Introduction - Database Processing: Fundamentals, Design, and Implementation - page 23

Another technique for structuring data relationships used data structures called networks.
The CODASYL Committee (the group that developed the programming language COBOL)
sponsored a subcommittee called the Database Task Group (DBTG). This subcommittee de-
veloped a standard data model that came to bear its name—the CODASYL DBTG model. It
was an unnecessarily complicated model (everyone's favorite idea made it into the committee's
design), but several successful DBMS products were developed using it. The most successful
was IDMS, and its vendor, the Cullinane Corporation, was the first software company to be
listed on the New York Stock Exchange. To the best of our knowledge, no IDMS database is in
use today.
The emergence and Dominance of the Relational Model
In 1970, a then-little-known IBM engineer named E. F. Codd published a paper in the
Communications of the ACM 3 in which he applied the concepts of a branch of mathematics
called relational algebra to the problem of “shared data banks,” as databases were then known.
The results of this work are now the relational model for databases, and all relational data-
base DBMS products are built on this model.
Codd's work was at first viewed as too theoretical for practical implementation.
Practitioners argued that it was too slow and required so much storage that it would never be
useful in the commercial world. However, the relational model and relational database DBMS
products became adopted as the best way to create and manage databases.
The 1977 edition of this text featured a chapter on the relational model (which Codd him-
self reviewed). Many years later, Wayne Ratliff, the creator of the dBase series of products for
personal computers, stated that he had the idea for dBase while reading that very chapter. 4
By The Way Today, there are as many opportunities for innovation as there were for
Wayne Ratliff in 1977. Perhaps you can read Chapter 11 and develop an
innovative product that integrates XML and DBMS processing in a new way or read
Chapter 12 and join the NoSQL and Big Data movements to help develop alternatives
to relational database technology. Just as in 1977, no product has a lock on the future.
Opportunity awaits you!
The relational model, relational algebra, and, later, SQL made sense. They were not need-
lessly complicated; rather, they seemed to boil down the data integration problem to a few
essential ideas. Over time, Codd convinced IBM management to develop relational-model
DBMS products. The result was IBM's DB2 and its variants, which are still very popular today.
Meanwhile, other companies were considering the relational model as well, and by 1980,
several more relational DBMS products had been released. The most prominent and impor-
tant of those was Oracle Corporation's Oracle Database (the product was originally just named
Oracle but was renamed as Oracle Database after Oracle Corporation acquired other products
and needed to distinguish its DBMS product from the others). Oracle Database achieved suc-
cess for many reasons, one of which was that it would run on just about any computer and just
about any operating system. (Some users complained, “Yes, and equally badly on all of them.”
Another, when asked, “Should we sell it to communist Russia?” responded, “Only as long as
they have to take the documentation with it.”)
3 E. F. Codd, “A Relational Model of Data for Large Shared Databanks,” Communications of the ACM , June
1970, pp. 377-387. A downloadable copy of this paper in PDF format is available at http://dl.acm.org/citation
.cfm?id=362685 .
4 C. Wayne Ratliff, “dStory: How I Really Developed dBASE,” Data Based Advisor , March 1991, p. 94. For more
information on Wayne Ratliff, dBase II, and also his work with FoxPro (now Microsoft Visual FoxPro), see
the Wikipedia article Wayne Ratliff at en.wikipedia.org/wiki/Wayne_Ratliff . For the history of dBase, see the
Wikipedia article dBase at en.wikipedia.org/wiki/DBASE.