See also: IRC log
<trackbot> Date: 02 August 2011
<scribe> scribenick: mhausenblas
<boris> hi
PROPOSAL: Accept the minutes of last meeting http://www.w3.org/2011/07/26-rdb2rdf-minutes.html
RESOLUTION: Accept the minutes of last meeting http://www.w3.org/2011/07/26-rdb2rdf-minutes.html
ACTION-141?
<trackbot> ACTION-141 -- Nuno Lopes to create a SQL/XSD data dataype mapping in the Wiki until next week -- due 2011-07-26 -- CLOSED
<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/actions/141
close ACTION-141
<trackbot> ACTION-141 Create a SQL/XSD data dataype mapping in the Wiki until next week closed
-> http://www.w3.org/2001/sw/rdb2rdf/wiki/Mapping_SQL_datatypes_to_XML_Schema_datatypes
Michael: thanks, Nuno, nice job
Nuno: Based on ISO/IEC
9075-14:2008 Part 14 XML-Related Specifications
... with the restrictions/facets
... question is how deep we want to go with this
... for example excluding the user-defined types
... or constructed types
... unsure if we should consider this
Michael: comments?
<cygri> http://lists.w3.org/Archives/Public/public-rdb2rdf-wg/2011Jul/0163.html
Richard: Really useful,
thanks
... above are my questions
... I understand you need some time to address my
questions
... there are situations where we want to map to plain
strings
... via casting
... toString (SQL-built-in)
... and we can use this as a fallback
Michael: does this always make sense?
Richard: impl. dependent
Michael: we need to keep this up
Richard: any volunteers?
<cygri> ACTION: nuno to look into how SQL 2008 handles CAST of user-defined types and constructed types to string [recorded in http://www.w3.org/2011/08/02-rdb2rdf-minutes.html#action01]
<trackbot> Created ACTION-149 - Look into how SQL 2008 handles CAST of user-defined types and constructed types to string [on Nuno Lopes - due 2011-08-09].
Eric: so, we look into this to figure if we have an RDF mapping?
Richard: we need to figure a way how to deal with it - simplest would be to relay on the SQL 2008 spec
<ericP> "Data conversions between a user-defined
<ericP> type and another data type are defined by a user-defined cast."
<dmcneil> +q
David: wondering where it is
heading
... we started out rather simple?
Richard: if there is a well defined translation to XSD (such as xsd:integer) use that
<ericP> i don't think there are any well-defined conversions from e.g. COMPLEX_NUMBER to CHAR(n)
Richard: if there is none, but the implementation knows how to do the toString use that
<ericP> that when you introduce e.g. COMPLEX_NUMBER, htat you ought to introduce the appropriate casts
<cygri> ericP, i think we just need cast to string
ACTION-124?
<trackbot> ACTION-124 -- Seema Sundara to implement decision re ISSUE-29 (bNodes identifier and URI expressions be of string types) -- due 2011-08-02 -- OPEN
<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/actions/124
ACTION-147?
<trackbot> ACTION-147 -- Richard Cyganiak to implement ISSUE-29 resolution by stating that conversion to string is done implicitly in any context where a string value is required, and is done according to the rules for SQL's CAST expression. Columns whose type cannot be CAST to string MUST NOT be used in a context that requires a string; and mark the issue as pending review -- due 2011-08-02 -- OPEN
<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/actions/147
<Seema> yes, it does
close ACTION-124
<trackbot> ACTION-124 Implement decision re ISSUE-29 (bNodes identifier and URI expressions be of string types) closed
<ericP> cygri, that makes sense, and if the UDT which provides a new type doesn't give you a cast, then you were going to lose anywyas
ACTION-137?
<trackbot> ACTION-137 -- Souripriya Das to create proposal for IRIs to resolve ISSUE-45 -- due 2011-07-05 -- PENDINGREVIEW
<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/actions/137
close ACTION-137
<trackbot> ACTION-137 Create proposal for IRIs to resolve ISSUE-45 closed
close ISSUE-45
<trackbot> ISSUE-45 IRIs instead of literals for rr:termType choices closed
ACTION-139?
<trackbot> ACTION-139 -- Boris Villazón-Terrazas to move the R2RML extended examples from the specification to the Test Cases document - ISSUE 52 -- due 2011-08-09 -- OPEN
<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/actions/139
close ACTION-139
<trackbot> ACTION-139 Move the R2RML extended examples from the specification to the Test Cases document - ISSUE 52 closed
ACTION-140?
