From SEQUEL to JSON: Tracing the Evolution of SQL Standards
This article chronicles the origins, standardization process, historical milestones, learning resources, and conformance challenges of SQL standards, highlighting key ISO/IEC versions, the roles of major vendors, and the growing complexity of the language over more than three decades.
SQL Origins
SQL, commonly known as Structured Query Language, originated from the SEQUEL language used in IBM's System R prototype, which explains why many pronounce it as “sequel.” Although some humorously expand it to “SQL Query Language,” it quickly became the de‑facto standard for relational database access, supported by commercial systems such as Oracle, DB2, Sybase, SQL Server, as well as open‑source databases like PostgreSQL and MySQL, and even smaller products like Access. The rise of NoSQL initially claimed to abandon SQL, but later adopted the “Not Only SQL” stance, acknowledging SQL’s enduring relevance.
SQL Standardization
The most influential bodies shaping the SQL standard are the major database vendors, while the actual specifications are drafted by non‑profit organizations such as ISO and ANSI. National standards often mirror ISO/ANSI specifications; China, as an ISO member, contributes translations of international standards. To avoid vendor‑specific terminology, the standards abstract many concepts, which can increase reading difficulty, especially in translated versions. Practitioners usually prefer the original English documents. Drafts are publicly accessible, and the final ISO standards, though not freely downloadable like RFCs, can be obtained through ISO’s website.
History of SQL Standards
Key milestones in the development of the SQL standard include:
1986 – ANSI X3.135‑1986 / ISO/IEC 9075:1986 (SQL‑86): first ANSI standardization.
1989 – ANSI X3.135‑1989 / ISO/IEC 9075:1989 (SQL‑89): added integrity constraints.
1992 – ANSI X3.135‑1992 / ISO/IEC 9075:1992 (SQL‑92, also called SQL2): introduced a tiered classification.
1999 – ISO/IEC 9075:1999 (SQL:1999, SQL3): major overhaul adding object‑orientation, regular expressions, stored procedures, Java support, and a new definition of conformance.
2003 – ISO/IEC 9075:2003 (SQL:2003): added XML support and window functions.
2008 – ISO/IEC 9075:2008 (SQL:2008): introduced TRUNCATE.
2011 – ISO/IEC 9075:2011 (SQL:2011): added temporal data.
2016 – ISO/IEC 9075:2016 (SQL:2016): added JSON support.
From SQL:1999 onward the standard’s naming switched from a hyphen to a colon, and the year format changed to four digits to avoid the Y2K issue. The size of the specifications grew dramatically: SQL‑86 was only a few dozen pages, SQL‑92 about 500 pages, and SQL‑99 exceeded 1,000 pages, making comprehensive mastery impractical.
SQL:2003, for example, is divided into nine parts, covering framework, foundation, CLI, persistent stored modules, external data management, object language bindings, schemata, Java‑based routines, and XML specifications.
Learning the SQL Standard
The technical committee JTC1/SC32 (joint ISO/IEC) is responsible for drafting the standards. Because the relational database market is mature and competition limited, the standards have become increasingly large and the development cycles lengthy, leading to debates about their relevance. Recommended reading strategies include a thorough study of SQL‑92 (the core language) followed by incremental reading of later versions. Technical reports such as the SQL Technical Reports (numbered 19075 with “TR” in the title) are freely downloadable from ISO and provide valuable insights.
SQL Standard Conformance
Conformance varies widely among products; a vendor’s claim of “SQL‑compliant” does not guarantee seamless migration. While major commercial systems (Oracle, DB2) and open‑source databases (PostgreSQL, MySQL) generally exhibit high conformance, many products only implement the core subset defined in SQL‑92. NIST once ran a compliance testing project, but after SQL:1999 the standard stopped using graded conformance levels, moving to core and feature compatibility without a formal certification program. Some efforts, such as MySQL 6.0’s attempt to improve compliance, were abandoned, and newer domestic databases have established their own testing bodies.
1 http://standards.iso.org/ittf/PubliclyAvailableStandards/index.html
2 https://www.wiscorp.com/free_downloads.html
3 http://www.jcc.com/resources/sql-standards
4 http://www.sigmod.org/publications/sigmod-record/1203/pdfs/10.industry.zemke.pdf
5 https://s3.amazonaws.com/artifacts.opencypher.org/website/materials/SQLStandards_2017-10-20.pdf
6 https://modern-sql.com/blog/2017-06/whats-new-in-sql-2016
7 http://itl.nist.gov/div897/ctg/sql_form.htm
8 http://mall.cnki.net/reference/detail_R201109078.htmlTencent Database Technology
Tencent's Database R&D team supports internal services such as WeChat Pay, WeChat Red Packets, Tencent Advertising, and Tencent Music, and provides external support on Tencent Cloud for TencentDB products like CynosDB, CDB, and TDSQL. This public account aims to promote and share professional database knowledge, growing together with database enthusiasts.
How this landed with the community
Was this worth your time?
0 Comments
Thoughtful readers leave field notes, pushback, and hard-won operational detail here.
