While Lewis says this book is about the use of CASE tools in the software engineering development process, it is actually narrower than that. It is about software engineering primarily on the Macintosh and primarily using a homegrown set of CASE tools.

The book covers the standard software engineering topics, but sometimes in an odd way. The chapter on “The Elements of Design” precedes anything on either requirements or the application. The topics of requirements, design, and implementation are strangely intertwined. Lewis discusses “design tradeoffs” in a section on the software requirements specification and even suggests that a question to be asked at the requirements review is “What is the overall design philosophy?” He suggests that object-oriented design is more of a “programming technique” than a “design technique” and, similarly, that “a module is much more closely associated with implementation than with design.”

Some thoughts are just plain odd. The discussion of “the software project” is almost entirely about configuration management and quality assurance. The notions of methodology and modularity are intermixed, as if the purpose of a methodology were to choose an appropriate modular approach. Object-oriented objects are treated as programming objects rather than as objects in the application domain. Further, objects are abstracted from a dataflow diagram rather than from the raw problem space.

The book states that “prototyping is a kind of CASE tool…” and asks “which is best, code review or CASE tool?” It also states, “software quality is the sum total of the features of a software product that satisfy the needs of the user.” It lists, under things that have little or no effect on cost, “documentation” and “walkthroughs.”

The reader feels that the author’s heart is in the right place, but that his mind did not pause long enough during the writing to state clearly formed and logical ideas.

This work does have some redeeming features. The book is at its best when it focuses on the use of CASE tools in the life cycle and the use of the wonders of the Macintosh to build user-friendly software. It makes some unique contributions, such as a table of compiler performance for a variety of applications. It takes a stab at some application-specific advice (“Jackson structured design is probably not the best for…interactive and real-time programming”) and presents “A Design Guide” to further help the software developer make appropriate choices.

Still, if the book were to really fulfill the promise of its title, it would need to step well beyond the author’s own toolset. The marketplace of CASE tools is now far richer than any one toolset, either academic or commercial. In a world full of competing software engineering books--many of them quite good--one has to care a lot about the Macintosh and the author’s own CASE tools to decide to buy this work.