RDB2RDF Working Group Teleconference -- 15 Mar 2011

<trackbot> Date: 15 March 2011

<ivan> member:Zakim, this will be 7322733

<betehess> ivan, take 2 seconds to look at https://dvcs.w3.org/hg/FeDeRate/file/66e9d9e8a090/rdf2rdf/src/test/scala/RDF2RDFTest.scala#l16

<betehess> you'll love it :-)

<betehess> I fixed your example btw ;-)

<ivan> does this do the transformation? Ie, is it some sort of a CONSTRUCT based transformation engine?

<betehess> it is

<betehess> look at the assertion at the end

<betehess> assert(expectedGraph === resultGraph)

<ivan> yes, I have seen it

<ivan> hence my question

<ivan> :-)

<betehess> it's actually doing the computation

<betehess> your example is just alive now ;-)

<ivan> :-)

<ivan> so what did you fix?

<betehess> some predicates where missing

<betehess> in the body of the WHERE

<mhausenblas> trackbot, start telecon

<trackbot> Meeting: RDB2RDF Working Group Teleconference

<trackbot> Date: 15 March 2011

<mhausenblas> scribenick: cygri

Admin

PROPOSAL: Accept the minutes of last meeting, see

<mhausenblas> PROPOSAL: Accept the minutes of last meeting, see http://www.w3.org/2011/03/08-rdb2rdf-minutes.html

http://www.w3.org/2011/03/08-rdb2rdf-minutes.html

<betehess> +1

RESOLUTION: Accept the minutes of last meeting, see http://www.w3.org/2011/03/08-rdb2rdf-minutes.html

Publication of the WD

mhausenblas: idea is to go through issues to see if they are addressed

<mhausenblas> ISSUE-27?

<trackbot> ISSUE-27 -- NULL graphColumn value producing triples in an unnamed graph -- open

<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/issues/27

souri: we addressed this
... document still uses "unnamed graph" in some places
... i'd prefer to keep that term

seema: i don't think we updated the document

ericP: SPARQL says, i'm querying a bunch of stores and one is the default graph
... we can define this directly like in SPARQL, or with a level of indirection

mhausenblas: but needs to be clarified

cygri: we should target RDF Dataset as in SPARQL

souri: but default graph in SPARQL can be constructed e.g. as union of several graphs with FROM <a> <b> <c>

<mhausenblas> Michael: seems this needs to be addressed in the text, propose to action Richard with it

souri: so the default graph can be synthesized or changed via protocol. you can change at query time

<Seema> we did add a special IRI rr:unnamedGraph to exlicitly refer to the unnamed graph

<ericP> test case: R2RML which asserts { <s> <p> <o> } into the "unnamed" graph and some other source <foo> { <s1> <p1> <o1> } : Query: FROM <foo> SELECT { ?s ?p ?o }

<ericP> does that give one triple or two?

<Souri> default graph in SPARQL => synthesized at query processing by UNIONing the graphs named in the FROM clauses

<Souri> default graph in SPARQL is different from unnamed graph in an RDF store

<Souri> I am worried about confusing people with these two distinct concepts being named by the same term

<ericP> PROPOSE that "unnamed graph" be called the "default graph" with the intention that the output of the R2RML mapping is the same RDF database described by SPARQL

souri: propose to mark it as open issue in the spec

<Zakim> ericP, you wanted to see if discussing the above test case will approach consensus

<Souri> I am ok to discuss it after the call

ericP: i have a test case; souri does this appeal to you?

(discussion of ericP's test case)

<privera> Zakim mute me

souri: maybe we can get some time to discuss this (after call)

ericP: can you put an action on souri and me to resolve this after the call

<mhausenblas> ACTION: Eric and Souri to resolve ISSUE-27 (note or text proposal) [recorded in http://www.w3.org/2011/03/15-rdb2rdf-minutes.html#action01]

<trackbot> Created ACTION-107 - And Souri to resolve ISSUE-27 (note or text proposal) [on Eric Prud'hommeaux - due 2011-03-22].

<mhausenblas> ISSUE-24?

