RE: URIs, names and well known RDDL names, was: Re: Quick edit

[Jonathan Borden <jborden@mediaone.net>]

Tim Bray wrote:
>
>
> At 10:20 AM 09/01/01 +0000, Henry S. Thompson wrote:
> >I strongly agree with Jason's analysis here.  I certainly want _some_
> >attribute of rddl:resource to _always_ be
> >http://www.w3.org/2000/10/XMLSchema if what's pointed to is in fact an
> >XML Schema!
>
> Hrumph.  I think Henry's right.  Maybe my minimalism was a bit too
> brutal?
>
> So, maybe
>
>  role= required canonical name for what kind type of thing it is, like
>        Henry says; we still provide a list of them in rddl.org/roles.html
>
>  arcrole= optional user-provided selector, also a URI, RDDL doesn't
>           provide any help here.
>
> Anyone have a better idea? -Tim
>

I also agree with Jason's analysis. Practically since RDDL was conceived,
I've been going back and forth on what should an arcrole vs. role be w.r.t
RDDL. Indeed one morning while driving to work, I suddenly became conviced
that we'd got it all wrong, and confused an arcrole for a role. At the end
of the day I'd conviced myself that arcrole was fine and that was that...

This is how I see it:

Software will use xlink:arcrole is to dispatch on a resource for a given
URI. In cases where there are more than one resource of a given arcrole,
xlink:role disambiguates, otherwise xlink:role defaults to
"http://www.rddl.org/#resource" (because every rddl:resource is at least a
RDDL resource, and this URI is simply the expansion of the qname
"rddl:resource").

Typing documents is a tricky business in that there is no one right way to
do it. I do think we need to provide examples in arcrole.htm ... these
aren't meant to be authoritative, rather just a convenient place to name a
few types of resources that will be commonly used.

It is great to hear that implementations of RDDL are underway. The behavior
of software provides great feedback on how practical this all is.

-Jonathan