Clean up more lint

[3noch]

Some more changes that my linter (hlint) was complaining about but feel free to close if you're not a fan. I sort of wonder: do we really need Monad m constraints in all these places? Could Applicative suffice?

[ivan-m]

I'm happy with Applicative + fmap instead of Monad + liftM if you want to make those changes. Just need to put a minimum bound on base for that then.

[ivan-m]

I prefer parentheses even for singleton constraints (so they don't have to get added when another constraint is added).

[3noch]

Ok I'll add those back and do the other changes.

[3noch]

So if we go through with this, the entire module is now incorrectly named! This is definitely a backwards breaking change as well.

[ivan-m]

I'm happy with it still being named "Monadic" (at least for backwards compatability reasons); the main reason I wrote this was for my own usage of having a Stateful pretty-printer and that still operates/requires monadic aspects.

[3noch]

How do you feel about using base-compat's Prelude.Compat in the Monadic module instead of changing the bound on base?

[ivan-m]

Sure, I'm happy with that.

[3noch]

Ok I've pushed changes. I used Functor instead of Applicative where possible. You may want to push our previous changes to Hackage before merging this since this will need to be a major version bump.

[ivan-m]

I'm OK with a major version bump, especially as for anyone using a relatively recent version of GHC it's not going to be a problem (unless they were relying upon type coercion from the Monad constraint somehow).