W3C

- DRAFT -

RIF WG TeleConf Minutes

11 Apr 2006

See also: IRC log

Attendees

Present
Peter Patel-Schneider, +4279aaaa, Chris Welty, Christian de Sainte Marie, Hassan Ait-Kaci, Markus Kroetzsch, Andreas Harth, Frank McCabe, Deborah Nichols, Ian Horrocks, Paula Patranjan, Jeff Pan, Igor Mozetic, Leora Morgenstern, David_Hirtle, Mohamed Zergaoui, Harold Bohley, John Hall, Sandro Hawke, Mike Dean, Mala Mehrotra, Jos De Roo, Giorgos Stoilos, Uli Sattler, Michael Kifer, Paul Vincent, Mohamed Zergaoui, Gary Hallmark, Alex Kozlenkov
Regrets
Chair
Chris Welty
Scribe
Uli Sattler

Contents


 

 

<ChrisW> updated minutes of last week's meeting can be found at: [http://lists.w3.org/Archives/Public/public-rif-wg/2006Apr/0036.html]

Hassan: points out that minutes don't make sense at some point(s) because the chronology is unclear

IanH: suggests that the chronolgy can be put back together like in a jiggsaw-puzzle

Hassan: doubts whether jiggsaw-puzzle can be accepted

Chris: explains how "+1" is always difficult,

<pfps> +1 :-)

Christian: advises scribes to add stuff as long as it's fresh in their mind.

<Deborah_Nichols> volunteers to send some amplification from my notes

Hassan no longer objects

<MoZ_> +1

<ChrisW> +1 to ian

IanH suggest to wait with "+1" until scribe has scribed

<MoZ_> +1 to make clearer +1

<Zakim> sandro, you wanted to comment on +1 in IRC

ChrisW: asks again for objections to accept the minutes?

Sandro: suggest to "annotate" the "+1"s with what you agree with (as in all the examples above)

<MarkusK> +1 to stating what one agrees to ;-)

ChrisW: minutes are accepted

Christian: raises issue that RIF telecon overlaps with SPARQL telecon because of Boston time/universal time difference

csma: has checked that there are no rules as to which time "pattern" should be used for WGs

csma: suggests that we could switch to UTC, but this will mean 1 hour later.

ChrisW: asks whether anybody wants to stay wrt constant?

<MoZ_> just prefer no overlapping

ChrisW: adds that this means "no summertime"

??? prefers this because otherwise, RIF members can't participate in SPARQL meeting

<csma> ACTION: Christian will investigate overlap with SPARQL [recorded in http://www.w3.org/2006/04/04-rif-minutes.html#action04] [CONTINUED]

JosDeRoo: finds summertime in general annoying

ChrisW: explains that there are two problems: changing to/from summertime and overlap with SPARQL telecon

<pfps> adds that there are places that don't go on summer time at all

ChrisW: isn't sure whether RIF WG's time is changeable

F2F

<csma> ACTION: chair to put design for extensibility and discussion of proposals on agenda for next telecon [recorded in http://www.w3.org/2006/03/28-rif-minutes.html#action14] [DONE]

ChrisW: repeats warning about filling flights to Dubrovnik

<Hassan> +1 to posting a page on travel

ChrisW: suggest to put up a page to register, Sandro?

<PaulaP> we can use the information found at http://www.w3.org/2005/rules/wg/wiki/F2F3/TravelTimes

Sandro: announces that, for registration and also participant's travel times (for co-ordination), he will set up a RIF registration page

<csma> ACTION: Sandro to set up registration page for F2F3 [recorded in http://www.w3.org/2006/04/11-rif-minutes.html#action03]

ChrisW: reminds that ESW conference organizers will organize transfer from the airport, and we need to co-ordinate with them

<AlexK> I can't access the travel times page--not allowed to view this page

<csma> it seems like the W3C site is down

<sandro> AlexK, you need to login to the wiki to get to that page. (BUt that page is now obsolete, if I understand correctly.)

<PaulaP> at moment it is not really clear whether the ESWC organizers will provide such a form

MKifer: are the visa requirements for Montenegro?

<AlexK> Croatia should be fine for most of us

MKifer: and visa for Croatia?

<MarkusK> Btw. Michael Sintek and I are going to Tivat

<PaulaP> there is information on the ESWC web page

<Deborah_Nichols> the country is Serbia and Montenegro. There is no Yugoslavia any more.

