Intelligent Enterprise

Better Insight for Business Decisions

RSS
Webcasts
Whitepapers
Subscribe
Home

May 31, 2003

Father Relational

The passing of a pioneer: Dr. E. F. Codd, the inventor of the relational model

by David Stodder

As we were putting this issue to bed, we heard the sad news that Dr. Edgar F. Codd had died on April 18 at his home in Williams Island, Fla. Dr. Codd started the relational revolution at the IBM San Jose Research Laboratory in the 1960s and 1970s. His paper for the Communications of the ACM, "A Relational Model of Data for Large Shared Data Banks," set out the relational model and became the manifesto for those passionate about creating a powerful new way of working with data that rested on a strong, if not ancient, foundation of predicate logic and mathematics.

From the moment of its publication, Dr. Codd's paper engendered heated debate. The relational model was not the only theory in play; its strongest initial competition came from proponents of the Cobol/CODASYL network database model, led by another legendary figure, Charles Bachman. Spirited debates between Dr. Codd and Bachman (with illustrious voices on both sides) spanned several years. Dr. Codd also fought fierce battles within IBM, which had already invested significantly in IMS, a hierarchical data management system. However, with excitement about the relational model brewing, IBM Research began its System R prototype, while over at the University of California, Berkeley, Michael Stonebraker led a team developing Ingres.

Relational Explosion

By the time I started covering this industry for Database Programming & Design in the late 1980s, relational databases were red hot; leading contenders included IBM's DB2 and SQL/DS, Informix, Ingres, Oracle, Digital Equipment Corp.'s RDB, Sybase, and numerous others, some of which were, shall we say, marginally relational. If it wasn't "relational," your database product faced tough odds, as object-oriented database partisans would soon find out. The relational model had become the singular foundation for a dynamic, multibillion dollar sector of the software industry.

Today, as Ken North's lead article in this issue illustrates, relational database providers continue to work with the model Dr. Codd provided as the theoretical basis for their expansion into XML, Web services, and grid computing. It is amazing that the relational vision continues to extend to all sorts of data types and users. The leading providers do not flinch at the grandiose vision of becoming the "center of the data universe."

Rate This Article

Comments:

Optional e-mail address:

Throughout the early boom years of the database industry, Dr. Codd, in association with C. J. Date, tried valiantly to keep IT developers' and managers' eyes focused on the power of the true relational model so that they wouldn't lose their way amid what he and Date saw as "fatally flawed" commercial SQL implementations. Dr. Codd continued to revise his model and published The Relational Model for Database Management Version 2 (Addison-Wesley, 1990), which stirred up even more controversy, this time among relational model partisans themselves. I remember well Dr. Codd's uniquely illuminating presentations at industry conferences — and the buzz of the debaters who would follow him out of the hall at the end.

We dedicate this issue of Intelligent Enterprise to Dr. Codd. Look for more coverage of his contributions in an upcoming issue and at our Web site, www.IntelligentEnterprise.com An A.M. Turing Award winner, he will forever loom large in the history of computer science. Few will forget his brilliance, determination, and uncommon insight.