RIF WG Teleconf Minutes

2 May 2006

See also: IRC log


Alex Kozlenkov, Allen Ginsberg, Axel Polleres, Chris Welty, Christian de Sainte Marie, Dave Reynolds, David Hirtle, Deborah Nichols, Donald Chapin, Francois Bry, Frank McCabe, Gary Hallmark, Harold Boley, Hassan Ait-Kaci, Igor Mozetic, John Hall, Jos de Bruijn, Jos de Roo, Michael Kifer, Leora Morgenstern, Markus Kretzsch, Michael Kifer, Mike Dean, Paul Vincent, Paula-Lavinia Patranjan, Peter Patel-Schneider, Said Tabet, Sandro Hawke, Uli Sattler
see Wiki page
Chris Welty
Mike Dean




<FrancoisBry> zakim ??P25 is me

<FrancoisBry> zakim please mute me

<pfps> but can you sing along? :-)

<sandro> FrancoisBry, you'll need to say "Zakim, Francois is me".

<FrancoisBry> zakim Francois is me

<sandro> You need the comma, FrancoisBry.

<AlexKozlenkov> AlexKozlenkov joined

<AlexKozlenkov> AlexKozlenkov joined

<AlexKozlenkov> yes I am

<AlexKozlenkov> yes I am

<Donald_Chapin> zakim P43 is me

<AlexKozlenkov> AlexKozlenkov zakim joined

<scribe> Scribe: Mike Dean

<scribe> ScribeNick: mdean


<ChrisW> http://lists.w3.org/Archives/Public/public-rif-wg/2006Apr/att-0112/01-part

RESOLUTION: accept minutes

<ChrisW> http://lists.w3.org/Archives/Public/public-rif-wg/2006Apr/att-0120/18-rif-minutes.html

includes revisions based on comments

RESOLUTION: accepted

<ChrisW> http://www.w3.org/2005/rules/wg/track/

now have an action tracker - will now be used for tracking actions

pfps: bug in action tracker: actions can have no topics
... e.g. action ID 1
... precludes actions by topic

Leora: still working on actions 1 and 2

<sandro> ack +87362aaaa

Leora: questions on FOL

ChrisW: defer discussion to later in agenda

Sandro: actions 3, 4, 6 continued

<LeoraMorgenstern> mdean, not questions on FOL, questions on what we'd like in the draft proposal for FOL

csma: action 9 and 10 continued

ChrisW: any amendments to agenda?
... no

F2F 3

Sandro: tomorrow is deadline for signing up for bus from airport

<PaulaP> no news


F2F 4

ChrisW: poll is now closed

<ChrisW> http://www.w3.org/2002/09/wbs/38457/f2f4dates/results

<sandro> 5-6 October, McLean, Virginia, US - MITRE Proposal 5 1 5 4 4 2

<sandro> 16-17 October, McLean, Virginia, US - MITRE Proposal 4 4 3 1 7 2

<sandro> 2-3 November, McLean, Virginia, US - MITRE Proposal 2 4 4 3 7 1

<sandro> 4-5 November, Athens, Georgia, US - At ISWC 1 4 6 10

ChrisW: 10 prefer ISWC colocation, 6 second highest - nobody objects - looks like clear favorite

csma: do not have clear local organizer

<PaulaP> how about funding?

pfps: volunteered to handle local arrangements
... or at least contact and interface with conference center

ChrisW: this will require each attendee to register and pay registration fee
... will accept credit cards

ACTION (pfps): set up page with information

ChrisW: email from Daniel Schwabe regarding arrangements for iswc

pfps: aware of at least 2 options for accepting credit cards

ChrisW: tutorials start on the 5th - workshops still not decided (1 day only)
... conference dates are correct - tutorials on Nov 5, not Nov 4


<AlexKozlenkov> how much is the registration fee?

<PaulaP> +1

ChrisW: thanks to mitre folks for nice proposal - will consider for future meetings

<sandro> +1 thanks for MITRE for their offier!

<AxelPolleres> +1 (for csma and me ;-) )


ChrisW: status of SPARQL?

Sandro: up in the air - Semantic Web Activity reorg hasn't happened yet

<JosDeRoo> it is PR

<josb> candidate

Sandro: SPARQL may be at CR for a while

<josb> http://www.w3.org/TR/rdf-sparql-query/

ChrisW: XQuery/XPath?
... CL in ISO equivalent of Last Call
... PRR?

csma: recent meeting

Paul: csma should have good presentation at F2F - hope to have update for F2F

<Donald_Chapin> Nothing new

ChrisW: SBVR?
... ODM?

