On a slightly different note, went to the American Museum (just down the hill from where Isobel lives at the University of Bath) last Saturday to see an open air production of Sheridan’s The Rivals (poor quality etext available from Bibliomania) by Bath Drama. I’d not seen any Sheridan before, and didn’t realise that he was responsible for Mrs Malaprop (of malapropism fame), nor that The Rivals was set in Bath (which made for interesting references to duels on Kingsmead Fields, and so on). Despite my usual hayfever, I greatly enjoyed the play (a `Restoration Comedy’, which seems to be… Read More »What kind of a place is this Bath?

Back in Southampton after the Bristol ER-SWAD meeting. Meeting was enjoyable and worthwhile, with some interesting demos of Semantic Web image annotation software, including Dan and Libby’s codepiction service. Rather nifty. Today’s going to be a bit busy, since I’ve now got to make up time on the presentations and demos that other people are going to be presenting on my behalf next week. Gah. Lots of work.

It’s been a busy couple of days. Spent most of Wednesday and Thursday preparing a presentation for a research group awayday (design brief: be funny and serious), then went up to Coventry on Friday with Isobel to visit friends and go on the annual all-day barcrawl of the Warwick SF society, shortly followed by Issy’s birthday on the Saturday. Tomorrow (Monday) I’m off to a workshop in Bristol which runs until Wednesday (and which clashes with the awayday, so I won’t actually to see the fruits of my efforts). Back in the office on Thursday to do some frantic preparation… Read More »

Bah. The Moz bug has been scheduled as ‘Future’ (i.e. we’ll fix it when we’ve got through all the numbered releases that we’re currently planning). I can’t entirely blame them, since the XUL template code is about 24,000 lines of C++ with little commenting. If there was some documentation on the algorithm that they’re using for matching and conflict resolution, I’d be very tempted to have a go at fixing it myself, but as it stands I can’t justify the investment of my time to piece through the intricacies of the code.

Filed my first mozilla bug today, having verified that the strange behaviour I was seeing was in fact an error. It’s probably an open question as to whether this is a bug, a feature of an enhancement given the comments in the relevant bit of source: // XXXwaterson Some day it would be cool to allow the ‘predicate’ // to be bound to a variable. On the plus side, I’ve got an easy workaround, even if it isn’t as neat as my original solution.

As I mentioned last week, I’ve been playing around with an add-on to Mozilla which annotates web pages against an RDF knowledge base. To cut a long story short, the software sends a request to a server for an RDF fragment, then uses Mozilla’s XUL templates to query the RDF fragment and transform it into a user interface. In theory, anyway. While I’ve been quite happy with Mozilla as a development platform in general terms (the combination of the javascript debugger and the DOM inspector being a formidable pair of tools), the RDF/XUL template jiggery-pokery has to be one of… Read More »Zen and the Art of Mozilla Hacking

In other news, very short WebOnt conference call this evening – barely half an hour, compared to the normal ninety minutes. Most surprising.

So, I finally gave in and succumbed to to the LiveJournal behemoth (thanks in no small part due to the inestimable h-l). I probably won’t update this with any great regularity, since I’m mostly here because I have a vague research interest in the design of more semantically rich weblogs. I’ve been working on a Semantic Web-based annotation client built into Mozilla for a while (nicknamed ‘Annonto’ until I think of something better, or publish), the purpose of which is to associate text strings found in web pages (eg. “Nick Gibbins”) with the entities from an RDF datasource to which… Read More »