W3C home > Mailing lists > Public > public-prov-wg@w3.org > October 2012

Re: PROV-ISSUE-565 (rdfa-check): Do the suggested way of pointing to provenance work in rdfa [Accessing and Querying Provenance]

From: Ivan Herman <ivan@w3.org>
Date: Sun, 7 Oct 2012 10:32:38 -0400
Cc: Paul Groth <p.t.groth@vu.nl>, Provenance Working Group <public-prov-wg@w3.org>
Message-Id: <4F9A3652-E53C-4133-8DE7-AC9652BC7F5C@w3.org>
To: Graham Klyne <GK@ninebynine.org>

On Oct 6, 2012, at 08:55 , Graham Klyne wrote:

> On 06/10/2012 12:09, Paul Groth wrote:
>> Hi Graham,
>> 
>> Can we use urls in the link headers?
> 
> Yes, I believe that's the preferred option for non-registered link relations.
> 

That is clearly the simplest approach. The fact that, in RDFa, there is a clean abbreviation mechanism for those URI-s is icing on the cake...

(<link rel="FULL URI"...> is perfectly fine with RDFa 1.1, too!)

I think we should consider adding those terms to the prov namespace. I know that prov is for prov-o, primarily, but I do not see any harm adding these object property in addition to the prov-o terms. Alternatively, we would have to define a separate namespace. 

