The book is synonymous with the Kennedy Carter iUML product it describes, which in turn is simply not able to support large scale, commercial software development. If your project is more complicated than the included examples of a Stop Light, or a Rail Way Station, you will find the book and the development environment lacking. I have been involved in one of the largest projects that attempted to use Kennedy Carter. The project was prominently featured on the KC website. But, in the end KC was not up to the task. There is no support for refactoring, for source control diff and merge, nor even a simple text search of the whole project. I can not recommend either the book or the tool.

This book provides and clear and detailed introduction to the method of developing complex software systems indicated by its title. It speaks with an air of authority clearly rooted in the authors' breadth and depth of experience yet retains a light and occasionally amusing style and includes many anecdotes that seasoned software professionals can readily relate to.

The book covers the full software development lifecycle and illustrates a pragmatic and proven approach to adopting a Model Driven approach to software development. The author's are clearly from the so-called `translational' camp that firmly believes in a detailed `Platform Independent Model' from which code generators derive up to 100% of a working system. While detractors might be sceptical of such a claim, the `Executable UML' provides the clue as this is made possible by an `action specification language' that augments standard UML and effectively provides an abstract programming language for the operations defined on the UML model.

While the books intends to be of general application to all types of software development its clear origin and emphasis is on real-time, control, embedded and safety critical software. However there are clearly good lessons that can be applied equally well to Enterprise IT systems and indeed one can see clear parallels and more similarities than differences between the advanced modelling approaches proposed here and in more Enterprise IT oriented texts such as `Domain Driven Design' by Eric Evans. What distinguishes the approach in this text say from Evans is its stated goal and claimed desirability of achieving complete separation of the domain model from the target software architecture.

Overall I thoroughly enjoyed reading this book. It tackled problems I recognised in a practical way and gave me a clear insight into how the method is used. I believe it represents a valuable contribution to the field of complex systems development and worthy of the time of anyone who wants to deepen their understanding of practical MDA.

This book provides and clear and detailed introduction to the method of developing complex software systems indicated by its title. It speaks with an air of authority clearly rooted in the authors' breadth and depth of experience yet retains a light and occasionally amusing style and includes many anecdotes that seasoned software professionals can readily relate to.

The book covers the full software development lifecycle and illustrates a pragmatic and proven approach to adopting a Model Driven approach to software development. The author's are clearly from the so-called `translational' camp that firmly believes in a detailed `Platform Independent Model' from which code generators derive up to 100% of a working system. While detractors might be sceptical of such a claim, the `Executable UML' provides the clue as this is made possible by an `action specification language' that augments standard UML and effectively provides an abstract programming language for the operations defined on the UML model.

While the books intends to be of general application to all types of software development its clear origin and emphasis is on real-time, control, embedded and safety critical software. However there are clearly good lessons that can be applied equally well to Enterprise IT systems and indeed one can see clear parallels and more similarities than differences between the advanced modelling approaches proposed here and in more Enterprise IT oriented texts such as `Domain Driven Design' by Eric Evans. What distinguishes the approach in this text say from Evans is its stated goal and claimed desirability of achieving complete separation of the domain model from the target software architecture.

Overall I thoroughly enjoyed reading this book. It tackled problems I recognised in a practical way and gave me a clear insight into how the method is used. I believe it represents a valuable contribution to the field of complex systems development and worthy of the time of anyone who wants to deepen their understanding of practical MDA.

Bad news first. For building executable UML models they use Action Specification Language (ASL). ASL is not a part of UML standards. Not so bad news - "Action Semantics for the UML" (OMG ad/01-08-04) imposed as an "unofficial OMG standard". Not by me, but by Anneke Kleppe, if you know, who is this person. And the "Action Semantics ..." is based on the ASL.

Currently ASL is not widely supported by popular UML tools (hah! how many of them really support OCL?). But the lite version of iUML CASE tool comes with the book. I don't want to discuss is this tool good, or is not. At least it gives you the ability to feel the concepts, not only read about them.

In any case, for building truly executable models, you need some kind of action language. And OCL is not that kind of a language. If you think Java is portable enough for PIMs development - use it. The authors did not recommend it, but they did not prohibit it also. You may use ASL, C++, Java, production rules, your own action language, etc. It's up to you.

If you know nothing about MDA and executable UML - don't worry. This book could be used as a start also.

When I sent this feedback, there was no ability to search inside the book. So, let me describe it. This book:
- gives some review of OO methods genesis
- gives an introduction to MDA and executable UML concept, including the difference between a method and a notation
- shows how to fit the method into the development processes
- shows how to gather the requirements with use cases
- shows how the method uses other kinds of UML diagrams to achieve the goal to make them executable, not just a pictures
- exlpains why some of UML diagram types are not used by the method at all, and brings only one extension to UML (state transition tables)
- describes how to build both state-independent and state-dependend behaviour models using UML and ASL
- describes how to integrate different PIMs
- describes how to generate code from models
- contains samples

What could I say. Maybe you have an opportunity to meet the standard of the future. Maybe not. But this book is not just a iUML or ASL tutorial. This is a good book. It was written by real professionals, followers by Shlaer & Mellor school. It's full of useful ideas. Strong, but not boring. Must have.