Blackwell Optimality in Denumerable Markov Decision Chains

We consider Markov decision chains with a denumerable state space E, compact metric action sets and the usual continuity properties of the transition probabilities and (possibly unbounded) immediate rewards.