Re: tag URIs (was Re: New Intro to RDF)

Aaron, 

I believe the distinction between URL and URN has been obsoleted by
reality - one can easily use an HTTP URI for a name and never put
anything on that HTTP address, or one can use URNs that are short-lived,
defying the purpose.

The added benefit of being able to resolve the identifier makes HTTP
URIs the preferred identifiers. Recent publications from the TAG (sorry
I have no link handy at the moment) talk about information resources, a
subclass or resources which are representable as documents (bits and
bytes) - these resources should be addressed by what you call URLs.
Other resources should not have direct representations, but if you use
an HTTP URI to identify them, you can redirect any browsers to an
appropriate documentation URI - http://jacek.cz/apartment1/bed1 could
redirect to http://jacek.cz/apartment1/description#bed1 where bed 1 is
described.

So generally again, URNs should be URLs so that a user can do a form of
discovery by dereferencing the URN, and URLs should be stable (URNs) so
that user confusion (bookmarks not working, unexpected results etc.) is
avoided. I prefer only to use the acronym URI for all the identifiers,
and not to distinguish overly between locators and names.

Best regards,

Jacek Kopecky

On Thu, 2005-10-06 at 09:47 -0700, Aaron Straup Cope wrote:
> Danny Ayers wrote:
> 
>   This reminds me, perhaps you can save me trawling the docs. The URI
> > for every property I've ever seen uses the http: scheme. But is the
> > http: scheme actually mandated anywhere in the specs?
> > 
> > It does make sense to use http: URIs, it's useful to have something
> > available on the Web for those URIs. (It looks like a RDDL doc at the
> > ns URI might be the something favoured by the TAG:
> > http://www.w3.org/2001/tag/issues.html#namespaceDocument-8 )
> > 
> > But is it The Law?
> 
> My understanding was that :
> 
> "[A] URL is a type of URI that identifies a resource via a 
> representation of its primary access mechanism (e.g., its network 
> "location") rather than by some other attributes it may have. Thus as we 
> noted, "http:" is a URI scheme. An http URI is a URL."
> 
> The source of which was the "URI Planning Interest Group, W3C/IETF" and 
> the link to which I've since lost. Perhaps this has been superseded by 
> another spec. (I hope not because the above is pretty straightforward to 
> my eyes.)
>

Received on Saturday, 8 October 2005 15:20:11 UTC