Open jariarkko opened 3 years ago
The problem is that I'm not sure all image files we point to would have the information. E.g., does a newspaper or wikipedia articles have that?
A quick run down the picture flies that I have, my own generally have stripped EXIF (or recent at ones least). Out of the external ones, some have a "photographer" property, but many do not.
I was thinking that this means we need to be able to include it as an extra JSON property where needed. Code could of course both look at the EXIF and the extra property. And photo credits are perhaps by default the people who were at the place, if no other information is provided.
IPTC Core metadata is of course independent of EXIF metadata, both of which can be embedded in image files, but your point is valid. People actively remove metadata. This also means that very few people actively use digital asset management software to type and embed extra info into the pictures. Given the type of image sources we have, these metadata probably see minimal use.
Let's implement the photographer/copyright holder/etc information as a separate JSON field as you suggested. Perhaps we could call it "image meta" or "image info" or "legend" or something like that to keep that field general and flexible.
Should we rely on standard embedded metadata then?
Documentation: https://iptc.org/std/photometadata/documentation/userguide/
The downside of that is that it is not editable in text editors or wikis but needs special (but common) tools.