RDF Data Access

Toward CR

<kendallclark> Hmm, I didn't realize that procedural point, Dan.

<ericP> Subject: [OK?] Re: minor technical on 11.2.1 Invocation

<ericP> Souri, can you respond to http://www.w3.org/mid/20060125181430.GC412@w3.org ?

<DanC> ... discussion of comments from individuals who are affiliated with organizations that are WG members

DanC: We would like closure responses from Oracle (Fred Z, Souri) on our responses to Fred's LC comments

<kendallclark> LeeF: I meant "illegal according to W3C policies and procedures", not juridically illegal.

<kendallclark> is that cwm w/ Pychinko dan? :>

<DanC> . ACTION PatH: draft response to @@

<patH> http://lists.w3.org/Archives/Public/public-rdf-dawg-comments/2006Jan/0065.html

<DanC> ACTION: PatH to draft response to http://lists.w3.org/Archives/Public/public-rdf-dawg-comments/2006Jan/0065.html and send to the WG list [recorded in http://www.w3.org/2006/01/26-dawg-minutes.html#action01]

Convene

<DanC> <ericP> DanC, meeting record 24 Jan 2006 sent --> http://www.w3.org/2006/01/24-dawg-minutes

<DanC> the http://www.w3.org/2006/01/24-dawg-minutes minutes are missing an action on me to publish proto-wd, but otherwise OK

<DanC> PROPOSED: to approve http://www.w3.org/2006/01/24-dawg-minutes ammended to show an action on DanC to publish proto-wd

<DanC> so RESOLVED

<DanC> # PROPOSED: to meet again Tue, 31 Jan 14:30Z;

<SteveH> sorry, cant

<EnricoFranconi> regrets for next tuesday meeting

Next meeting -- Tue, 31 Jan 14:30Z, scribe SerT

<SerT> fine

<AndyS> We agreed to meet twice-weekly in Jan - are we continuing to Jan 34?

<SerT> I'll scribe

Toward updated protocol WD (and results format WD)

<DanC> this one? > http://www.w3.org/2001/sw/DataAccess/sparql-xml-syntax/

DanC: Can we satisfy the link checker by using example.org for this namespace?

<DanC> this is in SOTD? http://www.w3.org/TR/2006/WD-rdf-sparql-protocol-20060125/

<DanC> no...

kendallclark: there's some other broken protocol link in the LC status page

<DanC> WG is OK with 10 Feb due date

<scribe> ACTION: DanC to noodle on status, publish [CONTINUES] [recorded in http://www.w3.org/2006/01/26-dawg-minutes.html#action02]

issues rdfSemantics, owlDisjunction

AndyS: outstanding rdfSemantics issue is if SPARQL defines G' = G

http://lists.w3.org/Archives/Public/public-rdf-dawg/2006JanMar/0293.html

<EnricoFranconi> FUB is neutral whether G=G' is in the spec ; we just believe that it is more consistent with the SPARQL philosophy if it is in the spec

<DanC> Pat on etc.

<DanC> Thu, 26 Jan 2006 10:06:13 -0600

<DanC> Pat's msg of Thu, 26 Jan 2006 09:55:06 -0600

<kendallclark> lots of x-talk

patH: the scope of answer bnodes has to be defined by G'

<DanC> (Pat, you already lost Andy)

<DanC> Andy asks to focus on http://www.w3.org/2001/sw/DataAccess/rq23/ [I wonder which section]

patH: in *our* SPARQL B is defined to be the identifiers in G'

rq23 states "The scoping set is the set of all IRIs and blank nodes in G'."

<AndyS> http://www.w3.org/2001/sw/DataAccess/rq23/#BGPsparql

patH: B is just a parameter in the general definition; for SPARQL we constrain B to be precisely the identifiers (terms?) in G'

EnricoFranconi: agree

patH: AndyS is asking whether G' has to be identical to G -- that is too strong -- should be graph equivalent

EnricoFranconi: if G is not identical to G' we end up with a SPARQL that is not purely syntactical

patH: I think that was always the intention.

DanC: can we observe this with a test?

patH: I think the tests assume that bnodes in the answers are independent of bnodes in the graph

<DanC> Andy reads "The scoping set is the set of all IRIs and blank nodes in G'." in 2.5.2.

