<ChrisW> /topic #rif 09 Jan RIF agenda: http://lists.w3.org/Archives/Public/public-rif-wg/2007Jan/0036.html

<ChrisW> dave hirtle, you there?

<ChrisW> are you joining us today, dave?

<csma> david, could you scribe today, please?

<ChrisW> dave hirtle are you thjere?

<ChrisW> john hall? are you joining us today?

<johnhall> trying to get a phone connection

<csma> scribe: Alex

<ChrisW> Scribe: Alex Kozlenkov

<ChrisW> scribenick: AlexKozlenkov

Next meeting 16th of January

ChrisW: actions review

Chris: no admin actions

Admin

<ChrisW> RESOLVED: accept Dec 19th telecon minutes

csma: December 19th minutes approved

Deborah: There are additional notes from Harold
... we should wait for the new version
... the minutes will be published tomorrow

csma: action for F2F5

Sandro: the meeting page is not yet set up
... should be there in the next couple of days

<ChrisW> action 201 continued

csma: Allen, any news about the meeting?

Allen: Hotel is the main issue. Comfort Inn is good
... free shuttle from Dalles airport
... the information will be put up shortly
... Dulles the bets place to fly

<ChrisW> despite attempts by the president...

Allen: the page will have to have nationalities registered on the meeting web

Deborah: we need to have nationalities for security procedures
... all requirements will be checked shortly

<ChrisW> ACTION: Allen to check information needed for foreign visitors, deadline for reg [recorded in http://www.w3.org/2007/01/09-rif-minutes.html#action01]

<rifbot> Created ACTION-206 - Check information needed for foreign visitors, deadline for reg [on Allen Ginsberg - due 2007-01-16].

<johnhall> SBVR - nothing new

F2f

csma: Liaisons, nothing new from OMG

Liason

Slots and constraints discussion

Technical Design

csma: Michael Kifer is not here yet so Technical Design discussion will be taken up later.

UCR

ChrisW: start first with the use cases and requirements
... proposes action review

csma: action 132

<ChrisW> action 132 closed

johnwall: finished as reported two meetings ago

<ChrisW> action 144 continued

csma: 144 continued

<ChrisW> action 167-168 closed

<Allen> yes

<igor> ok

csma: 167/168 closed: definition of "covers"
... 169: glossary

<ChrisW> action 169 continued

Hassan: continued, but proposes other should contribute

<ChrisW> action 169 closed

csma: closes the action waiting for a new responsible
... 197

Allen: cleared and ready apart from small details in UC1

csma: takes up an action on UC1 motivation for linking to requirements

<ChrisW> ACTION: Christian to clean up UC 1 requirements motivation [recorded in http://www.w3.org/2007/01/09-rif-minutes.html#action02]

<rifbot> Created ACTION-207 - Clean up UC 1 requirements motivation [on Christian de Sainte Marie - due 2007-01-16].

<ChrisW> action 197 closed

<ChrisW> action 205 closed

csma: 205, new definition of covers added to UCR

<Allen> http://www.w3.org/2005/rules/wg/wiki/UCR/Introduction

csma: question: any objections tro the proposed definition linked above

<ChrisW> "One of the critical factors for a successful RIF is that it be useful for interchange of rules among the set of rule languages it is intended to cover. Section 5, Coverage, deals with the issue of how to characterize the space of rule languages in such a way that clear and principled decisions as to what the RIF will (and will not) cover can be made. We note that in this document we deliberately refrain from defining the notion of "coverage" in a rigorous manner,

<igor> didn't Sandro propose an alternative?

<ChrisW> Proposed: Accept definition of covers and close Issue-22

Allen: altrernative Sandro's proposal is actually included based on e-mail exchange

csma: definition is approved and issue closed

<Deborah_Nichols> chris, yes, I can

<ChrisW> RESOLVED: Accept definition of covers and close Issue-22

ChrisW: it is approved by consensus

<ChrisW> ACTION: deborah to update issues list to reflect resolution [recorded in http://www.w3.org/2007/01/09-rif-minutes.html#action03]

<rifbot> Created ACTION-208 - Update issues list to reflect resolution [on Deborah Nichols - due 2007-01-16].

<csma> sandro, are you here? We are discussing issue 12

Dave: Sandro's objections are not critical

csma: RIF is the base of SWRL
... is the core question

DaveReynolds: can live with that

csma: we prefer that RIF will be the basis of SWRL

Sandro: OK with that, perhaps a third path could be found

csma: we would not work on that
... RIG WG is not responsible for this
... proposes to accept Sandro's proposal

ChrisW: proposes to put it back to next week

<ChrisW> ACTION: Christian to put resolution of issue-12 on next weeks agenda [recorded in http://www.w3.org/2007/01/09-rif-minutes.html#action04]

<rifbot> Created ACTION-209 - Put resolution of issue-12 on next weeks agenda [on Christian de Sainte Marie - due 2007-01-16].

csma: the text will be added to UCR

Dave: probably into the Introduction

Sandro: we could put it into the UCR now

csma: still let us wait for a week before adding it
... close to the third version of UCR

<ChrisW> Who is IVML?

