Review By: Matt Gelbwaks
10/17/2008John Hunt's 2006 book, Agile Software Construction, is both a survey of popular agile development methodologies and a set of recommendations for his preferred approach--agile modeling with XP. The book is directed towards those who consider themselves a step above novice. The coverage of FDD, XP, and agile modeling are thorough and well composed. Other approaches (DSDM, Scrum, Prince2, RUP) are all briefly mentioned as well. The bottom line is this: If you think that agile modeling would help your team and your product, then reading this book should be a great start. Mr. Hunt clearly has both an affinity and a depth of experience in this domain and it shows in his prose. The book is sprinkled with little gems that push forward the state of the art and the art of the practices and definitely make it worth reading! A couple significant examples are his depiction of the concept of modeling walls to support pair-modeling and the expansion of trust-based planning.

Mr. Hunt's recommendation for a combined XP/Agile Modeling emphasis is both viable and remains current, even though the spectrum of agile methods, being agile themselves, continues to evolve and mature. The book's message is clear and well constructed, though I suffered through two disappointments.

My first issue had nothing to do with either the writer or the words, but rather the medium. Surveys carry the inescapable problem of being a data point stuck in time. Since this book was written in 2005 and published in 2006, there are several trends that are not covered and several recently published experience reports that would have strengthened his arguments. Though the monetization would be diminished, perhaps a better forum for this type of material would be a continuously evolving (or at least updated) Web site that focused on Mr. Hunt's principal message. It would also help to have links to all the background and ancillary information.

My second issue is probably a disappointment that Mr. Hunt most likely shared: the editing on the book is rather poor. I remember Springer-Verlag (the publisher) from my college days. Most of my textbooks were published by them. They have always struck me as one of the old standbys in the industry. Therefore, as I was reading Agile Software Construction I was initially bemused by the occasional obvious typos in the text. But as these typos continued to surface, with a rather substantial density, it became distracting and ultimately off-putting. Hopefully, this minor construction issue will not stand in the way of many readers.