Databases Reference
In-Depth Information
tems. Improvement in personal productivity can be achieved, and applica-
tions backlogs can then be alleviated to some extent in relational systems.
Relational systems require much less maintenance
activity and thus reduce the cost of maintenance. Relational databases,
designed and implemented appropriately, can achieve data independence
when the applications do not depend on storage structure or access strat-
egy. Data independence, the use of 4GLs, and better relational program-
ming allow programmers to engage in less maintenance activity. However,
file processing methods may be appropriate if the solution operates as an
independent application.
Less Maintenance.
Relational technology can enhance a competitive posi-
tion for organizations. Relational data structures are becoming an increas-
ing technological choice, and almost all of current DBMS products are rela-
tional. Other emerging information technologies (e.g., 4GLs, SQL,CASE,
Decision Support System, executive information systems, and expert sys-
tem) have been developed using relational databases. Some organizations
even choose relational databases simply because they want to use 4GLs for
applications development.
Competitiveness.
PROGRAMMING IN RELATIONAL AND NONRELATIONAL
ENVIRONMENTS
During the past three decades, database software has evolved through
four phases: file, hierarchical, network, and relational. From the data
access perspective, programming with relational products(e.g., Oracle,
SQL/DS, database 2, and INGRES) is simpler and easier than programming
using nonrelational products (e.g., Integrated Data Management System,
Internet Multicasting Service, and traditional file systems). This section
compares programming in relational database environments with pro-
gramming in file processing and nonrelational database environments.
Relational Versus File Programming
In a relational database environment, it is the DBMS, not the program-
mer, that deals directly with the operating system's access method. A typ-
ical DBMS provides features that shield file organization from application
programs. Therefore, knowledge of how the data is actually stored is not
required for relational database programming. By contrast, in a traditional
file environment, knowledge of file organization (e.g., sequential file,
indexed files, and relative files) and access techniques (e.g., sequential,
random, and dynamic) must be built into the program logic.
In relational database programming, SQL statements are embedded in
programs primarily to perform I/O tasks. A nonprocedural language, SQL
Search WWH ::
Custom Search