Glossary · 3 · Publishing SDKs and structured content
Standard Generalized Markup Language (SGML)
Also known as: SGML, ISO 8879
In document processing, the Standard Generalized Markup Language (SGML), ISO 8879:1986, is an international standard meta-language for defining markup languages: it specifies how document types are declared in a document type definition (DTD) and how documents mark up their structure and content independently of how they are formatted.
- Publishing SDKs
- Standards
In one sentence
SGML (ISO 8879:1986) is the meta-language for defining markup languages through DTDs; it grew out of IBM GML and is the ancestor of XML and HTML.
Example
A defense supplier of the 1990s delivered its maintenance manuals as SGML documents valid against the customer's DTD, so the customer could publish them on paper and in its own electronic viewers.
How it applies
- Publishing SDKs: SGML established the principles that today's publishing toolkits still rest on: content is marked up by what it is rather than how it looks, document types are declared formally and checked by a parser, and formatting is applied afterwards by separate processing. These are the foundations of structured authoring and of DITA, which is defined in XML.
- Document type definitions: The DTD syntax of XML comes from SGML. A DTD declares the elements, attributes and entities of a document type; XML DTDs are a simplified subset.
- Legacy content: Large SGML collections remain in sectors with long product lifecycles, such as aerospace, defense and publishing. Converting them to XML is usually possible because XML is a subset of SGML, but SGML features that XML dropped (for example omitted tags) must be resolved first.
- Developer documentation: Many documentation formats used for software, such as DocBook, started as SGML applications and moved to XML.
History
SGML grew out of the Generalized Markup Language (GML), which Charles Goldfarb, Edward Mosher and Raymond Lorie developed at IBM from 1969 so that text editing, formatting and information retrieval systems could share documents. Standardization work led to ISO 8879, published in 1986 and maintained under ISO/IEC JTC 1/SC 34. HTML was originally defined as an SGML application, and XML, a W3C Recommendation since February 10, 1998, is defined in its own specification as “a subset of SGML”.
SGML vs. XML
SGML is highly configurable: its declaration lets a document type change delimiters, allow omitted tags and use other minimization features. XML fixed one syntax, removed most of these options and made documents parseable without a DTD (well-formedness). This simplicity made XML easier to implement and to use on the web, and it replaced SGML in most new work.