Home from XML 2006

New things for the future, interesting things from the past.

Since my last posting, some weblogs have mentioned that I was blogging the XML 2006 conference, so I feel bad that I haven’t gotten to my second posting about it until after the end of the conference. Most of my time sitting at a computer in Boston was spent on a project for a client, and there was enough of this that I had to skip several talks that I wanted to see. (For a little multi-tasking, I reviewed some project documents while Jason Hunter discussed Web Publishing 2.0. Jason was more interesting.)

There were plenty of presentations and conversations that gave me a lot of good ideas. Fabrice Desré’s talk Building Dynamic Applications With Mozilla, REX and XQuery has me looking forward to trying to build an application around Mozilla and the other components he described, and the discussion after my presentation gave me ideas about ways to build onto my OWL/RDBMS integration demo. An interesting point that came up in the questions after my talk was when Claudia Lucía Jimenez-Guarin (who had spoken on a panel about Agile XML Development that I chaired) asked about trying to integrate data from a different ontology into your data. At first I wasn’t sure what to say, and then I remembered that much of the point of my talk was that OWL’s ability to describe relationships between data from different sources is much of its power, because (with the right query engine or other software that can understand those relationships) it’s the key to using data from different sources as an aggregate whole that is greater than the sum of its parts. Once the paper and slides for my talk are available on the web I’ll include links here.

I’m also going to stay in touch with Ken Sall and Ronald Reck to learn more about their Applying XQuery and OWL to The World Factbook, Wikipedia and Project Gutenberg project. Unfortunately, their presentation on integrating different data sources with OWL took place at the same time as my presentation on integrating different data sources with OWL. I’ll write more about theirs here once I learn more.

The annual Docbook dinner organized by Norm Walsh was a lot of fun. When someone there mentioned that he wasn’t editing XML in Emacs with nxml mode because of his fondness for the keystrokes in the psgml mode for editing SGML (I’ve written a full book chapter about this, and that chapter is available for free) I said that I had written some Emacs macros to make nxml fill in some of the psgml gaps. After reviewing my .emacs file, I’m not sure which macros those are, so I just posted several candidates if anyone’s interested.

In Jon Bosak’s closing keynote of the conference, he told some great stories about XML’s birth from the inner core of the SGML community as he set the stage for a discussion of the current state of markup technology innovation. One story, concerning Charles “Father of SGML” Goldfarb’s insistence that even documents with no DOCTYPE declaration have an implied DTD, included references to Kant and Ben Jonson and had the old school SGML people laughing so hard that IBM’s Sharon Adler nearly choked on her drink. (When working on my XML: The Annotated Specification book for Charles’ Prentice Hall series, I remember long battles over SGML-rooted concepts that he insisted were implied in XML but that I kept pointing out were never mentioned in the specification that I was annotating. I certainly learned a lot from him though, while working on that and, before that one, a book on free SGML software that included the chapter on psgml Emacs mode.) There was lots of laughter and knowing nods from the veterans as Jon brought up SGML complexities that were painfully factored out by the working group. I wonder if those present at the dinner who hadn’t been working with XML as long were a bit puzzled by these reactions to talk of debates about DOCTYPE declaration syntax. They certainly didn’t double over at the mention of debates about whitespace handling like the people at the front tables did.

In general, it was great to catch up at the conference with other old friends and some former and current LexisNexis employees, and to get confidential opinions from key players in the field on new developments. Some regulars and semi-regulars who were missed at the conference this year included Eve Maler, Tim Bray and Lauren Wood, Paul Prescod, Uche Ogbuji (although it was great to finally meet his brother Chimezie), Dale Waldt, Eric Freese (who has a grander project under way), Edd Dumbill, Sean McGrath, Rick Jelliffe, Henry Thompson… it was a smaller conference, and with the opening and closing theme being the tenth anniversary of the announcement that XML even existed, this conference reminded me of those days, when the same annual gathering (then named “SGML 19yy”) was so much smaller than it got during the dot com boom.