I really appreciate they have a "reproduce the results" section. We need a lot more of that kind of thing in CS research.