RE: [aggregators] Re: Making feed subscribing easier...
- Looks to me like so far we've got at least three alternate approaches :
1. New mime type, something like
The file itself contains a list of the URIs of available 'feeds' (RSS, CDF
but also FOAF etc) and the file is linked from the site home page through a
single button. The application that is registered with the browser to launch
with the discovery mime will in the simplest case be the default aggregator.
The aggregator pick out its favourite format(s) and displays and/or
Further down the road it should be possible to make a simple router app,
which picks up this file and forwards the feed URIs to a load of different
I'm not sure, but I've a feeling RDDL could be used directly for this -
although the RDDL spec looks complicated, only a very small subset would be
needed for feed discovery.
2. Standard URI for readers to listen at, e.g. : http://127.0.0.1:8109
the single button on the page is linked to a URI of the form :
the path part can be used to provide the location of the blog home page.
When clicked, the aggregator goes to that path on port 80, lifts the URIs
from the <link> elements and subscribes/displays as appropriate.
the aggregator(s) use a minimal proxy to get the URI of the page being
viewed, then pull out the <link> material as 2.
The first two approaches are very browser oriented - a (single) button will
be on the web page. This does follow what's been suggested as the user
expectation of how such a UI might work.
The third approach could go either way - you could subscribe by clicking on
a button in the browser or a particular aggregator.
In all three cases a bookmarklet could (probably) be used to avoid having
the button on the page, though how cross-platform this would be is another
Not mentioned here is the simple dragging of the html page address from the
browser to the aggregator, that could be done already - not a popular
gesture, it would seem. A bookmarklet could also be used to pass the data
across, but presumably it would have to be aware of the specific aggregator
Any more approaches? Revisions of these?