Datasheet

SQL and Relational Database Management Systems 1
The development of relational databases was driven by the needs of big businesses for a medium
to gather, preserve, and analyze data. In 1965, Gordon Moore, the cofounder of Intel, made his
famous observation that the number of transistors per square inch on integrated circuits (IC)
doubles every year. Surprisingly, this rule still holds true. More powerful machines made it feasi-
ble to implement and sell RDBMS; cheap memory and powerful processors made them fast; and
perpetually growing appetites for information made RDBMS products a commodity, drastically
cutting their price down. Today, according to some estimates, less than 8 percent of the mar-
ket is being held by the database legacy ‘‘dinosaurs’’ — mostly because of significant investment
made by their owners more than 25 years ago, and their market share is constantly decreasing.
For better or for worse, relational database systems have come to rule on planet Earth.
Other DBMS Models
At the end of the 1980s, the buzzword was object-oriented programming (OOP). For very similar
reasons (memory requirements and processing power) as those preventing widespread adoption
of RDBMS, object-oriented programming did not take off until well into the 1990s. OOP lan-
guages are based on the notion that a programming (or business) problem could be modeled in
terms of objects.
While the code of the program remained practically the same, the way the code was organized
changed dramatically. It also changed the way programs were constructed, coded, and executed.
For programming applications that communicate with the databases, it would be only natural to
store objects on an as-is basis instead of disassembling them into text or into their component
fields and putting them back together when needed.
A modern RDBMS has the ability to store binary objects (for example, pictures, sounds, and so on).
In the case of object-oriented (OO) databases, they need to store conceptual objects: customer, order,
and so on. By the end of the 1990s, everything indicated that object-oriented databases were going to
be ‘‘the next big thing.’’ Surprisingly, since then, the object-oriented database hype has been gradually
diminishing. Even though there are several products on the market for pure OODBS (object-oriented
database systems), none of these has met with a widespread adoption, and it looks like the emphasis
has gradually moved toward OORDBMS (object-oriented relational database systems) that combine
object-oriented features within traditional RDBMS attributes.
Object-oriented features became part of SQL standards about 1999. All ‘‘big three’’
RDBMS vendors are at least partially compliant with the object-relational standards.
The other development worth noticing is a wide adoption of eXtensible Markup Language
(XML). XML was developed as a logical simplification of SGML (Standard Generalized Markup
Language), partly because plain static HTML (HyperText Markup Language) was focused almost
entirely on presentation, and a markup language that could accommodate structural and semantic
markup was needed. An XML document contains self-describing data in a platform-independent,
industry-standard format that makes it easy to transform into different types of documents, to
search, or to transfer across heterogeneous networks.
19