Ivan

> Cf.
>  http://tools.ietf.org/html/rfc5988#section-5.3
>  http://tools.ietf.org/html/rfc5988#section-5.5
> 
>> ... I think that would be the
>> clearest but there is a concern about the overhead of the header.
>> 
>> But I think consistency of use would be good.
> 
> Ack.  I don't see the overhead here being significant.  If there are very many such headers, I think some other mechanism would probably come into play.
> 
> #g
> --
> 
>> cheers
>> Paul
>> 
>> On Sat, Oct 6, 2012 at 8:58 AM, Graham Klyne <GK@ninebynine.org> wrote:
>>> (I'm not following this discussion as closely as I should, so please disregard
>>> if my comments miss the point.)
>>> 
>>> The original intent in PROV-AQ was that the new link relations will be
>>> registered - the primary goal here was HTTP Link: headers.  The introduction of
>>> corresponding URIs was intended to support RDF representations - I assumed that
>>> the RDF URI would also be applicable to RDFa.
>>> 
>>> RDF representation was a somewhat secondary consideration, as the information
>>> provided by these link relations was primarily intended to be used operationally
>>> rather than semantically.  Initially, we were trying to follow what POWDER was
>>> doing, but that seems to have become less relevant as our usage has developed.
>>> 
>>> I think the simple answer may be to just use full URIs everywhere (Link header
>>> rel=, RDF, HTML <Link...>, etc.?   It's not too late to change PROV-AQ.
>>> 
>>> ...
>>> 
>>> Ivan,
>>> 
>>> On a related point (not for PROV), you suggest the form:
>>> 
>>>   <html xmlns="http://www.w3.org/1999/xhtml" prefix="http://www.w3.org/ns/prov#">
>>>      <head>
>>>        <link rel="prov:provenance" href="provenance-URI">
>>>        <link rel="prov:anchor" href="target-URI">
>>>        <title>Welcome to example.com</title>
>>>      </head>
>>>    :
>>> 
>>> What is the primary reference you use to support this use of qnames/curies in
>>> link relations? (XHTML 2 says something about this, but that seems a bit out of
>>> favour these days).
>>> 
>>> Also, is this example complete, or should there be something that specifically
>>> associates the prov: prefix with the given URI (like xmlns:zzz)?  Can you point
>>> me to the relevant reference?
>>> 
>>> Thanks.
>>> 
>>> #g
>>> --
>>> 
>>> On 05/10/2012 18:07, Ivan Herman wrote:
>>>> Ah, sorry! I presume you meant:
>>>> 
>>>> <html xmlns="http://www.w3.org/1999/xhtml">
>>>>    <head>
>>>>      <link rel="provenance" href="provenance-URI">
>>>>      <link rel="anchor" href="target-URI">
>>>>      <title>Welcome to example.com</title>
>>>>    </head>
>>>>    <body>
>>>>          ...
>>>>    </body>
>>>> </html>
>>>> 
>>>> and, well, this is a problem. The issue is the @rel values without any namespace prefix.
>>>> 
>>>> I think that what we have to consider is HTML5; this is clearly the version of HTML that counts in future (although the example here is XHTML). In HTML5 it is still unclear (afaik) what the decision procedure is to accept and define @rel values like those ones above. Looking at
>>>> 
>>>> http://www.w3.org/TR/2012/WD-html5-author-20120329/links.html#linkTypes
>>>> 
>>>> at section "4.12.3.14 Other link types" it says that new link types may be registered via the microformats wiki page
>>>> 
>>>> http://microformats.org/wiki/existing-rel-values#HTML5_link_type_extensions
>>>> 
>>>> That is all fine, and can of course be done.
>>>> 
>>>> However, for RDFa: RDFa+HTML5 understands only a very limited number of unprefixed @rel values as part of the default context:
>>>> 
>>>> http://www.w3.org/2011/rdfa-context/rdfa-1.1.html
>>>> 
>>>> has only three of those for RDFa Core (see last table), and HTML5+RDFa does not accept any more of those.
>>>> 
>>>> I see two possibilities.
>>>> 
>>>> 1. Either the advice is to use prefixed terms:
>>>> 
>>>> <html xmlns="http://www.w3.org/1999/xhtml" prefix="http://www.w3.org/ns/prov#">
>>>>    <head>
>>>>      <link rel="prov:provenance" href="provenance-URI">
>>>>      <link rel="prov:anchor" href="target-URI">
>>>>      <title>Welcome to example.com</title>
>>>>    </head>
>>>> 
>>>> (which means that the attributes should be added to the prov namespace). That is clean and works with RDFa. As Prov becomes a Rec, the 'prov' prefix becomes automatically recognized so, in practice, the @prefix declaration may also be dropped.
>>>> 
>>>> 2. See with the RDFa WG whether updating the initial context by adding the new rel values is a possibility. I could see that happening, but I am  not sure it is worth the trouble. There may be some pushback, because that means all RDFa implementations will have to update their own tables, and that is always a source of possible trouble (that is also why I would advice to keep the @prefix above if we go down that route).
>>>> 
>>>> Cheers
>>>> 
>>>> 
>>>> Ivan
>>>> 
>>>> 
>>>> 
>>>> 
>>>> from the prov-aq document.
>>>> On Oct 5, 2012, at 06:47 , Paul Groth wrote:
>>>> 
>>>>> Hi Ivan,
>>>>> 
>>>>> This is nice but it doesn't test the predicates introduced by prov-aq,
>>>>> which is what I was looking for.
>>>>> 
>>>>> cheers
>>>>> Paul
>>>>> 
