Re: [ISSUE-22] Provenance and Agents

Hi Yves, all,

2012/10/26 Yves Savourel <ysavourel@enlaso.com>

> > I understand, I just couldn't see how if you already support
> > both global and local selectors and for one you want
> > to repeat attributes, why that would not be the case
> > for the other. It just seemed a bit inconsistent, but that
> > thought is probably a bit abstract seeing I can't recall the
> > use case offhand.
>
> I think stand-off markup can be used in two cases:
>
> a) To set on the same span of content several instances of the same data
> category without overriding each other. An example of this is Localization
> Quality Issue where the same content can have several issues.
>

So far I see the application of this really wanted (by *both implementors
and use case(s)*) only for Localization Quality Issue and translation
provenance. So I very much hope that we will not introduce this for other
data categories.

Just my two cents,

Felix



>
> b) To allow a format to assign assign data categories that have more than
> one attribute to a content where you can only use a reference. An example
> of that is XLIFF 2.0 where <mrk> doesn't allow non-XLIFF attributes.
>
> I think a) applies only to few data categories. For b) it would be logical
> to allow it to all data categories that have more than one attributes. For
> example, that could be used to set the Storage Size data category on a span
> of content. But I'm not very happy with this. I think as XLIFF users ITS
> really need to assess the restriction of non-XLIFF attribute on <mrk>. This
> causes a lot of headache. That why resolving the mapping soon is important:
> if <mrk> suddenly allows custom extensions, the use case b) disappear for
> ITS.
>
> Cheers,
> -yves
>
>
>


-- 
Felix Sasaki
DFKI / W3C Fellow