Preface iv
these projects, as well as the kind contribution of many individuals who have taken the time
to communicate with us, we have found ways to improve our method, in terms of better
articulating the process, adding and clarifying certain semantics otherwise missing or difficult
to express in the notation, and simplifying the notation where possible.
During this time, many other methods have also appeared, including the work of Jacobson,
Rumbaugh, Coad and Yourdon, Constantine, Shlaer and Mellor, Martin and Odell,
Wasserman, Goldberg and Rubin, Embley, WirfsBrock, Goldstein and Alger, Henderson-
Sellers, Firesmith, and others. Rumbaugh's work is particularly interesting, for as he points
out, our methods are more similar than they are different. We have surveyed many of these
methods, interviewed developers and managers who have applied them, and where possible,
tried these methods ourselves. Because we are more interested in helping projects succeed
with object-oriented technology rather than dogmatically hanging on to practices solely for
emotional or historical reasons, we have tried to incorporate the best from each of these
methods in our own work. We gratefully acknowledge the fundamental and unique
contributions each of these people has made to the field.
It is in the best interests of the software development industry, and object oriented technology
in particular, that there be standard notations for development. Therefore, this edition
presents a unified notation that, where possible, eliminates the cosmetic differences between
our notation and that of others, particularly Jacobson's and Rumbaugh's. As before, and to
encourage the unrestricted use of the method, this notation is in the public domain.
The goals, audience, and structure of this edition remain the same as for the first edition.
However, there are five major differences between this edition and the original publication.
First, Chapter 5 has been expanded to provide much more specific detail about the unified
notation. To enhance the reader's understanding of this notation, we explicitly distinguish
between its fundamental and advanced elements. In addition, we have given special attention
to how the various views of the notation integrate with one another.
Second, Chapters 6 and 7, dealing with the process and pragmatics of object-oriented analysis
and design, have been greatly expanded. We have also changed the title of this second edition
to reflect the fact that our process does indeed encompass analysis as well as design.
Third, we have chosen to express all programming examples in the main text using C++. This
language is rapidly becoming the de facto standard in many application domains;
additionally, most professional developers who are versed in other object-oriented
programming languages can read C++. This is not to say that we view other languages - such
as Smalltalk, CLOS, Ada, or Eiffel - as less important. The focus of this book is on analysis and
design, and because we need to express concrete examples, we choose to do so in a
reasonably common programming language. Where applicable, we describe the semantics
unique to these other languages and their impact upon the method,
