The relational model of data was invented by one man, E. F. ("Ted") Codd, during his employment in the late 1960s and early 1970s at the IBM Research Laboratory in San Jose, California. Today, virtually all general purpose database systems are based on that model. But those systems are still a long way from perfect-and a major reason for this sorry state of affairs is that they none of them abide 100% by the prescriptions of the model. To address this problem, Codd subsequently wrote a book, The Relational Model for Database Management Version 2, in which he defined a hugely expanded version of his original model. His aim was to spell out, in vastly more detail than before, exactly what a relational database system is supposed to look like.

Given his track record, anything Codd has to say regarding database matters clearly deserves very careful attention and analysis-and that's exactly what the present book provides regarding RM/V2 in particular: It describes, explains, and critiques Codd's new model in all of its myriad aspects. The author, Chris Date, was a close colleague and supporter of Codd's for many years and was instrumental in getting the word out regarding the original relational model.