The definitions of classes, more exactly the properties used on those classes, along the two documents (model and vocab) is not in sync. This makes reading the documents a bit difficult...
I have checked only the Policy
class in more details:
uid
, type
, conflict
, undefined
, inheritAllowed
, and profile
. By virtue of being a subclass of Asset
it can also use scope
, and via some UML tricks (that I do not really understand) the inheritFrom
also appears.scope
. O.k., that is still fine, but the fact that inheritFrom
does not appear here bothers me.@id
term which, I would think, corresponds to the uid
by virtue of the JSON-LD encoding, and to @type
for type
. This is mentioned in a Note but I have not seen it clearly documented that the general terms map to these and these features of RDF. This is important to make the serialization in other syntaxes clear.permission
or prohibition
. This comes out of the blue for the reader; of course, one can understand that these refer to a Permission
or Prohibition
, respectively, but on the spec level that isn't o.k.conflict
, inheritAllowed
, inheritFrom
, permission
, profile
, prohibition
, undefined
. Note that permission
and prohibition
is present, which is good. It does not explicitly refer to uid
and type
because those are RDF notions anyway; see above. Note the presence of inheritFrom
which is fine.assigner
, action
, target
, and probably others, that are not listed anywhere.(I have not checked the other classes but there may be similar issues...)
There are too many places defining properties and it is not clear which is the authoritative one. To be perfectly honest, the UML diagram does not really talk to me, and I just see it as an extra source of possible problems. But if we choose to keep it, things should be in sync...
I wonder whether it would not make it easier to the reader if we placed a link to the relevant entry in the vocab document for each class (e.g., a link to 4.1.1. in the vocab from 3.1 in the model). The latter should also include the 'assigner', etc., if we want to keep them really (I am not 100% sure, it may be cleaner to use only expanded rules...). The vocab document should be the authoritative definition.
RetroSearch is an open source project built by @garambo | Open a GitHub Issue
Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo
HTML:
3.2
| Encoding:
UTF-8
| Version:
0.7.3