@brianleroux I continue to feel that GraphQL (or rather, something like it) would have been a much better fit as a server-side data layer abstraction, that itself never goes over the network (and is just a library function call); because collating different data sources is the part that the model actually good at!