IIIF/discovery

Are names of content state attributes SHOULD or MUST?

tomcrane opened this issue · 3 comments

In the current content state draft, iiif-content as a GET or POST parameter is a MUST, but data-iiif-content as an HTML attribute is a SHOULD.

Typically, the owner of the process that is generating the HTML knows what viewer they are embedding and how it is configured. In some scenarios, the content state will have arrived at the page as a GET parameter (must), e.g., ?iiif-content=blah, then supplied to a viewer’s data-iiif-content attribute in dynamically created HTML.

Is this even in the spec at all? Is it a cookbook recipe only?
It would aid providers and implementers of viewers if they didn't have to think of a name for this attribute, but it won't break much if they don't use it.

Removed the warning and pushed updated text.
IIIF/api@f6f01cb

Close?

Done