Allen: make a note about specific features in RIFRAF

csma: could be for the next draft
... it is for the time when RIFRAF is complete
... Version 3 should be released before the end of the month

RIFRAF

csma: action review on RIFRAF quickly before returning to the Technical Design
... All ontologizing actions are continued

<ChrisW> all RIFRAF actions continued

Technical Design

csma: all RIFRAF actions are continued so back to the Technical Design

MichaelKifer: summary on the issues
... two styles: relational and OO
... relational is less general, no explicit object id
... relational can be converted to OO

MichaelKifer, in general, the opposite may not be possible

MichaelKifer, OO uses binary and unary predicates with object ids as first arguments

MichaelKifer: this means going baack to relational notation is not possible

<ChrisW> Michael, please stop breathing into the phone

csma: confused because all the information in the OO slotted notation can be expressed in relational notation by combination of binary predicates
... conversely, adding information is not possible that is introducing object id that is not in the relational model

MichaelKifer: going back is problematic because there is no place for object-id

csma: refers to his example in e-mail
... object id can be made explicit
... in relational notation

MichaelKifer: relational slotted notation is more restricted
... id is uniquely implied there by the values

<Harold> Michael, what Christian seems to say is oid:Class{s1-v1,...,sN->vN} can be simulated via Class{s0->oid,s1-v1,...,sN->vN}.

csma: is of opinion it is an important issue

<csma> acq francois

<Zakim> Francois, you wanted to translation oriented to relational.

Francois: OO has implicit ids vs. the ones that should be made explicit in the relational case

<MichaelKifer> Harold, the point is that Class{s0->oid,s1-v1,...,sN->vN} in the relational notation is an object with id that is different from oid.

FrankMcCabe: object is the query itself
... no handle as blank nodes in RDF

<ChrisW> Adrian, are you here?

<Harold> Michael, the (relational) 'key' is a local id only.

Harold: perhaps the dimension for slotted notations should be reconsidered for RIFRAF

<Hassan> I agree with Frank. Michael seemed to agree. Slotted notation should be out of CORE and left to each dialect to be specified as constraints.

<GaryHallmark> all the rule languages I'm interested in are slotted because they bind to "real world" data -- relational data, XML data, or Java data

<GaryHallmark> -1 for not addressing this up front in a common way in CORE

MichaelKifer: including slots or constraints affects roundtripping
... based on this understanding it should be agreed on where it is

<GerdWagner> q

GaryHallmark: better to have a common way of representing slots

<MichaelKifer> my understanding is that slotted or not slotted impacts only the roundtrip point. in fact, any syntactic feature beyond plain unsorted predicate calculus (including constraints) is a round trip issue

<Harold> Closed slots can be introduced without introducing oids at the same time, which is what we need for Phase 1.

csma: action on GaryHallmark t oprovide examples with rules where slotted notation is useful

<agiurca> There are many examples of rules with slots. See for example JBoss Rules

<csma> +1 to chrisw

GerdWagner: refers to his previous e-mails with examples of JBoss Rules

<agiurca> Must be a common understanding of what is a slot

<agiurca> F-Logic examples: http://oxygen.informatik.tu-cottbus.de/rewerse-i1/?q=node/22

<GerdWagner> see http://oxygen.informatik.tu-cottbus.de/rewerse-i1/?q=node/24

<sandro> The nature of the Core -- whether it should have Option Features -- is a different discussion!

<Francois> Sorry, Friends, I have to leave.

<Francois> bye.

<Zakim> ChrisW, you wanted to say i'm not sure having objects in rules and slotted notation are the same

ChrisW: having named roles is not the same as OO representation

<Harold> Chris, I agree calling them 'keyword arguments' can remove the confusion.

<Hassan> Yes Chris !

ChrisW: these are separate ideas possibly due to the term "slotted" used for both

<agiurca> In F-Logic : X:person[ancestor->>Y:person] . How this translate to RIF?

<Harold> What I called 'closed slots' could be called 'keyword arguments'.

<csma> acq hassan

<agiurca> JBoss Rules Column: i : Item(actualDeliveryDate : actualDeliveryDate, scheduledDeliveryDate : scheduledDeliveryDate ) is another example in favor of slots

Hassan: agree on the same datastructure and then sugar it into any form )in dialects)

<Harold> Hassan, the nice thing with 'keyword arguments' is we don't need extra semantics: this simple case we need in Phase 1 can be 'de-sugared', as you say.

