W3C home > Mailing lists > Public > public-rdf-wg@w3.org > August 2012

Re: graham's comment re IRI

From: Eric Prud'hommeaux <eric@w3.org>
Date: Tue, 7 Aug 2012 13:12:43 -0400
To: David Wood <david@3roundstones.com>
Cc: Jeremy Carroll <jeremy@topquadrant.com>, public-rdf-wg@w3.org, Richard Cyganiak <richard@cyganiak.de>
Message-ID: <20120807171241.GA26844@w3.org>
* David Wood <david@3roundstones.com> [2012-08-07 11:56-0400]
> Hi Jeremy,
> 
> Yes, we do owe a formal response and your message below covers the relevant points nicely.  

I think a strong motivation for adopting IRIs is that SPARQL did without hearing a murmer from the RDF community. Given the bredth of adoption of SPARQL (and probably many of the implementation use the grammar verbatim from the spec), we probably have to look in dark corners for folks who are more than mildly inconveninced by removing spaces from RDF IRIs\b\bURL\b\bweb identifiers.

I note that the 2004 version of Turtle permitted spaces (and everything except '>') <http://www.dajobe.org/2004/01/turtle/2006-12-04/#uriref> but in case we're thinking that the ontology folks are in that dark corner, Protégé's Turtle parser barfs on spaces (though not the RDF/XML parser).

At any rate, I think the SPARQL argument is very persuasive; you can't match IRIs with embedded spaces into graph patterns or filters.


> Richard, would you like to respond to the authors of the thread in a single message since you are the Concepts editor?  Thanks.
> 
> Regards,
> Dave
> 
> 
> 
> 
> On Aug 7, 2012, at 09:49, Jeremy Carroll <jeremy@topquadrant.com> wrote:
> 
> > http://lists.w3.org/Archives/Public/public-rdf-comments/2012Aug/0019.html etc.
> > 
> > I am not sure if we are intending a formal response, my preference would be along the lines of:
> > 
> > a)  RDF 2004 has an explicit note anticipating the (then) forthcoming IRI spec (now RFC 3987)
> > b) Some current software, essentially ignoring that note, complies with RDF 2004 but not with the IRI spec (and allows spaces etc.)
> > c) The RDF 1.1 spec will, as chartered, align these two, and software that complies with RDF 1.1 will also comply with IRI; and yes implementations which wish to also support the older RDF spec may have some (not insurmountable) difficulty - e.g. be liberal with what you accept and conservative with what you publish
> > 
> > Jeremy
> > 
> 
> 

-- 
-ericP
Received on Tuesday, 7 August 2012 17:13:16 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 22:02:06 UTC