Skip to content

Latest commit

 

History

History
92 lines (56 loc) · 6.31 KB

index.md

File metadata and controls

92 lines (56 loc) · 6.31 KB

Welcome to the (unfinished) world of Statecharts

What is a statechart?

A statechart can be explained in many ways, and we'll get to those explanations, but essentially, a statechart is a drawing. Here's a simple statechart:

A simple statechart

However, this drawing might not be so useful for software engineers who want to reap the benefits outlined elsewhere on this site, so let's dive into some other ways of describing what a statechart is. The original paper that defines statecharts bills them as "A visual formalism for complex systems" (Harel, 1987). With that out of the way, let's try to explain statecharts.

Introduction to statecharts

Put simply, a statechart is a beefed up state machine. The beefing up solves a lot of the problems that state machines have, especially state explosion. One of the goals of this site is to help explain what statecharts are and how they are useful.

Why should you use statecharts?

Statecharts offer a surprising array of benefits

Why should you not use statecharts?

There are a few downsides to using statecharts that you should be aware of.

What are the main arguments against statecharts?

There are a few common arguments against statecharts.

The benefits outlines above should make it clear that the introduction of statecharts is generally a net positive.

How do you use statecharts?

First of all, know that a W3C committee spent 10+ years (2005 to 2015) standardizing something called SCXML (yes, Statechart XML), and that it defines a lot of the semantics and specifies how to deal with certain edge cases. There are tools to read, author and even execute statecharts written in SCXML, in various languages. There are also some derivatives that support the same model as SCXML, but using a different syntax.

Additionally, there are statechart libaries for a variety of platforms, that in varying degrees support the semantics described by SCXML. You should consider using these libraries just to get those edge cases taken care of. The libraries generally perform entry and exit actions in the right order and so on.

With that out of the way, read on!

Executable statecharts

In addition to just using statecharts to model the behaviour in documents separte from the actual running code, it's possible to use one of various machine formats, both to design the behaviour, and at run-time to actually be the behaviour. The idea is to have a single source of truth that describes the behaviour of a component, and that this single source drives both the actual run-time code, but that it can also be used to generate a precise diagram that visualises the statechart.

This carries along some different pros and cons:

Why should you use executable statecharts?

  • No need to translate diagrams into code
  • No bugs introduced by hand translation of diagrams
  • The diagrams are always in sync
  • The diagrams are more precise

Why should you not use executable statecharts?

  • The format and tools for executable statecharts is limited
  • Type safety between statechart and the component is hard to enforce

How do you use executable statecharts?

In essence, if you have any definition of a statechart in your code, all you need to do is to take that representation and automate the generation of the visual statechart. This is of course simpler when the definition is in a separate file, e.g. in a JSON or XML file.

Anything else?

If you feel like chatting to someone about statecharts, you should go to spectrum.chat/statecharts, where you'll find a community of like minded developers that can help you understand and reap the benefits of using Statecharts.

Quite a few people have written books or held presentations that deal with statecharts in various ways, and they're included in our resources page.

There are some pages that haven't found any place in the web of documents, so they're honourably mentioned here:

Acknowledgements