See also: IRC log
<juansequeda> It seems that I can't dial in on a voip phone
<scribe> scribe: hhalpin
<MacTed> is anyone else getting that crazy echo?
scribenick hhalpin
<scribe> chair: ashok
trackbot, start meeting
<trackbot> Meeting: RDB2RDF Working Group Teleconference
<trackbot> Date: 14 September 2010
<Ashok> 1. Admin PROPOSAL: Accept the minutes of last meeting, see http://www.w3.org/2010/09/07-rdb2rdf-minutes.html
+1
<juansequeda> +1
APPROVED: Meeting minutes http://www.w3.org/2010/09/07-rdb2rdf-minutes.html
2. Name of standard
hhalpin: the name suggestion was half in humor but it is catchy.
ashok: name doesn't really matter but would be happy to hear other opinions
juan: could be confused with SQL
<Souri> +1 on waiting to think more ...
ashok: but ends in QL not RL.
hhalpin: would be happy to follow the spec editors here
ashok: no strong opinions, it can wait
3. First public working draft update
souri: have started setting up
cvs
... would like to share one point
... as regards xml syntax rdf
... in xml, we have cardinality restrictions
... richard has created a rdf syntax mapping
... we need to get it at the schema level
... for xml schema we'd like to transfer every bit to RDF
schema
... which means to get cardinality we'd have to with OWL
... but then we can do mapping
... validation
<Ashok> Harry: Advantage of XML schema is that it provides validation
<Ashok> ... RDFS and OWL does not give us that because they are based on open-world assumption
Turtle syntax
even something really drafty is fine.
feel free to e-mail me as well if eric doesn't get back quickly re cvs issues.
<Seema> sure thanks!
4. Semantics debate
ashok: there should be one
default mapping document
... could you guys get together and unify things.
juan: I think this document does
it.
... folks who have read and understood Eric's document
... would see that it would expense
... I've also ran it by Microsoft, but the reaction was
... was to make sure the notation is something people would be
comfortable.
... hoping to get more feedback from Microsoft.
ashok: how about other companies?
juan: I'm working with IBM now
ashok: can help with contacts here.
<juansequeda> http://www.w3.org/2001/sw/rdb2rdf/wiki/Default_Mapping
ashok: would like to run
<Ashok> I suggest Juan, Eric and Marcelo to go off and create a single default mapping document
Juan: we use a datalog
notation
... we have some predicates that build these URIs
... stemURI/name of relation, URI for key of that relationship,
you can concat all the primary keys
... the precise default generation is up for further
discussion
souri: you include foreign key URIs in attribute URIs for that scheme
juan: we have that as a tuple
URI
... we could have that as a tuple URI
souri: a foreign key is like a
virtual column
... more like an attribute URI so to speak
... four subsections
... x is rdf:type y
mapping template
[juan goes through document in detail]
juan: these rules seem to cover
all the defaults
... will write it all down in R2RML syntax
ashok: the difference is you guys use Datalog and Eric uses a set-oriented syntax, that's it?
juan: the rules he are using are
like a BNF, and we are using production rules
... only 4 cases
... eric has 20 cases
... we all should have done that
... and it's taken me time
... two weeks working on this
<juansequeda> http://www.w3.org/2001/sw/rdb2rdf/wiki/Default_Mapping_to_RDFS/OWL
juan: schema = ontology, we need both.
souri: that document is fairly
simple to understand
... they can understand it.
... they may not want to spend more than 5 minutes
... so they can see the different kinds
... if we have an example, then we get it.
... if there is an accompanying document like this, then if we
can get a formal semantics.
<juansequeda> http://www.w3.org/2001/sw/rdb2rdf/directGraph/
souri: I don't remember Eric's document too well...
LeeF, you there?
iv_an_ru?
MacTed: datalog seems better overall.
<juansequeda> http://www.w3.org/2001/sw/rdb2rdf/wiki/Semantics_of_R2RML
nuno: fine with datalog semantics, but can't speak for implementers
souri?
juan: IBM and Microsoft seem to prefer datalog as well.
souri: datalog seems simpler to me, a bit more precise
<Ashok> Harry: We should produce FPWDs first then create some test cases that people can play with
<MacTed> +1
ashok: we seem to get have consensus with.
<MacTed> OpenLink is likely to implement both.
souri: we already have a
prototype that works with both, including R2RML syntax
... a few months
<Ashok> Souri: Oracle will implement
juan: we will do mapping language, but we plan to implement the mapping languge
<Souri> At Oracle, we plan to implement the mapping language
ashok: probably d2rq
... what about microsoft?
juan: will ask
<MacTed> crazy talk!
anyways, just a thought MacTed :)
<Souri> I am at risk for next week's telecon
trackbot, end meeting
This is scribe.perl Revision: 1.135 of Date: 2009/03/02 03:52:20 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/aschok/ashok/ Succeeded: s/Mareclo/Marcelo/ Found Scribe: hhalpin Inferring ScribeNick: hhalpin WARNING: No "Topic:" lines found. Default Present: +3539149aaaa, nunolopes, +1.512.471.aabb, Ashok_Malhotra, +1.781.273.aacc, juansequeda, MacTed, +1.603.897.aadd, Souri, hhalpin Present: Souri Seema Juan Lee Harry Ted Ashok Nuno Found Date: 14 Sep 2010 Guessing minutes URL: http://www.w3.org/2010/09/14-rdb2rdf-minutes.html People with action items: WARNING: No "Topic: ..." lines found! Resulting HTML may have an empty (invalid) <ol>...</ol>. Explanation: "Topic: ..." lines are used to indicate the start of new discussion topics or agenda items, such as: <dbooth> Topic: Review of Amy's report[End of scribe.perl diagnostic output]