No subject


Sun Dec 13 20:02:41 CET 2009


'Chain properties. This is an interesting extensions which allows OWL
2 to derive uncle from the combination of parent and brother. You can
say that a chain of properties, when composed together, imply another
property.'

Or is there actual value for the SemWeb in having this kind of
distributed redundancy, since claims can be confirmed by virtue of a
weight of evidence. If 1000 pages all have the same name for the
organization, is that more persuasive than if just one page links that
name to the org? (maybe not if those pages were all created with the
same tool / process...)

cheers,

Dan


ps.
> The problem with #!! is that it subverts the naming authority. If I'm
> publishing an RDF document to a domain that I own, *I* get to choose
> what fragment IDs mean.

Yes, I acknowledge that. This is why we choose a namespace that is at
least obscure, and cannot overlap with names chosen for the massively
most common situation: HTML anchors. Since microformats, microdata,
and various 'well known URL path' efforts also subvert formal naming
authority, I can live with borrowing '#!!*'. I don't think I've ever
seen a # view identifier deployed that begins with '!'.


More information about the foaf-dev mailing list