- From: Robert Sanderson <azaroth42@gmail.com>
- Date: Mon, 24 Aug 2015 11:44:40 -0400
- To: Web Annotation <public-annotation@w3.org>
- Message-ID: <CABevsUF5ajmMYkRsUUNXFxQ_MV6Z6moYkzf33-mn_kxHjRd=uA@mail.gmail.com>
The (very valid) concern raised by Jacob and Stian is that the proposal as it stands does not mention multiplicity at all, being Choice, Composite and List. If role should be only on SpecificResource, then the content/source for it would have to be the multiplicity construct, such as: { "type": "Annotation", "target": "http://example.com/target", "body": { "role": "commenting", "content": { "type": "Choice", "members": [ "http://comment.example.org/en", "http://comment.example.org/fr" ] } } } Which doesn't seem too bad to me. Would be great to get other folks' reactions to this particular case. It becomes more complex for Composite or List, as there would be the specific resource for the Choice/List within the Annotation, and then the specific resource to associate the role with the actual resource. That said, the pattern is consistent and reliable, if deeply nested. { "type": "Annotation", "target": "http://example.com/target", "body": { "content": { "type": "List", "members": [ { "role": "commenting", "content": "http://comment.example.org/en" }, { "role": "tagging", "content": { "text": "tag" } } ] } } } This structure does make the lack of type more concerning, as the members of the Choice or List could be anything, and not necessarily consistent within any individual construct. Rob -- Rob Sanderson Information Standards Advocate Digital Library Systems and Services Stanford, CA 94305
Received on Monday, 24 August 2015 15:45:11 UTC