W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > August 2003

Re: On I18N issue - for cr request

From: Jeremy Carroll <jjc@hplb.hpl.hp.com>
Date: Wed, 20 Aug 2003 14:52:05 +0100
Message-ID: <3F437D05.3040001@hplb.hpl.hp.com>
To: Patrick.Stickler@nokia.com
CC: w3c-rdfcore-wg@w3.org



Patrick.Stickler@nokia.com wrote:

> Overall, looks great Jeremy.
> 
> A few questions:
> 
> 1. In addition to changes related to octets, did we not also address
>    the concerns brought forward by I18N about the nature of the
>    value space of rdf:XMLLiteral, making clarifications to the
>    definition of rdf:XMLLiteral to that end, as well as address
>    the relationships between XML literals, plain literals without 
>    lang tag, and xsd:string? (i.e. the idea of conversion versus equality)


Yes adding that would improve it.

> 
> 2. In your table at the end, it seems to me that the value for
>    "[avoid] an RDF-specific solution [to the problem of] XML [...] context"
>    for "Normal Untyped Literal" and "Special typed literal" should be "No", 
>    since for both, the form of expression of contextual information is 
>    RDF-specific.  ???
> 

I believe those two options are adequately ruled out by other rows.
I think arguments could be made for Yes, and prefer to concede those in 
advance rather than be drawn towards a rathole.

The only serious other contender, in my opinion, is the wrapper hack in 
syntax, possibly only when the xml:lang is anything other than "". I think 
we should have that specified as the alternative design and go to CR asking 
for implementor feedback between those two designs, using the 'at risk' 
phaseolgy. While that does not seem to be precisely what the process 
document has in mind, that is what OWL have done.

Jeremy
Received on Thursday, 21 August 2003 20:17:47 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:59:40 EDT