Thad McIlroy - The Future of Publishing

By Thad McIlroy, Copyright © 2007
Written on December 20, 2007

Attending the world's most important annual XML conference is to be reminded that the XML club is still a fairly exclusive one. There are the insiders, folks who were mostly around back in SGML's heyday, and many can claim some degree of authorship of XML. Then there are the adherents. They might not have watched the pregnancy or attended the birth, but have applied themselves so diligently to mastering XML (and its many related standards, technologies and processes), that they're often tough to distinguish from the progenitors of the standard.

And then there are the Johnny-come-latelies of the XML world. They're still trying to get a handle on the beast, don't have insider status, and have not yet become sufficiently proficient to cavort with the XML masters.

That may be why the world's most important annual XML conference draws only 300 people, 50 of whom traveled to Boston from abroad in early December last year, including from Brazil, Australia, Taiwan and, of course, Japan (if there's no one from Japan in attendance at a technical conference you might as well go home: you've come to the wrong place).

The sponsorship of the event points to another aspect of the enigma which surrounds XML's current status. Microsoft was the "exclusive premiere sponsor" of the event. IBM joined as a "platinum" sponsor; Intel as a "gold" sponsor. This is impressive. But they represented 30% of the mere ten exhibitors and sponsors. Something is wrong with this picture. When three of the largest technology companies in the industry sponsor an event, why are there only seven other vendor participants? The answer was unclear.

Certainly IDEAlliance, the organizers of XML 2007, hinted at the problem when designating this year's conference theme as "XML in Practice." As conference chair David Megginson pointed out in the program guide: "XML has been with us long enough that we can stop singing its praises and, instead, take time to look at a few of the thousands of projects that use XML and other markup technologies on the web, in the enterprise, at the publisher, and nearly anywhere else you look in the technology world." There certainly has been lots of time devoted to abstract discussion of XML. Perhaps if people had a better handle on the benefits of the standard, and pragmatic ways to put it to work, this event would draw more attendees and sponsors.

The opening keynote entitled "Does XML Have a Future on the Web?" featured panelists C.M. Sperberg-McQueen (one of XML's some 200 parents), Doug Crawford (creator of JSON), and Michael Day, from a firm called YesLogic. The session was not the kind of opener that suited the conference theme; the issue seemed obscure. Crawford stated unequivocally that "XML is not going to replace HTML on the Web," while Sperberg-McQueen maintained that he did not want to have a "lossy transformation from XML to HTML. Browsers do support XML and XSLT, even if you don't see much of this in the wild."

The title could just as well have been "Does XML Have a Future Powering Automobiles?" XML is well-entrenched. Surely its core value lies in what it adds to text and data classification. How this gets displayed on the Web is of some lesser concern, if of any real concern at all. As a bookend to the discussion, MarkLogic's Jason Hunter provided the wrap-up session called "You're Darn Right XML Has a Future on the Web." But his retort went back to XML's primary value, and didn't spend much time with the distraction of how XML-encoded information is displayed in browsers.

There's a great deal of interest (and curiosity) about how Microsoft is deploying XML in Vista and in Office 2007. Instead of using its premiere sponsorship to further illuminate this complex question, Microsoft's theme was "Interoperability by Design," which appeared to be nothing more than a public relations efforts to calm those who oppose the OOXML standard as another attempt by Microsoft at world domination. One of sponsored sessions worked at demonstrating the interoperability of OOXML and Oasis ODF (Open Document Format). Another examined Ajax with XML and JSON. It took a third-party speaker to address the question of what's important in OOXML.

Matt Turner, a consultant to MarkLogic, led a session called "First Encounters with Office Open XML." The session audience was mostly composed of the curious. Turner pointed out what a huge improvement OOXML was over Office's previous attempt to "save as XML," which produced "the most verbose XML possible." In Office 2007 the user gets the real thing: "You can query it, transform it, the whole enchilada," he said. One of the intriguing benefits is that Microsoft Word is now an XML editor, something that is not loudly acknowledged by vendors whose businesses have been built on home-grown editors. During Q&A Turner's enthusiasm suggested a brighter future for new vendors entering the XML field. "I don't see anything in the building blocks that would prevent more powerful applications from taking full advantage of what Microsoft is providing," Turner proclaimed.

There was a great deal of coverage of the W3C standard XForms, which although some suggest will over time have a broader impact than just upon Web forms, seems still very much focused on this specific issue. That being the case I did not endeavor to get some shut-eye at the XForms sessions.

The conference focused also to a significant degree on DITA (Darwin Information Typing Architecture). There's an element of the "flavor of the month" about DITA, but equally a hint of what the future simplification of XML implementations might look like. By all accounts it's less complex and restrictive than DocBook (the flavor of the month for the past several years). At the same time, it's still directed mainly at technical documents, although its architects are starting to look beyond this restricted format. Markup expert Eliot Kimber's excellent fast-paced tutorial on DITA offered a range of pragmatic and esoteric endorsements of the format (which he has been very involved in developing and supporting), but perhaps as succinct a blessing as any was his statement that DITA "enables blind interchange of content without requiring a single, invariant DTD." Just as importantly DITA "makes Web-based delivery as easy as possible."

XML 2007 offered over 70 sessions and tutorials in four different tracks, far too many to cover in this report. Let's hope that IDEAlliance's increasing focus on the practical aspects of XML adoption will begin to attract the larger number of attendees and sponsors that this fine conference merits.


An edited and abbreviated version of this article appeared in The Seybold Report, January 10, 2008, copyright © 2008 by RISI Inc. Subscriptions (including the first four issues free) are available at http://www.seyboldreports.com/.