Communicating Statecharts (CSC)

Abstract

Concurrency is increasingly gaining importance due to the rapid development of networked applications. However, concurrency comes with some complexities, including handling race conditions or deadlocks. Therefore, learning this concept and understanding its correct implementation is challenging, even for experienced programmers. This problem arises from the current practices of teaching concurrency because of the focus on confusing details instead of necessary concepts. We propose a new concurrency paradigm called Communicating StateCharts (CSC) to simplify the teaching of concurrency to beginner programmers. CSC preserves five main principles, aiming to make concurrency easier to learn and use for novices: software visualization, Model-Driven Development (MDD), pure functions, separation of concerns, and raising abstraction levels. In this regard, CSC adapts features from existing concurrency models that aligned with our principles, namely process calculi, the actor model, and Harel’s statecharts. This synthesis led to CSC’s atomic statecharts, communicating through messages transmitted via channels. To make CSC accessible for beginners, a visual MDD tool called CSCDraw is designed and developed. The main requirements that guided the design of CSCDraw include enforcing CSC principles, considering beginner-friendly features, ensuring faithful code generation, supporting conditional branches, and channel cardinality. We also present the design of a pilot study that investigates the most effective way of teaching CSC to beginning programmers. This study serves as a prelude to a more rigorous experiment to compare the effectiveness of CSC with the existing paradigms.