Components of a Hypermedia System

A chapter from the book Hypermedia Systems that lays out the theoretical foundation of the web as a hypermedia system. It decomposes a hypermedia system into four components — hypermedia (HTML), network protocol (HTTP), hypermedia server, and hypermedia client — then dives deep into Roy Fielding's REST architectural constraints as formalized in his doctoral dissertation.

The chapter walks through HTTP methods and response codes, the uniform interface constraint and its four sub-constraints (resource identification, representation manipulation, self-descriptive messages, HATEOAS), and the practical consequences for web application architecture. A key contribution is the HTML-vs-JSON comparison for the same endpoint, demonstrating concretely why self-describing hypermedia messages eliminate API versioning and client-server coordination overhead.

The authors argue that most JSON APIs are not RESTful in Fielding's original sense — and structurally cannot be — because they lack hypermedia controls. The chapter closes by positioning scripting (Code-On-Demand) as an optional but native part of REST, provided it augments rather than replaces the hypermedia model. An appendix on HTML5 semantic elements rounds out the practical advice.