While this results in way more requests needed, it has the great benefit of using
the same data structure for reading and writing. This saves us a lot of headaches
in the data modelling and we don't have to care that the info endpoints looked a
bit different.