August 2000, Vol. 11, No. 8

XML-men

The markup language becomes almost as popular as the movie

Yuri Rubinsky would be pleased.

The late advocate of the Standard Generalized Markup Language was always perplexed that SGML hadn't been adopted more thoroughly throughout publishing.

Developed by IBM in the ’70s, SGML became an international standard in the ’80s. Designed to allow for "generalized" markup -- that is the enhancement of computer text files with additional information both about structure and format, independent of any specific computer or typesetting system -- SGML kind of poked along in the long-document space throughout the ’90s. The users were primarily governments -- or governmental-sized businesses -- as well as scientists and educators.

SGML was designed specifically to be flexible about what it did; the result was the need for heavy-duty computing power to translate an SGML document into a typeset document (see The Cole Papers, August 1993).

When creating the World-Wide Web in the early ’90s, the scientists at the Cern laboratory in Switzerland gravitated toward SGML to create a markup language that would not only transmit structure and typographic information, but also handle hyperlinks. The result was HTML or Hypertext Markup Language.

HTML is a specific instance of SGML; it is limited in that it is designed specifically to run in a web browser; a code like <H1> will always mean a big headline (though each browser can define typographically an <H1>).

As the World-Wide Web became more complex, it became evident that HTML was too limiting. The boys in the back room therefore came up with XML or eXtensible Markup Language.

XML is the middle ground between the free-form SGML and the rigid HTML. It is designed specifically to deliver information over the Web. It supports many of the constructs of HTML but it also is adaptable insofar as it supports a document type definition (DTD).

Which is a long way of saying: it’s for the Web, so it’s hot.

Newspapers, that particularly independent of publishing media, evidenced little or no interest in SGML in the ’90s. Of course they began to support HTML as they began to develop web sites, but that was that. They could see no need to transmit application-independent, computer-independent structured data.

The advent of XML, though, has brought about a revolution in daily publishing.

First, the International Press Telecommunications Council (IPTC), based in the United Kingdom and populated by newspaper and wire service executives from around the world, elected to use XML as the basis for it’s "next-generation" wire service standard (to supplant the infamous ANPA 1312, which was first issued in 1979).

The work of the IPTC bore fruit last month when it delivered its News Markup Language (NewsML) for evaluation. Editorial systems suppliers all have an answer for the question of how they would handle NewsML.

Inside, you'll read about the other newspaper initiative that uses XML -- the classified markup language. Correspondent Steven E. Brier chats with members of the task force writing the new standard as well as some discussion of how suppliers are planning to implement this new standard for classified advertising.

The third major need for XML comes in the arena of archiving -- now known as digital asset management (DAM). Correspondent Marion J. Love takes us through what the newspaper supplier industry is doing in the world of DAM -- or at least what they were saying they were doing at NEXPO 2000, held in San Francisco June 17-20.

Another acronym that’s been in your face a lot lately is PDF -- or Portable Document Format. Correspondent George Powell looks at what was new at NEXPO in using PDF as a pre-press workflow tool.

And if those aren't enough acronyms for you, try this: CTP. Copy Editor Aimee Beck takes us through the various aspects of computer-to-plate as it was laid out at NEXPO.

Rubinsky, who devoted a big portion of his life to SGML, died in 1996. He would be pleased to see that an SGML offspring had done so well.

-- David M. Cole, dmc@colepapers.net