Focus on model

["Florian Reuter" <freuter@novell.com>]

Dear collegues,

may I suggest a way to approach consensus in our discussion. To me we mix the encoding problem (RDFa, IDs, Styles) with
the problem of the underlying metadata statements we would like to allow.

I would suggest to first try to get a consensus about which statements we would like to allow and then talk about the
way we would like to encode them in OD.

The first question is whether we can get an agreement which kind of statement we would like to allow and at the very
beginning whether this should be RDF (subject, predicate, object) statements or something different.

If we agree on RDF triples as the underlying model, then I guess we need to talk about what special OD items we would
like to allow beeing a "suject". E.g. candidates would be "paragraphs", "spans", etc.

If we agree on this and our use cases can be solved using the underlying metadata model we have in mind, then we can
talk about how we can encode then in OD either by using an RDFa like approcach, a style-based approach or an ID-based
approach.

Does this makes sense or am I off thead?

~Florian