This note is just an observation about myself. I don't think *relationally* - in the sense that I don't tend to see datasets as [[Relational Databases]], but instead see them in my head more easily as [[JSON]]-like trees... or even [[Graph]]s. For example - the dataset: ```json [ { "id": 1, "title": "A Game", "releaseDate": "2015-01-01", "platforms": [ {"name":"Xbox"}, {"name":"Playstation"} ], "rating": { "code": "E", "name": "Everyone" } }, { "id": 2, "title": "B Game", "releaseDate": "2016-01-01", "platforms": [ {"name":"PC"} ], "rating": { "code": "E", "name": "Everyone" } } ] ``` Fits pretty easily on one screen in JSON format... but to represent that as tables you need to define & visualize at least **2 if not 3** separate tables... then manage [[SQL]] joins to pull them together in a way that's readable. My brain just prefers to think of things hierarchically. I guess most applications I'm aware of just don't have enough scale to justify the switch. **** # More ## Source - self - example from [https://docs.flatterer.dev](https://docs.flatterer.dev/)