* Implement threading.
    - cache the article meta-data in a local database of some sort (DBI,
    Berkeley DB, etc.)
