Joshua KerievskyBesprekingen

Unfortunately I can't escape that I believe patterns are bullshit. As an IT professional (though perhaps an ignorant one) I have never been convinced by patterns. In my experience they are not directly reproducible in code, or tied to concepts or limitations for particular languages such as Java, or practised to an obsessive degree by enthusiasts who see patterns in everything ("patterns happy" as the author of this book describes it). So with this bias in mind, on with the review.

The book is divided into two sections. The first of these explains the author's motivations and suggested code smells and why one might want to refactor code. This was all very readable and sensible.

The second part was a long catalogue of patterns for refactoring, with step by step instructions and suggestions. This was where the book lost me. The code examples are in Java; many of the patterns aim to rectify the object-oriented house of cards from coding in that language. After a while I just got bored. The book seemed worthy and I wanted it to open my mind about patterns, but it stayed shut.

This is a thoroughly useful book. It breaks no truly new ground but can be treated as an extended meditation (with highly practical expresdion) on the GoF patterns and Fowler's Refactoring, bringing more detail to the context for the former - especially the situations which call for the deployment of the patterns - and longer examples of the process covered in the latter.

Interesting read after experimenting with the approaches in the Gang of Four book. This one chooses a somewhat more hands on approach, but I feel that the motivation behind the refactorings becomes a bit blurry.