Stephen’s Statements A little bit of everything from Stephen Duncan Jr, a Software Developer in Portland, Oregon

Wednesday, February 8, 2006


Scoble asks Why don’t you use a memetracker?. Check out the comments, as I agree with most of the answers. Summary: not enough new stuff, not a useful format, and, finally, why would I?

What would a useful “Memetracker” look like? First, it will have to integrate with my feed-reader/aggregator (do we have a single agreed upon name for this yet?) data. That way you can exclude the stuff I already know about. Plus you could tailor it to be interesting to me. Bloglines could pull this off, as could any other aggregator that centrally stores feed-reading data. Or, if attention or some other feed-reading-data-sharing mechanism were available, it could be done by a third party.

Second, it would be provided primarily as another feed. Not having to go out to another website and hitting refresh over-and-over is exactly the reason we love RSS and Atom so much.

So, what I’d want is an Atom feed of new items that I have neither read yet, nor have I read anything that linked to it yet. I want to see a summary of the main content, and possibly some links to the most interesting commentary on the item. An algorithm to determine interestingness like Flickr does would be the killer feature. A combination of popularity and making sure it includes some significant commentary instead of just being a another quote-and-link would be a good starting point. And, of course, I want “interestingness” to be somewhat based on what I already read, so that I can get more Java and less Microsoft, more economics and constitutional law and less party politics, and soccer and basketball, instead of just technology and politics.