Re: Atom-Owl

[Henry Story]

Very good comments below.

On 28 Feb 2007, at 04:20, James Leigh wrote:

> Hi Heny,
>
> I have some more comments about atom-owl.
>
> * Some properties I think may need to be functional. I noticed:
> #rel,

yes.

> #body,

that makes sense

> and #uri that I think should be functional;

No:
[[
The atom:generator element MAY have a "uri" attribute whose value
MUST be an IRI reference
]] in section 4.2.4 of rfc4287

> * There is no relationship between #Link and #to, either #Link
> needs
> to extend rdf:Statement or #to have a domain of #Link;

Ah yes. That's what we used to have. I had Link extend rdf:Statement,
with
the following rule

{ ?link :subject ?version;
:rel ?rel;
:to ?content . } => { ?version ?rel ?content . } .

ie, that a link is a reified statement, that is true.

> * If #Link extends rdf:Statement it could cause some confusion each
> #Link instance (likely) will not have a predicate.

No. I think they always do. The default relation is iana:alternate.

> * If the properties of atom:Link are moved to the target object of
> #Link, there needs to be a restriction on the target that says
> allValuesFrom an rdfs:Class that has the hreflang and length
> properties.

which would be the awol:Content class, right.

I suppose this could be done by creating a property awol:relationType
that had those restrictions?
Then one could say that all the iana relations were subproperties of
that relation?

Thanks for the comments.

>
> Best,
> James

[Henry Story]

I have incorporated some of James Leigh's suggestions, and published
them.
You can see the diff at:

https://sommer.dev.java.net/source/browse/sommer/trunk/www/atom/
2006-06-06/AtomOwl.n3?r1=text&p1=trunk%2Fwww%2Fatom%2F2006-06-06%
2FAtomOwl.n3&tr1=87&r2=text&p2=trunk%2Fwww%2Fatom%2F2006-06-06%
2FAtomOwl.n3&tr2=169&diff_format=h

or http://tinyurl.com/2jrz9t

Henry

Home page: http://bblfish.net/
Sun Blog: http://blogs.sun.com/bblfish/
Foaf name: http://bblfish.net/people/henry/card#me