Autodiscovery is the mechanism used by RSS aggregators to find a feed given a weblog URL.
Usually this is done by the presence of a link element in your HTML that has a href to your feed.
About 5% of the time this fails. Might not be as high as 5% but it happens on a non-trivial percentage of blogs.
You can fail back to using <a> links but this can be prone to error.
There are a few tricks though.
- if it’s a link to http://feeds.feedburner.com you should prefer this feed over the item in their autodiscovery section. This is of course only if the link of that FeedBurner feed is the same link as this weblog. A non-trivial percentage of FeedBurner users don’t configure their HTML template correctly.
- You can use the ‘add to Rojo’ or ‘add to bloglines’ links to discovery the feed. Usually they include the feed URL and are highly unique. For example here’s NewsGator’s ‘add to newsgator’ link:
http://www.newsgator.com/ngs/subscriber/subext.aspx?url=http://feeds.feedburner.com/Feedblog
The URL parameter includes my feed URL.
This can be used if you’ve failed to discover a feed for a given weblog. This is part of the ‘parse at all costs’ school of thinking.
I’m about to add this to the Jakarta FeedParser this week.
If anyone has any other ideas please leave a comment.
-
Spinn3r News- Spinn3r Hiring Five new Engineers (and growing rapidly)
- Spinn3r Hiring Support Engineer
- NYTimes, on Memetracker, and Spinn3r
- Spinn3r at Real Time Stream Crunchup, RSS Speedups, and Startup Discounts
- Spinn3r Hiring Crawl Engineer
- Spinn3r 3.1 - Now with Twitter Support and Social Media Ranking
- Blogfa.com (Iranian Blog Network) Down Due to Iranian Protests and Civil Unrest?
- Papers using Spinn3r Data from ICWSM 2009
- Spinn3r talk from ICWSM 2009
- Spinn3r 3.0: New Features, New Architecture, New APIs - More Goodness
Subscribe
-
Flickr
More Photos
January 4, 2006 at 3:53 pm
You may also want to look for anchor tags with a defined “alternate” relationship and a MIME type your indexer can understand such as text/xml, application/rdf+xml, application/rss+xml or application/atom+xml.
January 4, 2006 at 5:37 pm
“A non-trivial percentage of FeedBurner users don’t configure their HTML template correctly.”
Hmm…count me among those. Autodiscovery picks up all my feeds, not just the Feedburner one. I haven’t had time to figure it out yet – what did I do wrong?
January 5, 2006 at 8:09 am
In addition to this problem, without proper use of the Feedburner feed tag in your header – Feedburner’s wonderful Pingshot pinging service is useless. It’s indexing engine has to be able to find the same feed URL in your header as it was pinged from in order to index your site – as I understand it. But to be honest I don’t understand it well enough. I’d love to help answer Pete’s question, for example! Pete, maybe we should put our heads together and take the time to figure this one out.
January 5, 2006 at 11:02 am
It’s just a matter of updating the ‘link’ elements in your head tag. I think think FeedBurner has docs on doing this.
They need to get native TypePad support.
This stuff shouldn’t be hard. Maybe someone can write a plugin to authenticate against TypePads service and manually edit your template.
Could be a GREAT paid webservice. $0.99 to correctly setup your FeedBurner service on TypePad (done by a 3rd party).
Kevin
January 9, 2006 at 10:47 am
Kevin, I’d like to ask you a question via email, but can’t find your address. Could you send me a message so I can respond to it?