<IanH> We couldn't be so lucky as to be unable to go there!

<AlexK> http://www.southtravels.com/europe/serbiamontenegro/visa.html

Uli: ask your travel agent!

<igor> http://www.mfa.gov.yu/Visas/VisasR.htm

<AlexK> Have a look above

<igor> visa info for Montenegro

<PaulaP> http://www.mfa.gov.yu/Visas/VisasR.htm

Christian: seems most citizens won't need a visa...

ChrisW: reminds that next week is deadline for proposals for F2F4, for which there are currently two proposals

Liason

<PaulV> OMG PRR: no news from liason

Use cases and requirements

ChrisW: suggests to start by going through last week's actions

<csma> ACTION: Christian to send email to propose resolution that decidability is a non-requirement and gather relevant arguments before next telecon [recorded in http://www.w3.org/2006/04/04-rif-minutes.html#action14] [DONE]

ChrisW: ...this action on the decidability proposal is continued

<csma> ACTION: Frank will produce an initial diagram with existing constraints [recorded in http://www.w3.org/2006/04/04-rif-minutes.html#action10] [CONTINUED]

csma: was not comfortable with WG deciding "non-requirements", prefers requirements, etc
... so far, nobody submitted a requirement on decidability,

<csma> ACTION: MickaelK to extend page on pure prolog and give a precise definition (according to standard publications) [recorded in http://www.w3.org/2006/04/04-rif-minutes.html#action13] [DONE]

csma: suggests to come up with a requirement that makes decidability a non-requirement

<ChrisW> http://lists.w3.org/Archives/Public/public-rif-wg/2006Apr/0005.html

The structure of Dave's proposal

<DavidHirtle> I think he posted regrets on the wiki

<csma> ACTION: sandro to clarify meaning of sound and what is the requirement on RIF [recorded in http://www.w3.org/2006/04/04-rif-minutes.html#action11] [CONTINUED]

<csma> ACTION: Sandro to clarify whether sound reasoning constraint with unknown dialects is a requirement or a critical success factor [recorded in http://www.w3.org/2006/04/04-rif-minutes.html#action12] [CONTINUED]

ChrisW: Dave's proposal is a good example of a requirement... any comments on this?

csma: if we agree on a requirement/goal/etc., then they should be linked with other things on the web page

<csma> ACTION: Evan to publicize to ODM [recorded in http://www.w3.org/2006/03/28-rif-minutes.html#action09] [DONE]

sandro: we could use Dave's proposal as a "structural example" and have the others follow it

<pfps> Here we are getting into a problem similar to that brought up by Hassan - the log is not close to the timeline of the call

csma: disagrees

<sandro> Chris: Do we like this CSF Methodology?

chrisW: explains that we only meant the "abstract structure", not the specific proposal

FrankMcCabe: Sees some up-side down thinking and suggests to concentrate on goals before we go to requirements

csma: reminds that requirements can only come with critical success factors

<FrankMcCabe> suggests to mine goals from the charter.

MKifer: seconds FrankMcCabe's suggestions, and observes that Dave's proposal is vague in parts: e.g., "effective" and "sufficiently few dialects", and that he doesn't understand implications of several points in Dave's proposal

<PaulaP> +1 to more clear and detailed constraints

csma: suggests to add details and comments on the wiki

MKifer: suggests to avoid "general words"

csma: suggests to add examples for implementation of a requirement

ChrisW: asks what the relation is between hierarchy and goals and design constraints?

csma: suggests that, in Dave's proposal, there are "dependencies" between these concepts to be added, and that these dependencies will be visualised

ChrisW: sees a representation problem: we would need levels in hierarchy, to have structure between constraints

<Zakim> sandro, you wanted to ask if it's a tree or a lattice?

FrankMcCabe: explains how adding more justifications for a requirement makes it stronger

<sandro> Frank: It's important to know all the reasons why a Requirement is important (ie a lattice)

ChrisW: wants to know how to indicate direction of link?

