Oleg Grenrus

Just a small comment, when you have ```haskell data P a b = P a b ``` and you try to iso between ```haskell magical :: Iso (P a b)...

You'd need to pull in `semigroupoids` for the `Traverable1` / `Apply`.

@kcsongor in fact it isn't that heavy, ~20 modules and no extra dependencies, everything is depended by `profunctors` already. 

Hmm, one problem will be ```haskell lowerCurried :: Applicative f => Curried f a -> f a lowerCurried (Curried f) = f (pure id) -- we need pure ``` I'm...

I spotted this too. The text in `histo.pdf` says > This set is reduced by retaining only centroids with a count less than _4 \delta q (1-q)n_ ... The algorithm...

I see, we need to end section by starting a new one. And I agree, that sometimes examples aren't something we want to hide by default. I'd make *instances* section...

nice!

I'd rather do more general instance: ```haskell instance (All SListI a, f ~ I) => Generic (SOP f a) where .... ``` I don't think that `Generic (SOP f a)`...

doctests are indeed broken due changed `Show`

> iToK exists as mapIK. It's not, compare: ```haskell iToK :: (a -> b) -> (I -.-> K b) a mapIK :: (a -> b) -> I a -> K...