Review By: Arvind Pal Singh
08/08/2011Software Project Management: A Unified Framework by Walker Royce is a classic. It starts with older approaches, highlighting their pitfalls, and later suggests the importance of an iterative development cycle with architecture first. I highly recommend this book to professionals in IT projects and consider that this book will stay in the market for a long time.

I like the way Royce questions the most common approaches for an IT project in larger organizations. Apart from offering constructive criticism, he also provides readers with usable and innovative ideas. Experienced professionals who have or have not used Rational Unified Process will be able to appreciate its beauty for an IT project using an iterative approach. I had to read chapter 4 a few times for all the details, because it gives me clear answers to many questions I get asked at various levels within my organization.

Royce has done a commendable job in keeping the material in an organized and readable structure. One place where I got stuck was where Royce mentions testing as an integral part of each phase. In the past, I have tried doing this in a couple of projects but ultimately had to switch to testing as a separate piece. It may be because some of the presented approaches work better for very large projects and not for the size of projects I've gotten into. But, I would love to learn more tangible techniques for interweaving testing. I enjoyed the great recollection of the different approaches, which sometimes remain vague in practice. Also, the book covers estimating techniques in good detail.

The top ten principles of modern project management are certainly worth pondering over. Putting them into practice for large projects sounds like a great fit. For small teams or small projects, the ideas may not make much sense. Overall, I like that this book is very well organized and well written with a good sense of humor. I will certainly recommend it to my coworkers and library.