A few days ago I teased that I have ideas for new architectures for feed reading. Here's the main idea. Users should be able to subscribe to OPML lists, not just import them. So I could use a different editor, or discovery mechanism, or who-knows-what to manage (some of) my subscription lists. Or use two different readers and have them share their lists. Having this architecture widely supported in the feed reader market would open up possibilities for higher level applications. It's a major upgrade for the RSS market, something a silo can't match, without ceasing to be a silo. The challenge of open networks is that they leverage their openness. This is a remarkably easy feature for feed readers to support. I've implemented it in every feed reader I've written since 2002. It never takes more than a couple of days. If you're committed to no lock-in you should be all over this feature. [A place to post questions or comments.]#

I've been peripherally aware that some browsers and operating systems support the idea of a feed: prefix for web addresses. However I don't know the history, where the idea originated, or how widely it is supported, and what are its intended semantics? When the user clicks on a link like this on a Mac, what's supposed to happen? This came up in a thread on feedBase work. #