It's Your Content. Treat It With Respect!

Printer Friendly Version

In the decades-long discussion of how to manage data, one branch of the conversation has always confused me. Since the latter 1960s, the text community has grappled with the inherent contradiction that lies in all textual material; it must retain its rich, flowing nature in order to retain its meaning and coherence, but it must also be viewed as a collection of overlapping logical structures in order to be properly automated. This isn't much of a challenge for a reasonably well-educated human--the brain just works that way--but it has proven a real challenge for the computer industry.

Since 1969, when Charles Goldfarb matured the idea of placing logical tagging in the text and then creating the logical access structures as companions to it, the automation industry has had the essential key to this riddle. Goldfarb's brainchild, SGML, and its progeny XML, were brilliant attempts to create a richly identified and highly accessible text resource without destroying its essential flow and elegance. SGML had its problems realizing that goal, primarily because the software industry never saw value in creating affordable tools to work with the complexities of SGML tagging. Consequently, the value of SGML content, locked in its tagging structure, never became as easy to leverage as it should have been.

In the early 90s, one vendor, Open Text of Canada, developed an elegant SGML repository known as "PAT" (named after the particular indexing scheme it used.) PAT could make SGML act exactly as if it were a database, searching, accessing and locking any SGML structure without changing the original text document at all. But PAT was a product before its time and it failed to achieve the business success it needed to remain in the market. Even so, the path was clear--computer processes could successfully leverage structured text without turning it into something other than text. This vision lived on in the development of XML through the participation of Tim Bray, a key Open Text thinker.

XML solves the problem.

Extensible Markup Language, developed by people who had lived through the SGML wars, has made richly tagged text accessible and affordable. Partly through its simplified structure, but more likely due to its association with the burgeoning Internet and World Wide Web, XML attracted the attention and resources of the global software industry. The industry and its luminaries responded by generating a blizzard of inexpensive or free software designed to make content tagged in XML as accessible as a database or text index, without changing its structure. Through concepts like the Document Object Model, Xpath, Xquery and their supporting software tools, text can now be treated interchangeably as a coherent flow of thought or a logical map to the structures within it, all without changing it.