It does indeed seem that vendors have spent the last 8 years pushing an Esperanto for integration in the shape of Web Services. Well actually they've built a series of weak agreements on how to mint, describe and exchange auxiliary languages, encourage customers to build their own vocabularies, then disenfranchise them when they happen to use something from a competitor.

Meanwhile the cool Web 2.0 kids have been playing and rolled out "good enough" patterns of use and pigin languages for integration on the Web in the shape of RSS/Atom and HTML/Microformats. OK, they don't have much of a story for authentication, yet, but it's only a matter of time before all this Tok Pisin evolves into a creole, proves white is black and the current computing establishment disappears in a puff of logic.

And if all that wasn't enough to scare the incumbent wisdom, up in the sixth form there's a bunch of cool science projects demonstrating how to put all your data into a single unified form. I'm enormously attracted to RDF, not least because I want to embed data flatly into a feed, geoRSS style. Open content is definitely the way to go for versioning. I also love the notion of triple stores. However after trying and failing to get triplr to work beyond the trivial, I returned to scraping HTML with dapper and regex'ing the text with Yahoo! Pipes. Show me some useful, fun, active, live triples out there in the wild, beyond RSS 2.0 and MusicBrainz and I'll buy 'em!