Hassan: agrees with Gerd that a convenience will be useful

<Harold> In CLP terminology, 'keyword arguments' are very special kind of constraints. In Phase 2 we can generalize this in the light of full CLP.

MichaelKifer: can Hassan clarify his proposal for the convenience in the core

Hassan: if 80% of clients agree on the model even if it is not perfect, this syntax will be good to have

csma: concerned about the remaining 20%. Would it be too bad for them

Hassan: the convenince will be ignored by the 20%

<GerdWagner> there is no 100% coverage goal!

Hassan: the convenience notaiton would be able to be converted to the universal constraints notation and then it can be accepted by the remaining 20% of the systems

<MichaelKifer> i don't understand hak's arg: what is the point of having slotted notation in the core, but not giving it a semantics

ChrisW: taking up again the relational to OO roundtripping

<Harold> Chris, I proposed a round-trip between positional and slotted arguments, or better between non-keyword and keyword arguments.

ChrisW: how do we lose the relational tuples if it went to the OO and back

<Harold> Round-tripping between relational and OO is much harder.

ChrisW: are we giving up anything in that LangA goes to Core then to LangB and back it could be problematic

csma: UC1 is such example

<agiurca> We need to use object oriented notation. Then the roundtrip is possible.

<Harold> oid:Class{s1-v1,...,sN->vN} ==> Class{s0->oid,s1-v1,...,sN->vN} ==> oid:Class{s1-v1,...,sN->vN} has some problems, as Michael mentioned.

MichaelKifer: will work with csma on his example on this roundtripping

ChrisW: the core will have a keyword syntax

Chris: ... available to it

<DaveReynolds> possibly, depends on what it says about signatures

csma: consensus on that: have keyword arguments
... ... in the core

csma; takes up notion of RIF compliance

<GerdWagner> alex please add "keyword arguments in the sense of OO slots"

<ChrisW> consensus that we should have "keyword" syntax with what Michael called "OO semantics"

<Hassan> What is an optional feature?

Will do Chris

<Harold> Take Gary's recursion discussion, as an example.

<ChrisW> "implementing the core" means translating in/out of it

<Hassan> Then it is necessary to have such options if we adopt the 80%/20% convenience slotted syntax.

<Harold> The core could have optional feature recursive="yes" vs. recursive="no" (I think recursive="yes" should be the default, so we would have a 'negative' optional feature).

MichaelKifer: implementing the core is not yet fully defined

ChrisW: the question is not about implementing but about translating

<ChrisW> we are out of time

<ChrisW> we are out of time

<csma> almost

<johnhall> Sorry, I have another meeting

csma: implementing a dialect and compliance

AlexKozlenkov: we need both defintions

csma: no consensus yet on this

<sandro> +1 adjourn

<Hassan> +1 to adjourn

<GerdWagner> bye

<agiurca> -agiurca

<agiurca> quit

<ChrisW> oops

<ChrisW> hits the wrong button

<csma> anything else you wanted to discuss?

<ChrisW> no

<ChrisW> see you tomorrow

<csma> let's talk tomorrow, then

<csma> bye

<ChrisW> ciao

This is scribe.perl Revision: 1.127  of Date: 2005/08/16 15:12:03  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/Dallas/Dulles/
Succeeded: s/kifer/Kifer/
Succeeded: s/CrisW/ChrisW/
Succeeded: s/thrid/third/
Succeeded: s/convenince/convenience/
Succeeded: s/CfhrisW/Chris/
Succeeded: s/conformance/compliance/
Succeeded: s/translatingf/translating/
Found Scribe: Alex
Found Scribe: Alex Kozlenkov
Found ScribeNick: AlexKozlenkov
Scribes: Alex, Alex Kozlenkov
Default Present: Harold, ChrisW, Francois, FrankMcCabe, Sandro, Hassan_Ait-Kaci, csma, Dave_Reynolds, Deborah_Nichols, Jeff_Pan, Allen_Ginsberg, AlexKozlenkov, Gary_Hallmark, StellaMitchell, agiurca, johnhall, [IVML], igor, Michael_Kifer, Gerd_Wagner
Present: Harold ChrisW Francois FrankMcCabe Sandro Hassan_Ait-Kaci csma Dave_Reynolds Deborah_Nichols Jeff_Pan Allen_Ginsberg AlexKozlenkov Gary_Hallmark StellaMitchell agiurca johnhall [IVML] igor Michael_Kifer Gerd_Wagner
Regrets: PaulaLaviniaPatranjan JosDeBruijn LeoraMorgenstern MichaelSintek
Agenda: http://lists.w3.org/Archives/Public/public-rif-wg/2007Jan/0036.html
Got date from IRC log name: 9 Jan 2007
Guessing minutes URL: http://www.w3.org/2007/01/09-rif-minutes.html
People with action items: allen christian deborah

WARNING: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.