For the most part normalizr has worked out as expected setting up schemas and using object IDs.
For brevity going to omit most everything that comes back from the API and