Hi Simon,

Yes, you can also use the W3C Distiller for most of the testing, except
that our validator also makes sure that you are using the right
vocabularies, and combine them in the way we suggest.

With respect to address data, unfortunately the vcard specification on
which our vocabulary is based doesn't contain hints as to what to do in
this case [1]. The question to me is whether this extra piece of
information (Hammersmith) is necessary to identify the place or not (try
Yahoo Maps, for example). If not, you may safely leave out, otherwise I
would add it to the location.

Cheers,
Peter

[1] http://www.faqs.org/rfcs/rfc2426.html


sjgibbs80 wrote:
>
>
> --- In searchmonkey-siteowners@yahoogroups.com
> <mailto:searchmonkey-siteowners%40yahoogroups.com>, Peter Mika
> <pmika@...> wrote:
>
> >
> > There seems to be a bug and one other issue here. The bug relates to
> the
> > online validator
>
> Right OK, I won't worry too much then. I will need to defer to the W3C
> distiller for testing though I suppose? At least for the time being?
>
> > You
> > can consider this a bug as well, but it's easily fixable on your
> side by
> > using a resource and either rdfs:label or the vcard properties
>
> I should be able to decompose the structure. I wonder if you can help
> with one very specific detail.
>
> My address data contains a notion of "post town" e.g. "London" and a
> second more specific term which I've labelled "locality" e.g.
> "Hammersmith", formally this is known as "dependant locality". See:
>
> http://en.wikipedia.org/wiki/Post_town
> <http://en.wikipedia.org/wiki/Post_town>
>
> There seems to be no predicate in your documentation for, post town,
> town or city specifically, yet the examples give "London" as an
> example of a locality. So I was wondering if *multiple* locality
> statements would be accepted and if the ordering of statements would
> be preserved?
>
> Of course, rdfs:label is fine for me in any case.
>
> Simon
>
>