>>>>> On Mon, Oct 1, 2012 at 10:39 PM, Ivan Herman <ivan@w3.org> wrote:
>>>>>> Let us see if this is what you were looking for. I attach two HTML files, ghosh.html and ghoshQualified.html. Both of them use HTML with RDFa 1.1. The text itself is of course silly, but this is the Turtle it represents:
>>>>>> 
>>>>>> [[[
>>>>>> @prefix dcterms: <http://purl.org/dc/terms/> .
>>>>>> @prefix foaf: <http://xmlns.com/foaf/0.1/> .
>>>>>> @prefix prov: <http://www.w3.org/ns/prov#> .
>>>>>> @prefix xsd: <http://www.w3.org/2001/XMLSchema#> .
>>>>>> 
>>>>>> <http://.../isbn/000651409X> a prov:Entity;
>>>>>>     dcterms:title "The Glass Palace";
>>>>>>     prov:wasAttributedTo <ghosh.html#AmitavGhosh>;
>>>>>>     prov:wasGeneratedBy [ a prov:Activity;
>>>>>>             prov:endedAtTime "1999-09"^^xsd:gYearMonth;
>>>>>>             prov:startedAtTime "1999-01"^^xsd:gYearMonth;
>>>>>>             prov:wasAssociatedWith <ghosh.html#AmitavGhosh> ] .
>>>>>> 
>>>>>> <ghosh.html#AmitavGhosh> a prov:Agent;
>>>>>>     foaf:name "Amitav Ghosh" .
>>>>>> ]]]
>>>>>> 
>>>>>> and
>>>>>> 
>>>>>> [[[
>>>>>> @prefix dcterms: <http://purl.org/dc/terms/> .
>>>>>> @prefix foaf: <http://xmlns.com/foaf/0.1/> .
>>>>>> @prefix prov: <http://www.w3.org/ns/prov#> .
>>>>>> @prefix xsd: <http://www.w3.org/2001/XMLSchema#> .
>>>>>> 
>>>>>> <http://.../isbn/000651409X> a prov:Entity;
>>>>>>     dcterms:title "The Glass Palace";
>>>>>>     prov:qualifiedAttribution [ a prov:Attribution;
>>>>>>             prov:agent <ghoshQualified.html#AmitavGhosh>;
>>>>>>             prov:hadRole <http://purl.org/spar/pro/author> ];
>>>>>>     prov:wasGeneratedBy [ a prov:Activity;
>>>>>>             prov:endedAtTime "1999-09"^^xsd:gYearMonth;
>>>>>>             prov:startedAtTime "1999-01"^^xsd:gYearMonth;
>>>>>>             prov:wasAssociatedWith <ghoshQualified.html#AmitavGhosh> ] .
>>>>>> 
>>>>>> <ghoshQualified.html#AmitavGhosh> a prov:Agent;
>>>>>>     foaf:name "Amitav Ghosh" .
>>>>>> 
>>>>>> ]]]
>>>>>> 
>>>>>> Is this what you were looking for?
>>>>>> 
>>>>>> Cheers
>>>>>> 
>>>>>> Ivan
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> On Oct 1, 2012, at 11:09 , Paul Groth wrote:
>>>>>> 
>>>>>>> Hi Ivan,
>>>>>>> 
>>>>>>> I don't have any current examples in rdfa using these properties. I'm
>>>>>>> pretty sure they are fine but I want proof. In particular, I'm
>>>>>>> interested in how complicated the mark-up is.
>>>>>>> 
>>>>>>> If you want to code up an example in RDFa that would be great :-)
>>>>>>> 
>>>>>>> Otherwise, this is a reminder for somebody (i.e. me or graham) to do it.
>>>>>>> 
>>>>>>> does that make sense?
>>>>>>> 
>>>>>>> cheers
>>>>>>> Paul
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> On Mon, Oct 1, 2012 at 4:04 PM, Ivan Herman <ivan@w3.org> wrote:
>>>>>>>> (Putting my RDFa hat on:) I do not understand the issue here. Could somebody provide some details?
>>>>>>>> 
>>>>>>>> Ivan
>>>>>>>> 
>>>>>>>> On Sep 27, 2012, at 11:03 , Provenance Working Group Issue Tracker wrote:
>>>>>>>> 
>>>>>>>>> PROV-ISSUE-565 (rdfa-check): Do the suggested way of pointing to provenance work in rdfa [Accessing and Querying Provenance]
>>>>>>>>> 
>>>>>>>>> http://www.w3.org/2011/prov/track/issues/565
>>>>>>>>> 
>>>>>>>>> Raised by: Paul Groth
>>>>>>>>> On product: Accessing and Querying Provenance
>>>>>>>>> 
>>>>>>>>> PAQ suggests using prov:hasAnchor, prov:hasProvenance etc. It is unclear that these work with RDFa this should be checked
>>>>>>>>> 
>>>>>>>>> 
>>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>>> ----
>>>>>>>> Ivan Herman, W3C Semantic Web Activity Lead
>>>>>>>> Home: http://www.w3.org/People/Ivan/
>>>>>>>> mobile: +31-641044153
>>>>>>>> FOAF: http://www.ivan-herman.net/foaf.rdf
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> --
>>>>>>> --
>>>>>>> Dr. Paul Groth (p.t.groth@vu.nl)
>>>>>>> http://www.few.vu.nl/~pgroth/
>>>>>>> Assistant Professor
>>>>>>> - Knowledge Representation & Reasoning Group |
>>>>>>> Artificial Intelligence Section | Department of Computer Science
>>>>>>> - The Network Institute
>>>>>>> VU University Amsterdam
>>>>>> 
>>>>>> 
>>>>>> ----
>>>>>> Ivan Herman, W3C Semantic Web Activity Lead
>>>>>> Home: http://www.w3.org/People/Ivan/
>>>>>> mobile: +31-641044153
>>>>>> FOAF: http://www.ivan-herman.net/foaf.rdf
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> --
>>>>> --
>>>>> Dr. Paul Groth (p.t.groth@vu.nl)
>>>>> http://www.few.vu.nl/~pgroth/
>>>>> Assistant Professor
>>>>> - Knowledge Representation & Reasoning Group |
>>>>>   Artificial Intelligence Section | Department of Computer Science
>>>>> - The Network Institute
>>>>> VU University Amsterdam
>>>> 
>>>> 
>>>> ----
>>>> Ivan Herman, W3C Semantic Web Activity Lead
>>>> Home: http://www.w3.org/People/Ivan/
>>>> mobile: +31-641044153
>>>> FOAF: http://www.ivan-herman.net/foaf.rdf
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>> 
>> 
>> 


----
Ivan Herman, W3C Semantic Web Activity Lead
Home: http://www.w3.org/People/Ivan/
mobile: +31-641044153
FOAF: http://www.ivan-herman.net/foaf.rdf
Received on Sunday, 7 October 2012 14:33:05 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Sunday, 7 October 2012 14:33:06 GMT