Re: Unresolvable URIs (in Atom link relations)

[Julian Reschke <julian.reschke@greenbytes.de>]

Davis_Cornelia@emc.com wrote:
> Hi Julian, all,
> 
>  
> 
> In your weekend post you make the point that
> 
>  
> 
> “ it also needs to be clear
> 
> that we have a URI for the link target. (Note that that URI doesn't
> 
> necessarily need to be resolvable.)”

I pointed out because you mentioned URI + fragment identifiers, but of 
course it's true for non-resolvable URIs as well.

  > While I, of course, understand that URIs are not URLs (that is, they
> needn’t be resolvable), and that the value of an Atom link relation href 
> is an URI (really an IRI), does the description of, for example, the 
> “service” relation given here 
> (http://www.iana.org/assignments/link-relations/link-relations.xhtml)
> 
> ‘rel="service" indicates a URI that can be used to *retrieve* an Atom 
> Publishing Protocol Service Document as defined by RFC 5023.’  (emphasis 
> added by me)
> 
> not add the requirement of resolvability to these URIs?  This is one 
> specific example, and the Atom specs do similar things – talk about URIs 
> used as addresses.

Yes, that's true. So yes, a link relation can be defined in a away so 
that only resolvable URIs make sense as targets.

> I understand that there will be uses of atom link relations where the 
> URI is NOT resolvable, for example, a URI value that represents me, and 
> that there is always the possibility that at any instance in time, a URL 
> may be not be resolvable.  The question is whether we can spec relations 
> where the href value really is a URL?

I would think so.

BR, Julian

PS: cc'ing Mark Nottingham, maybe he's got more advice.

-- 
<green/>bytes GmbH, Hafenweg 16, D-48155 Münster, Germany
Amtsgericht Münster: HRB5782