W3C home > Mailing lists > Public > public-owl-wg@w3.org > November 2007


From: Jeremy Carroll <jjc@hpl.hp.com>
Date: Mon, 26 Nov 2007 10:13:21 +0000
Message-ID: <474A9C41.7010205@hpl.hp.com>
To: Alan Ruttenberg <alanruttenberg@gmail.com>
CC: "Web Ontology Language ((OWL)) Working Group WG" <public-owl-wg@w3.org>

Alan Ruttenberg wrote:

>> - there are an excess of WDs with a claimed normative defn of CURIEs 
>> (the CURIE WD, an RDFa WD, and two XHTML 2.0 WDs)
>>     + a copy/paste solution would make this worse
> Agreed, at least if they are not the copy/pasting the same thing. 
> Haven't checked.

Even if the text is identical - the review cost of having N copies of 
the same text increases probably with N^2. (Reflecting deficiencies in 
one's ability to review - arguments concerning owl:sameAs, show that it 
should be O(N))

>> - the prefix binding issue that is not addressed with CURIEs is 
>> non-trivial (e.g. early versus late binding).
> "When CURIES are used in a non-XML grammar, the grammar MUST provide a 
> mechanism for defining the mapping from the prefix to an IRI."
> So, in our case, for the functional syntax, this choice is delegated to 
> us, and we have an existing mechanism.

Hmmmm - we need to think about what happens when "eg:name" moves from my 
document to yours, or is imported from my documented into yours etc. 
scoping ... Not difficult problems, but that hasn't stopped people 
getting it wrong.

>> - tension with the SPARQL not-a-CURIE construct
> Why an issue: Because of  cut and paste from functional syntax to 
> SPARQL? SPARQL's choice is upward compatible with CURIES.

[considered silence]

>> - if we are going to use CURIEs in XML documents (such as the XML 
>> serialization of the functional syntax), then
>>     + we should be asking the XML Schema WG to extend the primitive 
>> base types of XML Schema 1.1 (Datatypes).
>>     + it is likely that XML tools, such as XSLT, will have difficulty 
>> in getting the prefix binding correct.
> The proposal was solely for use in functional syntax.

That's probably OK - although I am unclear how abstract or concrete the 
functional syntax is meant to be. If it is merely an internal notation 
for the WG, then whatever we do is OK, but in such a case we wouldn't 
really have an issue (I don't think). I take it that this is intended to 
impact some software, but which software?

>> - this TAG issue  
>> http://www.w3.org/2001/tag/issues.html#abbreviatedURIs-56

Received on Monday, 26 November 2007 10:13:54 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:42:00 UTC