Rating: 
Summary: Pragmatic Fundamentals -- distilling the accumulated wisdom
Review: Who should read this book: Senior Programmers and Systems Analysts.This book can be very valuable anyone who builds Object-Oriented computer programs, and anyone building computer programs either is or soon will be using Object-Oriented tools. Although it focuses on the Unified Modeling Language, a standard for most Computer-Aided Software Engineering tools, I found the real value of the book to be in the lucid explanations of principles of good software analysis and design, even more than in the nuts and bolts of UML. Dr. Page-Jones' style continues to combine well-researched information with down-to-earth pragmatism and a delightfully irreverent tone towards those who take this business (or themselves) way too seriously. Who am I? I am a computational physicist turned systems analyst, with almost 20 years experience developing complex codes for scientific modeling and analysis, now working on real-time defense systems. I have been technical lead and mere contributor, subcontractor and lead contractor. My passion is for tight modular designs that facilitate high-reliablility code. Part I, Introduction, gives excellent working definitions of the main concepts generally considered part of "Object Orientation" in a way that should be useful even to those beginning to use an OO language, and a historical perspective that helps explain why some issues are still messy. Part II gets into the "nuts and bolts" of UML itself, of necessity illustrating many key concepts along the way. Even if you never use a CASE tool, the ability to discuss design issues using accepted "standard" diagrams will help you think through the key issues, communicate your ideas more clearly, and ultimately develop better designs. Better designs will not only avoid headaches in the implementation, but also guide the coding effort by making it clearer what the goals are. Part III, the Principles of Object-Oriented Design, is worth the price of the book all by itself. Meilir's style helps you get through the big words by reminding you that that they are generalizations of common sense. For example, "connaissance" is basically old-fashioned "coupling": the more Module X's correct execution depends on details in Module Y, the more chances for mistakes. The OO quality of "encapsulation" helps you minimize it, whatever you call it. The old-fashioned concept of "coherence" becomes "cohesion": if everything a module does supports a well-understood single purpose, it is less likely there will be confusion or duplication. Do you hate a utility because you need to know so many stupid implementation details to use it properly? Don't lamely whine "I just don't like it!" when you can impress everyone by showing that it has an unjustifiably high "encumbrance." The book is full of sound principles illustrated with real-world examples. If your gut has ever told you a proposed action was a bad idea, but you didn't know how to express your misgivings, the sound principles in this book will help you put your finger on what is wrong and why. Even if you don't care to use the impressive academic-sounding terms for them, the principles will help you become a more effective, persuasive member of your team. And even if you ARE the whole "team," they will keep you out of trouble.
Rating:
Summary: Top class
Review: Written with style, clariity and a sprinkling of humour, this is an excellent introduction to both OO concepts and UML. It is vastly more readible than the competion (e.g. Fowler). and lends itself well to teaching yourself OO/UML. The exercises and answers at the end of each chapter are particularly useful in this respect. Be warned - it gets tougher as it goes along, but that's because the stuff he's putting across IS tough. All in all, highly recommended.
| 
