ABSTRACT
In this chapter, we build a context for Part IV, in which we examine integration and system testing for traditional software. Our immediate goal is to identify what we mean by these levels of testing. We took a simplistic view in Chapter 1, where we identified three levels (unit, integration, and system) in terms of symmetries in the waterfall model of software development. This view has been relatively successful for decades; however, the advent of alternative life cycle models mandates a second look at these views of testing. We begin with the traditional waterfall model, mostly because it has enormous acceptance and similar expressive power. To ground our discussion in something concrete, we switch to the automated teller machine example.
