Isotone optimization is formulated as a convex programming problem with simple linear constraints. A R implementation of a particular active set strategy is discussed, and applied to various isotone optimization problems important in statistics. The implementation is user-extendable, and handles a great many convex loss functions and partial orders.