McBreen points out that the discipline of Software Engineering was invented for extremely large software development projects -- large teams and long timescales -- and that is totally unsuited to smaller projects -- involving a few people and lasting a few years. Yet those smaller projects are the norm today. He instead proposes that such software projects should follow a craftsman approach, using the classic structure of apprentices, journeymen, and master craftsmen. He claims that this approach produces better software -- robust, high-quality applications that meet the customers' needs and will be used for years.
The book is filled with examples, supporting evidence, and advice. There are a lot of wise words about the importance of people, how they are all different, and how they should be trained and developed, respected, and made responsible for their work. In particular, the chapter on personal reputation versus licensing is thought-provoking. More surprisingly, perhaps, there is a lot about the importance of the developed software itself -- McBreen says it should be treated as a capital asset, something to be maintained and developed over a long time -- and he explores the consequences of this view.
The argument is a little repetitive in paces, but on the whole this is a refreshing look at a way better software could be produced, from a real practitioner who has thought deeply about his craft.
In this further book in the XP series, XP "outsider" McBreen takes a critical look at the whole XP debate, examining the hype, and the reality. He questions the process and the detractors equally. His answers may not satisfy the most extreme XPers, as he argues that their process is not appropriate for every software development project. But his answers will satisfy the detractors even less, as he argues quite forcefully that where it does work, it works very well indeed.
A good, thoughtful addition to the series.