The American National Standards Institute (ANSI) National Committee on Information Technology Standards (INCITS) Technical Committee on Database, DM32.2, was founded in April 1978. This committee meets six or more times per year.
Information Technology Standards are successful only when three conditions exist:
- A significant market share vendor community
- Publicly developed, available, and evolving standards
- Enforced conformance tests
These conditions are described in the paper, The Essential Paradigm for Successful IT Standards
Starting in 1978, DM32.2 was initially named, X3H2. It was renamed to just H2, and finally renamed to DM32.2. These name changes were merely administrative.
Upon its inception, DM32.2 considered hundreds of technical change proposals to the CODASYL 1978 JOD, and in August, 1984, finished work on the first American National Standard language for the network data language (NDL) databases. This standard, modeled after the CODASYL data model, was released for public review in January, 1985. It was accepted as an American National Standard in 1986. The NDL standard specifies the syntax and semantics for both schema and subschema of network structured databases, and specifies the semantics for a data manipulation language.
In late 1982, ANSI DM32.2 began to standardize a version of the relational data model through the IBM donated language, SEQUEL. Renamed SQL by DM32.2, basic SQL was completed and became an American National Standard in 1986. DM32.2 completed two “relational-like” data model revisions in 1989 and in 1992. Thereafter, DM32.2 expanded its focus from just relational-like data models to address the needs for object-oriented and non-relational data models. Those version were released starting with the SQL:1999, and SQL:2003 standards and have issued additional functionality for the SQL 2008, 2012, 2016, and is now working on the SQL:2020 version.
Since Database Languages were first introduced in the late 1950s, there mainly were three popular data models: Network (child has many parents), Hierarchical (Parent has many children), and Relational (independent simple-data structured tables). By the end of the 1970s these three data models become four with the introduction of the “Independent Logical File” data model. The ILF “files” had multi-levels of data structures.
Because of the popularity of ANSI SQL, almost all “non-relational” DBMSs disappeared from the market place.
However, with the advent of SQL:1999 and beyond, the data model capabilities of ANSI SQL have expanded to include the majority of the key data model features of Network, Hierarchical, and the ILF DBMS Systems.
The overall architecture of the SQL standard includes:
- Interfaces to compiler based languages such as Ada, C, COBOL, Fortran, JAVA, and through ODBC and JDBC to a wide variety of computer languages.
- Full referential Integrity.
- Multiple-language support, transaction management, two-phase commit, etc.
- Abstract data types, nested data structures within columns, and a number of object oriented facilities.
- DBMS generated “pointers” to greatly accelerate relationship processing.
- Codasys Data Model Sets among “records.”
- SQL/Multi-Medial “application support” for BLOBS, text processing, and libraries of scientific, geologic abstract data types, and the like.
- Remote Data Access protocols for generic distributed processing and for SQL specializations.
- SQL/PSM, the SQL programming language for persistent stored modules that are defined and controlled by SQL.
- SQL/CLI, a common call level interfaces that standardizes the CALLs interface between products from multiple vendors in the SQL family.
For a comprehensive description of SQL:1999 download SQLBOM application from the free downloads section of this website.