Several months ago, I reviewed a book that was supposedly about the Unified Modeling Language (UML), or at least its title contained this name. In my review, I regretted that the book did not explain at all what UML is, and thus I could not learn all that I needed to know about this new fad. A clever CR editor jumped at the opportunity, and soon I was given this book to review.

This book’s title is not a misrepresentation: most of its contents are about UML. Still, I have several complaints about the contents. The first is a problem of style, and the second is about the information the book tries to provide.

At slightly less than 300 pages, this book could be considered rather short, at least according to current standards. It offers the pleasant and careful design that is customary for O’Reilly books. All in all, the editors, illustrators, and designers have done a fine job.

I found several defects, however. Chapter 5 is called “The Unified Modeling Language” in the table of contents, but “UML Overview” in the text. Second, the introduction to Part 3, on pages 127 to 128, seems to be a joke: we read that “Chapter 6, Diagramming and Model Organization, presents information on diagramming and model organization.” The summaries of the next 10 chapters then simply repeat each chapter title.

These are details. The problem of style is more severe. As the author states, “The book makes extensive use of bulleted lists.” They are extensive indeed, to the extent that the first 60 pages probably contain more than 100 bulleted lists or sublists. Many lists contain more than ten items, and one may wonder whether this is not simply the skeleton of the real text to be written--a simple draft of many ideas that should be developed later. This continues, ad nauseam, throughout the whole book, to the point of being a real hindrance. When a page without any bulleted list appears, it is a welcome relief.

The book is divided into three parts: Part 1, “Introducing the Unified Modeling Language” (3 chapters, 60 pages); Part 2, “Using the Unified Modeling Language” (2 chapters, 58 pages); and Part 3, “The Unified Modeling Language Quick Reference” (11 chapters, 134 pages). These titles are misleading. The first part does not deal with UML at all, except in its first chapter, which is only 12 pages long. The rest is mainly about problem solving and object orientation, and introduces a lot of concepts and vocabulary. Either this part is useless, since any reader already accustomed to the object paradigm should know its contents, or it is too short and almost exclusively declarative, without real explanations.

The second part contains two very different chapters, and only the first one is about using UML. In fact, it is called a tutorial, and it is indeed a quite readable example of using UML, developed over 40 pages. My only negative remark is that it is not always easy to separate the description of the example from that of UML itself. The second chapter of this part tries to give an overview of UML, but it is surprising to find this at the beginning of the second part of the book. It would be more natural if the overview was the first chapter of the book; maybe then, the book would be more “a desktop quick reference.”

The last part, finally, is indeed a quick reference to UML. Once again, bulleted lists are overused, and the style is mainly purely declarative--there are generally no explanations or justifications. The last chapter, “The Object Constraint Language”, seems to have been added as an afterthought and is not clearly linked to the rest of the book. The book is completed by references to URLs and books, and an index.

All in all, I find it unfortunate that the author did not devote most of his energy to the main contents of the book, the material found in the third part. Moreover, I strongly recommend that he not use any more bulleted lists, at least not for his next ten books.