Add a data layer honestly

Adding data is not just fetching a URL. In this project a new layer ships with a license check, a credit, and an accuracy statement. This guide is the checklist.

1. Vet the license before you write code

Licensing is a gate, not a footnote

If the license is unclear, non-commercial only, or no-derivatives, do not use the dataset. The project only ships assets that are clearly usable and creditable.

Accepted sources tend to be:

• Public domain: NASA, USGS, NOAA (US federal works).
• CC BY 4.0: ESA/Hubble, ESA/Webb (attribution required, keep the exact credit line).

Rejected: anything CC BY-NC, CC BY-ND, or with unclear terms.

2. Add the source with attribution

1. Record the dataset in docs/DATA_SOURCES.md (or a world-specific *_DATA_SOURCES.md): the URL, the license, the exact credit line, and the date you verified it.

2. If you ship a file (a texture, a JSON snapshot), place it under public/ and keep the credit next to it.

3. Keep the transform pure and tested

Put the parsing and any physics in a lib/ module with tests and the null-safety contract, exactly as in Read real data in the browser.

4. Write the accuracy note

Every world has a *_PHYSICS.md that states what is computed, from which inputs, and how accurate it is. Add or update it. Be specific about the bounds, for example: “valid for low redshift, not for high-redshift cosmology.”

5. Label anything illustrative

If part of the layer is an artist’s rendering or a stylised value, label it as illustrative in the UI and say so in the note. The honesty rule is that a reader should never mistake an illustration for a measurement.

Checklist

• License verified and clearly usable
• Exact credit line recorded with the source
• Transform is a pure, tested lib/ module
• Accuracy note written with explicit bounds
• Illustrative elements labelled

See the Data sources reference for the current set.