Actually the abstractions they are talking about are explicitly for most of your use cases: - Transforming data from one structure to another. - Moving data a

Actually the abstractions they are talking about are explicitly for most of your use cases: - Transforming data from one structure to another. - Moving data around. - Writing and re-writing similar business rules and domain logic in different libraries and languages. - Writing 'adapters' to get one library to work with another. - Trying to find 'the right' data structure or algorithm. There are answers to these problems using mathematical structures common in Haskell: Functor, Applicative, Monad, and so on.