Re: Action-48 text: a New Plan for plain literals

On 24/05/11 13:10, Steve Harris wrote:
...
> I agree with Andy, this is not going to play well with existing
> systems, unless rdf:LanguageTaggedString exists purely in the
> abstract syntax, you can still end up with literals with both a
> datatype, and a language tag.
>
> - Steve

I still in two minds over the proposal.  If we were starting where 
RDF-2004 started, it's a better proposal, but we're not starting from 
there.  The test for for me is whether it's sufficiently better.

It's better for the modelling possibilities.  Pat said you can now 
restrict queries to them but I don't think that's a new feature (you can 
do that anyway in SPARQL).  Old code may behave differently to new code.

The latter is a cost, not a problem although people running platform 
services to customer applications (not web pages) may disagree as 
when/if they make the change the applications may be affected.

There seem to be three classes of proposals now:

1/ datatypes: rdf:LanguageTaggedString being the current proposal in the 
area

2/ Convert xsd:string to simple literals (parsing etc). Deprecate 
xsd:string in the ^^xsd;string usage.

3/ Do nothing.

	Andy

Received on Tuesday, 24 May 2011 13:15:56 UTC