Columns

Can EJB be fixed?

• By Michael W. Bucken
• June 1, 2004

In the early days of the Java revolution, Enterprise JavaBeans (EJBs) promised to end complaints that the new hot language was good only for client-side development. The EJB architecture was designed by Sun engineers for building server-side Java components, a job previously thought to be too complex for the Java language.

Unfortunately for corporate development teams, the EJB model has yet to meet expectations and hasn't spread nearly as fast as its backers promised it would in the mid- to late-1990s, points out veteran technology journalist Johanna Ambrosio in this month's Cover Story, 'Simple EJB: Is it ready yet?' Ambrosio finds that EJB has long required a steep learning curve, one that kept the technology from the vast majority of corporate developers. At the same time, new tools emerged from other quarters that promised to deliver the promises of EJB more easily.

The lack of widespread use and the long learning curve for EJBs prompted Sun and its partners in the Java Community Process (JCP) over the past year or so to work on simplifying the EJB model in its next iteration (Version 3.0) in an effort to make the technology more mainstream. In her research, Ambrosio talked to members of the JCP team simplifying the EJB model and to top developers who fear that an easier version will eliminate some key features. Clearly, the JCP team faces some difficult choices.

The story also looks at the status of EJB 3.0 development and what JCP developers hope will result from the work. Clearly, the goal is to make the technology attractive to far more developers than those that use EJB today, thus keeping it from becoming irrelevant to corporate IT. The good news for developers is that hard work is underway. The bad news is that the JCP won't say when the new version will be complete, only that the technology will first be available with J2EE 1.5, whose ship date has yet to be disclosed.

This issue also features a look at the increasing use of platform-independent Web services to access apps running on traditional mainframe systems ('Web services and the mainframe' ). Freelance writer Richard Adhikari points out that significant progress has been made in the difficult task of bringing mainframes into the Web services world, thus making the latter emerging technology an important ingredient of corporate development operations.