Possible improvements people could work on:

- [OPTIMIZATION] Connection pooling. Right now we open for every link a
  new connection.
  Must be thread-safe, must handle timeouts and connection expiration
  (HTTP: only pool persistent (ie. HTTP/1.1) connections; Keepalive
   header parsing).

- [USAGE] rethink intern/extern stuff

- [BUGFIX] when an URL is found in the cache and it has a broken anchor,
  the broken anchor name is not displayed as a warning

- [OPTIMIZATION] Python 2.4 introduced an "usenetrc" option to ignore
  ~/.netrc for NNTP. Thus we could get rid of our custom NNTP class that
  ignores ~/.netrc completely and use "usenetrc=False" instead.

- [USAGE] make a nice GUI for linkchecker

- [OPTIMIZATION] SF bug #992389 bit me when I wanted to do absolute imports
  at the checker classes :/
  When the bug is fixed we can import absolute classes

- [FEATURE] warn if overall size of page (including images/flash/etc.) is
  too big right now, the page size is only html content

- [FEATURE] save downloaded pages
