Details entered in prosumer UI are overridden by JSON content

[nick-shaw]

It may be arguable which should take precedence if there is a JSON file in the archive, and then the user enters different details in the UI. Currently what is entered in the UI is completely ignored. That is confusing for a user who may have spent time entering details and expect those to appear in the resulting CLF. Of course, in theory it is the same user who made the package with the JSON, so they should know that they have already included details there and can leave the UI blank.

Ideally when you drag in an archive with a JSON file the fields in the UI should be populated with the content of the JSON, so the user immediately knows they don't need to enter that information. It also potentially gives them an opportunity to edit it, in which case the updated information should replace that from the JSON in the generated IDT.

This is perhaps less of an issue in a real situation, but when testing with the synthetic test data the user is less likely to be familiar with the content of that data, and may be confused that the information they enter in the UI does not appear in the IDT. I was!

[nick-shaw]

In fact the values from the JSON appear to override pre-entered UI values. So for example under Advanced Options the optimisation space defaults to Oklab, and that's what you see in the UI. But synthetic_001.json says CIE Lab. So the resulting CLF says CIE Lab in the comments. Which one was actually used? It is unclear.