Regression tests will be written. (Current coverage is around 30%, and that's largely due to POD coveage.)
It should be possible to create users automatically upon login attempt. When coupled with HTTP authorization, this should allow intranet users to begin using a local Rubric seamlessly.
Rubric::WebApp::EntriesQuery should only parse URIs and send result sets to the pager. (A renaming may be in order, too.) Another module should take natural parameters and produce result sets. It should be able to return errors on invalid queries and empty sets. (Future revisions should turn a query into English for display on error or as a title.)
Rubric now includes a class (Rubric::WebApp::Login) that defines logic for accepting logins; subclasses for login via HTTP and via posted form data are also included. This change should make intranet Rubrics easier to implement.
There was an 0.03 milestone to allow querying by date and time ranges. To implement this, a new query builder was written to replace Rubric::Entry->by_tag. Right now it's got its web bits and SQL bits all intertwined, but that will be fixed. The /user and /tag run modes were eliminated as such.
Rubric now sends fatal error messages to the end user. This won't persist forever, but it will continue until Rubric is more mature and less likely to fail during normal use.
Release 0.02 fixed a HTTP header bug.
New accounts are verified via email. Registration can be disabled.
Entries have a title and tags. They might have a link, description, and a long text body. A user only has one entry per link.
Anyone can search for entries based on the user who created them or the assigned tags.
Any search that can be performed via the normal HTML interface can be retrieved as RSS. The RSS links are announced on the HTML results.