's hard to give this a review... is this a typeclass generalisation of corecursive data structures like corecursivelist?

Sweet. All we need now is CanActuallyBuildFrom

I agree that Travis is **Problematic**.

I'd quite like it if there was tighter scalaz integration to be honest, e.g. if the relationship between a type and its generic repr was an `Isomorphism` and we could...