Re: Comments on RDFa Core 1.1 Working Draft

On Jan 11, 2012, at 17:15 , Andy Seaborne wrote:

> On 11/01/12 13:34, David Wood wrote:
>> Hi all,
>> This message records my comments on RDFa Core 1.1 WD dated 15 Dec
>> 2011 [1].  This review completes my action from the RDF WG [2].
>> I was pleased to see the RDFa Core WD refer to IRIs in place of URIs.
>> This will ease transition of the RDF documents to using IRIs and not
>> require a revisiting of the RDFa Core document.  However, there are
>> still several places in the document that refer to URIs (even when
>> not in direct relation to CURIEs); those should be revisited for
>> consistency.
>> Most of the RDFa Core WD discusses syntactical issues.  As such, I
>> (and hopefully the RDF WG) are agnostic.  I agree with Ivan that we
>> should think of RDFa 1.1 as "just" another standard RDF serialization
>> syntax.
> One nearly-syntactic issue from RDF 1.1 is plain literals/xsd:string and rdf:langString.

This was actually discussed in the RDFa WG. But the problem is that this is not yet a standard, only part of an editor's draft in RDF. It would not be appropriate for the RDFa WG to use something that would be a draft only...


>> However, we don't want to encounter any syntax or usage patterns
>> within RDFa that results in a conflation of IRIs for both resource
>> names and resource contents, as with SPARQL's named graphs.  This is
>> especially important to the RDF WG, given our resolutions related to
>> this problem in SPARQL:
> This should include pointers to what your talking about.
> I would suggest
> (SPARQL 1.0)
> but they do not back up using the word "conflate" which comes more from common practice.
>> - At the RDF WG's first FTF [3], we resolved, "Named Graphs in SPARQL
>> associate IRIs and graphs *but* they do not necessarily "name" graphs
>> in the strict model-theoretic sense. A SPARQL Dataset does not
>> establish graphs as referents of IRIs (relevant to ISSUE-30)".
>> ISSUE-30 is at [4].
>> The RDFa Core WD refers to RDF Concepts in relation to graph
>> definition (sec. 3.7) [5].  This is very helpful and will allow us to
>> update RDF Concepts as required.  However, there is possibly a
>> confusion between a graph's name (a g-snap) and its contents at a
>> point in time (a g-text) in RDFa Core, depending on the
>> interpretation of the base IRI.
> Minor: "its contents"
> I read it as "its" refers to the graph, and so its "contents" risk mixing the contents of a graph container (g-box) and the graph (g-snap - contents are triples as it is a set of triples).
> Suggest: "its representation" (AWWW term).
>> The RDFa Core WD, section 7.2 [6] specifies the base IRI of a
>> document this way: "The base. This will usually be the IRI of the
>> document being processed, but it could be some other IRI, set by some
>> other mechanism, such as the (X)HTML base element. The important
>> thing is that it establishes an IRI against which relative paths can
>> be resolved."
>> That means, to me at least, that the base IRI is often going to be
>> the *same as* the document IRI, thus resulting in conflation of
>> denotation of the graph and the document.
> FWIW: I read the RDFa doc as restating the usual handling of base IRIs.
>> I don't think that needs to change in the RDFa Core document, but it
>> does mean that our named graphs discussion should take note.  The
>> following RDF WG issues relate (and possibly more):
>> - ISSUE-14 "What is a named graph and what should we call it?" [7] -
>> ISSUE-15 "What is the relationship between the IRI and the triples in
>> a dataset/quad-syntax/etc" [8] - ISSUE-17 "How are RDF datasets to be
>> merged?" [9] - ISSUE-29 "Do we support SPARQL's notion of "default
>> graph"?" [10] - especially ISSUE-32 "Can we identify both g-boxes and
>> g-snaps?" [11] - ISSUE-38 "What new vocabulary should be added to RDF
>> to talk about graphs?" [12]
>> Andy Seaborne has pointed out [13] that, "the doc URI is not the
>> graph name in every case -- it is in the web cache pattern."  (By
>> "doc URI" he meant "the RDFa document's base IRI")  That is a good
>> point and should be considered.
> As part of that discussion, didn't conclude that RDFa does not talk about named graphs (any sense) at all?  One document -> some triples.
> A quick check of the processing model suggests to me that RDFa just generates triples, no clumping into any units, so it's a single graph, just as if it were a Turtle or N-triples document.  But it is a quick check (and needs checking!).
>> Dan Brickley asked [13], "what about # URIs in RDFa: can they
>> identify any resource?"  The RDFa Core WD has a note on this [14]
>> which basically says "unfortunately not".
>> Regards, Dave
>> [1] [2]
>> [3]
>> [4]
>> [5]
>> [6]
>> [7] [8]
>> [9]
>> [10]
>> [11]
>> [12]
>> [13]
>> [14]

Ivan Herman, W3C Semantic Web Activity Lead
mobile: +31-641044153

Received on Wednesday, 11 January 2012 16:32:26 UTC