SGML Enables "Sustainable" Web Publishing

by Dick Shippee

At a recent Bay Area Roundtable discussion sponsored by the University of California at Irvine, an audience of about 30 IT professionals heard Robert Glushko present an enlightening talk on the role SGML (standardized general mark-up language) plays in providing a sustainable or revisable Web publishing environment.

Glushko, the cofounder of Cupertino, CA-based Passage Systems, stated that "handcrafted" Web publishing--and by that he meant primarily HTML--is the norm today. This kind of Web publishing he calls the "Up-and-Running Strategy." Publishers get started quickly, using ad hoc tools without standardization or automation in order to create professional-looking Web pages. This strategy is low on up-front costs and high on immediate results. That is why it has found favor with so many Web publishers.

A problem arises, however, when Web publishers want to reuse, revise or retarget their material. With the "Up-and-Running Strategy" there is no document database and no way to tap into a standardized mark-up that allows for documents to be reused. This is where the strategy of "sustainable" Web publishing comes in, based on an SGML backbone. SGML, ISO 8879, defines syntax for representing the structure and content of textual information, complete with "hooks" for nontext. It defines the grammar (or mark-up language) that specifies logical rules for a given document, and defines the set of elements that encodes the key distinctions in textual information.

Glushko feels that using SGML is the only way publishers can profitably tackle large-scale Web publishing projects. Automation is the goal, and Glushko's mantra, which he repeated over and over during his entertaining talk, was "Standardization and validation are the keys to automation."

Using this mantra, Glushko set out a "Sustainable Strategy" for a Web publishing project::

Glushko went on to provide key factors to help pinpoint cases where the "Sustainable Strategy" of using SGML-based Web publishing is called for:

  1. The information being published has a long useful life and will bereused, revised and/or retargeted.
  2. The information is valuable because of its content, not its appearance.
  3. The information is highly structured and can be thought of as a "document database."
  4. The information is in large volume and high frequency, both of which suggest that automation will equal a payback.
  5. The information does not have to be up on the Web right now, this minute, and therefore the publisher can make an investment today that will pay off over time.

SGML at UniForum

Many Web publishers, including UniForum, have need of both strategies. The Association produces UniNews in text on the Web and with Adobe Acrobat hooks. The goal is to get the information out to as many people as we can in a quick turnaround time. We also publish the 1,800-page, text-heavy Open Systems Products Directory, an ideal candidate for automated SGML-based Web publishing. We will be revising the document as new listings come in and reusing it in new editions. We can also retarget it by locating specific segments of the open systems audience and sending tailored portions of the Directory to them.

In future issues of UniNews, we will report on the various Web publishing projects of the Association and how they fit into the strategies of "Up and Running" or "Sustainable Web Publishing." We'd also like to hear how your company is managing its Web publishing strategy. For instance, are you taking reuse and revision into account for your marcom materials (being mindful of the up-front costs of SGML), or is it full-speed ahead to get Web pages up? Tell us your stories and we will share them with the UniForum family.

For more information about how the "Sustainable Strategy" works, contact Robert Glushko at glushko@passage.com or at http://www.passage.com.