Modern Byzantine fault-tolerant state machine replication (BFT) protocols involve about 20.000 lines of challenging C++ code encompassing synchronization, networking and cryptography. They are notoriously difficult to develop, test and prove. We present a new abstraction to simplify these tasks. We treat a BFT protocol as a composition of instances of our abstraction. Each instance is developed and analyzed independently. To illustrate our approach, we first show how, with our abstraction, the benefits of a BFT protocol like Zyzzyva could have been obtained with much less pain. Namely, we develop AZyzzyva, a new protocol that mimics the behavior of Zyzzyva in best-case situations (for which Zyzzyva was optimized) using less than 24% of the actual code of Zyzzyva. To cover worst-case situations, our abstraction enables to compose AZyzzyva with any existing BFT protocol, typically, a classical one like PBFT which has been proved correct and widely tested. We then present Aliph, a new BFT protocol that outperforms previous BFT protocols both in terms of latency (by up to 30%) and throughput (by up to 360%). Development of Aliph required two new instances of our abstraction. Each instance contains less than 25% of the code needed to develop state-of-the-art BFT protocols.