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

Re: RDF 1.1 IRIs and %-escaping

From: Richard Cyganiak <richard@cyganiak.de>
Date: Sat, 11 Aug 2012 14:02:16 +0100
Cc: public-rdf-comments Comments <public-rdf-comments@w3.org>, Graham Klyne <gk@ninebynine.org>
Message-Id: <F03E82C5-0F6B-43A2-AD15-F99BA7B20386@cyganiak.de>
To: Jeremy Carroll <jeremy@topquadrant.com>
Hi Jeremy,

On 10 Aug 2012, at 23:03, Jeremy Carroll wrote:
> On 8/8/2012 4:01 AM, Richard Cyganiak wrote:
>> 1. RDF 2004 is the only standard on the planet that allows spaces in URIs/IRIs
> I believe that the plural in spaces is crucial in making that sentence true ...

Well, having checked a few more references, I'll have to correct my statement above to: RDF 2004 is the only standard on the planet that allows spaces in URIs/IRIs without strongly discouraging their use.

> <http://www.example.org/Two  Spaces>
> conforms with RDF 2004 but not XML Schema 1.0
> and
> <http://www.example.org/One Space>
> conforms with both ....

That is rather funny. Also, in XML Schema 1.0, one can put tabs and line breaks into URIs, but they will be converted to a single space.

To be fair, XML Schema 1.0 says: “Note:  Spaces are, in principle, allowed in the ·lexical space· of anyURI, however, their use is highly discouraged (unless they are encoded by %20).”

> Hmmm, java.net.URI is from the same era, and has similar issues I think.

It doesn't allow spaces. (It allows non-US-ASCII characters, in deviation from RFC 2396.)

> Checking out the XML Schema ref [1], I see there is a second spec - which really does seem to have very similar rules to RDF 2004, XLink [2].

Well, the rules for XLink href are indeed very similar to RDF 2004, but XLink doesn't claim that the value of href is a URI, IRI, URIRef, or URL. It simply says it contains data that allows finding a remote resource, and defines an algorithm for encoding that data to an RFC2396-compliant URI before the data is passed to an URI resolver.

My main point is that, whatever the situation in 2001-2004 was, today we have the IRI spec, and there's no benefit in trying to perpetuate the various subtle deviations from RFC 3987 that were written up in various specs back then.

(And in 2020, a bunch of kids who are editing RDF 2.5 will ask what we were smoking in 2012 when we based RDF 1.1 on IRIs rather than on the obvious correct choice, HTML5 URLs.)


> I think one
> Jeremy
> [1]
> http://www.w3.org/TR/2004/REC-xmlschema-2-20041028/#anyURI
> [2]
> http://www.w3.org/TR/2001/REC-xlink-20010627/#link-locators
Received on Saturday, 11 August 2012 13:02:42 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:29:53 UTC