Statecharts as a Logic of Effects

Do your programs only compute pure functions of data, or do they also perform effects such as dynamically reading input, writing output, transitioning database state, making network requests, etc.?

One sense of the term “logic” is as a general subject, i.e. the study of how to draw valid conclusions. Another sense is as a particular language, i.e. a logic for representing information along with a set of inference rules that specify exactly what conclusions can be drawn from what premises.

To ensure the reliability of your functions and of your data at rest, you may use type annotations, assertions, database system affordances such as schema declarations, etc. Ideally, these tools help you express both your domain logic and your application-specific logic.

Statecharts1 are a formal extension of finite state machines for modeling stateful, reactive systems. Extensions include guarded transitions, managed effects, extended state (context), orthogonal (parallel) states, and hierarchical (nested) states. The formalism has been standardized as SCXML. XState is a popular implementation that adheres to SCXML while providing a JSON syntax and execution via JavaScript/TypeScript.

Statecharts help to ensure a logic of effects. Effects are managed and executed by the system; you needn’t execute them within your functions, i.e. as “side-effects” of a function’s transformation of input to output. In the taxonomy of statechart effects, there are “fire-and-forget” effects such as actions and activities, as well as “invoked” effects such as promises, callbacks, observables, and other machines that can send and receive events to/from the invoking statechart.

Crucially, a statechart interpreter cancels the execution of / discards the results of effects when there is a finite state transition – this logic is managed by the system. And unlike finite state machines, a statechart can have “extended state”, i.e. context. In other words, you can have an “infinite” state space as long as you carve your universe of state into a finite number of partitions; you might say that a statechart is a partitioned-state machine. This partitioning facilitates a logic of managed effects.

This post was adapted from a note sent to my email list on Scientific Data Unification.
I'd love for you to subscribe.

  1. Harel, David. “Statecharts: A Visual Formalism for Complex Systems.” Science of Computer Programming 8, no. 3 (June 1, 1987): 231–74. ↩︎