TODO

In Progress
===========

* Docs on contributing.
* Lucene backend.
* Formal packaging.
* Thread-safety (ongoing).


Future
======

* Make schema generation more flexible.
* Distributed read/write setup?
* A QueueSearchIndex (add to queue for updating instead of immediately firing update)?
* ModelIndexes to streamline index setup.
* Spelling support?
* A pure Python dummy backend for testing purposes? (Whoosh)
* SearchManager to be attached to Models?
* A means to register multiple ModelIndexes with a Model?


Complete
========

* Move to "site" setup a la NFA
* Research various backend capabilities
* Convert to SearchQuerySet
* Prevent forking (better extendability)
* Rip out query parser
* Rip out paginator
* Fix ModelSearchForm to get site's registered indexes
* Work on initial docs
* Refactor ModelIndex
* Test coverage for ModelIndex
* Finish Solr SearchQuery
* PySolr branch on GitHub for JSON
* Better "no-template" handling for ModelIndex
* Support for stored (non-indexed) fields
* Support for More-Like-This
* Provide a management command to create a Solr schema
* Highlighting
* Better ModelIndex support
* Support for faceting
* Support FQ to initially limit the results
* Figure out how to load sites regardless
* Tests for backends
* Good Test Coverage
* Good Docs
* Several complete backends   
* A management command to update (last 24 hours).
* setup.py
* Why Haystack/History documentation.
* A means to control commit frequency on updates.
* Better automatic site loading.
* Docs on SearchIndexes (prepare's).
* Handle missing dependencies.
* Docs on backends.
* Docs on faceting.
* Docs on forms/views.