8 Jun 2006

See also: IRC log


Present (remote): Mike Dean


Christian de Sainte Marie
Paul Vincent


PaulV is scribe for this session

F2F3 Budva now in progress

Agenda discussion

Goals and CSF

See http://www.w3.org/2005/rules/wg/wiki/UCR/Critical_Factors_Analysis

Christian is chairing session

Strawpoll on who agrees on leaving goals and CSFs as are...

poll: 15 agree, 1 against

<ChrisW> discussing the third goal: "compatibility with W3C standards"

Dave: another option is a goal ie a SW rule lang
...modifiy 3rd goal, or add 4th goal --- "Basis for a Semantic Web Rule Language"
...For example -- RDF Deduction rules. Not really motivated by W3C Consistency.

Sandro: would this be covered if RIF covered existing SW rule langs, eg N3 and JenaRules?

JosB: support need for discussion on SW rule lang

Dave: reply: RIF does not solve *which lang* to choose for SWeb
...people need to know which language to write rules in (for the SemWeb)

<josb> RIF should recommend THE language to use for the SemWeb

<sandro> Mike, if you want to talk, you might include my name to help make sure I notice and let the chair know.

Michael: building SWeb langs requires other groups involvement...

Christian summary: option is 1 a new SW rule lang or 2 interchange with a SW rule lang

Dave: maybe a SW rule lang is not a part of RIF, but specified langs could be defined as a subgroup of RIF; RIF WG needs to be clear on SW rule lang ambitions
... I'd like UCR to be clear that we have this ambition -- of being SemWeb Rule Language Basis

Christian: opinion: specifing SW Rule Lang is out of scope for RIF

Hassan: a SWeb rule lang needs to be RIF compliant, WHEN it is produced

Christian: if a semantic is defined as appropriate for SWeb rules, and semantic is covered in RIF, then RIF could specify this semantic (not language)

JosB: semantic = language (!)

<ChrisW> syntax&semantics = language

Sandro: (such a semantic) might not be called a "language" = small issue; problem with RIF as SW rule lang is that RIF != a single dialect

JosB: you propose a particular way to implement a SW Rule Lang - is this a RIF goal? SWeb may have other semantics...

<sandro> Sandro: Having RIF-WG recommend one selected dialect as the Semantic Web Rule Language -- would that address the need here? (Dave and Jos nod)

Sandro: Having RIF-WG recommend one selected dialect as the Semantic Web Rule Language -- would that address the need here? (Dave and Jos nod)

Christian: RIF has not been chartered to select the semantic for SW Rule Lang

Michael: dialect framework may be basis for SW rule lang requirement

Harold: could have a group discussing interchange vs human use etc; call for implementations would be different eg translators rather than execution engines

Sandro: general W3 is call for implementation (translator / executor not specified)

<sandro> Proposed Goal or CSF -- "Single dialect recommended for interchanging Semantic Web Rules" or "Recommend Semantic Web Rule Language"

<sandro> "Provide S.W.R.L" ?

Axel: Requirement is a subgoal with translators

Gary: RIF also used for exch between editors etc, eg consumer = human

Axel comment correction: I meant: Requirement for Translators/Wrappers from to systems vs. Requirement for implementations to accept RIF natively boils down to the same thing.

JosB: SW rule lang != fixed semantics; RIF in SWeb activity means RIF should be a SWeb rule lang.

JosB comment correction: SW Rule lang not necessarily a single fixed semantic

Harold: Michael's comments could be dialects for different semantics, translation via RIF: propose glossary entry for distinctions between human vs machine level etc

Hassan: agrees with Harold; no syntax in RIF, not human readable, not a language but a representation for languages

<sandro> (many disagree with Hassan)

Dave: use cases include human readable rules
... decision needed on dialects for SW rules lang
...It is not black-or-white, there is a whole spectrum: RIF will support all languages <----> one selected language

<josb> +1 to Dave!

<AxelPolleres> +1 to Dave

<mdean_home> +1 to Dave

Dave: charter calls for compatibility with SWeb: do we interpret this for SW Rule Lang?

Sandro: Proposed 4th goal: "Support Semantic Web"
...but that's implied by Use Cases, I think.

<josb> +1 to Sandro

Christian: computation characteristics eg features related more to inference mechanism than language are out of scope of RIF

Sandro: call for poll for 4th goal "support for SWeb"

ChrisW: ... but Dave asked for "SWeb rule lang"

<ChrisW> I said, "Dave has asked for a goal: the BASIS for a semantic web rule language"

Axel: Support for SemWeb and SemWeb Rule lang - these are the same...

Hassan: I support goal "Support Semantic Web", but not "Foundation for Semantic Web Rule Language"
...SWeb rule lang way out of scope / unfeasible

ChrisW: proposal needs further discussion; accepting goals does NOT preclude adding more

Christian: proposal: ACCEPT these goals NOT precluding adding new ones

<josb> so what is the use of accepting them?

<DavidHirtle> to settle on at least some

<ChrisW> Goals: Exchange of Rules, Widescale Adoption, and Consistency with W3C spec

<PaulaP> and get the UCR document ready soon :)

Dave: clarify: not ruling out vs objective for next draft?

<sandro> ChrisW: let's talk about the 4th goal more tomorrow

<sandro> (and maybe during breaks and stuff)

<scribe> ACTION: Dave to describe proposal for 4th goal in more detail by email [recorded in http://www.w3.org/2006/06/08-rif-irc]

Poll to accept these goals ...

Objections - none
Abstraining - FZI, DERI Innsbruck, JosefStefanInstitute

Approve: remaining (incl. DERI Galway)

RESOLUTION: accepted three goals as proposed above as goals for the RIF WG


Summary of Action Items

[NEW] ACTION: Dave to describe proposal for 4th goal in more detail by email [recorded in http://www.w3.org/2006/06/08-rif-irc]
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.127 (CVS log)
$Date: 2006/06/17 20:13:49 $