<sandro> Sandro: I like "Motivation"

<sandro> paste link to what you're loking at, folks.

csma: explains how dependencies work between requirements and critical success factors...

sandro: suggest to make things easier by using an indented list for requirements

<Zakim> sandro, you wanted to suggest a simple way to do the structure

sandro: and to add critical design factors for each requirement explicitly

ChrisW: and link requirements to detailed goals? This would make structure more apparent

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

ChrisW: explains that we are using this requirement because it's complete, and reads out an example

csma's proposal at http://www.w3.org/2005/rules/wg/wiki/RIF_must_define_for_all_RIF_elements_a_default_behaviour_for_compliant_applications_that_do_not_know_how_to_process_it

ChrisW: comments on requirement that RIF-compliant applications must be able to handle rules in a predictable way

csma: suggests that we start discussing requirements with a less complex one

A simpler proposal at http://www.w3.org/2005/rules/wg/wiki/RIF_Core_must_cover_pure_Prolog

ChrisW: invites comments on "Rif core must cover pure prolog"? And reminds that we already discussed difficulty with "pure prolog"

csma: suggests to remove "core" from this requirement

<sandro> +1 to csma, it's too early to argue that RIF **Core** should cover pure Prolog

csma: would prefer that this is "Rif standard or extended Rif"

<GaryHallmark> +1 to remove core. Core seems by definition the intersection of the other requirements

csma: would prefer, in general, to be a bit more open or global wrt different RIFs

pfps: points out that even pure prolog is "very complicated", and suggests to decide whether/where we need to cover pure prolog, and suggests to replace "pure prolog" with "X", for a simpler logic. Also points out that it is inappropriate to talk about ISO standards here, and thus pure prolog is difficult

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

<MarkusK> http://www.w3.org/2005/rules/wg/wiki/Pure_Prolog

Igor: explains that all information is found at the prolog wiki page

<PaulaP> +1 to pfps on finding the motivation of using pure prolog here

pfps: questions whether enough information about occurs check can be found there

Igor: suggests to accept MKifer's suggestion to drop pure prolog

ChrisW: suggests to Igor to explain which rules we mean on the wiki, and points out that it's up to the requirement's authors as to whether they require prolog or Horn or...

Igor: suggests to replace pure prolog with "horn clauses"

ChrisW: points out that Horn clauses is not unambiguous either

sandro: asks for new suggestions for a new name to replace "pure prolog" in this requirement?

MKifer: repeats from his email: that pure prolog is horn rules plus ordering

csma: clarifies his previous remark, that requirements shouldn't target specific RIF variants, and that such a discussion should be left to a later point in time because we need to get a complete picture first

<csma> ACTION: Sandro and Igor to find a name+definition for the "pure prolog" requirement that does not mention "pure prolog" [recorded in http://www.w3.org/2006/04/11-rif-minutes.html#action10]

Harold: agrees with MKifer, and suggests "Ordered Horn clauses" because, in pure prolog, both the order of literals and rules is important

<ChrisW> ach harold

<pfps> An interesting page on ISO Prolog is http://pauillac.inria.fr/~deransar/prolog/docs.html. The page seems to indicate that the occurs check is somehow optional, at least in some situations.

<ChrisW> ach\k harold

<Harold> We can now define positively what we converged to mean by "Pure Prolog": "Ordered Horn clause".

<csma> ACTION: Sandro to discuss on email whether the "pure prolog" requirement can be replaced by a "horn logic" requirement or if we need both [recorded in http://www.w3.org/2006/04/11-rif-minutes.html#action11]

JosDeRoo: agrees with Harold, and points out how complicated prolog is (with occurs check and such like)

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

ChrisW: next design constraint: extended RIF ...

The next proposal at http://www.w3.org/2005/rules/wg/wiki/Extended_RIF_must_cover_FOL

FrankMcCabe: wants to know critical success factor of FOL

<Harold> "Ordered Horn Clauses" could be the result of annotating "Horn Clauses".

<AlexK> I'll just merely say about the FOL that it captures requirements in some applications

Leora: explains that FOL is important because of its expressiveness

<csma> +1 to FOL being impotant because of expressiveness

ChrisW: reminds that we are talking about "critical success factors"

<sandro> perhaps: goal = RIF must be useful for KR ?

Leora: will formulate critical success factor for FO and its expressiveness

<AlexK> do we have meta-definitions of Goal, Requirement, CSF?

MalaM: seconds "RIF should cover FOL"

MKifer: proposes that we apply the same standard to FOL as to pure prologue, and hence to explain exactly what we mean by FOL

<Hassan> +1 with Mike

csma: makes two comments: (1) "extended RIF" will cover everything since it will be extensible; and (2) agrees that FOL is important, and he thinks that it is a critical success factor

<MalaMehrotra> +1 to csma

Leora: asks (procedural) how we come to an agreement on what we mean by FOL

ChrisW: the champions decide what they mean, and then we revise it

MarkusK: wondered in how far "rule set" is compatible with FOL? Do we find "rules in" FOL? That is, can we view any FOL theory as a rule set?