<JosDeRoo> re SPARQL http://www.w3.org/News/2006#x20060406a

ChrisW: presentation at OMG meeting - in OMG equivalent of Last Call

Use Cases and Requirements

ChrisW: Leora questions on actions

Leora: what do we mean by FOL?
... specify requirements not syntax
... but some syntax needed for examples
... Question 2: should we just use Common Logic?

ChrisW: question previously raised by Michael
... CL is fine
... may add or subtract features as necessary

<josb> how is CL the standard for RDF?

Leora: will start with CL

Sandro: some concerns about differences between CL and FOL from Peter?

pfps: don't know

<Zakim> sandro, you wanted to say Any FOL, and not CL

pfps: RDF is based on non-standard FOL that shares some features with underlying semantics of CL

<sandro> Peter: RDF is based on a non-standard view of FOL, which it shares with CL

ChrisW: no standard FOL other than CL
... CL designed to be very flexible, and allows fragments

<FrancoisBry> +q

ChrisW: start with CL, get refinement from WG - can narrow if needed

josb: what is motivation for using CL?
... why diverge from textbook definitions
... can substitute symbols as needed for web

pfps: CL is not yet an ISO standard

Leora: can use vanilla FOL and CL and then pick

Sandro: pick vanilla FOL that works with some theorem prover

Frank: disagree that you can't define FOL without syntax
... enumerate features
... versions of CL have been in standards process for almost a decade
... why do we need to define FOL if it's not on our roadmap?

<sandro> (Note syntax for TPTP which is a sort of de-facto FOL interchange format, tptp.org

<josb> Some textbooks which define FOL:

ChrisW: stop discussion - group can respond to Leora's action

<josb> http://www.amazon.com/gp/product/0387942580/qid=1146584341/sr=2-1/ref=pd_bbs_b_2_1/104-3762612-0543939?s=books&v=glance&n=283155

<josb> http://www.amazon.com/gp/product/0122384520/qid=1146584361/sr=2-1/ref=pd_bbs_b_2_1/104-3762612-0543939?s=books&v=glance&n=283155

<josb> http://www.amazon.com/gp/product/0387945938/qid=1146584328/sr=1-2/ref=sr_1_2/104-3762612-0543939?s=books&v=glance&n=283155

Leora: describe critical success factor
... confused by new document from Paula
... what is CSF for FOL?

csma: earlier discussion about whether FOL is CSF or requirement

<PaulaP> http://www.w3.org/2005/rules/wg/wiki/Towards_a_Rule_Interchange_Format%3A_Goals%2C_Critical_Success_Factors%2C_Requirements

<PaulaP> here FOL is a requirement

csma: if requirement, need to identify CSF on which it depends

ChrisW: a little vague - needs to be clarified

<ChrisW> q>

PaulaP: got it

<ChrisW> http://www.w3.org/2005/rules/wg/wiki/Towards_a_Rule_Interchange_Format%3A_Goals%2C_Critical_Success_Factors%2C_Requirements

CSF Goals and Requirements from Paula

ChrisW: published Friday, not extensive time to review

Frank: interesting list, but didn't meet Frank's requirements
... most of list is technical requirements about product
... for CSF, consider whether you addressing builder of spec or users
... too much internal, not enough external
... wrote own CSFs
... goal: wide scale adoption
... CSF: relevant to users
... not FOL or whatever

<ChrisW> Frank's list: http://www.w3.org/2005/rules/wg/wiki/Goals%2C_Critical_Success_Factors_and_Requirements

Frank: may have subrequirements
... e.g. support for production rules

<FrancoisBry> Is the difference between 'critical sucess factor' and 'requirement' worth being discussed at length?

Frank: may well have multiple? CSFs
... try not to pre-guess solutions
... Paula's list nice as requirements, but not CSFs
... want to avoid confusion soundness and completeness
... write from customer's point of view

<Hassan> I agree with Frank that ultimately, RIF's success will depend on its usability

Frank: would add Paula's requirements to this list, linking to use cases important, desiderata are also nice
... borrowed some from Paula, but somewhat different style

Paula: wide-scale adoption implied by expressiveness
... sounds a bit too general
... applicable to all W3C work not just RIF

Frank: good point, but could see situation where someone has a tantrum saying feature X must bein RIF
... how does the group semi-rationally decide on including that feature in RIF?
... strong basis: if too expensive or nobody understands it

<sandro> q>

Frank: motherhood and apple pie, but useful as background to reinforce future discussions
... [explanation of motherhood and apple pie]

<sandro> "motherhood" and "apple pie" are, idiomatically, things which are so good that no one would ever argue against them.

