An addendum to my spec for how I'd do a new feed reader. At least at first, we'd approve feeds the way Apple approves apps for their app store. We wouldn't let through feeds that look bad in the reader. Let me amend that. If it's due to a bug in my software, or my reader failing to support a feature of RSS correctly, then the feed would be put in a different pile. One of the problems in the RSS world was people who made feeds didn't know that they had to care how their feeds behaved. Or they only cared about one reader. That was a prescription for the mess we're in now. My hypothetical feed reader would not work like that. Bugs in feeds are not compensated for by workarounds in readers. #
I wrote this piece in 2013, after Google Reader had its plug pulled. I'm guessing most people who read that piece have no idea what I'm talking about, because most don't think about online systems as defining a base of content they can move around. Think of it this way. Imagine if you couldn't fly into one city, you had to take a train, but to another city you had to fly, there was no train. The obvious thing would be to build a train route to the city without one, and do the same for air. #