We finally discuss Otis' workflow. Some time is spent on the dreaded correlated error terms. Some abuse is hurled at econ departments and stata.