<sandro> Markus: does using the word Ruleset mean we're only talking about part of FOL?

ChrisW: so we need to clarify the relation between "rule set" and FOL

sandro: points out that "FOL" should mean "arbitrary FOL theories"

<sandro> Frank: CSF might be "you have to support KR", FOL is not a CSF

FrankMcCabe: thinks that FOL is not a requirement because it is KR?

<sandro> Frank: as in, "if you can't do FOL, you can't do KR"

FrankMcCabe: doesn't think that FOL is a requirement (but a critial success factor?) since "if you can't do FOL, you can't do KR"

<sandro> Frank: Which aspects, eg universally quanitied variables, etc.

csma: clarified that Frank said that FOL was too specific to be a csf but not specific enough to be a requirement

<AlexK> FOL is quite important for KR appplications, exchanging those between companies is very useful

Frank: suggests that, for example, "we need to express existentially quantified variables" would be more like a csf

<csma> I clarified that Franck said that FOL was too specific to be a csf but not specific enough to be a requirement

<sandro> Frank: it's a short circuit to jump to FOL

Frank: "we need FOL" is to short since FOL is a technology

<Zakim> sandro, you wanted to argue that things like "sorted" don't matter for this purpose

<sandro> Sandro: I think "FOL" is the right level of granularity for this year.

Hassan: is concerned about people claiming that "FOL theories" are "rules"

<LeoraMorgenstern> +1 with sandro

Hassan: there are too many ways to describe the same theory (gentzen, sequents, etc), and wants to know whether RIF is about "(deduction) rule interchange" or "theory interchange"

<sandro> Sandro: I'm just talking about standard, textbook, FOL here, as something we need to support in an extension.

<sandro> LeoraMorgenstern: FOL is a cohesive whole, too, to it's a good point of granularity.

Leora: sees a risk in Frank's suggestion to chop FOL up into lots of "little" requirements since it enables people to choose subsets of these and obtain unintended such subsets

ChrisW: points out importance of clarity in the requirements formulation

Leora: points out the risk of mini-requirements

AOB

Summary of Action Items

[NEW] ACTION: Sandro and Igor to find a name+definition for the "pure prolog" requirement that does not mention "pure prolog" [recorded in http://www.w3.org/2006/04/11-rif-minutes.html#action10]
[NEW] ACTION: Sandro to discuss on email whether the "pure prolog" requirement can be replaced by a "horn logic" requirement or if we need both [recorded in http://www.w3.org/2006/04/11-rif-minutes.html#action11]
[NEW] ACTION: Sandro to set up registration page for F2F3 [recorded in http://www.w3.org/2006/04/11-rif-minutes.html#action03]
 
[PENDING] ACTION: Christian will investigate overlap with SPARQL [recorded in http://www.w3.org/2006/04/04-rif-minutes.html#action04]
[PENDING] ACTION: Frank will produce an initial diagram with existing constraints [recorded in http://www.w3.org/2006/04/04-rif-minutes.html#action10]
[PENDING] ACTION: sandro to clarify meaning of sound and what is the requirement on RIF [recorded in http://www.w3.org/2006/04/04-rif-minutes.html#action11]
[PENDING] ACTION: Sandro to clarify whether sound reasoning constraint with unknown dialects is a requirement or a critical success factor [recorded in http://www.w3.org/2006/04/04-rif-minutes.html#action12]
 
[DONE] ACTION: chair to put design for extensibility and discussion of proposals on agenda for next telecon [recorded in http://www.w3.org/2006/03/28-rif-minutes.html#action14]
[DONE] ACTION: Christian to send email to propose resolution that decidability is a non-requirement and gather relevant arguments before next telecon [recorded in http://www.w3.org/2006/04/04-rif-minutes.html#action14]
[DONE] ACTION: Evan to publicize to ODM [recorded in http://www.w3.org/2006/03/28-rif-minutes.html#action09]
[DONE] ACTION: MickaelK to extend page on pure prolog and give a precise definition (according to standard publications) [recorded in http://www.w3.org/2006/04/04-rif-minutes.html#action13]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.127 (CVS log)
$Date: 2006/04/11 16:31:44 $