Reconsider the Datasette first-run experience

[simonw]

Had a really interesting conversation today about how hard it is to get from "I installed Datasette" to "I've done something useful with it": https://news.ycombinator.com/item?id=33216789#33218590

Spending some time focusing on that first-run experience feels very worthwhile.

[simonw]

I think this mechanism could go a long way towards helping here:

• "datasette insert" command and plugin hook #1160

It was part of a larger idea I was exploring around ensuring Datasette could be used to start interacting with CSV/JSON data out-of-the-box, without needing to first convert that data into SQLite using separate tools.

[ocdtrekkie]

I think there's probably some core plugins a new user should just "get" maybe. I feel our Sandstorm package has really reasonable defaults, the ability to bring data in for instance. However, you also have to know that's in the hamburger menu, it's not necessarily intuitive where to start.

[kindly]

It was part of a larger idea I was exploring around ensuring Datasette could be used to start interacting with CSV/JSON data out-of-the-box, without needing to first convert that data into SQLite using separate tools.

This would be great. My organization deals with very nested JSON open data and I have been wanting to find a way to hook into datasette so that the analysts do not have to first convert to sqlite first.

This can kind of be done with datasette-lite.

From this random nested JSON API:
https://api.nobelprize.org/v1/prize.json

You can use the API of https://flatterer.herokuapp.com to return a multi table sqlite database:

https://lite.datasette.io/?url=https://flatterer.herokuapp.com/api/convert?output_format=sqlite%26file_url=https://api.nobelprize.org/v1/prize.json

This is great and fun, but it would be great if there was some plugin mechanism that you could feed a local datasette a nested JSON file directly, possibly hooking into other flattening tools for this.