W3C home > Mailing lists > Public > public-lod@w3.org > April 2011

Re: Quick reality check please

From: Christopher Gutteridge <cjg@ecs.soton.ac.uk>
Date: Sun, 03 Apr 2011 12:12:56 +0100
Message-ID: <EMEW3|5900c2fa499c819b58da810117d80eacn32CEJ03cjg|ecs.soton.ac.uk|4D985638.7000904@ecs.soton.ac.uk>
To: Hugh Glaser <hg@ecs.soton.ac.uk>
CC: Richard Cyganiak <richard@cyganiak.de>, Mischa Tuffield <mmt04r@ecs.soton.ac.uk>, "<nathan@webr3.org>" <nathan@webr3.org>, Linked Data community <public-lod@w3.org>
It really depends if you put the work on the producer or consumer. 
Perhaps there needs to be a short document on best practice for ^^ and @?

- Only use @en (etc) if your dataset contains data in multiple languages.
- If you have and rdfs:label "some label"@en please also provide a 
default "some label"
- If in doubt, don't use data types for anything other than "date", 
"datetime", "float" and "int". With the exception of skos:notation which 
explicitly tells you to use a datatype for the notation being used.
- The benefit of setting a datatype is to allow the value to be queried. 
ie. it changes what < > and ORDER BY does with that field. That's all.

What benefit do I get using "foo"^^xsd:string vs "foo"? I get a pain in 
the neck from having a mixture -- maybe a quick regexp on import?
s/^^<http://www.w3.org/2001/XMLSchema#string>\s*\.\s*$/ ./;
... or can someone tell me why that's a bad idea?

Hugh Glaser wrote:
> On 22 Mar 2011, at 12:37, Richard Cyganiak wrote:
>> On 21 Mar 2011, at 13:05, Hugh Glaser wrote:
>>> So I guess I need to do four patterns just to find all the exact "World Wide Web Consortium" English phrases (with and without @en and with and without datatype string).
>>> Is that really right?
>> Three -- you can't have both a datatype and a language tag on a literal.
> Well that's good to know.
> Mind you, since it is not a syntactic constraint (I think), that doesn't mean we couldn't find it, I suppose.
>> This suggests two things:
>> 1. xsd:string in RDF must die. It's one of those completely and utterly useless pieces of rubbish that litter the RDF specs.
> Perhaps you could tell us what you really think :-)
>> 2. If you publish in multiple languages, then perhaps it's a good idea to include a plain literal in a “default language” without a language tag, to make SPARQLing easy.
>> If publishers did that, we'd be back to one pattern.
>> Best,
>> Richard
> So I would guess from this that it could be that some documents could be adjusted to recommend this sort of thing.
> Certainly for 2; is it the case for 1 that technically there should be a type?
> You are a good editor - can we do a little something?
> Best
> Hugh

Christopher Gutteridge -- http://id.ecs.soton.ac.uk/person/1248

You should read the ECS Web Team blog: http://blogs.ecs.soton.ac.uk/webteam/
Received on Sunday, 3 April 2011 11:14:53 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:16:13 UTC