- made ClientStorage and BlobStorage create the blob root directories automatically

 - replaced "DIRTY" with "LOADING" for ClientStorage

 - fixed problem handling data manager registrations for non-default 
   transaction managers

Formatting changes.

- Made the ClientStorage cache directory the same structure as BlobStorage

up on the system. Also add some tests for the configuration of BlobStorage.

Add a test to show that our cleanup code (currently in "_abort") doesn't get called.  The complexity of the proxied storage, the connection as a datamanager, and the transaction indirections make it unclear exactly how to hook a transaction abort for cleanup purposes.

Override tpc_begin and tpc_abort because methods found on the proxied object aren't rebound to the proxy. :-(

Add (badly implemented) methods to TmpStore to allow blob import to work again.  This will need to be redone, but it gets us to the point where all the tests pass.

Also provide a __repr__ to the BlobStorage proxy which lets us know that we're using a proxied storage.

Add some missed merges to Connection.py and add note about cluelessness about what to do in the face of savepoints when blobs are involved.

First crack at merging ctheune-blobsupport into a recent trunk checkout (tests have not yet been run).

now activates obj.

Re-add the zeoctl module to ZEO, and rewrite runzeo so
it can run as a main program (the scripts created by
mkzeoinstance.py expect to be able to do this).