[terrycojones]

I guess I'll get the ball rolling. We're building (and close to releasing) an app that's focused on Twitter users. It has FluidDB objects that each correspond to a Twitter user. Because the FluidDB objects aren't owned, other apps can put their tags onto the objects too. The query language lets you search on objects based on any tags (that you have read permission for), so we're hoping we'll soon see a wide variety of tags on these objects, which will give us a bunch of new ways to identify users on Twitter, to tag them arbitrarily, etc. The are obvious things like ratings, or people you've met, but there are also lists (a tagged set of users is just a list, more or less) including queries that cross lists, and things like http://tunkrank.com scores about Twitter users. FluidDB gives a place to put all that stuff - without requiring you to ask permission or for your needs to be anticipated - and it gives you a way to query on it too.

That's just a first app. We'll open source it. I can think of many more, as you might hope :-)

Terry Jones (disclaimer: I work for Fluidinfo, makers of FluidDB)