RDB2RDF Working Group Teleconference -- 13 Apr 2010

<trackbot> Date: 13 April 2010

http://www.w3.org/2001/sw/rdb2rdf/wiki/ScribeList

ivan?

<iv_an_ru> I'm at chat only, no phone

Convene RDB2RDF meeting April 13th 2010

roll call

<scribe> chair: Ahmed

<scribe> scribe: hhalpin

Seema, MacTed, Michael, Richard, Nuno, EricP, Ashok, Juan,

<mhausenblas> re http://lists.w3.org/Archives/Public/public-rdb2rdf-wg/2010Apr/0011.html

<mhausenblas> please do not post files to the mailing list

we are in the rdb2rdf working group

<juansequeda> mhausenblas, our bad. Dan doesn't have access to the wiki yet

http://www.w3.org/2010/04/06-rdb2rdf-minutes.html

<juansequeda> I'm resending everything now

<ericP> juansequeda, i never saw a google invite either

<ericP> could you send to eric+google@w3.org ?

<juansequeda> ericP, I already did. Didn't you get it?

<mhausenblas> ACTION-36?

<trackbot> ACTION-36 -- Eric Prud'hommeaux to review the UC and report back to the WG with requests for more details -- due 2010-03-30 -- OPEN

<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/actions/36

APPROVED: Meeting minutes approved for April 6th 2010

<mhausenblas> ACTION-38?

<trackbot> ACTION-38 -- Eric Prud'hommeaux to review the UC and report back to the WG with requests for more details -- due 2010-04-06 -- OPEN

<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/actions/38

<mhausenblas> close ACTION-36

<trackbot> ACTION-36 Review the UC and report back to the WG with requests for more details closed

<mhausenblas> close ACTION-38

<trackbot> ACTION-38 Review the UC and report back to the WG with requests for more details closed

going through open actions

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

<mhausenblas> ACTION-39?

<trackbot> ACTION-39 -- Harry Halpin to begin a draft of Working Group Note on use-cases, ordering them from most complete to least complete -- due 2010-04-06 -- OPEN

<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/actions/39

<mhausenblas> ACTION-40?

<trackbot> ACTION-40 -- Juan Sequeda to will make ER diagrams for use-case document -- due 2010-04-06 -- OPEN

<trackbot> http://www.w3.org/2001/sw/rdb2rdf/track/actions/40

ericP: I am interested mostly in expressivity
... but others are interested in user stories

<mhausenblas> close ACTION-40

<trackbot> ACTION-40 Will make ER diagrams for use-case document closed

don't worry the expressivity requirements will become very important

juan: angela has updated wiki with intro and motivation
... daniel and I were updating the spreadsheet

<juansequeda> http://www.w3.org/2001/sw/rdb2rdf/wiki/Use_Cases_and_Requirements#Introduction

juan: but we're just giving write access to people now

everyone has had a chance to try to fill out their use-cases

Daniel: these actions are still open, people haven't really updated their use-cases
... I am happy to update the spreadsheet.
... but I have to nag people

Use-Case document

Ahmed: the structure is motivation, use-case, and then each requirement deduced from the spreadsheet
... Is anyone working on the requirements

ericP: I've been pooling away looking at the difference between the various requirements
... but I have my own feelings about the characteristics of the expressivity

daniel: I'm pushing down that path as well, but from top-down as opposed to bottom up
... we should meet at some point
... a lot of these pieces are already written
... for the use-cases, from the original workshop and the charter
... if we take the time to look at the community, we can bring and borrow them into this report

Ahmed: Can you send us URLs?

<mhausenblas> http://www.w3.org/2007/03/RdfRDB/

<mhausenblas> and the report at http://www.w3.org/2007/03/RdfRDB/report

daniel: I will mail these links out

Ahmed: can we pin down dates for the various sections?
... because otherwise we run the risk of missing april

juan: the whole document needs to be done by end of April

Ahmed: So this week we really need to give input
... I can see we have a lot of material for motivation and use-cases, but little in the way of requirements
... We need a lead person for the requirements

Michael: Note we only have two more telecons, the 20th and 27th
... the 27th will be during WWW2010.
... Do you work everything out ericP including pubrules

ericP: The fact that there's two of us working on it is better

Michael: So do we work via XML

ericP: or directly via HTML

Michael: then let's use XMLspec
... let's draft something asap.

Ahmed: Yes, you should create a skeleton, and then extract from the wiki and spreadsheet into the document.

<mhausenblas> ACTION: mhausenb to draft an Use Case WD with xml-spec [recorded in http://www.w3.org/2010/04/13-rdb2rdf-minutes.html#action01]

<trackbot> Created ACTION-49 - Draft an Use Case WD with xml-spec [on Michael Hausenblas - due 2010-04-20].

angela: I am done with the motivation section mostly

Ahmed: so give feedback to angela this week

Daniel: We should kill a few that's easy, but it's harder to determine if we are missing a use-case

Ahmed: Every use-case should go by the end of the spreadsheet

Daniel: by today

<scribe> ACTION: jsequeda to nag each of the use-cases owners to finish off the spreadsheet [recorded in http://www.w3.org/2010/04/13-rdb2rdf-minutes.html#action02]

<trackbot> Created ACTION-50 - Nag each of the use-cases owners to finish off the spreadsheet [on Juan Sequeda - due 2010-04-20].

