Re: the i18n paragraph

> In general I'm ok with the new paragraph. The last sentence may not
> necessarily be  in parentheses. 

Eric Prud'hommeaux pushed for the parentheses, to help people realize
they don't need to understand this part to understand the document.

> Also note that in RDF/XML, it is
> possible to use xml:lang without the explicit mentioning of
> rdf:XMLLiteral, e.g.,
> 
>  <rdfs:label xml:lang="en">Some text</rdfs:label>

Right, that's how you express the language tag of a a plain literals in
RDF/XML.

Note that such a language tag has no affect on XMLLiterals.  For them,
you have to put the xml:lang attribute on some XML element inside the
literal.  To me, this seems a rather awkward design, since it means you
can't use the design pattern of providing one value in your graph for
each language and then querying for the value that's in the language you
want.  (Or, at least, you can't do that with SPARQL; you'd need a query
mechanism that looked inside the XML Literal.)  This issue is part of
why I didn't want this paragraph, and why I left xml:lang out of the
current version.

     -- Sandro

Received on Tuesday, 2 June 2009 12:20:25 UTC