The Fathers of Modern Databases
When taken together, Bachman, Codd and Stonebraker sound like some 1970s-era arena rock supergroup. Separately, each man contributed to the development of the contemporary database. (Here, “contemporary” is defined as a database management system [DBMS] or a methodical structuring of virtual information accessible to any authorized user who has a computer and connection to a network.) What follows is an overview of who these men were and their respective roles in the evolution of the database.
In 1960, an industrial researcher named Charles Bachman jumped ship from Dow Chemical to General Electric, which was about the same time that the idea of a DBMS began to emerge. Bachman’s research during this period led him to the conclusion that newly developed, direct-access storage devices could change data processing, which had been based on punch cards and magnetic tape.
He was a catalyst for getting the DBMS concept on the ground, as he developed the integrated data store (IDS) and multiprogramming access to it during his tenure at General Electric. He established Bachman Information Systems in 1983, which offered the BACHMAN/Data Analyst as its flagship product. Essentially, it provided graphic support for building and maintaining data structure diagrams.
In his work and life, Bachman touted the power of the navigational database, which combines network and hierarchical configurations. This position made him something of a rival of the next individual.
Edgar Frank “Ted” Codd
Ted Codd, whose defense of relational database models put him at odds with Bachman, performed his groundbreaking work while at IBM during the 1960s and 1970s. In fact, he more or less created the relational database management theory as an employee for the Big Blue when he issued the now-famous “A Relational Model of Data for Large Shared Data Banks” paper in 1970.
IBM, however, did not act on the recommendations he laid out in this opus because of a desire to preserve the position of its IMS/DB (which was influenced by Bachman’s hierarchical database management formulations) in the market. This was a big disappointment for Codd, but he didn’t sit around and wait for his company’s leaders to change their minds. Instead, he took his case to IBM’s clients, who then requested that the organization enact his ideas.
IBM eventually did, sure enough, but kept Codd out of it by handing the System R relational model project over to a group of developers with whom he had no meaningful contact. The result was SEQUEL, a programming (and, incidentally, nonrelational) language that Larry Ellison “borrowed,” renamed SQL and used to turn his new company Oracle into an IT powerhouse.
Codd also made a key contribution to the database field when he coined the term Online Analytical Processing (OLAP) and built on that concept with his own ideas. In addition, he continued to advocate for the relational database model all his life, and his efforts greatly influenced the next giant of the database world.
One of the earliest implementations of a relational database was Michael Stonebraker’s Ingres, which he developed at Berkeley after reading works on information management theories produced by Codd and other IBM employees. (Interestingly, his Ingres came out in 1976, the same year as IBM’s System R project, but the first relational databases designed for commercial consumption wouldn’t come out for another four years.)
Stonebraker’s influence was felt directly in the market through the companies he founded such as Ingres Corp., Illustra and Streambase Systems. He also made an impact in academia, with faculty tenures at Berkeley and MIT. In fact, while at Berkeley in the late 1980s and early 1990s, he launched the Postgres (post-Ingres) project, which would later serve as the coding cornerstone for today’s open Postgre SQL database. For his ingenuity and hard work, Stonebraker was recognized with the first Edgar F. Codd Innovations Award.