AndyS: looking at 2.5.2 ( http://www.w3.org/2001/sw/DataAccess/rq23/#BGPsparql ) -- 2nd paragraph says "The scoping set is the set of all IRIs and blank nodes in G'." -- is that right?

patH: should also include literals

AndyS: for consistency, in the def'n of BGP uses the word "identifiers" -- should that be RDF terms?

patH: yes

<DanC> ("RDF term" includes bnode, yes?)

EnricoFranconi: yes

<DanC> (is it feasible to give andy 3 minutes to do s/identifier/RDF term/ now?)

<DanC> "up to bnode renaming" makes sense to me

EnricoFranconi: it is enough to change paragraph 3 of 2.5.2 by adding "up to bnode renaming" at the end of the last 3 sentences
... semantic web colleagues have trouble understanding wording of definitions -- both for precision but mostly for understanding

<DanC> PatH: s/the set of all the pattern solutions is unique./the set of all the pattern solutions is unique up to bnode renaming./

EnricoFranconi: discussion of precise wording ...

<DanC> I see Revision: 1.619 of Date: 2006/01/26 16:41:24

<AndyS> http://www.w3.org/2001/sw/DataAccess/rq23/#BGPsparql

<DanC> type that, pat?

<DanC> maybe andy got it

EnricoFranconi: more wording discussions of where to put "up to bnode renaming" and the like ...

<DanC> 1.620...

<SerT> I'm not sure a pattern solution is unique, it's the answer set which is unique

<patH> text says 'set of all'

<DanC> I read "set of all the pattern solutions is unique up to blank node renaming."

<DanC> Ser/PatH: s/the uniqueness property of pattern solutions/a uniqueness property/

patH: caterpillars, mushrooms, and the like.

<DanC> 1.621

<DanC> Enrico: why do we repeat defn graph equivalent?

<DanC> AndyS: because rdf abstract syntax doesn't have variables

<DanC> ... discussion of where some defns should go...

AndyS: moving 3rd paragraph from 2.5.2 to before 2.5.1 ...

EnricoFranconi: might need to be in a box (definition)

<DanC> yup, 1.623 $ of $Date: 2006/01/26 16:52:30

patH: 1.623 looks good

<DanC> PROPOSED: that SPARQL QL editor's draft 1.623 addresses issue rdfSemantics and is sufficient to postpone issue owlDisjunction.

patH: maybe not.
... whence BGP'

EnricoFranconi: that was from mymessages that allowed getting rid of the OrderedMerge

DanC: is the definition correct as stated, modulo elegance

patH: this is to handle told bnodes, right?

EnricoFranconi: exactly

<SerT> BGP' allows the server to change bnodes in the queries

AndyS: I've been concentrating on SPARQL v1 and am surprised that so much indirection is necessary -- if it's needed for extension points then that's fine but should have text explaining that

<EnricoFranconi> Pat, please read emails !

<EnricoFranconi> http://www.w3.org/mid/992AE607-17D4-44AF-BF5C-31E3E56C857F@inf.unibz.it

<DanC> PROPOSED: that SPARQL QL editor's draft 1.623 addresses issue rdfSemantics and is sufficient to postpone issue owlDisjunction.

<DanC> PROPOSED: that SPARQL QL editor's draft 1.623 section 2.5 addresses issue rdfSemantics and is sufficient to postpone issue owlDisjunction.

<EnricoFranconi> +1

<SerT> +1

<DanC> RESOLVED, Hayes objecting

<EnricoFranconi> Pat, we can convince you offline!

<DanC> ACTION: AndyS to propose a publication plan [recorded in http://www.w3.org/2006/01/26-dawg-minutes.html#action03]

<patH> You can try, but I doubt it. Ive alrady given way on a lot I don't like in the name of consensus, but this is too far.

test suite maintenance

<scribe> ACTION: EricP to fix test schema to match manifest with negative tests [recorded in 09/27-dawg-minutes.html#action16] [CONTINUES] [recorded in http://www.w3.org/2006/01/26-dawg-minutes.html#action04]

<scribe> ACTION: DanC to follow up re optional test based on op:dateTime triple [CONTINUES] [recorded in http://www.w3.org/2006/01/26-dawg-minutes.html#action05]

<scribe> ACTION: DaveB to add to test suite the temperature case from comment on truth tables in commentor's message [CONTINUES] [recorded in http://www.w3.org/2006/01/26-dawg-minutes.html#action06]

ADJOURNED.