-
Notifications
You must be signed in to change notification settings - Fork 57
Type scoped context upstream change #1846
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Comments
Editors agree: See if @w3c will publish a 1.1 context for the Web Annotation Data Model if one is written up and if yes, then write one up for review. If we need to publish before it's live on w3c, then we can temporarily host on iiif, and rewrite our context when it is available. This is the right long term approach |
This was resolved upstream, and now unblocks us here. There's now a way to have the scoped contexts propagate with an explicit declaration: https://www.w3.org/TR/json-ld11/#context-propagation We need to update our v3 context documents to the latest patterns. |
Delaying on this until the playground is updated so that we can check that it works as expected. But as the contexts aren't normative or semantically versioned, this shouldn't delay 3.0 release |
Seems to work as expected in the playground: :) |
A proposal from a different WG (Verifiable Credentials) has led the JSON-LD WG to conclude that type scoped contexts should NOT be inherited outside of the instance of the type.
The issue: w3c/json-ld-syntax#174
This means that our use of Type Scoping to import the Annotation context will NOT work in JSON-LD 1.1, as the effect would be limited only to the specific type, and not its child nodes.
It also means that our current pattern of importing the contexts via type scoping of Services doesn't work.
To make this work, we would need to do one of the following:
items
onRange
, where the referenced class can be either aRange
from IIIF's context or aSpecificResource
from the annotation context.2b. There is a shortcut suggestion to allow type scoped contexts to have a method to say that all properties of the instance should be treated as if they were property-scoped to the same context. This seems to be the right fix.
Thankfully this is not a normative change due to contexts not being semantically versioned, but it is a potentially very big one.
The text was updated successfully, but these errors were encountered: