UML Applied: A .Net Perspective

The critics of this book seem to be most negative about Mr Shoemaker's failure to link Dot Net and UML. While it is true that there is not a great deal in the book about this specific topic, that is certainly the top half of the glass. The bottom half is the Five-Step-UML and requirements gathering with UML. The Five-Step-Process, DRADR (define, refine, assign, design, reiterate) has been used by this writer with great success and is worth the price of the tome by itself. In studying and obtaining a certification in UML many books and articles have been perused but this was the first found that took UML beyond a set of thirteen diagrams to an executable process. Not only is the process defined but also detailed to the point where a process template has been developed from the ideas outlayed. It is a cliche that the down fall of most failed business process projects, including many SDLC projects, is the failure to properly elicit, define, analyze and communicate process requirements. This fact makes Mr. Shoemaker's three chapters, five, six and seven, on requirement gathering, categorization, analysis and refinement a trove of process knowledge.

I have several UML Books. If the other books had done their job, I would not have had to buy this one, This is the ONLY one that get's the point across without making you nod off. It's hard to put down. Probably because I finally found someone who can clearly explain this stuff. The author has experience TEACHING this class and has learned from the students in how to do just that. What's with the reviewer making fun of stick figures? That's UML from the three amigos, not this author. This is a UML book not a dot net book.

A curious thing about the title is its mention of .NET. Strictly, UML is independent of any environment or operating system or language. Those are implementation level details. But here, Shoemaker brings .NET into his UML discourse to show how UML can be well integrated with a .NET development process. And indeed, that is part of the book's value. Having said this, most of the book can be read, ignoring .NET. So those of you into C++ or Java can still gain from the book. He gives lengthy, detailed explanations of defining and refining requirements. And then mapping these to components and interfaces and using these results to design architecture and components. Not a book for the impatient. Shoemaker takes time to carefully expound on the basic ideas. If you're new to UML, a deliberate slow reading might be best, to gain the most from what he is saying. Another virtue of the book is that apart from the .NET details, it can be used years from now. (Barring any major advances in our understanding of the design process.) It is not one of those computer books that you have to replace in 2 or 3 years time.