Re: Determination of subjects/objects (was: ISSUE-42)

Niklas has already touched upon the problem of @id, and I was also a bit
surprised to see it in the list. I there is a genuine danger that an
'id' used for something not really  RDF-ish (eg, if I want to have a
table of content for my document, I will have to add 'id'-s to some
elements, but this may have no relevance to any generated RDF).

However... I presume it is true that a change in the current 'RDF
identity' as an effect of an @id happens _only if there is an RDFa
related attribute on that element_, ie, if there is an @rel, @property,
etc, around. Other than that, the current RDF identity remains intact
while going down the XML tree. Is that so? If yes, than an @id appearing
on one of the elements in the tree may not have any effect. Is that
correct? With that additional rule, the usage of @id may not be that
dangerous after all.

Having said that, I wonder whether it is not simpler to remove @id from
that algorithm altogether. It may be a little bit more convoluted for
the user but certainly safer.

Ivan

Ivan Herman wrote:
> I follow here Niklas' advice[1] on starting a new thread. It is indeed a
> central issue and it is better tracked that way. Reminder: there was an
> algorithmic description on what happens in Ben's mail[2] and an older
> one from me[3] as well as my proposal for anonymous nodes[4].
> 
> Let me also reproduce here, for easier reference, Ben's algorithm[2]. I
> believe it is cleaner than mine[3] because it does not give a different
> behaviour to @resource (apart from a priority issue). Niklas has already
> started to match it against various use cases and we should probably
> follow that. So here it is:
> 
> Consider the concept of an HTML element's "corresponding RDF identity".
> 
> - the "corresponding RDF identity" is determined as follows:
>   - @resource takes precedence
>   - if no @resource, @href is next
>   - if no @href, then @id.
>   - if no @id, then @about.
>   - if all else fails, a bnode.
> 
> - the element's corresponding RDF identity is the subject of an rdf:type
> triple when instanceof appears.
> 
> - the element's corresponding RDF identity is also the object of any
> @rel, or the subject of any @rev.
> 
> - the subject of a @rel, or the object of a @rev, is determined using
> the normal @about rules.
> 
> - the subject of an element's *content* is the element's corresponding
> RDF identity when @rel, @rev, @href, @resource, or @instanceof appears.
> Otherwise, it's @about resolution.
> 
> 
> I will react on some of the issues here in a separate mail, to have a
> clean thread.
> 
> I hope this helps
> 
> Ivan
> 
> 
> 
> [1]
> http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2007Jul/0159.html
> [2]
> http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2007Jul/0158.html
> [3]
> http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2007Jul/0153.html
> [4]
> http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2007Jul/0144.html

-- 

Ivan Herman, W3C Semantic Web Activity Lead
Home: http://www.w3.org/People/Ivan/
PGP Key: http://www.ivan-herman.net/pgpkey.html
FOAF: http://www.ivan-herman.net/foaf.rdf