W3C home > Mailing lists > Public > www-rdf-interest@w3.org > May 2004

Re: Incompatibilities in RDQL implementations

From: Dan Brickley <danbri@w3.org>
Date: Fri, 07 May 2004 12:08:17 +0100
Message-ID: <409B6E21.30601@w3.org>
To: Libby Miller <Libby.Miller@bristol.ac.uk>
Cc: Jeremy Carroll <jjc@hplb.hpl.hp.com>, Jeen Broekstra <jeen@aduna.biz>, www-rdf-interest@w3.org, Arjohn Kampman <arjohn.kampman@aduna.biz>

Libby Miller wrote:

> 
> 
> On Thu, 6 May 2004, Jeremy Carroll wrote:
> 
> 
>>
>>>Steve Harris wrote:
>>>
>>>
>>>>On Thu, May 06, 2004 at 10:16:35 +0100, Phil Dawes wrote:
>>>>
>>>>
>>>>>SELECT ?family , ?given
>>>>>WHERE  (?vcard  vcard:FN "John Smith")
>>>>>      (?vcard  vcard:N  ?name)
>>>>>      (?name   vcard:Family  ?family)
>>>>>      (?name   vcard:Given  ?given)
>>>>>USING  vcard FOR <http://www.w3.org/2001/vcard-rdf/3.0#>
>>>>>
>>>>>
>>>>>Unfortunately both Sesame(1.0) and RAP(0.71) require '<' and '>'
>>>>>around qnames for the query to parse.
>>>>
>>>>
>>>>FWIW 3store requires both <>'s and commas, the commas thing is a bug and
>>>>will be fixed. I'm be reluctant to remove or stop requireing
>>>
>>>the <>'s as I
>>>
>>>>think it makes the queries more readable.
>>>
>>>ObAOL: Ditto for Sesame.
>>
>>What about the potential confusion between <vcard:Family> qname and
>><vcard:Family> new URI scheme?
>>
>>While I don't use RDQL much, I note that N3 has the same convention: <>
>>indicate URI, no-<> indicates qname.
>>
>>
> 
> 
> 
> We had several discussions about this during the design of Squish and
> related discussions with Andy about RDQL. Dan Brickley and I argued that
> 
> * the number of cases where the qname and uri schemes were likely to
> clash was small and could be avoided with sensible queries, and

Sometimes machines write queries; they're famous for being senseless. 
Sometimes sensible people don't keep up to date with all the new URI 
schemes out there, nor whether proposed new schemes are properly 
registered or just deployed anyway. We really shouldn't use heuristics 
here, it leaves a nasty hole for confusing errors, makes the language 
harder to explain/document and harder to use. Jan Grant or maybe Andy I 
think proposed having a precedence order, eg. first try as namespace 
then fall back to URI if no namespace with that prefix. That's better 
than leaving it open. I much prefered using '::' (see my Squish fork :) 
but that never really caught on.

Dan


> * potentially, being able to embed RDF queries in XML meant that it
> was probably best to avoid < >
> 
> Libby
Received on Friday, 7 May 2004 07:09:34 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:52:07 GMT