Keep project metadata clean/healthy, keep re-use strong. Get more Gump hosts (for more cycles, smaller time windows). Ensure the subtle differences of Python Gump are correctly noted. Make it easier for folks to install/get started with local/private Gumps. Port Gumpy to pure Python for Windows implementations. Make distributed management easier (e.g. manually, for legal reasons, rsynch packages from central repository.) Allow cascading Gumps, that 'download' jars from each other to save cycles & increase "reach" to a wider community. Mail general@gump.apache.org with feedback on usability (of the outputs, or providing the metadata). Identify 'prebuilt' projects (visually, and in lists like 'Packages'). Allow Cascading Gumps (downloading JARs via Ruper) Add 'server' entries, see mailing list for proposal... Document gumpy.py and the other gump/*.py scripts. Add more interfaces to source control, and clean up the XML metadata to be SCM agnostic. Historical database (probably RDBMS, not just a simple DBM like now). Created