XML 2003

Volume 6, Issue 129; 12 Dec 2003

XML 2003 was a great conference again this year! It kept this markup geek happy.

Mankind are always happy for having been happy; so that if you make them happy now, you make them happy twenty years hence by the memory of it.

Sydney Smith

[n.b. The URIs for the online proceedings may not be available until after 20 Dec 2003. Sorry about that.]

XML 2003 was a great conference again this year. The US edition of the annual XML conference is aimed at a fairly broad crowd and I’m sure it’s a tough balancing act to keep the suits and the markup geeks equally happy. I think the organizers succeeded; at least they kept this geek happy.

I saw a lot of familiar faces, old friends and I met some folks for the first time, too. I met David Megginson at Tim and Lauren ’s soiree. I met Sam Ruby and Betty Harvey and Dare Obasanjo and probably a bunch of other folks, too. (Those are just the names that run through my head now, sitting here on the floor of the Amtrak station slurping power out of the wall; please take no offense if we met this week and I’ve failed to single you out.)

I saw David’s Strange Creations talk on the first day and predicted confidently that his text-adventure XML visualization tool would be the most entertaining thing I saw all week. James Clark talked about Incremental XML Parsing and Validation in a Text Editor (i.e., how he implemented nxml-mode). It satisfied the “yes, of course” metric that’s the hallmark of elegant solutions. (Elegant is not the same as “easy to implement,” it’s no less a marvel.)

Eric van der Vlist gave an overview of the ISO DSDL (Document Schema Definition Languages) effort and Makoto Murata talked about combining multiple vocabularies which is the goal of Part 4 of DSDL. I wish I could justify participating in that effort to my employer. James gave a related talk, Namespace Routing Language (NRL) that I definitely need to spend some time investigating.

On the last day, Uche Ogbuji gave me more reasons to learn Python and Henry Thompson described a re-interpretation of the XML Pipeline work that Sun submitted to the W3C last year. I think his assertion that the XML Pipeline work requires a processor to serialize intermediate results is wrong, but he’s probably right that the dependency graph approach is overkill. Most exciting: Henry says that a public release of their pipeline processor is in the works.

Finally, if your brain wasn’t full yet, Sam Ruby presented Atom in Depth a detailed introduction to Atom, the son-of-RSS. I think we can agree to disagree on escaped markup, he’s doing the right thing under the circumstances.

And that’s just a few of the talks that I saw. There were, and this is a good if unfortunate thing, talks that I had to miss because of conflicts. Independent sources tell me the most important talk I missed was Programming with Circles, Triangles and Rectangles a presentation from Microsoft Research that explored extensions to programming language type systems that provide native support for XML. Circles, triangles, and rectangles completely failed to convey how important the talk was, alas. But naming talks is hard. My own attempt at cleverness, Caching in with Resolvers , was in retrospect trying to hard to be clever.

Oh, and somewhere in there, I gave a tutorial on XPath 2.0 and XSLT 2.0, delivered my resolvers paper, and encouraged some bright folks to share their RDF projects at a town hall.

And if even that doesn’t seem like enough, the TAG announced the Last Call Working Draft of Architecture of the World Wide Web, First Edition at another town hall.

KISS and Kiss the Frog!

I’m not really very good at identifying the themes of a conference, picking up on the meme’s du jour. But I’m often asked, so I try to think of some sort of answer.

One of the themes was a backlash against complexity: an admonition to “Keep It Simple Stupid”. I don’t think this is really a new theme, I think there’s been growing anxiety about the complexity of XML specs for a while, but it turned up in several places this year. In the opening keynote, Jon Udell encouraged us to think about the social aspects of XML, which in retrospect seems harmonious with a desire for less technological complexity, Adam Bosworth’s keynote was bluntly critical of complexity, and Ludo van Vooren (whom I also just met at this conference) made a related point in his closing keynote: sometimes you should kiss the frog.

Ludo told the anecdote of a technologist who finds a talking frog. “Kiss me,” says the frog, “and I will turn into beautiful princess and be with you forever.” The technologist scoops up the frog and puts it in his bag. “Hey,” shouts the frog, “what are you doing. Kiss me and I’ll turn into a beautiful princess and be with you forever.” “No way,” says the technologist, “a talking frog is way cooler.”

We aren’t building technology for the sake of building technology. Sometimes it’s important to do the useful thing, solve the simpler problem, instead of always trying to do the biggest thing, the hardest thing, the coolest thing.

Good advice, I think.

Anyway, great to see you all! And see you all next year!


Great stuff. I've just seen the namespaces pic at Tim Bray's - frog or princess?

Point of comment - the links to papers seem to be 404'ing.

—Posted by Danny Ayers on 14 Dec 2003 @ 04:09 UTC #

Those are the links from the CD-ROM proceedings. Maybe the online proceedings aren't quite in place yet. If I get better information, I'll update the links.

As to the frog or princess, I'm not sure. That big screen of namespace declarations is just a consequence of aggregation. None of the actual data I work with has anywhere near that many.

—Posted by Norman Walsh on 14 Dec 2003 @ 09:09 UTC #

Hi Norm, Thanks for pointing to the Xen paper - really good stuff. James says (see comments to my blog entry) that Groovy is going to implement some of its features. Jon PS. You don't seem to have trackback implemented yet?

—Posted by Pixilated Jon on 19 Dec 2003 @ 11:26 UTC #

What would implementing trackback entail? How would it work and what would it do?

—Posted by Norman Walsh on 19 Dec 2003 @ 04:50 UTC #