Support discovery of feeds for WebFinger identities

[voxpelli]

Would be cool if one could discover the feed of a "acct:[email protected]" URL as well (and for the more unofficial "[email protected]").

Just to make this issue complete, I think you're aware of these already, but here's the WebFinger spec I'm refering to: http://tools.ietf.org/html/draft-jones-appsawg-webfinger-00

And the link relation to look up would probably be the "http://schemas.google.com/g/2010#updates-from" one according to: http://code.google.com/p/webfinger/wiki/CommonLinkRelations

[superfeedr]

@voxpelli That is an amazing idea! How would you see the API and algorithm for this?

API: GET http://[email protected] ?

Algorithm: Do the magic dance and find all urls for this webfinger, then, for each of them, grab the feeds?

[voxpelli]

@superfeedr I would do it like this I think:

API: GET http://feediscovery.appspot.com/?url=acct:[email protected]

And in the backend do:

1. Fetch host-meta for the host
2. Look for LRDD file in host-meta
3. If LRDD is found, then fetch LRDD for the specific URL
4. Assemble LRDD and host-meta
5. Look for all defined feeds in there (in theory the host-meta could define a global feed that's valid for the entire host) and return them.

[superfeedr]

This sounds quite doable! On my todo list!