improving OpenID support

Sam Alexander of MyVidoop emailed me last week to say that they’d be rolling out some new features in regards to OpenID support. Sure enough, you can see on the OpenID Support table that they’ve added support for xrds-header, yadis-html, and most importantly, content-type. Additionally, IDtail, a Korean OpenID provider, added support for content-type as well since I last updated the table. That’s very exciting to see providers beefing up support, as everyone really does win. For those not aware of why I harp on content-type support so much, it is the only XRDS discovery method that can be done in a single request. It allows a consumer to include a header in the request and immediately receive the XRDS document. The other two methods (using a response header, or embedding HTML code) simply advertise the location of the document, which requires a second request to actually retrieve. Sure it’s not that much more work, but every bit helps, and it reduces the risk of parser problems when using embedded HTML.

I really wish I had the time to setup some automatic detection of when provider support changes and make a blog post or something… work that benefits the community should be recognized.