Re: [office-metadata] Rough notes (I won't call them minutes just yet)

[Bruce D'Arcus <bruce.darcus@OpenDocument.us>]

My answers/questions to some of these ...

On Jan 31, 2007, at 4:17 PM, Patrick Durusau wrote:

> Open Questions:
>
> ...
>
> 3. Use of both path and global IRI in a manifest file

Right now, I say yes, with the latter being optional. Elias should be 
back soon, so maybe we could run this by him.

> 4. xml:id required?

Required in what sense? That we allow it?

> 5. Preservation of all metadata? Means content not understood must be
>    preserved.

Yes.

> 6. How to handle metadata in file plus metadata in separate file? If
>    we want applications to honor the original storage of the
>    metadata? (Florian) Need to say that applications must support
>    both ways.

I agree with Florian.

> 7. Bernd: Structured content, what about table content? Span would not
>    work between rows. Svante: can use RDFa or XForms approach. Need
>    metadata on row, column, and sets of rows and columns.

Yes.

> 8. Bernd: To enable storage of files by document characteristics and
>    retrieval by the same means rather than tree directory
>    structures. An add-on value for ODF metadata. Perhaps a vocabulary
>    that could be understood by all applications. Makes the document
>    self-characterizing.

This is unclear. Are we asking to be able to type a metadata file? E.g. 
to then be able to associate it, for example, with a plug-in? If yes, 
in general I think the manifest needs this.

> 9. Svante: Patient medical record example - RDF statement about the
>    doctor - someone wants all the documents where this doctor is
>    taking care of this patient - vocabulary dependent - Search RDF
>    only.

Document repositories are likely to be using an abstraction anyway, so 
I don't think it matters whether the statements are in RDF/XML or not 
(in-content attributes). It'd be easy to create a script that indexes a 
collection of documents anyway.

Bruce