W3C home > Mailing lists > Public > public-rdf-wg@w3.org > May 2011

Re: Proposal for ISSUE-12, string literals

From: Richard Cyganiak <richard@cyganiak.de>
Date: Fri, 13 May 2011 18:22:03 +0100
Cc: public-rdf-wg@w3.org
Message-Id: <73D5B9C5-1616-4A16-A7BF-D35013659471@cyganiak.de>
To: antoine.zimmermann@insa-lyon.fr
Antoine,

On 13 May 2011, at 17:57, Antoine Zimmermann wrote:
>>>> 2) of the built-in datatypes, only xsd:string can have non-empty language tags;
>> 
>> What is silly about that?
> 
> Well, I don't see what you improve by saying that all tags are empty except for language-tagged strings, as opposed to no literals have language tags except strings. In both cases, you have to bother with an exception, while the goal was to simplify things by having a more uniform description of literals.

The current situation is that we have (ignoring rdf:PlainLiteral) two kinds of literals:

  "typed"^^ex:type
  "plain" with optional @lt

Afterwards, we'd have one:

  "typed"^^ex:type with optional @lt

with the exception that the language tag can be non-empty only for xsd:string.

That is a simplification compared to the current situation.

The alternative proposal from Alex was to have this:

  "typed"^^ex:type
  "plain" with compulsory @lt

That's also a simplification, but it's weird that there's this major difference in model, syntax and behaviour between a string on the one hand, and a string with a language tag on the other hand.

> So, maybe it's not silly but it's unnecessary. Moreover, one can always decide to implement literals like this, even with the current spec.

No, one cannot, because that leaves no way of distinguishing between "foo" and "foo"^^xsd:string, which are supposed to produce two different triples.

Best,
Richard
Received on Friday, 13 May 2011 17:22:32 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 16:25:42 GMT