After the various buzz about Clay Shirky's article (which I haven't read directly) and spreading the social software quiz, I thought I'd try making a program to convert LiveJournal userinfo pages into FOAF documents. The HTML-scraping LJ part was bad enough, but compared to that, building the FOAF is a brick wall. While Paul Ford makes a convincing argument, that doesn't make rdflib any better documented or more understandable (or less than 168 KB; I hoped TRAMP would help lighten it up but it's a layer over rdflib itself).
So I'll probably get back to my Blackjack tournament player for the AI final instead.
