The idea of Object-oriented COBOL may seem like the ultimate oxymoron to many who have briefly studied COBOL, only to move on to favorites such as PASCAL, C, or C++. Now comes a new breed of COBOL, with objects and many of the familiar concepts of object-oriented programming (OOP), such as classes, methods, inheritance, and messages. This book, as stated in the preface, has three major goals: to provide a firm understanding of object-oriented concepts; to introduce the syntax and semantics of Object-oriented COBOL; and to provide enough background in object-oriented analysis and design so that readers can begin to implement their own small to medium-sized systems. The intended audience has a background in procedural COBOL.

I found the book lacking for its stated purpose in several respects. The one undeclared purpose of the book that became evident early on is to evangelize for Object-oriented COBOL as a programming language. Setting this feature aside, the book fails to meet its stated goals. It is my experience that making the transition to object paradigms in programming requires significant effort and adjustment in how problems are solved. An audience consisting of COBOL thinkers will have a hard time using this book as the source of the motivation and inspiration needed to make such a transition. The book mixes a discussion of the history of objects, definitions, and object-oriented languages with COBOL syntax. The issue with learning the object-oriented approach is not syntax, but meaning, and understanding objects and how they interact to produce a problem solution. It is hard to imagine that the COBOL people I know would be able to understand OOP ideas based on the explanations in this book.

One of the most daunting challenges in doing real work in today’s world of objects is learning the application programming interface to the host environment. Yet there is nothing in this book describing how a real program would connect to and interact with a database. There is nothing describing how one would build a meaningful user interface with any popular graphical user interface. There is no discussion of any set of objects or classes that could be used to leverage such interfaces.

I would recommend this book for those interested in seeing how an object-oriented structure can be imposed on a flat language like COBOL. However, for those intending to learn and use Object-oriented COBOL for real work, this book will only provide a small start. You may wish to begin implementing your own systems (the third stated goal), but you probably will not be able to do so using only the information contained in this book.