derek told me about how the quasi-familiar concept of "ehresmann connection" was apparently created by taking a more involved concept of "cartan connection" and isolating a less involved part of it (or something...) ... but on the other hand we seem to be running into certain situations (for example in connection with an ordinary "rolling ball") where an ehresmann connection (or something...) can be used to create a distribution... so i wonder how that might relate to the idea of the "cartan-rolling distribution" of a cartan connection... ??or something...

hmm, these ideas aren't fitting together for me at the moment... i'm getting confused about the idea of a torsoroid vs the idea of the configuration space for a "rolling" system... ???...