HTML's Surprising History


Click here to change the theme.

This article provides some surprising information about HTML's ancestry that few people are aware of. Most people know that, until version 5 of HTML, HTML is an application of the Standard Generalized Markup Language (SGML). This article is actually a history of SGML.

Script/VS commands provide the details of how something is done and the GML macros specifed what is to be done.

Standard Generalized Markup Language (SGML)

You might however be unsure of what SGML is. SGML can be defined as a markup language for documents. So what is a markup language? I will get to that soon. SGML is ISO standard 8879.

SGML's parent is IBM's Generalized Markup Language (GML) developed by Charles Goldfarb, Edward Mosher, and Raymond Lorie. Although Charles Goldfarb helped develop IBM's GML, he also participated in the development of the SGML standard and wrote the book "The SGML Handbook". IBM called their software the Generalized Markup Language but actually it is not a language. So why did they call it a language? I will get to that soon. IBM's GML is (was) a part of their SCRIPT/VS word processing software.

The parent of IBM's GML is SCRIPT/VS. SCRIPT/VS is an IBM Mainframe word processing system. SCRIPT/VS documents were edited in a batch format in the sense that they were not edited interactively. What you edited was not what you would see. You would execute the SCRIPT/VS word processor to create a final document. The equiavlent for HTML is a parser; SCRIPT/VS was like a parser that took in an unformatted document and created a formatted document. GML was originally developed to provide many of the features that the original HTML does.

SCRIPT/VS includes a macro language. IBM's GML is a set of macros for SCRIPT/VS. The GML macros made it easier to do things like making lists, much like lists (such as the HTML OL tag) in HTML. Something else that GML was designed to do is to make it easy to format a document for various devices, and this is something that HTML alone does not do. CSS has been added to HTML to satisfy that requirement but CSS was not part of the original HTML and I think CSS does not do it as well as IBM's GML does. Using GML it was possible to format a document for a screen then later format a document for a printer then later create a different format for the same printer or a different printer and the details of the formatting was in GML; it was generaly easy to generate a variety of formats. To the extent that this capability exists for HTML using CSS, it took the HTML people a long time to get there whereas it was explicitly built into IBM's GML before the internet existed.

What did GML of IBM's GML first stand for? I did indicate that GML stands for Generalized Markup Language and that is IBM's official name for it, but originally it was named for Goldfarb, Mosher and Lorie (its developers). When IBM wanted to adopt GML as an official part of SCRIPT/VS they needed to come up with an official name. So they called it Generalized Markup Language but it is actually a set of macros, it is not a language.

References

Current Text ofISO 8879 (SGML)