Re: Valid values for @type in @context

Thanks for your explanation, Markus. I agree that the spec is technically
clear on the point, if you read it all. My point was just about helpfulness
to someone who has not yet digested the whole text. I write from the point
of view of a fan of the idea of JSON-LD -- I would like to use it as one
binding for InLOC [1].

To me (and possibly others) it would be helpful either
(a) to put a note in section 5.4 in any case -- not talking about type
coercion, but simply noting that @type arrays cannot be used within
@context (you don't have to say why at that point), or
(b) to dis-overload (?) @type, so that within @context there is
"@valuetype" or something, with @type reserved for nodes outside the
context.

But if people reckon I am in an insignificant minority of people who would
be helped by this, fine...

Thanks

Simon

[1] http://wiki.teria.no/display/inloc/Home


On 4 October 2013 11:37, Markus Lanthaler <markus.lanthaler@gmx.net> wrote:

> On Friday, October 04, 2013 12:15 PM, Simon Grant wrote:
> > Following up this thread for values of @type in @context...
> >
> > Would I be right in thinking that within the @context one cannot
> > have @type as an array? I can see problems in type coercion if
> > multiple types were allowed in the @context.
>
> You are right.
>
>
> > If I am right, might it be a good idea to add a note in section 5.4
> > to the effect that an array cannot be used within the context?
>
> Section 5.4 would be the wrong place to do so because it doesn't talk
> about type coercion it would belong into section 6.5. I think that section
> is already clear enough:
>
>   Type coercion is specified within an expanded term definition using
>   the @type key. The value of this key expands to an IRI
>
> Furthermore, the grammar in section 8.7 says
>
>   If the expanded term definition contains the @type keyword, its value
>   MUST be an absolute IRI, a compact IRI, a term, null, or the one of
>   the keywords @id or @vocab.
>
> Does this address your concerns?
>
>
> --
> Markus Lanthaler
> @markuslanthaler
>
>
>


-- 
from Simon Grant +44 7710031657 http://www.simongrant.org/home.html