I started my visit to Montreal for Extreme Markup Languages 2006 very early, and at the moment (although probably not by the time I've finished this) I'm sitting in on a continuation of today's daily polemic (opinion statement) that has a significant subset of the conference's attendees rather stirred up. In his polemic, "XML for publishing", Joe Gangemi argued that XML technologies do not adequately support the needs of publishers. He clearly wanted more time to vent, so we all decided to get together for a nocturne (late impromptu session). I went to get food first, which turned out to be a good idea.

Interestingly, we've gotten somewhat off topic onto another issue in this space: usability of markup by actual authors and the degree to which software interfaces (for XML editing) are usable by these authors. Earlier in the day, Peter Flynn gave a talk analyzing the current state of usability of XML editors, concluding that none are adequate for the markup novice. I'm hoping someone points him to XMLmind and Serna, which I think deserve some attention. XMLmind even specifically implements one of the features (a drop-down menu from a formatting button) that someone specifically mentioned as lacking. Of course, XMLmind has a lot more to offer, as well. It doesn't do that nifty multi-enter ancestor-walk-and-element-cycle thing that someone else mentioned; I wonder if I could implement it as a macro? At the moment there's a lot of discussion about the nature of authors and whatnot, so I decided to start this entry until they get back to tech issues.

Eventually we got back to the tech. Joe wants some sort of functionality like exclusions from SGML and entity support at a level above core XML. The group finally managed to convince Joe that reincorporating SGML features directly into the XML spec was not going to happen, and everyone generally agreed that the current reality of the universe is that markup space has moved from a large black box that solves a number of problems (SGML) to a set of small black boxes, one of which forms the basis for the rest (XML), and the rest of which can be used in a modular fashion with core XML. Personally, I think they're more transparent boxes than black, which is another reason why I prefer XML to SGML.

Other than that excitement, it was a generally good day. Dimitre Novatchev kicked things off with a bang with his exposition of how to (ab)use XSLT 2.0 as a fully functional language (no pun intended, I think). I'm excited to look at his older work with XSLT 1.0 as well, which he reviewed briefly. Steve Pepper made a convincing argument that mapping between RDF and Topic Maps is tenable, which I find to be an exciting result. Allen Renear gave an interesting talk about fitting XML documents into the FRBR model that made me want to read Guarino & Welty. Norm Walsh brought us up to date (that date being just a few days ago) with the very latest on XProc, which I think will become an important member of every markup geek's toolkit. With the other activity described earlier, that about wraps up my day.

Can I sleep now?