 |
|
Enterprise JavaBeans (EJBs) may be the clumsiest name that our industry has come up with recently -- but the technology is anything but. Enterprise JavaBeans combines distributed components with transaction monitors to produce an infrastructure that can manage database-style transactions, security, object persistence, and resource management in a standardized way. Enterprise JavaBeans has nothing in common with plain ol' JavaBeans except its name, and it is much more similar to the Microsoft Transaction Server (MTS).
Richard Monson-Haefel's Enterprise JavaBeans starts by describing the historical and technological background to EJB, including the development of transaction monitors over the last 30 years, and the underpinnings of distributed-object technology. Chapter 2 then looks at the EJB architecture, while Chapter 3 discusses resource management. As part of this discussion, Monson-Haefel looks at how EJB handles the six primary services defined by the OMG: concurrency, transactions, persistence, distributed objects, naming, and security. This discussion is informative, but noncritical -- the book never suggests that EJB has weaknesses, or that a different approach to some technical problem (or more rigorous standardization) might have been more flexible or easier to implement.
Chapter 4 describes how to implement some simple EJBs. As with other examples in the book, most of this material is taken from a system to manage bookings for a holiday cruise line. Chapters 5 through 8 then discuss how EJBs are used by client-size applications, the development of entity beans (that is, components that describe real-world objects such as cruise ships), session beans (which describe business logic such as the booking of a cruise) and EJB's support for transaction management. The last chapter, "Design Strategies," then gives some tips on how to go about developing EJB-based applications. The book closes with three appendices: One on the EJB API, one (which I found very helpful) that gives UML state and sequence diagrams for various operations, and one (which is already out of date) that provides information about EJB vendors.
Enterprise JavaBeans is well written, and well edited: I did not notice any awkward sentences, typographical errors, or ugly diagrams. In addition, Monson-Haefel's examples are easy to follow. His discussion of technical issues is clear, but would have been better, in my opinion, if it had been a little more critical, or if it made more comparisons between EJB and other systems. (One of the things that makes Szyperski's Component Software such an excellent book is the way its author does precisely these things.) Overall, Enterprise JavaBeans is a good starting point for anyone who is interested in the subject, and a reasonable reference for anyone who is already in the middle of an EJB project.
-- Gregory V. Wilson
| Readability |
|
| Originality |
|
| Organization |
|
| Accuracy |
|
| Consistency |
|
| Depth |
|
| Timeliness |
|
| Editing |
|
| Design |
|
| Overall Value |
|
Explanation of ERCB rating scale: No stars = unacceptable, 1 Star = marginal, 2 Stars = average, 3 Stars = above average, 4 Stars = exceptional.