W3C Working Groups in Florida

By Michael Kay on February 03, 2006 at 02:18p.m.

I'm just back from spending most of the week in St Petersburg, Florida, at meetings of the W3C XSLT and XQuery working groups. Someone asked on xquery-talk how the meetings went: the answer is that at this stage of the end-game, you have to be a bit of a masochist to take part (and this no doubt explains why attendance is dropping at each meeting). Essentially the specs are finished -- one might say they were finished years ago -- and it's largely a question of fixing bugs in obscure edge cases. I probably cause a certain amount of friction by being the one who seems to find a lot of these bugs -- or in some cases, to pass them on from Saxon users.

There were a couple of more meaty technical subjects on the agenda. One was the question of the "xdt" namespace, which holds definitions of types such as xdt:dayTimeDuration and xdt:yearMonthDuration. The question on the table was whether these types should be moved into the XML Schema namespace. This is as much a procedural/political question as a technical one, which has significant implications for users both short-term and long-term. I think we made the right decision, but I won't reveal it until the public announcement is made.

The other substantive question concerned the interaction of the different options for controlling how text gets serialized: that is, the order in which operations like character mapping, URI escaping, XML escaping, and Unicode normalization take place, and whether data that's subjected to one of these operations then continues to be processed by the others. I doubt that many users will notice the differences (although it's probably one of those cases where they are likely to notice if we get it wrong). The interactions are pretty subtle and it's going to require a fairly careful look at the Saxon code to see if it reflects the spec correctly.

It was nice to have some sunshine, though there was none on the Monday which was my day off! Oddly, with temperatures in the twenties Celsius, the flight back to the UK was delayed for an hour or more by ice on the wings of the aircraft.