New Databrowse Notes
--------------------

- Databrowse cetnered around library operation... generating XML representations of files and folders
- Canonical naming to reference files
- Generalized Caching Functionality
  - Probably out of directory, but we need to figure out how to deal with security
    It's easy when running as user, hard when running as web server
- Should we use xmldoc?
  + Provides Namespace Management
  + Easy Handling of Locking, Contexts
  + Numpy Matricies
  - Too much overhead?
- Provenance Tracking for Generated XML
- Lazy Loading, Cross Links?