Search in Book Reviews

Most of the book is taken up by an 11-step plan for testing. It covers the whole software lifecycle, starting with testing the project plan and estimates and ending with testing the effectiveness of the testing process. Each step has a chapter to itself and the author has attempted to make each chapter self- contained. Unfortunately this results in a lot of duplication: I am left with the feeling that the material could have been covered in half the space.

The book is packed with checklists, worksheets and N-step procedures for each stage of testing. At first sight it seems impressively detailed, but a closer inspection shows that most of the material is just a repetition of generalities. Where the book does go into specifics, the practices seem to be straight out of a large-scale 1970s Information Systems development project. A few chapters at the end attempt to cover more modern material, such as client-server, RAD, Web applications, off-the-shelf packages, multi-platform and data warehouses.

I turned with interest to the section on testing web applications: it told me that I should test security, performance, correctness, compatibility, reliability, data integrity, usability and recoverability. I agree that these tests are necessary, but they are equally necessary for most other types of application. Just about the only piece of advice specific to web-based applications was that you need to test with different browsers; any book on HTML will tell you this and will probably go into far more detail about the differences to test for.

I cannot recommend this book. Even if the level of bureaucracy it recommends is suitable for your organisation, you should still be able to find another book which has a higher ratio of specifics to generalities and which takes into account at least some of the developments of the last 20 years.