W3C

- DRAFT -

RDB2RDF Working Group Teleconference

26 Oct 2010

Agenda

See also: IRC log

Attendees

Present
boris, mhausenblas, juansequeda, Souri, MacTed, Ashok_Malhotra, +49.322.222.0.aaaa, +49.322.222.0.aabb, soeren, Seema, EricP, Marcelo, Alexandre, hhalpin, nunolopes
Regrets
Chair
Michael
Scribe
juansequeda

Contents


<trackbot> Date: 26 October 2010

<mhausenblas> scribenick: juansequeda

https://docs.google.com/document/pub?id=1e9rnIW5vos1mtxBtD4M06APi3uE0pTvX-fQeUjl-NzY

http://www.w3.org/2001/sw/rdb2rdf/directGraph/

http://www.w3.org/2001/sw/rdb2rdf/directGraph/alt

<soeren> > Zakim, +aaaa is soeren

<Souri> The document is still not ready for publishing, but we can finish it by tonight. For one thing, we need a slightly altered set of diagrams (due to introduction of a new property: rr:propertyColumn yesterday and some more change we will do in the next hour or so in the list of properties. (If we get Boris's help, that will be very good.)

Admin

<mhausenblas> PROPOSAL: Accept the minutes of last meeting, see http://www.w3.org/2010/10/19-rdb2rdf-minutes.html

+1

<soeren> +1

<mhausenblas> http://www.w3.org/2001/sw/rdb2rdf/track/actions/open

RESOLUTION: WG has accepted minutes of last meeting

Reminder, next week, Nov 2, there is no telcon

FPWD R2RML

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

<mhausenblas> $Id: Overview.html,v 1.18 2010/10/25 18:28:06 sdas2 Exp $

souri: noticed a few things. we introduced a new property. the diagram doesnt have the new property
... this needs to be updated
... and to make things more understandable, we need to change the name of some properties, given soeren's comments
... we can finish by tonight
... if boris can help us with the diagrams today, then we can have this out today

boris: yes I can help with the diagrams

mhausenblas: the editorial notes @@ should be removed or have the correct editorial notes

<Souri> I prefer "default mapping"

soeren: i think arbitrary mapping #218 and default mapping are different
... we don't need to solve this for the FPWD
... but we should discuss this

mhausenblas: if we use default mapping in the R2RML document, what do we do with the Direct Mapping document

ashok: what is the difference between direct and default mapping

mhausenblas: I see it as the same
... what does ericP think?

ericP: I prefer calling it a arbitrary mapping #218 because it is the default configuration

<betehess> +1 to Eric's explanation about Direct vs Default

MacTed: i agree with the relationship of direct and default, but I prefer the use of default

juansequeda: the default mapping is the arbitrary mapping #218

mhausenblas: we came up with these turns

ericP: default came from inside

<ericP> sorry, i misspoke: "direct" came from outside

marcelo: I don't see the term "arbitrary mapping #218" used

<mhausenblas> http://www.w3.org/2001/sw/rdb2rdf/directGraph/

<mhausenblas> http://www.w3.org/2001/sw/rdb2rdf/defaultGraph/

<mhausenblas> "A Default Mapping of Relational Data to RDF"

ericP: can I check before changing

mhausenblas: in the introduction, we call the document "Direct Mapping..." but if we use in the document "default mapping", people will be confused
... that is why I propose to change the document to "Default Mapping of Relational Data to RDF"

ericP: I wanted to check with the people who have been referring it to arbitrary mapping #218 before
... sandro, and somebody from lily use the term direct
... The R2RML uses a default mapping that is the direct

cygri: we need to make a decision what to write in the R2RML doc. Even if we call the doc "Default Mapping..." we can still make a distinction in that doc between default and direct

<Zakim> betehess, you wanted to give my understand of direct vs default

betehess: for me it is fine to say that the default mapping is the arbitrary mapping #218
... this is a deeper conversation
... the arbitrary mapping #218 is a wrapper
... and a standalone specification

hhalpin: wondering if ericP wanted to keep the term Direct Mapping?

ashok: would like to publish with the default mapping name

MacTed: the distinction is between a direct map and a non-direct map

ashok: if you do not have any customization, then you want the default mapping

<betehess> the real question is: do we want the Direct/Default Mapping to be defined in terms of R2RML? I don't think so, hence Direct Mapping

Macted: which is then the direct map

soeren: make a poll

Macted: is there a case where a default can be something different than a direct

betehess: is the arbitrary mapping #218 able to be defined in its own spec, or is it defined with R2RML ?

