W3C home > Mailing lists > Public > public-rdfa-wg@w3.org > February 2011

Re: ISSUE-84 (Cool URIs and HTTPRange-14): The W3C TAG has asked us to mention that the use of fragment identifiers can be problematic [LC Comment - RDFa Core 1.1]

From: Nathan <nathan@webr3.org>
Date: Tue, 08 Feb 2011 10:39:57 +0000
Message-ID: <4D511D7D.2010605@webr3.org>
To: Ivan Herman <ivan@w3.org>
CC: RDFa Working Group WG <public-rdfa-wg@w3.org>
agree, good thinking re @vocab too!

We could define Absolute-URI pointing to [1]

   absolute-URI  = scheme ":" hier-part [ "?" query ]

[1] http://tools.ietf.org/html/rfc3986#section-4.3

We actually may have an issue around this, that may be worth raising (if 
you agree I can raise it) - because currently we use the phrase "AbsURI" 
(e.g. TERMorCURIEorAbsURI) through the spec to refer to a "URI" (not an 
absolute uri), and now it seems need to define a real AbsURI.

I'd be happy to detail the issue and run through the spec pointing to 
the areas where things are ambiguous, with proposed text to clear up the 
issue.

Best,

Nathan

Ivan Herman wrote:
> I would agree with this restriction. @profile values should indeed be absolute URI-s, and I have not found any explicit text in the document that says so. I guess this should be added to, eg, section 9.
> 
> I would not care too much whether the @profile has a fragment or not, as long as it is clear that the fragments are, essentially, stripped by the HTTP rules.
> 
> I would actually add the same restriction to @vocab.
> 
> 
> On Feb 8, 2011, at 11:18 , Nathan wrote:
> 
>> Ivan Herman wrote:
>>> There may be one more additional feature we may want to add to the text, beyond what the TAG might give us: avoid using fragid-s in profile URI-s. That is the only URI that an RDFa processor will dereference, and two different URI-s differing by a fragid only will return the same graph. On the other hand, using two different URI-s for the same graph my make the local caching process inefficient (unless clients would strip the fragid part before caching but I would not expect them to do that...)
>> Hmm, should we be clarifying the definition of @profile to accept a whitespace separated list of absolute-URIs (never relative, never with fragment) to avoid /some/ unexpected behaviour?
>>
>> * I say some, because people could still use "mailto:bob@example.org", but I figure we shouldn't cater for people who like to do things like that!
>>
>> Best,
>>
>> Nathan
>>
>>>>>> ISSUE-84
> 
> 
> ----
> Ivan Herman, W3C Semantic Web Activity Lead
> Home: http://www.w3.org/People/Ivan/
> mobile: +31-641044153
> PGP Key: http://www.ivan-herman.net/pgpkey.html
> FOAF: http://www.ivan-herman.net/foaf.rdf
> 
> 
> 
> 
> 
Received on Tuesday, 8 February 2011 10:40:49 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:19:51 UTC