[annotator-dev] Open Annotation support in Annotator

Fri Oct 10 21:17:33 UTC 2014

On Fri, Oct 10, 2014 at 2:15 PM, Robert Sanderson <azaroth42 at gmail.com>
wrote:

>
> Hi Ben, Randall,
>
> On Fri, Oct 10, 2014 at 2:08 PM, Benjamin Young <bigbluehat at hypothes.is>
>  wrote:
>
>> > 2.  User holds the user name of the account, and thus the appropriate
>>> mapping is foaf:nick
>>> >For our use at Hypothesis, the user field is a URI, not a string
>>> literal representing a foaf:nick.
>>>
>> Could the value of oa:annotatedBy not directly be the user URI in this
>> case?
>> "The object of the [oa:annotatedBy] relationship is a resource that
>> identifies the agent responsible for creating the Annotation."
>> Is that overly simplified?
>>
>>
> Yes, it could be ... and indeed that would be exactly what the model
> expects.
>
> The distinction would be between:
>     { "oa:annotatedBy" : "http://server.net/people/handle" }
> and:
>     { "oa:annotatedBy" : {"foaf:nick" : "handle" } }
>
> Where the first is clearly globally unique, and the second is dependent on
> the user list of whichever system is managing the 'handle' user.
>
> It'd be great if this was able to be relied upon in Annotator 2.0?  Or is
> that asking too much? :)
>

I think I'm not understanding the ask.

What I thought you were asking was for Annotator 2.0 to always assume the
foaf:nick case and to serialize things that way.

What I'm asking is for Annotator 2.0 not to take a stance on the presence
or absence of an annotatedBy relation or whether its r-value is inlined or
a URI and what its properties are.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.okfn.org/pipermail/annotator-dev/attachments/20141010/4443c5b8/attachment-0004.html>