Re: Looking at the time element (again) (ISSUE-97)

Sorry, some input from a lurker ...

we had a similar issue in the media annotations working group: what actual
time related data type to choose if you are not sure about the actual
values? We came up with the same solution: analyze the lexical space and
fall back to no data type (or string). See the discussion at

http://lists.w3.org/Archives/Public/public-media-annotation/2011Sep/0069.html
http://lists.w3.org/Archives/Public/public-media-annotation/2011Sep/0070.html
http://lists.w3.org/Archives/Public/public-media-annotation/2011Sep/0072.html
http://lists.w3.org/Archives/Public/public-media-annotation/2011Sep/0082.html

FYI, scenario here is to convert XMP multimedia metadata to instance data
of the ontology for meda resources, following these mappings from XMP to
the ontology http://www.w3.org/TR/2011/CR-mediaont-10-20110707/#xmp-table

Felix

2011/11/10 Ivan Herman <ivan@w3.org>

>
> On Nov 10, 2011, at 15:54 , Shane McCarron wrote:
>
> > What if it matches none of them?
>
> I would propose: no datatype then
>
> Ivan
>
>
> >
> > On 11/10/2011 8:52 AM, Gregg Kellogg wrote:
> >> In my version of the proposals, I perform lexical analysis of @datetime
> against xsd:date, xsd:dateTime and xsd:time and choose the datatype based
> on the match. It's quite simple.
> >>
> >> Gregg Kellogg
> >>
> >> Sent from my iPad
> >>
> >> On Nov 10, 2011, at 4:48 AM, "Ivan Herman"<ivan@w3.org>  wrote:
> >>
> >>> Now that<time>  seems to be back into the picture, I have looked at
> ISSUE-97 again[1].
> >>>
> >>> The issue, as raised by Stéphane, proposes to understand the
> '@datetime' property of the<time>  element. Essentially, if the source
> contains this:
> >>>
> >>> <time property="something" datetime="2009-05-10">May 10th 2009</time>
> >>>
> >>> we should, implicitly, consider this as being
> >>>
> >>> <time property="something" datetime="2009-05-10" content="2009-05-10">May
> 10th 2009</time>
> >>>
> >>> and then let the core RDFa processing go. That is of course easy.
> >>>
> >>> However... do we want to add a datatype to this? One would think so,
> but then we get to a very slippery slope. Which datatype? Looking at
> >>>
> >>> http://www.w3.org/TR/xmlschema-2/#date
> >>>
> >>> we do have quite a lot of possibilities... There is of course
> xsd:dateTime (this is what Stéphane used in his original mail for the
> issue). This would mean the transformation of the<time>  element into:
> >>>
> >>> <time property="something" datetime="2009-05-10"
> content="2009-05-10T00:00:00-00:00" datatype="xsd:dateTime">May 10th
> 2009</time>
> >>>
> >>> but there are a bunch of others, like gYear, gYearMonth, etc.
> >>>
> >>> Personally, I would propose to use xsd:dateTime only. But that has to
> be decided by the group.
> >>>
> >>> However, nothing with time is simple... If the author puts in the
> whole ISO format, then are of course fine. But I would expect that in the
> vast majority of cases the hour and minute and the others will all be
> missing. Is it all right to just add the 0 hour, as Stéphane did it? Again,
> I can live with that, but this is something to be decided and known for
> interoperability reasons...
> >>>
> >>> Minor things, but should be cast in stone:-)
> >>>
> >>> Ivan
> >>>
> >>>
> >>>
> >>> [1] http://www.w3.org/2010/02/rdfa/track/issues/97
> >>>
> >>> ----
> >>> 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
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >
> > --
> > Shane McCarron
> > Managing Director, Applied Testing and Technology, Inc.
> > +1 763 786 8160 x120
> >
> >
>
>
> ----
> 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
>
>
>
>
>
>
>