The framework of promise theory offers an alternative way of understanding programming models, especially in distributed systems. We show that promise theory can express some familiar constructs and resolve some problems in program interface design, using fewer and simpler concepts than the Unified Modelling Language (UML).