What Every Programmer Should Know About Object-Oriented Design

Rating:
Summary: Updated version available under a different name.
Review: I already had a book from this writer (Meilir Page-Jones). So, I expect no less than high quality. I was not disappointed. This is one of the best books that I know of that can be bought about object-orientation. The problem with most other books, that is all other books that I have seen, is that they usually sink very quickly in the mud pool of the teeny weeny details of some object-oriented language. And before you know it the language syntax is discussed and the object-oriented principals are totally forgotten or at its best difficult to extract.

This writer never looses track of the subject he is trying to explain, namely: object-oriented design. As far as I can judge this is about the only book that succeeded to do this. This despite a lot of other books and highly acclaimed writers. On top of that this book is well structured, well built, and, God forbid, actually fun to read!

My advice, run to the bookstore and buy this book!

Rating:
Summary: Don't Buy This Book!
Review: I bought this book solely based on the greatness of another Page-Jones' book:"Practical Guide to Structured Systems Design". I figured that if this book was half as good as the "practical guide" was, my money would have been well spent.

This book is very inaccessible. It keeps you so deep in "buzz-word soup" that reader has to work very hard to understand his points. The examples are very abstract hard to follow.

I've been programming and designing systems using object-oriented languages for years. I can safely say that OO programmers do not anything in this book to be successful.

Rating:
Summary: Good, but dated
Review: There is a lot of good in this book. It offers a wide variety of basic design principles important to OO design: the Law of Demeter, a form of Liskov's Substitutability, stability of interfaces and behavior, and a fair discussion of coherence within a class and coupling between classes. It approaches Design by Contract and formal verification, without getting close enough to scare a beginner. It's first appendix anticipates Beck's "funny smells" and patterns for refactoring. It really does have large amounts of practical advice.

I truly wish I could recommend this book, but I can't. Its notation and vocabulary get in the way of its many messages. Every book has some boxes-and-arrows notation, and UML has become the accepted standard. This book predates UML, and offers one of the most ornate menus of different boxes I've seen. (The word "menu" comes to mind because so many of the diagrams look distractingly like a Big Mac to me.)

Page-Jones also edged into neologism, creating vocabularies for ideas that already had an established terminology. I know that, strictly speaking, he did not invent "connascence" or most of its friends (contranascence, disnascence, ...). Still, he seemed to cut the words out of their original context and wedge them, uncomfortably, into usages distant from their accepted meaning. That far a stretch is about the same thing as making the word up from scratch. It may be OK for advanced mathematical usages; a book for mainstream readers should stick to the main stream of common terms.

The technical advice in this book deserves much better than three stars. Its obsolescent notation and opaque vocabulary interfere with a modern reader's understanding, though. Beginners, the ones most likely to benefit from the advice, would have the hardest time with it. I like the book and will come back to it, but I can't recommend it.

Rating:
Summary: Updated version available under a different name.
Review: This is a good book, but it's been obsoleted by a newer version that uses the UML instead of the non-standard notation in this book. The new book is "Fundamentals of Object-Oriented Design in UML".