license pages do not respond to Accept headers (ex. for requesting RDF)
Opened this issue · 9 comments
Expected Behavior
Sending an Accept header with the desired content-type (e.g., RDF/XML) should result in CC returning the RDF/XML for a license, rather than the HTML:
curl -iL -H 'Accept: application/rdf+xml' https://creativecommons.org/licenses/by-nc/4.0/Actual Behavior
The server ignores the Accept header and returns HTML.
This is a separate issue from the following issue (in that case it was a separate URL that was down):
I've checked the code and we do not currently implement this behaviour.
Currently, the rdf files are treated as distinct documents rather than different content types for the same document. Since the rdf is not simply a different marking-up of the same text, I think it is reasonable to keep them as distinct documents.
If we want RDF, should we just be appending /rdf to the URIs?
Out of curiosity, was this behavior implemented at some point in the past? At least one RDF library uses the non-RDF URIs, apparently assuming sending the Accept headers will work: https://github.com/curationexperts/oargun/blob/master/lib/oargun/vocabularies/cclicenses.rb
I believe I remember that there was some content negotiation as some point in the past. The reason that I believe this was the case is twofold.
- you recommend it in section 5.2 Microformats of the ccREL: The Creative Commons Rights Expression Language.
- Over at rightsstatements.org a lot of the insights of CC have been mimicked and that platform has content negotiation.
Also I would argue the RDF is the same document as the deed, not as the legal text, and could be dealt with Accept headers.
Dunn, I'm not a CC-developer. I just been around a long time around cc tech. I am member of the tech. working group at rightsstatements.org
👍 to adding support for content negotiation. I think the RDF and HTML are different formats of the same information, and the fact that the RDF subject URI is the same as the HTML URI supports this view.
Note
Any further consideration about (re)adding this behavior?
We're adding this to our queue of issues to consider/work on but we can't promise a timeline.
Anyone here have opinions on the implementation in dev:
Looks elegant to me. Do we have a fallback when the rdf does not exist?