====
TODO
====

Meta
====

Copyright (c) 2003-2006 gocept gmbh & co. kg
See also LICENSE.txt

:Valid for:  gocept.linkchecker 3.0
:Author:     Christian Theune <ct@gocept.com>,
             Christian Zagrodnick <cz@gocept.com>
:CVS:        


Distant future
==============

Note that some of those features are unlikely to be implemented unless we
(gocept) get funding to implement them.

- Save the availability history of a link

- "Self healing": Recover links to documents from other Plone sites that also
  have the linkchecker installed (Leverage 30X for that, or use UIDs to
  communicate identities of objects). Have to look out for the special border
  cases (renaming an object and adding a new one under the old name. which one
  is the referred one?)

- "Self healing": Recover links to documents from non-Plone sites.

- Avoid that users can delete documents that are referred to
  from within the same plone site.

- Make links in the HTML code tell the visitor that the target is
  currently down

- Acknowledge links that are down for a while so they don't
  get annoying.

2.1
===

- Important: Move from hashing of urls and links to encoding. Encoding needs
  to be url compatible (which makes the code long :/)

- Bug: getting notified about link status might not update links with the same
  url that are retrieved in between: normalize the dependency between link and
  url and status

- Allow users to update links centrally

    - Search and replace on links

    - Replace one URL globally on all documents

    - Edit a single URL for a document from the list of links

- Clean up statistics / central view

- Store relative and absolute urls for comparison

- Use stored relative links for updating and the in-site link highlighting

- Global statistic: optionally ignore private objects

- Refine global statistics view (remove useless numbers) / divide global view
  and additional statistics

- Metadata for links (Title of the link, probably context) 
  -> eliminate link tab
