Datasheet

SQL and Relational Database Management Systems 1
computer language to access the data. The problem was that IBM already had declared its own
product, called IMS, as its sole strategic database product — the company management was not
convinced at all that developing new commercial software based on a relational schema was
worth the money and the effort. A new database product could also potentially hurt the sales of IMS.
In spite of all that, a relational database prototype called System R was finally introduced by IBM
in the late 1970s, but it never became a commercial product and was more of a scientific inter-
est, unlike its database language, SQL (first known as SEQUEL), which eventually became the
standard for all relational databases (after years of evolution). Another relational product, called
Ingres, was developed by scientists in a government-funded program at the University of Califor-
nia, Berkeley, at about the same time, and had its own nonprocedural language, QUEL, similar
in some respects to IBM’s SQL.
The first commercial relational database was neither System R nor Ingres. Oracle Corporation
released its first product in 1979, followed by IBM’s SQL/DS (1980/81) and DB2 (1982/83). The
commercial version of Ingres also became available in the early 1980s. Sybase, Inc. released the
first version of its product in 1986, and in 1988, Microsoft introduced SQL Server. Many other
products by other companies were also released since then, but their share of today’s market is
minimal.
A brief history of SQL standards
The relational database model was slowly but surely becoming the industry standard in the late
1980s. The problem was that even though SQL became a commonly recognized database lan-
guage, the differences among major vendors’ implementations were growing, and some kind of
standard became necessary.
Around 1978, the Committee on Data Systems and Language (CODASYL) commissioned the
development of a network data model as a prototype for any future database implementa-
tions. This continued work started in the early 1970s with the Data Definition Language
Committee (DDLC). By 1982, these efforts culminated in the data definition language (DDL)
and data manipulation language (DML) standards proposal. They became standards four years
later — endorsed by an organization then known as American National Standards Institute’s
Technical Committee X3H2 (Database).
The name has changed several times. In the advent of the Internet, the characters
‘‘X3’’ (which stood for nothing at all; it was only an identifier) became part of many
people’s effort to locate pornographic (‘‘XXX’’) films, so ‘‘X3’’ has been changed to NCITS
(National Committee for Information Technology Standards). Just to be sure that nobody got it all
figured out, they changed names again just a couple of years later to INCITS, replacing ‘‘National’’
with ‘‘International.’’
SQL-86/87 (SQL1)
X3H2 was given a mandate to standardize the relational data model in 1982. The project ini-
tially was based on IBM SQL/DS specifications, and for some time followed closely IBM DB2
23