<Souri> if possible, it will be nice to see a list of <usecase, owner> pairs

<scribe> ACTION: hhalpin to create WBS poll for use-case document [recorded in http://www.w3.org/2010/04/13-rdb2rdf-minutes.html#action03]

<trackbot> Created ACTION-51 - Create WBS poll for use-case document [on Harry Halpin - due 2010-04-20].

The final decision lies in the hands of the editors

who are Michael, Angela, Daniel, Juan, and Eric it seems

<mhausenblas> Editors are Eric and Michael

Ahmed: I'd be comfortable if the editors do first clean-up, and then we review next week

<mhausenblas> Authors are Angela, Daniel, Juan, Soeren and Lee

Ahmed: We need two dates, one for editors to clean up use-cases, and then another so we don't lose the dates.

Daniel: early Monday morning is the earliest deadline

Ahmed: Early monday morning is fine

+1 not missing first deadline!

Ahmed: You own the requirement section
... ericP will be the main owner of the requirement sections.

ericP: works for me
... I can do Sunday night.

Michael: how we get from XML document to period when it's done.

<Souri> current use-cases list: <HCLS, ericP>, <BioPAX, ericP>, <integrating RDBs for tax control, Angela>, <WebApps-wordpress-and-oscommerce, ???>, <UniProt, ???>, <rCAD, juan-and-dan>, <many-to-many-join-tables, ???>, <value-based-type-spec, ???>

ericP: We have faith in the editors

hhalpin: there will be some time
... at the end of the meeting

michael can be able to make that meeting

<mhausenblas> http://www.w3.org/2001/sw/rdb2rdf/use-cases/Overview.xml

Ahmed: So the entire XML document will be available to us a few days before the 27th, so we can draft it up quickly after the 27th to send the final version to the W3C.
... no objections

ericP: let's have all the editors talk to Dan over the spreadsheet

Going over use-case spreadsheet

ericP: descriptive text?

<mhausenblas> Michael: the Working Draft will appear at http://www.w3.org/2001/sw/rdb2rdf/use-cases/

<mhausenblas> (note that ATM there is just a minimal skeleton)

ericP: you would want first
... more text
... than one appears right now

<Souri> Could someone update the owner info in the use-cases list I posted above?

ericP: the point I interpreted is to expand the use-case

daniel: we could have 12 pages and then several appendices.

<Souri> Specifically, who submitted the "UniProt" use case? the "many-to-many join tables"? and the "value-based-type-spec"?

<juansequeda> Souri, I just gave you access

daniel: two kinds of readers, one that wants all the details to run the code and the other that doesn't.

ericP: Can the person read it on the plane? How useful is UML diagram?

daniel: I live and die by them.

<MacTed> +1 for details in appendices

<MacTed> hrefs to #frag-ids which are within-URL appendices seems OK to me...

<juansequeda> Souri, Uniprot is from Eric and the last two examples are from Lee

<Souri> updated list just to be complete (all the info is in the spreadsheet): <HCLS, ericP>, <BioPAX, ericP>, <integrating RDBs for tax control, Angela>, <WebApps-wordpress-and-oscommerce, Soeren>, <UniProt, ericP>, <rCAD, juan-and-dan>, <many-to-many-join-tables, LeeF>, <value-based-type-spec, LeeF>

daniel: I'd prefer to keep UML diagrams in but we have to see it first
... before making the decision

angela: do you prefer me to generate a UML for the example diagrams I gave?

Daniel: yes

ericP: intended queries
... what kinds of queries to expect
... wanted it to be clear that we are actually querying the database, not just doing a dump and then querying that dump
... do we not want to show that people?

orri: I think it's important that we show this
... it should figure prominently in the intro that we do the mapping on the fly without an ETL

ericP: Happy to say that in words
... but do we just say that it words

orri: we should have fairly typical queries
... rely on SPARQL 1.1 regardless

thinking that we need to check how our schedule interacts with SPARQL 1.1.

Ahmed: I think we need to show that it can work this way

ericP: we should have example query for each use-case

Souri: it's key we can't materialize on triple form
... that's important it almost comes from there
... so we don't replicate the database
... but how that is done is implementation-specific

orri: a query for each one
... may take up too space
... so we can say it may translate into some SQL
... although the exact SQL will only be may be translated
... since we are not sure of which kind

daniel: we should not worry about implementation details

ericP: we can not worry about this now

Ahmed: We don't want to put in...

angela: I did put it in and then took it out
... remember that I introduced the URIs in order to join two tables

Daniel: we have to adjust when making the working language

MacTed: We are turning use-cases into case studies
... we don't need to do exhaustive example work-throughts

thats the actual test-cases

angela: we will put as much as possible and then the editors can cut

Meeting Adjourned

trackbot, end meeting

- DRAFT -

RDB2RDF Working Group Teleconference

13 Apr 2010

Attendees

Contents

Convene RDB2RDF meeting April 13th 2010

going through open actions

Use-Case document

Going over use-case spreadsheet

Summary of Action Items

Scribe.perl diagnostic output