RFC 5 coordinate transformations

[d-v-b]

It would be great to start implementing some of the data structures defined in the coordinate transformations proposal, assuming there won't be too much churn in those data structures. I think the models we create should be based on the proposed spec changes in this PR, @bogovicj please correct me if I'm looking in the wrong place.

in terms of scope, we should only aim to model the structure of the transformations, not their behavior.

[bogovicj]

Yup, that's the place to look.

Looking forward to watching this repo!