1. Enhance importing library dataset into the current history by now allowing the user to select from a list of existing histories or create a new named history ( similar to copying datasets ).  This feature is available for a specific library dataset, a folder ( or root folder ), when selecting multiple datasets from in a library,  and on the results page for searched library datasets.
2. Add Brad Chapman's latest patch for fixing the lucene search implementation
3. Eliminate searches on dataset state for both whoosh and lucene searches
4. Clarify the objects being searched in the regular search and advanced search boxes in the data libraries grids.
5. Add the ability to perform certain actions on 1 or more selected library datasets on the results page for searched library datasets.
6. Add a "select all" checkbox to the browse_library.mako template.
7. Fixed library functional tests accordingly.
The ui is Very rough, at best.
Libraries can be accessed by users through a new tool under datasources.
Libraries can be created and accessed for modification via a link on the admin page.
Currently, all library datasets are currently associated with the public group.
Tags have been defined (tables and mappings) and can be associated with a dataset or a folder, but there is currenty no interface for displaying or creating these.
Creation and editing of a library should be an admin only function, however this check is not currently present.
The creation/editing page for a dataset is currently hard coded, it should be dynamic (i.e. datatypes and dbkeys)
Brief description of a library:
A library is composed of a name, description and a root folder.
A folder is composed of a name, description, and a parent folder (none when is the root), where a folder can have many or none child folders and datasets.
A library dataset is composed of most of the same attributes as a history dataset, but it is linked via a folder_id to a folder (instead of a history_id to a history).
Ordering of the library is maintained by the 'order_id' attribute of each folder and dataset object (which is populated by the item_count attribute of a folder)
peterjc