* Still need to support feed expiry.
* Need to store page author metadata and include it in the rss feed.
* Some rss feeds contain relative links or relative urls to inline images,
  which break when aggregated. Do I need to parse the html and make them
  all absolute?