Closed elsdvlee closed 5 months ago
Explanation of the CORS problem:
But...
So... what's happening here? Maybe some extra testcase info will help.
This happens on the https://pod.playground.solidlab.be/, logged in as user1@pod.playground.solidlab.be
:
(The Accept header in the request is as intended)
Info: same exercise on https://css11.onto-deside.ilabt.imec.be works. Is this an issue of the playground?
This happens on the https://pod.playground.solidlab.be/, logged in as
user1@pod.playground.solidlab.be
:(The Accept header in the request is as intended)
What's the actual result when you're logged in? nquads or jsonld? Maybe the playground for some reason can't handle graphs?
On the Solid Playground, the returned content type is application/ld+json
.
On our CSS it is application/n-quads
.
Same code calling in both cases, same request headers etc...
Many details in illustrations below.
Questions:
The only thing I can think of is that the playground for some reason doesn't support quads, and you could try accept: text/turtle before jsonld to see whether that solves it. turtle is a MUST for LDP, so we should at least assume every server can handle that one
More observations:
original (goes wrong, returns application/ld+json)
Accept: application/n-quads,application/trig;q=0.95,application/ld+json;q=0.9,application/n-triples;q=0.8,*/*;q=0.1
original modified: delta between q factors 0.1 (works, returns application/n-quads)
Accept: application/n-quads,application/trig;q=0.9,application/ld+json;q=0.8,application/n-triples;q=0.7,*/*;q=0.1
original modified: delta between q factors 0.1 and replaced application/ld+json with text/turtle (works, returns application/n-quads)
Accept: application/n-quads,application/trig;q=0.9,text/turtle;q=0.8,application/n-triples;q=0.7,*/*;q=0.1
The problem is that in contrast with the CSS's (v7.0.2) on https://....onto-deside.ilabt.imec.be/
, the CSS in the Solid Playground https://pod.playground.solidlab.be/ (v7.x.x) does not follow q factor priorities as expected with the q values in the original Accept header we were using.
This seems solved by changing the q factors so that they differ at least 0.1, see the second Accept header above.
I filed an issue for that in the CSS repo.
Finally, for the viewer I'll use the third Accept header, not mentioning application/ld+json
at all.
No results for queries over jsonld resources with VC.