Testing the Machine in the World

Abstract

A central aim of software testing is assurance of functional correctness and dependability. For many software-intensive systems, including administrative, embedded, enterprise and communication systems, functional correctness means achieving the desired effects in the physical world, and dependability means dependability of those effects. For example, an administrative system for a lending library is required to ensure that only members can borrow books; that a member who has reserved a book and has been notified that the book is now available for collection in the library will not find that it has been lent to another member; that the catalogue gives reliable information about what is currently on the shelves, and so on. Similarly, a system to control a lift must ensure that the lift comes when summoned and takes the user to the desired floor; that the building manager can specify service priorities in terms of express lifts and time-dependent needs such as rush hours at the beginning and end of the working day; and that failure of the mechanical equipment does not endanger life.