Foreword

Java Data Objects (JDO) is an important innovation for the Java platform. At a time when developers were using JDBC almost exclusively for database access, and expert groups from major enterprise vendors were devising the much-touted Enterprise Java Beans APIs for entity beans and container-managed persistence, Craig Russell and David Jordan had the courage to take a different course. With a handful of others, they looked for a simpler way to provide persistence in the Java platform, something that would be both natural and convenient for programmers. This book describes the result of their work: JDO.

The key, unique idea behind JDO is to provide database persistence in Java with a minimum of extra stuff for the programmer to do. The programmer doesn't need to learn SQL, doesn't need to tediously copy data into and out of their Java objects using JDBC calls, and can use Java classes, fields, and references in a way that is natural to them, without lots of extra method calls and coding that is extraneous to the programmer's focus and intent. Even queries can be written using Java predicates instead of SQL. In other words, the programmer just writes Java; the persistence part is automatic.

In addition to this transparent persistence, code written to JDO benefits from binary compatibility across implementations on different datastores. JDO can be used with an object/relational mapping, in which JDBC calls are generated automatically to map the data between Java objects and existing relational databases. Alternatively, the JDO objects can be stored directly in file pages, providing the functionality and performance of an object-oriented database.

The hard work on JDO paid off: the idea of transparent persistence has proven quite popular. JDO has its own community web site, www.JDOCentral.com, and on enterprise Java discussion sites such as www.TheServerSide.com, developers praise the simplicity and utility of JDO. Many developers use JDO as a replacement for entity beans, by using data objects from within session beans. Others use JDO as a convenient high-level replacement for JDBC calls in JSP pages or other Java code. JDO has come a long way from the JDBC interface I defined in 1995 with Graham Hamilton, and JDO is quite valuable in conjunction with J2EE.

I can't think of two individuals better qualified to write a book about JDO. Craig is the specification lead for the JDO expert group, and Dave was one of the most active members of that group. But their qualifications go far beyond that, and JDO was well designed as a result of those qualifications. Both have over a decade of experience with issues in programming language persistence, including subtle transaction semantics, different persistence models, relationships between objects, caching performance, interactions between transient and persistent objects, and programming convenience in practice. Both had extensive experience with C++ persistence before they applied their experience to Java. Both were key members of the Object Data Management Group (http://www.odmg.org) for years. And, most importantly, both were developers who appreciated and needed the functionality that JDO provides.

Craig and Dave have put together a thorough, readable, and useful book. I hope you enjoy it as much as I did.

—Rick Cattell, Deputy Software CTO Sun Microsystems, February 16, 2003