Re: [office-metadata] Persistence of the relation betweentext:meta-field and it's metadata.

[Svante Schubert <Svante.Schubert@Sun.COM>]

Bruce D'Arcus wrote:
>
> On May 16, 2007, at 10:11 AM, Svante Schubert wrote:
>
>>> Perhaps if Elias is there today he just take a few minutes to 
>>> explain his position again? I really don't want to spend the entire 
>>> call talking about this though. We need a tight agenda where we can 
>>> wrap our work (at least WRT to concalls) up today if possible.
>> Do we really need further implementation experience to specify 
>> informations to give ODF applications the opportunity to relate an 
>> xml:id of the text:meta-field to metadata files?
>
> If there's debate about the best way to do this (or even that "this" 
> is the goal), maybe yes.
>
>> This argument to wait for implementation experience could be given 
>> for the whole data of the metadata manifest.
>
> True, but so far we've had no disagreements about those details.
>
>> Why allowing odf:type on files and not on elements?
>> We should be able to draw a line based on arguments traceable by 
>> outsiders.
>
> I agree we should clarify. Perhaps a short discussion during the call 
> and then a resolution in the minutes?
>
That seems adequate.
I suggest as solution to give at least the possibility to add odf:type 
to an odf:Element, without specifying that this is the way the relation 
will be established, leaving room for more sophisticated solutions (e.g. 
reference to a namespace).

OWL question:
In this context I am uncertain if we can specify odf:type twice once 
domain odf:Element and once with a domain odf:File.
We might specify two properties odf:element-type and odf:file-type or 
create a super class odf:Thing for all ODF nodes as odf:Element / odf:File.

Svante