If a Document has a known encoding, is it stripped before entering the Docseq?

Check Collection: are all subobjects working correctly?

Add support for query caching server configuration.

Add module for doing encoding conversions.  Should try out what Perl modules
are out there for attempting to do the conversion.  Build in the capability of
doing that automagically by the objects in the future.

Add module "Issue" that is capable of keeping multiple issues of an indexdir
and have new issues made (either incrementally or from scratch), possibly
together with an "Issues" module, which would allow you to specify how many
issues to keep.

Check all of the docs: create section of inherited methods in each submodule.
Add more examples.

Add module "HTML" for creating "Document" object from an HTML file.  Use
ntvhtml2ntvml as a basis for this functionality.

Add module "Message" for creating "Document" object from a mail or a newsgroup
message.

Add module "Mailbox" for interfacing with the Perl Mailbox:: modules.

Add ultralite support to NexTrieve::Resource?

Why is t/querylog not removed when it _is_ empty?
