Re: Exploring the change from Ref to Uri

Hi Yves, all,

I'm not sure if the change really makes things clearer. Looking at other
vocabularies which require an URI data type, I don't find one with the
*name* "URI". For example, HTML has the href attribute, XLink as well, ...

should we really name the attributes after their data type, or isn't
naming after their function the common way? I am not sure if the term
URI is common enough for that purpose.

Cheers,

Felix



Yves Savourel wrote:
> Hello everyone,
> 
> My action item
> _http://www.w3.org/2006/07/24-i18nits-minutes.html#action05_ was to look
> at the possibility of renaming of all our 'xyzRef' and 'xyzRefPointer'
> to 'xyzUri' and 'xyzUriPointer'.
> 
> This was a thought from Christian to clarify better the value held by
> the "Ref" attributes when we looked at the isssue #3494 during today's
> call (_http://www.w3.org/2006/07/24-i18nits-minutes.html#item06_).
> 
> 
> -- Rational:
> 
> The content/values of the nodes pointed by all xyzRef attributes are
> always URI. We might as well make this clear by using 'Uri' in the name
> of the attributes. Using the more generic 'Ref' could possibly lead to
> thinking that other types of value (e.g. an ID) could be used.
> 
> 
> -- Things we would have to rename:
> 
> locNoteRef                      ->      locNoteUri
> locNoteRefPointer               ->      locNoteUriPointer
> termInfoRef                     ->      termInfoUri
> termInfoRefPointer      ->      termInfoUriPointer
> 
> 
> -- Pros:
> 
> - More specific, and therefore clearer.
> - Last chance to make that change.
> - It would be consistant with uri in <its:ns> where we call the URI 'uri'.
> 
> 
> -- Cons:
> 
> - May could be seen as a substantive change by some(?)
> - Lot of references and examples, to change.
> 
> 
> -- Personnal opinion:
> 
> I would think it's probably a good idea to change the attribute names
> from 'Ref' to 'Uri'.
> 
> 
> Cheers,
> -yves
> 
>