OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

office-metadata message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Re: [office-metadata] finishing up



Bruce D'Arcus wrote:
> A reminder to people: we need to be done with the proposal this week. 
> So tomorrow really needs to be our last meeting.
>
> To that end, here are my final comments on the latest proposal (though 
> am not sure I have the right file; see below), with the last one being 
> critical:
>
> 1) p4, section "Metadata related Files"
>
> We need to confirm with Elias that the substitution of the 
> dcterms:hasPart property is appropriate. I think it is.
I hope as well Elias is able to review these changes, esp. as they have 
been introduced after his comments on the meta manifest two weeks ago.
>
> If it is, remove the OWL fragment there, since it is not our job to 
> define its semantics. The same with other reused properties from DC.

Your objection seems correct, we should not redefine existing 
properties. But by removing this section completely we would lose the 
definition of the relation between odf:Package and odf:File.
What would be the common way to solve this problem when using OWL?
The creation of a subproperty of dcterms:hasPart (e.g. odf:contains) 
could be a possible solution, like:

contains
    Indicates a File that belongs to the Package.
    Sub-property of: dcterm:hasPart
    Domain: Package
    Range: File

>
> 2) Section 1.3
>
> "This element should be kept as well by application that not support 
> metadata" is missing a "do".
Has been changed. I will update the document right away. We used change 
tracking so you can easily review the changes.
>
> Also, do we need discussion here of how one associates a field with a 
> plug-in and graph?
By the metadata manifest.
>
> 3) Am I looking at the wrong file ("07-04-18-2nd-ODF-MetaData")? 
> Where's the binding stuff?? Its application to the field?
You mean the examples? They have been moved into a separate document, as 
ISO like to have examples being separated from the normative sections.
>
> Finally, FWIW, I have decided that citations really ought to be 
> implemented using this generic field. We talked about this last week, 
> and Rob suggested we perhaps issue standalone notes of sort of how to 
> implement different use cases like this, so that we get the benefits 
> of the generic approach, but also of a standard way to implement 
> specific functionality.

Svante


[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]