Free and Public news servers are visited once a day and their information
is added to my index. In addition another task visits servers on a
rotating basis and their article headers are indexed and random messages
are tested to see if they are really on the servers. This information is
used to calculate the binary and text group article retention figures.