>> http://bbc-hackday.dyndns.org:2825/ >> - the whole thing's stitched together with MusicBrainz artist ids
> Theoretically, it should be possible to stitch www.bbc.co.uk/music/ into > this, too. we're currently working to make this so. the later + totp stuff is really just a prototype of how we could handle linked music data across the beeb > That uses Musicbrainz data, ish. the tool used to generate the /music site has brainz ids as foreign keys but these aren't present across the full data set. this relationship will be reversed soon > but I've no idea where the odd IDs > come from. The Coral, for example, is > http://www.bbc.co.uk/music/artist/m3qv/ but if you Google "m3qv" and > "Coral", it just returns that page, so it looks like an ID arbitrarily made > up by the BBC. when we were developing the music site a view was taken (can't remember by whom) that butt ugly 36 character uuids might scare users. so new shorter opaque ids were introduced - kinda like pip keys (http://www.plasticbag.org/files/misc/pips_etech.pdf). these are made up on the fly - not programmatically derived from brainz ids and have no meaning outside /music personally i'm kinda fond of brainz ids and would love to see them across /music, radio1, radio 3, bbc 2, bbc4, anywhere in the beeb with music content > I shall endeavour to find the relationship out if you'd be interested; > benefits are that /music links to all music content throughout the BBC, > which is therefore a Good Thing. better still - if other users of brainz like last.fm (i'm looking at you russ ;-) ) could be persuaded to also expose their music resources using brainz ids, all music across the web could be linked up u might call me a dreamer ;-)
<<winmail.dat>>