Frank: cost of implementation is important consideration

<Hassan> Frank's making sense to me ...

Dave: ... [missed it - sorry]

ChrisW: goal 1 identical

<sandro> Dave: Frank's "1.2 Widescale Adoption" includes much of what I said in my e-mail to Paula

ChrisW: expressiveness is way to achieve widescale adoption
... consistency with W3C specs important goal for RIF
... would like to see more discussion at goal level

csma: Paula's goal equivalent to Dave Reynold's interchanges can be meaning preserving

<ChrisW> DaveR's proposal: http://lists.w3.org/Archives/Public/public-rif-wg/2006Apr/0005

csma: what is perceived difference between Dave's proposal for foundation and Frank's compatibility with W3C standards?

Sandro: pretty similar - perhaps W3C standards and vision
... include things that other Working Group have hinted at
... middle ground - may be too nebulous
... comments from group?

Dave: isn't this covered by logic preserving?
... semantics preserving and conformance covered under rule exchange

ChrisW: what about extensibility?

<PaulaP> +1 to Chris' comment on conformance

ChrisW: not just about soundness, but anticipating different extensions

Hirtle: perhaps need more goals - OWL had 8

Paula: need concrete goals, CSFs, and requirements

<DavidHirtle> one other thing: "Level of expressiveness" doesn't much sound like a goal, but I'm sure it could be rephrased

Frank: trying to avoid prejudging the outcome - focus on problem
... somewhat fuzzy, but otherwise get lost in weeds

<FrancoisBry> What are your "right questions", Frank????

csma: covering production rules is probably too fuzzy
... concrete version would be like concrete syntax for PRR

<DavidHirtle> (for comparison, OWL's goals: http://www.w3.org/TR/webont-req/#section-goals)

Frank: working with Paula - views are complementary and can be merged into 1
... hoping to merge

csma: statement requirements in such a way that they are useful for design

<sandro> I wonder if it's time to have Frank and Paula go off and come up with a consensus document between them? Do they have enough guidance from us?

<sandro> (Or do they need a 3rd?)

Hassan: also don't see contradiction - external usage is important
... reduce search space
... willing to help if needed

Dave: quite a few are compatible
... but emphasis on simplicity vs. specificity

<Hassan> It will require work

Dave: trivial merge loses criteria for specific expressivity requirements

Frank: buy-in from everybody needed as we progress
... some requirements may not fit into overall picture

csma: need more complete picture
... to identify incompatibilities

ChrisW: can Frank and Paula agree on goals?
... without requirements for now

Sandro: also include requirements that have been discussed

ChrisW: agree first on initial set of goals

ACTION (Frank, Paula): propose goals - within next couple days if possible - for discussion next week

<sandro> Frank/Paula goal recorded as http://www.w3.org/2005/rules/wg/track/actions/11

Technical Design

<ChrisW> http://lists.w3.org/Archives/Public/public-rif-wg/2006Apr/0068.html

Chris: design for extensibility
... out for over a week

csma: like the proposal quite a lot - authors did a good job
... orthogonal to extensibility mechanism - could be complemented by it
... staging design
... extensibility should also be discussed

ChrisW: doesn't provide extensibility

Hassan: like grammar-based, family of languages approach
... achieve extensions by adding new (composible) grammar rules

Harold: discussion 2 months ago: modular extensions not appropriate for layers of logic languages
... different than programming languages
... some ideas for adding negation, etc.

csma: extension mechanism should be added or clarified

<sandro> Harold -- there are two kinds of programming language extensions -- in unix Volume 2 and Volume 3. Volume 3 are just more C code and anyone could do -- they do not add expressive power. Volume 2 require system changes -- they do add expressive power.

<FrancoisBry> Xor is good for bewtter expression and better processing.

Kifer: hint at extensibility mechanism in taxonomy of semantic and syntactic features

<Uli> where would this attribute be attached to?

Hassan: how are we proceeding with this idea?

ChrisW: discussing, silence generally interpreted as agreement
... discuss via email, will add to agenda next week
... seem to be moving to a menu approach

Kifer: may have initial proposal for a taxonomy in a couple weeks

csma: email lists good for discussion, including counter proposals

<Harold> Sandro -- right, so I guess we meant Volume 2: in logic "system changes" correspond to transit to more expressive logics by adding syntax and semantics.

Hassan: working on new proposal


<PaulV> Bye


Summary of Action Items

See Action Tracker

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.127 (CVS log)
$Date: 2006/05/02 16:28:20 $