b'
I continue the discussion of Mitchell's paper Type Inference with Simple Subtypes.\\xa0 Coming soon: a discussion of semantics of subtyping.