Bright Year In Prospect For XML

January 16, 2002

Edd Dumbill

And Now... The Good News

Having spent the last column happily making fun of XML in 2001, I thought I would turn my hand this week to predicting some of the forthcoming year's happy XML moments. It's often all too easy to predict gloom, but there are some really promising areas in XML that I hope to see blossom this year.

Many encouraging developments are happening in the area of user interface and presentation technologies. Readers of this column will recall my positive feelings towards SVG, which appears to be going from strength to strength. Thanks to Adobe's bundling of their SVG viewer with Acrobat Reader, and their arrangement with Real, there's a good chance that SVG support will be increasingly installed by default on users' machines. As a consequence of integration with the DOM and scripting, SVG images can play a real part in web pages, as opposed to the "drop in" nature of Flash animations and most Java applets. XSL Formatting Objects are also poised for rapid growth this year. XSL-FO solves a key problem: getting data from the Web and onto paper while preserving a high degree of fidelity.

The trouble with XHTML has always been that it's not offered enough that is new to make it worthwhile implementing in the browser. Now, with a plurality of client-side XML vocabularies joining the worthy but not-quite-rock-star-material MathML, the benefits of XHTML start to mean something. It is perhaps not a vain hope that the excellent rendering engine from the Mozilla browser could be used to implement XHTML modularization, and add support for rendering ad hoc vocabularies for which it is given a CSS stylesheet. Along with many, I've previously emphasized non-PC devices such as phones or set-top boxes as major areas of deployment for XHTML and SVG, but this year I hold out the (perhaps irrational) hope that desktop browser technology can start to move forward again.

Perhaps the key technology to kickstart browser improvements will be XForms, a richer widget set and interaction model for browser-to-server communication. Half the crazy Javascript hackery of today wouldn't be a problem if XForms had been around a few years back. Anyone who's written a non-trivial web application will find XForms a welcome addition.

Another area of XML that is becoming increasingly encouraging is that of grassroots innovation. For better or for worse, we've spent a large portion of the last two years with eyes glued to the machinations of the large companies and the W3C. Last year, I spoke of the need for "XML heroes" and that many of those from XML's early days appeared to have tired. At the XML 2001 conference in Orlando last December, I had a strong sense of a renaissance of individual contributions and thought. This was led, not surprisingly, by James Clark and his prominent re-emergence into the XML hurly burly through the TREX and RELAX NG schema languages.

This renaissance coincides with the tightening of the focus of the software big boys -- IBM, Microsoft, and, to some extent, Sun -- on "web services" as their primary expression of XML. Now these companies have found on which portion of the field they want to play their game, the rest of it is open again for those who got squashed before. RELAX NG is the poster child for grassroots power in XML and is proving an inspiration for the community.

One sign of this confident thinking is that people are willing to reconsider the definition of XML itself. Folk such as Tim Bray are joining the swell of voices seriously considering an XML version 2.0 that, far from playing the usual version numbers game, goes back, cleans up, and creates a more long lived-foundation for XML. Such notions are worthy indeed, and usually right at the bottom of any purely corporate agenda.

XML has been around for five years now, and a lot of the hype is dying away. I hope that the willingness to re-evaluate and learn from experience will bring a new dimension to XML this year and deliver further on the original promises and vision. One high point is the work done by Rick Jelliffe, who has consistently challenged conventional thinking in the XML world. Rick may well be on the verge of a big improvement to one of the thorniest areas for users of XML, editing. His company's editor, previewed at XML 2001, accepts the reality of the editing situation, that documents don't start off or necessarily remain valid and well-formed, even though that is the ultimate aim. The quickest route from unordered thoughts to a complete document may not take you through well-formed or valid paths.

During the short dot-com boom years, we all fooled ourselves that it would take two years to know all there is to know. Despite the harder circumstances many now find themselves in, the bursting of the bubble is not bad news for XML: we may have time to do things properly after all. What has become apparent is that XML is a long game, and one that has certainly not lost its ability to deliver surprises.

XML at the O'Reilly Open Source Conference

O'Reilly XML editor, Simon St. Laurent, has just issued the call for participation for XTech 2002, the XML stream at the O'Reilly Open Source Conference. The conference is being held from from July 22-26 in San Diego, California. Simon writes:

We're hoping to continue last year's success with presentations which either highlight Open Source work within the XML community or demonstrate XML's usefulness to the broader Open Source community. (There is also a separate track devoted to the Apache XML Project's work.)

Information on submitting proposals is available on the O'Reilly conference site.

There are a few key areas that would be especially interesting, though the conference will certainly go beyond these areas:

• Schemas - W3C XML Schema, RELAX NG, Schematron, etc.
• Presenting XML - SVG, XSL-FO, CSS, XHTML
• Modularization - XHTML, SVG, SMIL, other approaches
• Web Services - too many acronyms to list
• i18n/l10n - Unicode, making XML work in non-Unicode environs, etc.

Presentations should be aimed at a 45- or 90-minute time slot, though full day and half day tutorials are another option. There's a definite bias toward Open Source at this show, but more general XML presentations are definitely welcome. Marketing talks will be rejected outright, and speakers must submit their own proposals.

Proposals are due March 1, and presentations will be due April 29.

Fragments

• January's Sensitive Corporation of the Month award goes to IBM, which sent me a press release bragging about their record achievement in filing patents: "patents related to IBM Software technologies were included in the record 3,411 patents awarded to the IBM Corporation last year." They go on to mention that "IBM's number surpassed the combined total of patents awarded to 12 of the largest U.S. companies in the IT industry -- Hewlett-Packard, Intel, Sun, Microsoft, Compaq, Dell, Apple, EMC, Oracle, EDS, Accenture and AOL." Well, that'll help IBM's image in the XML world then.

• A warm welcome to the W3C's new Chief Operating Officer, Steve Bratt. I received notice of his appointment from the kind people at the W3C, who invited me to "see a photo of our new COO." I hope Mr. Bratt will forgive me for saying that he does look rather startled at his newfound fame. Reading his brief bio, I find that the W3C's COO has an eminently suitable job history, having formerly worked at the Comprehensive Nuclear-Test-Ban Treaty International Data Centre in Vienna.

• Happily, for me at least, folks in the XML world took the 2001 Anti-Awards in good humor. Well, almost. One correspondent writes to inquire if the <taglines/> logo for this column implies an empty content model, while VoiceGenie, worthy winners of the Bluestone Award for Aggressive Press Releasing, has written to pledge that I will never face a Monday morning without a communication from them.

As usual, comments and suggestions for <taglines/> columns are welcome, send them to me at edd@xml.com.