While catching up with some old blog posts, this post on WebFinger by Dare Obasanjo caught my attention. WebFinger is a very cool project. It reminded me of the initial RSS days where we were trying to extract all the blogs (blogroll) a user follows.
WebFinger would serve as the anchor point to discover all the social network of a user. As Dare points out, this makes certain workflows in social applications easy. I think, if they can put some “discoverable” (like Atom) structure around this, it would lead to much better search results. Remember “Wolfram|Alpha” ? Today, given my email id, no one can find out whether I have facebook account and what is the Id as there is no correlation (most of the time). A service like WebFinger can bridge this gap.
Haven’t looked at the spec, but looks like we can approach this in two ways:
- Multiple centralized authorities that can register and serve WebFinger- could be public/private. No correlation to email ids, domain names etc
- Same as above but correlate to email ids/domain names via DNS system itself. Much like MX records for email servers, one should be able to determine WebFinger server given email id