Re: JSON-LD context URI

> On 03 Dec 2014, at 23:06 , Robert Sanderson <azaroth42@gmail.com> wrote:
> 
> 
> All,
> 
> Also not as chair*, the opinion I quickly expressed on the call this morning in more detail:
> 
> 1.  We must create a new context URI.

I agree.

> I'm not sure if this needs to be done before FPWD or not?

We should try. See below

> 
> Right now, we have placeholder text in the context appendix:
>         "... and can be referenced as http://www.w3.org/xxx/yyy."

Yep, it is seeing that in the text that triggered my comments:-)

I do not remember whether this URI is 'active' in the text, ie, whether it is an <a> element. If yes, than it must be a dereferenceable URI even for the FPWD; if it is only text, then it is not a show stopper, procedurally. That being said, I think it would be better to do it for real; I am not sure how long it will take for a next version of the model document (our focus will probably shift for a while) and implementers may already experiment with JSON-LD. Better provide something already now. (In which case the URI in the text should also be 'active'.)


>  We also do not anywhere actually give the @context key in any of the JSON-LD examples, including the "complete" example in appendix C.

That is fine. But we do say it is JSON-LD; if people really want to experiment with it, they will need the @context.

> Even if we hadn't changed anything in the model or vocabulary, the decision to use more developer friendly keys in the JSON serialization would require a new context document.

Yep. And whatever we do today is not cast in concrete.


> 
> 2.  It would be beneficial to NOT create a new namespace URI and instead continue to use the /ns/oa# URI.  The community group baked this notion in from the beginning, such as not having a version or date in the URI on the very sensible recommendation of Dan Brickley and his experience with FOAF, amongst others.

That works with me.

> 
> 
> 3.  The discussion previously that resulted in the ontology being included in the context document was around the rule that /ns/ must only be used for ontology specifications.   Thus the workaround was to include both context and ontology in the same document.  As JSON-LD is increasingly popular, I'm sure that the issue as to context URIs could be solved in the general case as a recommended best practice.
> 
> Is this something that the chairs+staff should raise more broadly within the W3C?  And to what extent do we need to solve it for FPWD or soon thereafter?
> 

Hm. I see your point. I can explore that, but there is a possibility around this: what about

http://www.w3.org/annotation/context.jsonld

as a URI. That can be created at any time.

I will explore the issue, and may come back to you before you get back at your desk. In any case, if we go down that route, somebody should gives me the JSON-LD source, and I am happy to create that file and the FPWD should be updated accordingly. Does this work?


> 4.   Paolo and I, as community group chairs, should also confirm this approach with the CG that clearly has a stake in what happens at the namespace and context URIs.  I personally would be very surprised if there's any pushback, but it would be both polite and encourage continued engagement to do so.

Please do. And another thing: I have not followed all the details, but has there been any change on the vocabulary itself (I do not think so, just checking...)? I mean: are the /ns/oa.* files up-to-date? In any case, the HTML page must be updated, referring to the current state; please provide Doug or I with a new version.

Thanks

Ivan


> 
> 
> Rob
> 
> * Please assume this for everything I say, unless I specifically say "As chair, ..." :)
> 
> 
> On Wed, Dec 3, 2014 at 8:12 AM, Ivan Herman <ivan@w3.org> wrote:
> 
> > On 03 Dec 2014, at 16:57 , Frederick Hirsch <w3c@fjhirsch.com> wrote:
> >
> > Ivan
> >
> > Thanks for noting this.
> >
> > It seems to me (personally not as chair) that the right thing to do is the following
> >
> > 1. Keep oa as 'http://www.w3.org/ns/oa for the Web Annotation Data Model we are producing, as in the FPWD
> >
> > 2. Update the landing page appropriately
> >
> > 3. Add a clear warning to the Community Group  Open Annotation Data Model ‘community draft’ that there is an update underway in the Web Annotation WG, with a pointer to our new draft and home page
> >
> > It seems better to focus on the new work with the same URL as implementations are changing along with the specification.
> >
> > JSON-LD is new work so we don’t have a backward compatibility issue, right?
> 
> Actually, there is a oa.jsonld file right now which is the @context AND the full vocabulary. In other words, it is a JSON-LD serialization of the full vocabulary. And what we need is a separate file for the @context only on, probably, a separate URI.
> 
> Ivan
> 
> >
> > regards, Frederick
> >
> > Frederick Hirsch, Nokia
> > Co-Chair W3C Web Annotation WG
> > @fjhirsch
> >
> >
> >
> > On Dec 3, 2014, at 4:37 AM, Ivan Herman <ivan@w3.org> wrote:
> >
> >> Rob,
> >>
> >> one thing I realized, while looking at the FPWD text: the text refers to oa as 'http://www.w3.org/ns/oa. At this moment that URI exists and refers to the Community Group's output, with the vocabulary set up as an HTML+RDFa and other formats like turtle or json-ld.
> >>
> >> If we decide to keep that URI, we should update its landing page. It also means that this WG will override the various vocabularies that are already there. Do we want to do that? If so, we should check whether a new version should be put at those places while publishing the FPWD.
> >>
> >> We should also put up on the site a proper JSON-LD @context file; the current document has a phony URI for that purpose, and http://www.w3.org/ns/oa.jsonld has the full vocabulary (I am not sure whether we want that to be the @context file, it is probably too big for that purpose)
> >>
> >> Cheers
> >>
> >> Ivan
> >>
> >>
> >> ----
> >> 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
> >>
> >>
> >>
> >>
> >
> >
> 
> 
> ----
> 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
> 
> 
> 
> 
> 
> 
> 
> --
> Rob Sanderson
> Technology Collaboration Facilitator
> 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

Received on Thursday, 4 December 2014 08:57:43 UTC