<trackbot> ISSUE-24 -- Clarify support for complex joinConditions -- open

<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/issues/24

souri: this has been addressed in section 3.9.1

<Souri> http://www.w3.org/2001/sw/rdb2rdf/r2rml/#RefObjectMapClass_joinCondition_Property

souri: this addresses ISSUE-24 and ISSUE-32

<mhausenblas> close ISSUE-24

<trackbot> ISSUE-24 Clarify support for complex joinConditions closed

<mhausenblas> ISSUE-22?

<trackbot> ISSUE-22 -- Support for database vendor specific SQL statements. -- open

<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/issues/22

seema: we added a note about this

<mhausenblas> http://www.w3.org/2001/sw/rdb2rdf/r2rml/#sql-conformance

souri: we extended the note that was already present in 1.3

<mhausenblas> close ISSUE-22

<trackbot> ISSUE-22 Support for database vendor specific SQL statements. closed

<mhausenblas> ISSUE-29?

<trackbot> ISSUE-29 -- Require blank node and IRI identifier expressions to produce strings -- open

<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/issues/29

<mhausenblas> ACTION-105?

<trackbot> ACTION-105 -- Eric Prud'hommeaux to provide text for ISSUE-29 -- due 2011-03-15 -- OPEN

<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/actions/105

ericP: can anybody hear me?

mhausenblas: yes we can. nice try!

ericP: this was about how SQL does datatype conversions/casting
... ACTION-105 continues

<Zakim> back, you wanted to the join condition

ivan: re ISSUE-22 changed text, are the curlies in the right place?

seema: we missed it in the last example

souri: yes this has to be changed

ivan: i don't understand this one. why do we need the curlies around the pseudo table names?

souri: so we can avoid parsing the SQL

ivan: i guess that's where i'll have to believe you

souri: (explains example)
... we tried to make the name of the pseudo tables very esoteric to avoid clashes

ivan: what would happen if i bound the rr namespace URI to bla: instead of rr: ?

souri: it would not change anything here. maybe we can make this simpler
... maybe we don't need the 'rr:' part in there
... so we'll take that out

<Souri> Issue-24, Issue-32

ISSUE-32?

<trackbot> ISSUE-32 -- Remove the use of curly braces in joinCondition -- open

<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/issues/32

<Souri> then they must escape the { and } with \

cygri: (wants either limited expressivity in the join, or allow full sql without curlies)

mhausenblas: so how to resolve?

souri: well limited expressivity might also be ok
... i'm not sure which one is better
... yes a bit hacky, you have to escape curlies
... but there are similar examples in other languages

<mhausenblas> PROPOSAL: postpone ISSUE-32, put a note in the R2RML text explaining the options (limited expressivity, current hack, full SQL)

RESOLUTION: postpone ISSUE-32, put a note in the R2RML text explaining the options (limited expressivity, current hack, full SQL)

<mhausenblas> ACTION: Richard to add note re ISSUE-32 to R2RML [recorded in http://www.w3.org/2011/03/15-rdb2rdf-minutes.html#action02]

<trackbot> Created ACTION-108 - Add note re ISSUE-32 to R2RML [on Richard Cyganiak - due 2011-03-22].

<Souri> a PL/SQL string with SQL query: select empno || 'Souri''s document' from emp -- notice the double single-quote

<mhausenblas> ISSUE-30?

<trackbot> ISSUE-30 -- Different blank node identifier than was defined in mapping -- open

<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/issues/30

ISSUE-30

<mhausenblas> ACTION: Richard to address ISSUE-30 [recorded in http://www.w3.org/2011/03/15-rdb2rdf-minutes.html#action03]

<trackbot> Created ACTION-109 - Address ISSUE-30 [on Richard Cyganiak - due 2011-03-22].

<mhausenblas> ISSUE-31?

<trackbot> ISSUE-31 -- Clarify reference to "blank node label" -- open

<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/issues/31

cygri: i said i'd modify the text, didn't do yet, can i have an issue?

ISSUE-32

souri: i clarified the text to consistently call it "blank node identifier"
... only required a change in section 5.1

<mhausenblas> http://www.w3.org/2001/sw/rdb2rdf/r2rml/#Blank_Nodes

<Souri> +1

s/Topic: ISSUE-32/Topic/ISSUE-31/

<ericP> tacit approval!

