Re: Body as Named graph

> On 05 Jun 2015, at 20:29 , Robert Sanderson <azaroth42@gmail.com> wrote:
> 
> Hi Giulio,
> 
> Apologies for not answering the question earlier, and thanks for pushing on it.
> 
> The current opinion of the WG, which is certainly able to change based on input and further discussion, is that a named graph body is outside of the scope of the WG efforts, but is not (and must not be) prevented by them.  In essence, as any resource can be the body of the annotation, and a named graph is a resource, it can be the body of the annotation and we don't need to specify it explicitly.

+1

> 
> The challenge is the turtle serialization if the named graph is embedded in the annotation's serialization, as turtle does not support named graphs. [Happy to be corrected if that's not the case!]  JSON-LD, which is our default serialization, does support them thankfully.
> 

True for Turtle. But TriG[1] is also a W3C Recommendation; in effect, though with a different media type, TriG may be viewed as an extension of Turtle that allows for the expression of graphs (and there may be triples without graph denomination, conceptually as a 'default' graph).

One approach may be to refer, possibly as a note, to TriG in the spec saying that implementations MAY accept TriG as a serialization syntax to accommodate Named graphs in annotations.

Ivan

[1] http://www.w3.org/TR/trig/


> If there is an issue that named graphs in particular bring up, then we could specify things more explicitly.  One point to keep in mind is that we would need two implementations and a test suite in order for the feature to make it through the standardization process.  If you would be willing to take the lead on that, it might help?
> 
> Many thanks,
> 
> Rob
> 
> 
> 
> 
> On Wed, Jun 3, 2015 at 5:32 AM, Giulio Andreini <info@giulioandreini.it> wrote:
> Hi,
> as far as I remember into Open Annotation model (http://www.openannotation.org/spec/core/publishing.html#Graphs) there was a particular configuration regarding body.
> In fact body could be identified as rdf graph.
> We are actually using this configuration in order to address our clients needs.
> 
> Basically they are doing a sort of semantic tagging and targets sometimes are associated to LOD resources by using a set of different properties.
> 
> <target1> <talksabout> <dbpedia:Paris>
> <target1> <customProperty> <object>
> 
> Are you planning to cover this case in Web Annotation also?
> 
> Thanks in advance for yr help :)
> Cheers
> Giulio
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> --
> ------------------------------------------------------------------------------------------------
> Giulio Andreini / UX designer, project manager
> info@giulioandreini.it, andreini@netseven.it
> www.netseven.it
> +39 050 55 25 74
> Mobile +39 347 8283609
> Skype giulioandreini
> Linkedin
> 
> Net7
> Via Marche 10 / 56123 | Pisa
> 
> P.Iva e CF 01577590506
> CCIAA di Pisa n. 01577590506 del 26/04/2001
> Capitale Sociale 10.000,00 €
> ------------------------------------------------------------------------------------------------
> 
> 
> 
> --
> Rob Sanderson
> Information Standards Advocate
> Digital Library Systems and Services
> Stanford, CA 94305


----
Ivan Herman, W3C
Digital Publishing Activity Lead
Home: http://www.w3.org/People/Ivan/
mobile: +31-641044153
ORCID ID: http://orcid.org/0000-0003-0782-2704