Here are the bugs I and others have found in the Ultra Gleeper.

* The static/gleeper.js file contains a reference to image files which includes
  the Gleeper installation directory as "gleeper". This may need to be
  changed depending on where you install it. If you are using the RSS
  feed interface, it needs to be changed to an absolute URL, so that
  the RSS reader knows where to get the images from.

* The link chooser is not random enough. I get clumps of links from the same
  sources. I hacked in a way to use mysql's rand() ordering, but it
  requires a change to SQLObject. See the NOTE in User.py for details.

* For some reason it's possible for links to get an estimated rating
  higher than 1. I've seen 1.009 and 1.7, and also negative estimated
  ratings. I could just cap it but I want to see if there's some deeper
  problem here.

* If a script finds a http://foo.com/ and there's already a
  http://foo.com, a second webpage record will be created. The reverse
  is not true.