<trackbot> ACTION-140 -- Boris Villazón-Terrazas to produce an RDF Schema representation of the R2RML vocabulary terms. -- due 2011-08-09 -- OPEN
<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/actions/140
<cygri> see http://www.w3.org/ns/r2rml#
Richard: we have updated the
R2RML namespace document, see above
... we have synced it
Michael: Thanks to Ivan for the support
Ivan: is it planned to make it data-in-HTML aware?
Richard: At the moment it is not clear who is responsible for the namespace document
Michael: is there an issue for this?
<cygri> this is ISSUE-53
ISSUE-53?
<trackbot> ISSUE-53 -- RDFS vocabulary reference for R2RML, and namespace document -- open
<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/issues/53
Michael: how about resolving it now with Boris as maintainer
Ivan: let's use http://www.w3.org/2001/sw/rdb2rdf/track/products
<cygri> ivan++
PROPOSAL: the WG considers LC-relevant issues only in the R2RML and DM product group of the tracker http://www.w3.org/2001/sw/rdb2rdf/track/products
<cygri> +1
<boris> +1
<juansequeda> +1
<ivan> +1
<Souri> +1
RESOLUTION: the WG considers LC-relevant issues only in the R2RML and DM product group of the tracker http://www.w3.org/2001/sw/rdb2rdf/track/products
ISSUE-54?
<trackbot> ISSUE-54 -- Simpler constant-valued term maps -- open
<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/issues/54
Michael: who is on this?
Richard: a proposal for syntactic sugar
Michael: is it doable in the time we have w/o distracting from the MUST-issues?
Richard: don't know in detail
<dmcneil> personally I lean towards "postpone" on this, because I feel like we need to push to get the last-call spec out
Richard: in terms of impl not a big deal but might induce discussion
Michael: let's leave it open for now
Richard: seems the same for ISSUE-57 - ISSUE-60
ISSUE-61?
<trackbot> ISSUE-61 -- Re-using public entity identifiers - look-up table -- open
<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/issues/61
<scribe> ACTION: Richard to come up with a proposal for ISSUE-61 (entity ID look-up table) [recorded in http://www.w3.org/2011/08/02-rdb2rdf-minutes.html#action02]
<trackbot> Created ACTION-150 - Come up with a proposal for ISSUE-61 (entity ID look-up table) [on Richard Cyganiak - due 2011-08-09].
Richard: IIRC there was a comment
from Souri that if we do this then it should be bijective
... I'd be interested in why?
Souri: If it is not 1:1 then, it is feels a little bit odd
Richard: In D2RQ we have
bijective mappings, error otherwise
... from implementation POV it's easy
... but in use it's inflexible
... can't report on experience with the more flexible mapping,
though
... unsure what the right way is
Souri: it's obviously much
cleaner if it is a 1:1 mapping
... the value of non-bijective mappings is there, and should be
possible to do
Seema: we can handle 1:M
Souri: what are the two sides of the mapping?
Richard: DB column and RDF
term
... type code for example 1... abc:accepted, 2 ...
abc:rejected
Michael: what's the insight now?
Richard: I'll propose a M:N
mapping then
... maybe I'll use SKOS?
<dmcneil> it doesn't seem to me that we can reach into unknown territory at this point, since we are trying to wrap things up for last call
<dmcneil> i am not referrring to SKOS, but to trying to do M:N lookups
David: M:N is too far reaching
<ivan> +1 to dmcneil
<Souri> many-to-many mapping => :AsianCuisine => "Chinese", "Japanese", "Vietnamese", "Indian" ; :SpicyCuisine => "Indian" ; I am not sure about implementation complexity (not for constants, but translating the DB result back to RDF terms)
Michael: Let's try to see how far we get with the M:N mapping, fallback is 1:1
Richard: I understand the concern
and I agree that none of the above issues should keep us from
going LC
... I invested quite some time to have the buffer now for
'optimizations' ;)
-> http://www.w3.org/2001/sw/rdb2rdf/test-cases/ Test Cases document
<boris> https://dvcs.w3.org/hg/rdb2rdf-tests
<dmcneil> from my perspective it seems to me that we barely have enough time to get to last call with the features that we have (there is no need to discuss it further, just wanted to get my view out there)
<boris> https://dvcs.w3.org/hg/rdb2rdf-tests/file/6a177981074b/D000-1table0rows/manifest.ttl
<boris> http://www.w3.org/2001/sw/rdb2rdf/test-cases/TCOverview.html
Michael: thanks a lot, David, for
your input - I appreciate it and very much value it!
... Thanks Boris for the hard work
Michael: next week we focus on DM
[adjourned]
trackbot, end telecon