Re: xml:lang and XML infoset: two new datatypes

Dan Connolly wrote:

> On Fri, 2002-09-20 at 10:15, Sergey Melnik wrote:
> 
>>I'm suggesting to treat strings with xml:lang specifiers as a new 
>>datatype (call it "language-tagged string"), disjoint with xsd:string. 
>>Similarly, XML infosets should simply be yet another datatype, disjoint 
>>with any other XSD datatype.
>>
>>These two datatypes were essentially defined as such in the original RDF 
>>spec. Now that we have a general-purpose datatyping mechanism, we can 
>>make use of it. The two datatypes should get their own URIs.
>>
>>If there is enough support for that, I'd like to put the above point for 
>>vote at the next telecon.
>>
> 
> Er... as far as making group decisions, we're supposedly done.
> We've disposed of all our issues.
> I'd really rather not put new issues in the ciritcal path.
> 
> Maybe you can just work with the editors to make this a
> natural consequence of the decisions we made on datatyping;
> it seems reasonably consistent with what we discussed.
> 
> i.e. in stead of suggesting this for the WG agenda, I suggest
> you just take it straight to the editors.


Fair enough. I interpret it as a voice of support, and I hope that the 
editors are taking notes...

 
>>The current proposal
>>
> 
> which proposal is that, exactly? I gather the status quo
> (i.e. what exactly the WG has decided)
> is rather sparsely documented (e.g. the whole discussion
> of tidy/untidy options was reduced to pretty much one bit
> in today's telcon record), and the editors have actions
> to figure out all the details.


I'm referring to the pair vs. value decision taken at that telecon. As 
far as I understand the issue, "pair" means that typed values are 
represented as pairs in the abstract syntax, e.g. (URI, string). "Value" 
means that typed values are opaque constants in the graph (applications 
are free to represent them however they like). The group voted for "pairs".

My observation is that there is no room in a (URI, string) pair to keep 
e.g., an xml:lang value, if the language-tagged strings is a separate 
datatype. My suggestion to the editors is to make the pairs more 
flexible, e.g., (URI, something).

Sergey


> 
> [[
> ACTION: 2002-09-13#2 Graham to update concepts doc to cover 1.1, 1.2,
> 2.2, 
> 2.3 to reflect datatype concensus
> 
> ACTION: 2002-09-13#3 Danbri to update schema to cover 2.1, 3.1, 3.2, 5
> to 
> reflect datatype concensus
> 
> ACTION: 2002-09-13#4 PatH to update model theory to cover 4 to reflect 
> datatype concensus
> 
> ACTION: 2002-09-13#5 FrankM to update primer to reflect datatype
> concensus
> 
> ACTION: 2002-09-13#6 Jan to create testcase to reflect datatype
> concensus
> ]]
>  -- http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Sep/0178.html
> 
> 
> 
> 
>>for representing typed values in the abstract 
>>syntax (URI + string) fails for the above datatypes. Therefore, I'm also 
>>suggesting that this overspecification is not required. In the abstract 
>>syntax, typed literals may be kept as opaque constants, whereas the 
>>applications may use their internal representation of choice.
>>
>>Sergey
>>

Received on Monday, 23 September 2002 11:32:16 UTC