SXSW: Microformats: A Quiet Revolution

Microformats panel

Jeremy Keith (blurred), Tantek Celik (standing), Glenn Jones, Karsten Januszewski.

Everyone is using microformats, even if we don’t know it.

Keith showed us huffduffer, a site that creates bookmarks for audio information. It is a consumer of microformats. He talked about rel=”me” for pointing out where you exist on the web. Shows relationship of linked to document to current document. Uses rel=”me” from XFN. If you give your URL to huffduffer, then anything tagged with rel=”me” can be called in from your URL and show as links for you. Then huffduffer goes through all your URLs looking for hCards and avatars and pulls them in. Huffduffer rechecks everything every 24 hours. So if you upload a new avatar to Twitter, huffduffer picks it up within a day. He says we already publish our information, so why make people upload it again for each site. He wants to know how people respond to this. Do they like it, have negative reactions, etc?

Karsten is associated with the MIX09 conference. They built that uses microformats to pull in information and get content. Firefox operator plugin shows when there is microformat information on a page. Wrote scripts that will work cross browser (without Firefox plugin) to pull in all micorformats on pages. Available at

Glenn works for He talked about how microformats can be used with privacy. Privacy can be implemented by allowing people to use rel=”me” with a system where they sign on and agree to disclose more or less information. Uses oauth, that lets you decide how much access you can get to private information. Microformats are a way to aggregate all the information you leave about yourself on sites like delicious, flickr, upcoming., twitter, linked-in,, and more. (Shades of the privacy discussion from the previous session.) It pulls in hCard information from all those sites.

Yahoo YQL. Tantek called it an API for the web. Send off a URL and get back all sorts of information.

Tantek. Challenges for microformats are localization and  accessibility.  Many issues of localization require hacks for language at this point. They are finding solutions to the localization problems. Screen readers have problems with reading out things like dates in a sensible way. They’ve used span elements to solve this. And they’ve begun to store date and time information separately. He wants people to write test cases and implementations and let them know the results. Links on the


Leave a Reply