I first started programming in the 1960s. Now in the 1990s, it still seems as though management just does not know what to do next. Systems seem to cost too much to develop, are always late, and rarely work the way they should. Such concerns gave rise to process-oriented structured methodologies, to data-driven approaches, and lately to object-oriented panacea. I confess, I believe that each development did improve the situation somewhat. And I confess to being part of the boom, first with my Structured systems development manual [1], and now with my research and seminars in object-oriented analysis and metrics.

Nonetheless, it is apparent that the object-oriented realm is still in confusion. While we have sold management on objects, we cannot seem to agree on how to find them, classify them, document them, implement them, or reuse them. Even so, lots of us are making money on our version of the path to success.

One variant to emerge from this confusion has been business oriented notation (BON). Its origins seem to be in the Eiffel language, rather than in C++ or Smalltalk. This text focuses on the BON methodology, which puts together a high level of abstraction, strong typing, and formal specification. The static and dynamic system models are detailed, and suggestions are given for project deliverables. Even so, this book is not an overview. It is a detailed guide, heavy on the theoretical basis for the BON methodology.

The BON model is detailed in chapters on the static model (classes and clusters), static relations (inheritance and clients), and the dynamic model. The next section details “the method,” with an excellent chapter describing the specific tasks and deliverables suggested by the authors for those implementing the BON methodology, along with the standard activities demanded by any BON user. The final section of the book offers two chapters of case studies (a conference management system and a video recorder); an illuminating discussion of the coexistence of relational and object-oriented systems; and ten exercises for those wishing to try out their new knowledge. Appendices detail the BON grammar and give references. The index is spotty and too self-referential. Although Rumbaugh is mentioned in passing in the text, his name does not appear in the index.

The comparisons with the more widely accepted object-oriented methodologies (such as those of Booch, Jacobson, and Rumbaugh) are superficial. These competitors seem to be mentioned only to say they have existed; a thorough analysis of why BON is better is avoided with the use of statements such as, “It is clear that diagrams of the type represented…are too weak for dynamic modeling” (p.110). Just one page suffices to dismiss these others on the matter of the dynamic model. Finally, the writing is somewhat dry and academic.

This book has a great deal to offer. It is thoughtful and informative. Despite some shortcomings, it provides a detailed look at an important approach to object-oriented analysis and design.