Re: FPWD comment - literals, data types and language tags

Hi Jacco,

Typically, specifications do not explain the rationale behind every
decision. So I disagree that it would not survive CR/PR as currently
stated. As a normative set of requirements, it's very clear what MUST and
MUST NOT be done.  I agree that not having to explain why there's strange
requirements would be better, of course, and that we should explore all of
the possibilities.

Secondly, while mapping value to @value is possible, it does not solve the
problem as we still need to have value and langauge as keys for a resource
with different semantics (rdf:value, dc:langauge rather than json-ld
literal constructs).

It wouldn't work the other way either (mapping @value to rdf:value), as
then there could be no literals anywhere in the annotation.

Rob


On Mon, Dec 15, 2014 at 1:36 AM, Jacco van Ossenbruggen <
Jacco.van.Ossenbruggen@cwi.nl> wrote:
>
> Dear all,
>
> I like to express my strong support for Antoine’s and Ivan’s arguments.
>
> Having to explain to developers with an RDF background why this triple
>
> > <> oa:body "hi"@en .
>
> is illegal in OA would be a royal pain.  I think a syntax variant like the
> one suggested by Ivan that avoids us having to explain this would be worth
> investigating. The current text, that just declares it illegal without
> explanations just raises questions without providing any understanding of
> the issue. (and I do not think it would survive the REC track process
> anyway).
>
> Jacco
>
>
>
>
>
>
>

-- 
Rob Sanderson
Information Standards Advocate
Digital Library Systems and Services
Stanford, CA 94305