[surface] Mutation information

[mchaamsler]

Is the mutation information (e.g. if a value is original, interpolated, automatically or manually corrected, aggregated etc.) or a value interesting for the user? If yes, in which format should we publish this data?

[lrosique]

yes knowing that is important, particularly for "original" so we can have our own AI models predict on those values. Imagine we use an interpolated one and in a year you change your interpolation: our models won't work well anymore, so it is better if we have the choice (by knowing what we use).
I'd say it all depends on your data:

• if the type of data applies to whole lists then you can have a separate field, if it is per value then it seems hard (unless you add another list). If it is by endpoint maybe a query parameter is enough but i don't think it will be the main usage