The strength of DDC and Library of Congress are their ready availability, to those who can or would pay dearly for them. UDC is also available, for a hefty fee.

For Joe Sixfootshelf, there is no easy way to legally and inexpensively classify his books aside from developing his own classification system… reinventing wheels for most of us.

Certainly, a simple enumerative classification system built around the basic architecture of DDC (disregarding jingoistic stuff like the 810-819 area) and offering faceting could be quite valuable to amateurs and unspecialised libraries. Conceivably, such a system could be expanded to fit the needs of specialised and larger libraries.

Or perhaps there is a way to do as software developers do and fork something that exists and modifying it instead of starting from stratch?

Credit to LibraryThing though for being interested in these issues, I’d like to see them hook up with the Open Library project, and semantic web developers.