Separating representation datasets

[tpluscode]

In v1.0 all requests get combined into a single dataset. The purpose of that was to have an easy way to combine different resource representations and especially their API Documentation(s) for the purpose of discovering operations

In hindsight this proved problematic when trying to access the dataset directly because everything got combined, potentially thousands of combined resources.

Instead, I would like to keep even response separated and only combine them with API Documentations when accessing the Hydra terms such as supported operations and properties