Re: Updated RDFa Core Editor's Draft

On Mon, 18 Oct 2010 22:19:06 -0500
Shane McCarron <shane@aptest.com> wrote:

>   Sorry for the confusion earlier.  There is an updated editor's
> draft at
> http://www.w3.org/2010/02/rdfa/drafts/2010/ED-rdfa-core-20101018/

2.2: the profile example is wrong dc/dcterms.

3.3-3.6, 7.3, 8.1-8.2: the URIs for Albert Einstein, the German Empire,
the United States, etc should not start with <http://dbpedia.org/page/>
but <http://dbpedia.org/resource/>.

3.10: mentions @src in the same breath as @resource and @href, whereas
@src actually behaves more like @about.

6: "An RDFa Processor must not use the XML 'default namespace' as the
'default prefix'." -- perhaps needs rephrasing, as is suggests that
something like this is illegal:

    <g xmlns="http://example.com/" vocab="http://example.com/">

8: Most of these examples are missing prefix mappings. For little
snippets which don't purport to be full documents, that's fine. For the
longer ones which include entire <html> elements, at least an @profile
should be there.

8.3.1.3: A lot of problems here. "It does not help to escape the
content, since the output would simply be a string of text containing
numerous ampersands" is simply nonsense. The following:

	<html property=":foo" content="&lt;bar&gt;" />

is parsed as:

	<> :foo "<bar>" . 

The phrase "To make authoring easier, if there are child elements and
no @datatype attribute, then the effect is the same as if @datatype
have been explicitly set to rdf:XMLLiteral" is no longer true. That
phrase and the following couple of examples with their explanations
need removing/revising.

-- 
Toby A Inkster
<mailto:mail@tobyinkster.co.uk>
<http://tobyinkster.co.uk>

Received on Tuesday, 19 October 2010 19:51:45 UTC