RE: roles, relationships, etc...

Hi Pierre-Antoine,

I also thought of option (a), and using comprehensive lists of the sub-properties (such as the EBU one for the roles of creators and contributors). We shortly discussed in Stockholm whether these lists of sub-properties should be normative or not, but afaik we do not have an agreement on this.

Best regards,
Werner

> -----Original Message-----
> From: public-media-annotation-request@w3.org [mailto:public-media-
> annotation-request@w3.org] On Behalf Of Pierre-Antoine
> Sent: Dienstag, 20. Oktober 2009 15:34
> To: public-media-annotation@w3.org
> Subject: roles, relationships, etc...
> 
> Hi again,
> 
> in several places in the API, we have qualifiers to the returned value:
>    Contributor.role
>    Fragment.role
>    Relation.relationship
> 
> by contributing to the API Open Issues wiki page, I realize that I'm
> not
> sure of what is inteded here.
> 
> Is it expected that a) we define in our API a list of possible roles
> for
> contributors (to keep this sole example), and define a mapping of this
> list with all the possible roles in the in-scope formats ?
> 
> Or do we expect b) to get here a string coming "directly" from the
> underlying format?
> 
> I was implicitly assuming (b) (e.g. in my mail [1]) but it seems to me
> that others are assuming (a) -- and the more I think about its, the
> more
> I prefer (a) myself ;) It seems to offer a greater interoperability and
> is homogeneous to what we do for 1st level properties (defining a
> "central" property, and a mapping from that property to any other).
> 
> Could anybody confirm that this is how they envision the definition of
> those "sub-properties" in the API?
> 
>    pa
> 
> 
> [1]
> http://lists.w3.org/Archives/Public/public-media-

> annotation/2009Oct/0038.html

Received on Wednesday, 21 October 2009 12:28:51 UTC