Formats for HTTP Proxy Caches

On today's telecon Jan mentioned that he needed a format for an 
HTTP proxy cache. I'd like to humbly suggest the one implemented 
in my Python code:

http://logicerror.com/archiverProxy

Which creates a directory structure like this:

http/www.w3.org/2001/foo/bar/baz/filename.nt/

with files:

1, 1.headers, 2, 2.headers, etc. corresponding to various 
versions of the document along with associated HTTP headers.

--
       "Aaron Swartz"      |              The Semantic Web
  <mailto:me@aaronsw.com>  |  <http://logicerror.com/semanticWeb-long>
<http://www.aaronsw.com/> |        i'm working to make it happen