I can't wait for a community-driven, open-source data model that we can take offline and plug into any software adapter. With a little version control and some voluntary data samples, a large enough community could get it going.