<mhausenblas> close ISSUE-31

<trackbot> ISSUE-31 Clarify reference to "blank node label" closed

ISSUE-7

<mhausenblas> ISSUE-7

<mhausenblas> ISSUE-7?

<trackbot> ISSUE-7 -- Linked Data aspects for mapping RDB2RDF -- open

<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/issues/7

<mhausenblas> ACTION-106?

<trackbot> ACTION-106 -- Sören Auer to review R2RML draft and tell editors where notes should be made -- due 2011-03-15 -- OPEN

<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/actions/106

mhausenblas: we lost soeren

<mhausenblas> soeren?

<mhausenblas> Michael: see discussion http://lists.w3.org/Archives/Public/public-rdb2rdf-wg/2011Mar/0044.html

<mhausenblas> close ACTION-106

<trackbot> ACTION-106 Review R2RML draft and tell editors where notes should be made closed

ISSUE-28?

<trackbot> ISSUE-28 -- Clarify whether blank nodes are equivalent based on their identifier -- open

<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/issues/28

<mhausenblas> http://lists.w3.org/Archives/Public/public-rdb2rdf-wg/2011Mar/0048.html

ISSUE-28

mhausenblas: there's a new test case

souri: there was discussion on how to generate class URIs ... we'll handle without ISSUE/ACTION

Direct mapping

<mhausenblas> ISSUE-10?

<trackbot> ISSUE-10 -- Hash vs Slash -- open

<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/issues/10

ISSUE-10

<mhausenblas> ACTION-104?

<trackbot> ACTION-104 -- Juan Sequeda to implement decision re ISSUE-10 (remove hash) and close ISSUE-10 -- due 2011-03-15 -- OPEN

<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/actions/104

ivan: last time i checked this wasn't done yet in the spec

<mhausenblas> ISSUE-9?

<trackbot> ISSUE-9 -- Generate Blank Nodes for duplicate tuples -- open

<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/issues/9

ISSUE-9

<Souri> Richard's email regarding rr:class: http://lists.w3.org/Archives/Public/public-rdb2rdf-wg/2011Mar/0063.html

mhausenblas: we also asked the RDF WG about this
... so i'd prefer to keep that open

<mhausenblas> PROPOSAL: postpone ISSUE-9 and take feedback from RDF WG into account

<ivan> +1

RESOLUTION: postpone ISSUE-9 and take feedback from RDF WG into account

mhausenblas: all editors, could the drafts be ready for review on monday morning?

ericP: should we line up reviewers?

<Souri> I have updated, based on Ivan's comment, the rr:joinCondition section and corr examples: new version 1.45

<ivan> Thanks SOuri

<mhausenblas> close ACTION-104

<trackbot> ACTION-104 Implement decision re ISSUE-10 (remove hash) and close ISSUE-10 closed

(discussing how to get the remaining actions done and till when)

<mhausenblas> ACTION: Eric to implement decision re ISSUE-10 (remove hash) and close ISSUE-10 [recorded in http://www.w3.org/2011/03/15-rdb2rdf-minutes.html#action04]

<trackbot> Created ACTION-110 - Implement decision re ISSUE-10 (remove hash) and close ISSUE-10 [on Eric Prud'hommeaux - due 2011-03-22].

ericP: (volunteers to take over ACTON-104)

<mhausenblas> PROPOSAL: Both R2RML and DM address all open action till Fr 18 March and report on the list

<ivan> +1

RESOLUTION: Both R2RML and DM address all open action till Fr 18 March and report on the list

<betehess> +1

ericP: is it acceptable to refer to the draft version of SQL?

mhausenblas: as fallback, yes
... adjourned

<mhausenblas> [meeting adjourned]

<Souri> FROM <g1> FROM <g2> ... FROM NAMED <ng1> FROM NAMED <ng2> ...

<Souri> default graph = UNION(<g1>,<g2>)

<MacTed> boo, calendar fail... :-(

<mhausenblas> trackbot, end telecon

- DRAFT -

RDB2RDF Working Group Teleconference

15 Mar 2011

Attendees

Contents