Overview

Structured analysis arose in the 1970s out of structured programming. The basic idea of structured analysis and structured programming is that the high-level structure of a software system should match the structure of the problem to be solved rather than the structure of the implementation technology that happens to be available. Despite its name, structured analysis offers techniques and guidelines to design a high-level system structure but not to analyze this structure. The most influential school of structured analysis is the Yourdon approach, in which DFDs, ERDs, and Mealy diagrams without local variables are used to represent the structure of the SuD.

Yourdon-style structured analysis does not include the environment in its models. However, for the design of reactive systems, environment models are crucial and, in this chapter, I discuss an updated version of structured analysis called Postmodern Structured Analysis (PSA), in which two environment models are produced: an ERD of the subject domain and an extended context model that includes the part of the world where the desired effects of the SuD are to be felt and the part that must be monitored by the SuD in order to achieve its desired effect. Another important difference between PSA and classic structured analysis is the particular brand of DFDs used, in which event flows can contain data. This is illustrated using the heating control example.

I call the version of structured analysis presented here "postmodern" because it combines design approaches from several schools, some of which oppose one another, namely Yourdon-style structured analysis, Jackson System Development, and object-oriented analysis. Section 20.1 lists the notations used by PSA and relates...