Extreme Markup Languages 2004 is behind us.
Let’s put the best news first: Extreme will happen again next year. It’s that good: every time I go, the very next thing I want to do is go again. (The conference is small, specialized, and very technical, consequently it struggles to meet its fiscal obligations; but it’s done so again this year!)
The venue hasn’t been negotiated yet, of course, but I hope we can return to the Europa. They’re doing some renovations, but I hope they don’t remove all of it’s delightfully bizarre charm.
As a group, we fit in well, and the hotel treated this assortment of a hundred odd markup geeks very generously: the wifi was fast and free and they even granted special requests.
Yes, that’s a bunch of us standing in the fountain. Yes, the photographer is standing in the fountain too. Someone else has a picture of him wearing Patrick’s hat standing in the fountain next to Liam wearing his hat. If you were at the conference, you’ll know why that’s funny. And if you took that picture, please send it to me!
We fit into the hotel intellectually, too. Consider, for example, their use of markup. My tutorial was in the “Incognita” suite. Let’s go find it, shall we?
This looks promising.
Ah, there’s “Mont Blanc”. There’s only one other door in this corridor: that must be “Incognita”.
And indeed it was. Last but not least, Cora’s, around the corner, makes a delicious and dramatic breakfast:
But seriously, it was another great conference. My general remarks from last year are true this year too, so I won’t repeat them. I also won’t try to give a blow-by-blow account, I think Elliotte Rusty Harold‘s got that covered (that’s not a permalink; where’s the permalink?)
That said, I can’t resist enumerating a few highlights. Tommie Usdin opened the conference with an admonition not to “pull the ladder up behind us”. Don’t try to stop other people from solving their problems just because they want to use a different solution than the one you’ve developed. A fair point, I think, though it’s a question of achieving balance among competing points of view, like most social problems.
James Mason talked about managing the resources (information, equipment, supplies, systems) associated with building…atomic bombs. It’s not a product that you get to test very often, so it’s really important to manage the process. Jim was using topic maps, something on my “must learn more about” list.
Bryan Thompson proposed a very clever way to implement server-side XPointer. It relies on a new kind of “range” at the HTTP level. It’s clever, I think it could work, and I don’t think it violates the principals of web architecture, but is it a good idea? I’m not sure yet.
Jeremy Carroll proposed a simpler RDF/XML syntax. ’nuff said. Last time I talked about RDF/XML, some folks thought I was just heaping abuse on it and that really wasn’t my point. On the topic of simplification, Eric Miller talked about extracting RDF automatically from other XML grammars using XSLT. Both Jeremy and Eric propose using (or in Jeremy’s case, I fear abusing) the XML stylesheet processing instruction. They really want a simple pipeline language. But I’ve already run that commercial a couple of times, so I’ll leave it alone for now.
Lots of folks said really interesting and exciting things about overlapping markup. I didn’t see all of it, but I was particularly delighted to see Wendell Piez showing some implementation.
Liam Quin gave a very fair and balanced account of the motivations for exploring “binary XML”. I’m not convinced, but I shouldn’t be trying to pull the ladder up behind me.
Matthew Fuchs talked about writing extensible and reusable XSLT 2.0 stylesheets, about coping with schema evolution. Interesting stuff. He also reminded us just how much of the design of W3C XML Schemas was influenced by some of the work he was doing at the same time. I’m not sure I’d have the courage to admit that in public if it was me. (Just kidding, Matt ;-)
Extreme was a two-track affair most days and one of the hardest choices I had to make all week was between David Birnbaum’s presentation of Interpretation beyond markup and Anne Brüggemann-Klein’s presentation of Balanced context-free grammars, hedge grammars, and pushdown caterpillar automata. I chose Anne, and I don’t regret it, though I also don’t claim to have understood all of it. There are some seriously technical papers at Extreme.
Lou Burnard, Syd Bauman, and Julia Flanders discussed the TEI’s next generation version of the ODD system. ODD (One Document Does it all) is the literate programming system that is used to build the TEI and its documentation. They’ve all got me thinking seriously again about doing DocBook that way, maybe even with the same markup.
On the closing day, Simon St. Laurent shined his flashlight into another dark corner, as he is wont to do. This time it was the continuing problems of dealing, in a post-DTD world, with what we used to do with entities. Simon’s not doing as much with XML as he used to and that’s a real loss. We need people who aren’t afraid to pull the skeletons out of our closet.
Sam Wilmott presented a Python library for doing pattern matching using a paradigm that’s very different from our customary, off-the-shelf regular expression techniques. Cool stuff. Also on my “must learn more about” list.
Michael Sperberg-McQueen’s closing keynote challenged us to think hard about what it means for XML to have a model. As the editor of the data model specification for the XSL/XML Query specifications, I was quite curious about what he’d say. But let’s face it, I’d have been curious to hear what Michael had to say about anything. What did he say? I’ll paraphrase: it is not only proper for there to be more than one model of XML, it is necessary. I hope someone transcribes his keynote, I know it was recorded. And for the record, I vote we make it a tradition for Syd to introduce him!
Hmm. That was more than a few, wasn’t it? Oh well.
I wanted to do an “Overheard at Extreme” essay, but I’ve come up a bit short. I missed a few things; it’s hard to write them down when your actively engaged in a conversation and hard to remember them after. Here’s the best of what I got.
“Would you like a smaller spiky one for formal occasions?”
“Do the snoopy-happy dance.”
“It's like toast.”
“It’s an attempt to evoke model envy.”
“Never use a more powerful mechanism than you need to.”
See you next year!