[rdfweb-dev] foaf & genealogy

[Danny Ayers]

>I reckon we can work around that, simply by adopting a document-centric
>approach to 'who said what' (just like the rest of FOAF/RDFWeb stuff does).
>
>gravestone_transcription.rdf
>...would hold an account of the world according to some gravestone, as
>transcribed by someone. Depending on how precise you want to be, you could
>have arbitrary levels of hairiness when representing the
>who/where/what etc of
>the transcription. There is a sub-tribe of the metadata world who
>work with 'Record Keeping Metadata' and concern themselves with
>accuracy about
>such things. We could probably learn a lot from them, while
>starting simple.
>Instead of trying to put everything in one super-sophisticated RDF
>file, we
>put 'claims from some perspective' in each file.

I can see how this could be workable in practice, though have a few doubts
about how it would fit with the theory. Can we really refer to the document
gravestone_transcription.rdf as a resource *and* relate the triples it
contains with the document itself?

>> I reckon the bits needed for simple genealogy should sit very nicely
>> alongside FOAF, and it would allow the foafnaut to do families too.
>
>That'd be good. There are some lists of candidate properties for
>foaf scattered
>around too, but nothing comprehensive yet.

Yep, I bet they're piling up - probably a lot more than would be appropriate
to go in foaf itself. But even if things like genealogical stuff went in
their own namespaces (friend of the family?), the same 'incubator' kind of
approach that seems to have been happening with foaf could be applied - play
with it and see where it goes ;-)