the list of concept types that each user chooses from when adding a new one is stored on every concept object. this should be abstracted at a higher level.

The lists of concept and resources are currently contained in a datastream in every object. This was done originally so that each concept type could have it's own set of allowed sub-concepts and resources (I think). Clearly, it is a pain from a system management point of view.

We should at some point soon take a look at this and figure out a better way. I will add this to content models for the moment but not sure where it really belongs.