<mhausenblas> PROPOSAL: call it default mapping

<MacTed> +1

betehess: if there is no dependency, then it is a simple wrapper

<MacTed> -1

<Souri> +1

<MacTed> simple wrapper = arbitrary mapping #218 = that's the name

<MacTed> "default" defines behavior

cygri: there is a problem with the term arbitrary mapping #218. it implies that somehow the arbitrary mapping #218 is special.

<mhausenblas> PROPOSAL: call http://www.w3.org/2001/sw/rdb2rdf/directGraph/ default mapping in both title and URI (but keep the possibility to talk about a arbitrary mapping #218 there)

cygri: what eric has written is one way
... how do you create uris?
... what to do with FK
... the specs will make decision
... but it is not the case that there is only one correct to do it

<Souri> +1 for Richard's comments

<betehess> I see a very deep mis-understanding in what is the Direct Mapping. I'm really afraid of this....

<hhalpin> Can you type that in?

<mhausenblas> PROPOSAL: call http://www.w3.org/2001/sw/rdb2rdf/directGraph/ default mapping in both title and URI (but keep the possibility to talk about a arbitrary mapping #218 there)

mhausenblas: lets change to default mapping, and if there is a discussion on direct vs default, having it in the other doc

<cygri> +1

<Seema> +1

<Souri> +1

<boris> +1

+1

<soeren> +1

<Ashok> +1

<Marcelo> +1

<MacTed> abstain.

<betehess> -1

<ericP> -¾

<hhalpin> abstain

mhausenblas: I understand eric's objection, but I don't understand alex's objection

<mhausenblas> http://www.w3.org/2001/sw/rdb2rdf/defaultMapping

<mhausenblas> "A Default Mapping of Relational Data to RDF"

<cygri> +1 to juansequeda

<MacTed> exactly.

<MacTed> I don't see it as just about the name.

betehess

hhalpin is proposing what mhausenblas already proposed

macted: default is a behaviour and not a definition
... the default behavior can change

mhausenblas: we should open an issue and let the semantics sub group come up with the arbitrary mapping #218 definition

<betehess> my understanding: Direct Mapping is a simple wrapper for RDB in RDF. I was expecting the WG to discuss about which features would be interesting to have in a Direct Mapping, as a definition. The Default Mapping is somehow defined in terms of r2rml.

<mhausenblas> ACTION: Hausenblas to create an issue for the default vs. arbitrary mapping #218 [recorded in http://www.w3.org/2010/10/26-rdb2rdf-minutes.html#action01]

<trackbot> Created ACTION-77 - Create an issue for the default vs. arbitrary mapping #218 [on Michael Hausenblas - due 2010-11-02].

RESOLUTION: call http://www.w3.org/2001/sw/rdb2rdf/directGraph/ default mapping in both title and URI (but keep the possibility to talk about a arbitrary mapping #218 there)

hhalpin talking about publication details

<hhalpin> its 4 days to be courteous :)

<mhausenblas> Michael: we note the three objections (Eric, Alexandre and Ted) regarding the resolution

<hhalpin> so this is already not very courteous

<hhalpin> but I think the Systems Team understands :)

betehess is talking about publication details

<betehess> http://www.w3.org/2002/01/tr-automation/techs.rdf

<hhalpin> I thought the short-name agreement was /TR/r2rml

<betehess> the Webmaster needs: the group(s) this spec belongs to. Please see: http://www.w3.org/2002/01/tr-automation/techs.rdf

<betehess> the Webmaster also needs a short description

hhalpin: can eric actually change the uri and agrees?

cygri: the R2RML uses default mapping everything. the only place that arbitrary mapping #218 is used is in the title of the doc and the uri

<Zakim> betehess, you wanted to speak about dependencies

mhausenblas: question that hhalpin has is if eric has the time and willing to do that

<hhalpin> Could we just copy an "English" version of Eric's document into the R2RML? Is that necessary

betehess: there might be the case that we will have to remove the reference to the Direct Mapping document. I
... I'm not sure but this can be the case

cygri: is this for normative or informative reference?
... does it make a difference?

betehess: I'm not sure. I need to ask

<Souri> I still do not think it is that incomprehensible without the default-mapping doc!

