Closed acoburn closed 3 years ago
Using a completely different URL, as the Web Annotation specification does, would be an excellent approach here.
The suggestions above make sense to me:
JSON-LD Context: https://www.w3.org/ns/solid/oidc-context.jsonld Vocabulary: http://www.w3.org/ns/solid/oidc
I do like the .jsonld
suffix because it makes this all the more explicit.
N.B. I've got a PR into jsonld
's http-client
. The most recent version won't resolve contexts from servers that don't give an application/json
content type on .jsonld
(e.g. raw.githubusercontent.com or most other legacy server configurations). If this doesn't get resolved, it might be better to use .json
to avoid rejecting contexts on most servers.
@csarven are you ok with merging it now ?
I think it's ready to merge 👍 Let's give everyone time until the call on Monday morning and we can merge it during that call.
@elf-pavlik , you've been pretty much saying "let's merge" almost at every step ... and about at every juncture, there's been stuff raised which made significant changes to the original PR. Please digest that first. It is not particularly reasonable - IMHO - to expect that everyone is going to review stuff on the weekend and have it ready by "Monday morning". The last commit in this PR was made on a weekend even.. Let everyone know if that's the kind of quality/timeline of reviews you expect... because I have some news for you.. :)
I am perfectly willing to let this sit longer. I won't be at the 5/31 meeting anyway (US Holiday)
Resolves #75
As discussed on 3/29/21, this changes the client identifier definition from being a JSON literal to being a first-class RDF-based serialization. This also defines a JSON-LD
@context
that can be used to bridge the RDF world of Solid and the JSON world of OIDC.Open questions:
This proposal suggests publishing the JSON-LD context file at
https://www.w3.org/ns/solid/oidc.jsonld
. I do not have any real opinion about that URL, other than that it should use HTTPS and use a.jsonld
extension. Should it instead be published underhttps://solidproject.org/TR/...
?This PR also refers to an
oidc
vocabulary which does not currently exist. That vocabulary (once published) would simply define URLs for the various OIDC-related field names. As with the URL for the JSON-LD@context
, should this be published underhttps://www.w3.org/ns/...
or underhttps://solidproject.org/TR/...
?