API dream for datasets connections

[ppKrauss]

Hi, I am not understanding the aims of Frictionlessdata (+Core Datasets) project... Discussing a dream is a way to see the horizons of the project.

Datasets, as "open knowledge resources", must be a set of connected datasets. All are CSV files, them all are relations, so can be connected by a relational albebra... The set of datasets form a SQL database. Where is that database and its server? Can I query it?

---

Map from JSON+CSV files to a connected database

The datapackage.json specifications not show how to map CSV files to a set of "connected tables" in a SQL database... I suppose that we can do it (see starting point here).

A dream?

PostREST-like APIs and GraphQL APIs are organized in terms of types, tables and fields, not endpoints. It give you full access of all datasets from a single endpoint.

Of course, we can use also SQL, but it is easy to do basic queries (SQL subset) by GET method and avoiding server crash. Only registered users will access the full capabilities of SQL.