<Ashok> http://www.w3.org/2001/sw/rdb2rdf/r2rml/Overview.html

macted: we have had external views stating that they don't understand R2RML doc without the default mapping

<ericP> will people understand r2rml FPWD without the arbitrary mapping #218 text?

<Souri> Have we really checked its readability with a real database person (as opposed to logicians), who is really a target audience?

<mhausenblas> A related specification is A Direct Mapping of Relational Data to RDF [DIRECT]. It defines a fixed “default mapping”. In the default mapping of a database, The structure of the resulting RDF graph directly reflects the structure of the database, the target RDF vocabulary directly reflects the names of database schema elements, and neither structure nor target vocabulary can be changed. With R2RML on the other hand, a mapping author can define high

<mhausenblas> customized views over the relational data.

<cygri> replace first two sentces with: “This WG will define a fixed “default mapping” from a database to RDF.”

<Souri> People in some Oracle groups understand the R2RML doc perfectly and thinks its too easy!

<MacTed> "too easy" is not a valid objection to anything.

<mhausenblas> PROPOSAL: replace the second paragraph, first two sentences of http://www.w3.org/2001/sw/rdb2rdf/r2rml/ “This WG will define a fixed “default mapping” from a database to RDF" and create and editorial note re discussion

cygri and mhausenblas are talking about who to rephrase the default mapping section of the intro

<cygri> +1

This new Proposal will make the previous proposal obsolete

+1

<Ashok> +1

<Marcelo> +1

<boris> +1

<hhalpin> abstain

<nunolopes> +1

<Souri> +1

<MacTed> +1

<Seema> +1

<ericP> i don't know who will understand this doc

<hhalpin> eric - is that an objection or an abstain?

<ericP> abstain

<hhalpin> I mean, I'm hoping we can link these two documents up.

<hhalpin> but if we can't do it in the FPWD we can do it when we agree things are baked.

RESOLUTION: replace the second paragraph, first two sentences of http://www.w3.org/2001/sw/rdb2rdf/r2rml/ “This WG will define a fixed “default mapping” from a database to RDF" and create and editorial note re discussion

no objections to this proposal. Only two abstains from ericP and hhalpin

<mhausenblas> [adjourned]

<mhausenblas> http://www.w3.org/2001/sw/rdb2rdf/track/issues/5

<mhausenblas> trackbot, end telecon

Summary of Action Items

[NEW] ACTION: Hausenblas to create an issue for the default vs. arbitrary mapping #218 [recorded in http://www.w3.org/2010/10/26-rdb2rdf-minutes.html#action01]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.135 (CVS log)
$Date: 2010/10/26 17:26:39 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
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/direct mapping/arbitrary mapping #218/G
Found ScribeNick: juansequeda
Inferring Scribes: juansequeda
Default Present: boris, mhausenblas, juansequeda, Souri, MacTed, Ashok_Malhotra, +49.322.222.0.aaaa, +49.322.222.0.aabb, soeren, Seema, EricP, Marcelo, Alexandre, hhalpin, nunolopes
Present: boris mhausenblas juansequeda Souri MacTed Ashok_Malhotra +49.322.222.0.aaaa +49.322.222.0.aabb soeren Seema EricP Marcelo Alexandre hhalpin nunolopes
Agenda: http://lists.w3.org/Archives/Public/public-rdb2rdf-wg/2010Oct/0075.html
Found Date: 26 Oct 2010
Guessing minutes URL: http://www.w3.org/2010/10/26-rdb2rdf-minutes.html
People with action items: hausenblas

[End of scribe.perl diagnostic output]