**** Release 0.017 *****

Wednesday 3 March 1999 - Ave Wrigley <wrigley@cre.canon.co.uk>

* uses HTML::TreeBuilder instead of the (depricated) HTML::Parse 

* uses HTML::LinkExtor to extract links

* minor documentation bugs

* IGNORE_UNKNOWN option allows the robot to ignore unknown HTML tags

* ACCEPT_LANGUAGE option allows the user to specify languages accepted by the
  USER_AGENT

* robot does a head request to check the MIME type of potential links, rather
  than relying on filename extension

**** Release 0.018 *****

Thursday April 8 10:52:31 BST 1999

* temporary filename munging changed to avoid problems on Win32

**** Release 0.019 *****

Tuesday April 27 10:57:08 BST 1999

* regex to determine which links are HTML links changed - m{/$} added

**** Release 0.020 *****

Friday May  7 09:54:41 BST 1999

* bug fix for javascript links (thanks to Petr Kristan <petr.kristan@epos.cz>)

**** Release 0.021 *****

* no longer supports $filename argument for invoke-on-contents hook
* add-url-test hook added
* listUrls method added
* unshiftUrl method added
* CHECK_MIME_TYPES attribute added