W3C home > Mailing lists > Public > public-rdf-dawg@w3.org > January to March 2011

Re: Draft answer KK-7

From: Axel Polleres <axel.polleres@deri.org>
Date: Tue, 1 Mar 2011 21:15:56 +0000
Cc: "Lee Feigenbaum" <lee@thefigtrees.net>, "SPARQL Working Group" <public-rdf-dawg@w3.org>
Message-Id: <FC6725FF-5194-45F1-975F-9C2058F269C9@deri.org>
To: "Andy Seaborne" <andy.seaborne@epimorphics.com>
Good for me, draft response changed accordingly.

If no more suggestions, I will send the response by tomorrow evening or so.
(Will be traveling then for a week, so regrets for next week.)

Axel

On 1 Mar 2011, at 20:57, Andy Seaborne wrote:

> 
> 
> On 01/03/11 10:49, Axel Polleres wrote:
> > draft answer is here:
> >
> > http://www.w3.org/2009/sparql/wiki/CommentResponse:KK-7
> 
> It starts:
> 
> """
> reuse of bnodes across multiple BGPs in a query may lead to unintuitive
> results:
> """
> but Kjetil isn't suggesting it's the same bNode - only that the same
> label can be reused (to mean a different bnode).
> 
> Maybe the response could just say that, on balance, the unique label per
> query is felt by the WG to be the clearer approach.
> 
> To mutate and add to Lee's words:
> 
> """
> There is a balance to be struck between potential confusion due to
> reusing labels to identify different things and the convenience in
> composition of queries.
> 
> On balance, the WG believe that the approach of SPARQL 1.0, which avoids
> the confusion possibilities, is the better choice.
> """
> 
>         Andy
> 
> 
> >
> > please let me know if that works for you.
> >
> > Axel
> >
> > On 1 Mar 2011, at 01:59, Lee Feigenbaum wrote:
> >
> >> On 2/28/2011 8:54 PM, Axel Polleres wrote:
> >>>
> >>> On 1 Mar 2011, at 01:46, Lee Feigenbaum wrote:
> >>>
> >>>> My personal feeling is that it would be _very_ confusing to allow the
> >>>> same bnode label in two BGPs but have it refer to distinct blank nodes.
> >>>> You'd have a situation where sometimes (within the same BGP) two
> >>>> mentions of _:a would be the same and other times (in two BGPs, perhaps
> >>>> separated by BIND or something like that)  they wouldn't.
> >>>>
> >>>> Please let me know if anyone feels otherwise. If there appears to be
> >>>> silence / consensus, then I will draft a response to Kjetil.
> >>>
> >>> That was my feeling as well, I just thought that this motivation was probably discussed in DAWG1 already s.t.
> >>> we can refer to it in the answer.
> >>
> >> I don't remember specifically discussing the option of allowing the same
> >> label in 2 BGPs but having it refer to different blank nodes.
> >>
> >> Lee
> >>
> >>>
> >>> Axel
> >>>
> >>>>
> >>>> Lee
> >>>>
> >>>> On 2/28/2011 8:15 PM, Axel Polleres wrote:
> >>>>> Hi all,
> >>>>>
> >>>>> in order to answer comment KK-7
> >>>>>     http://lists.w3.org/Archives/Public/public-rdf-dawg-comments/2011Jan/0009.html
> >>>>>
> >>>>> I am pretty sure that this has been discussed in depth and there is some DAWG1-discussion
> >>>>> about this issue somewhere back in the archives... If anybody from our DAWG1 members
> >>>>> feels like pointing me to it, I'd be grateful!
> >>>>>
> >>>>> Axel
> >>>>>
> >>>>
> >>>
> >>>
> >>
> >
> >
> 
Received on Tuesday, 1 March 2011 21:16:31 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 16:15:45 GMT