As soon as I receive a book I have a quick flick through it just to see what I can expect. It soon became apparent that had I flicked through this book in a bookstore I would not have purchased it. Not because the book is no good but due to its irrelevance to the way I currently develop software.

The book is, in fact, extremely well written and very clear with easy to follow guidelines and I thoroughly enjoyed reading it. The whole unified process is explained from start to finish with pointers on what to look out for and some useful examples. I found the sections on testing to be extremely useful and full of common sense.

So why is this book not for me? The whole unified process is geared towards large software developments. You need architects, testers, coders, etc. Although these can be the same person the overhead of the process is probably too great for my one man, one-month stand alone executables. If you run a large software department, however, I can see great benefits in adopting the process. I will hold on to this book, however, as I will definitely refer back to it if my projects increase in size.

My one bugbear with the book is that it constantly refers to one of the author's previous books 'UML Explained'. I don't buy a book to be told I really should buy another one as well. Some of the diagrams are explained fully in the other book and only referenced in this book. If the explanation is necessary for a full understanding then it should be reproduced in full.