W3C

- DRAFT -

RIF Telecon 14 Aug 07

14 Aug 2007

Agenda

See also: IRC log

Attendees

Present
ChrisW, +95362aaaa, josb, csma, +1.443.539.aabb, Harold, Hassan_Ait-Kaci, AllenGinsberg, +1.512.773.aacc, PaulVincent, Jeff_Pan, Gary_Hallmark, DougLenat, Leora_Morgenstern, Stella_Mitchell, Sandro, IgorMozetic, Dave_Reynolds, MichaelKifer
Regrets
Chair
Chris Welty
Scribe
Paul Vincent

Contents


 

 

<ChrisW> scribenick: PaulVincent

<ChrisW> Scribe: Paul Vincent

<csma> continued

Action 331 continued

Action 323 continued

<ChrisW> http://lists.w3.org/Archives/Public/public-rif-wg/2007Jul/0197.html

<scribe> ACTION: minutes from June still required [recorded in http://www.w3.org/2007/08/14-rif-minutes.html#action01]

<Harold> Action 253 Experiment with modularization of Condition library (ARCH)

<DaveReynolds> 258 is similarly "pending discussion"

Minutes from July 24 accepted

Admin

<Harold> http://www.w3.org/2005/rules/wg/wiki/Arch/RIF_Components/RIF_Dialect_Structure

<csma> "to discuss"

Action 330 is pending discussion

Liason

No liaison reports

F2F

F2F: need registration page

<josb> +1 for Manhattan

<csma> after dinner, people should work!

F2F: dinner options: (1) around the mtg location (2) downtown to Manhatten -- preferences to Chris

<DaveReynolds> +1 for local with easier logistics

<AllenGinsberg> +1 Manhattan

F2F weekend activities / social --> comments to Chris

UCR

Chris: should revisit use cases / requirements addressed viz use cases
... UCR doc may need to be revised in light of changes to RIF Core / dialects

Harold: Core references are still valid

<DaveReynolds> We did discussion compliance, but we didn't agree on it

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

<DaveReynolds> No, we discussed metadata arch, but again no agreement

<DaveReynolds> No, it is not in the published version

<LeoraMorgenstern> This was one of the requirements that came up from UC 6.

<LeoraMorgenstern> But it arose from other use cases as well, as I remember.

Requirement: Merge rulesets discussed as requirement - new? published?

Allen: accepted at Athens F2F

<LeoraMorgenstern> It was discussed at the break-out session.

<LeoraMorgenstern> I don't remember if it was resolved at the plenary.

Sandro: was resolved in Nov F2F

<sandro> (from my irc log) px-#rif.log:Nov 05 09:44:50 <sandro> RESOLVED: RIF should support the ability to merge rule sets

<csma> see also http://www.w3.org/2005/rules/wg/wiki/F2F4

<sandro> (not the official record, but should make it easy to find in the official record, if needed.)

Chris: RIF team should review requirements

<LeoraMorgenstern> I just did mine (6)

<AllenGinsberg> http://www.w3.org/2005/rules/wg/wiki/D._Evaluation%3A_Use_Cases

Allen: see worked out examples (see above URI) for UC 2, 9

<LeoraMorgenstern> My worked out case is at http://www.w3.org/2005/rules/wg/wiki/UC6_Worked_Example

<LeoraMorgenstern> Harold's are at similarly named web pages.

<DaveReynolds> I did the UC8 example a long time ago and it predates any of the current BLD spec

<LeoraMorgenstern> Also Harold did 3 and 10 as well

Chris: all UC authors to link UC examples to UC page ready for discussion: 2, 3, 6, 7, 8, 9, 10
... these examples NOT for the UCR doc

<Harold> The "D" comes from the lettering in the earlier http://www.w3.org/2005/rules/wg/wiki/Extensible_Design

Chris: will add this page as a link to UC work in progress

BLD-RDF Compatibility

<ChrisW> http://lists.w3.org/Archives/Public/public-rif-wg/2007Aug/0012.html

<josb> direct link: http://www.w3.org/2005/rules/wg/wiki/Core/RIF-RDF_Compatibility

http://www.w3.org/2005/rules/wg/wiki/Core/RIF-RDF_Compatibility

Jos: explanation: 3 parts to discussion: (see index on page)

<josb> "a"^^int

<ChrisW> "ill typed" literals

Michael: ill typed treated as illegal syntax before now

<josb> "a"^^int

Christian: which RIF? Jos: based on BLD

Jeff: illegal syntax - differences between OWL and RDF for types

Michael: yes should be treated as illegal syntax

Dave: simple entailment is NP complete? Jos: Yes

<josb> S |= G

<josb> skolemize S

<josb> encode G as a query

Jos: syntactic subset but standard RDFS semantics

<josb> S |= G

<Zakim> csma, you wanted to ask what "illegal syntax" means for RIF

Christian: compliance issue for RIF translators: need to check for illegal syntax?
... should be added to compliance discussion at some point

Action 256 to be updated with syntax compliance

<Harold> Jos, in "Embedding Symbols", did you want tr to look like an identity here? tr maps typed literals "xxx"^^type-uri to symbols with the datatype type-uri: tr("xxx"^^type-uri) = "xxx"^^type-uri for every "xxx"^^type-uri in VTL.

<josb> yes

<josb> I did not propose to change the meaning

BLD-Frames

Chris: some reservations on syntax extension...

<DaveReynolds> Not *everyone* - I didn't

<JeffP> +1 agree with Dave and csma

<MichaelKifer> To the contrary: borrowing a construct from a foreign language that has a different semantics makes no sense

<IgorMozetic> +1 agree with Michael

<MichaelKifer> why restrict rdfs?

Michael: why restrict RDFS in RIF - may confuse users?

Jos: restricting no of axioms will remove complications of RDFS...

<ChrisW> what about rif:subClassOf rdfs:subPropertyOf rdf:Property

<Harold> Couldn't we define, in RIF, a reflexive R subClassEq S as Or( R subClass S R = S) ?

<ChrisW> Harold, reflexiveness is not the issue here, it is pretty minor

<ChrisW> oops, I meant rif:subClassOf rdfs:subPropertyOf rdfs:subClassOf

<Harold> Chris, didn't you worry about the reflexiveness of RDF's subClassOf, which RIF's ## doesn't have?

<Hassan> I agree with MK's argument...

Chris: need examples of both language issues. [None are to be claimed as intuitive]

<Hassan> Yes MK you are right!

Michael to propose in email the reason/example for specifying this behavior in RIF spec

<rifbot> Tracking ISSUEs and ACTIONs from http://www.w3.org/2005/rules/wg/track/

<Hassan> Dave - you are essentially saying that we should abstract all data models from RIF - just as I proposed originally using CLP

<Hassan> +1 with Dave's statement re. what is a good level of abstraction of data models vs. rules

<GaryHallmark> I like frames, classification, and datatypes in RIF because it makes RIF "closer" to my rule language and thus makes translation, esp. round-tripping, easier

<GaryHallmark> -1 for minimal. It makes translation too hard.

Chris: will send email to summarize discussion - will continue next call

<JeffP> +1

<AllenGinsberg> +1

<DaveReynolds> This comes down to implementing v. conforming - it makes conforming easier but implementing harder

<JeffP> bye

<DaveReynolds> bye

Summary of Action Items

[NEW] ACTION: minutes from June still required [recorded in http://www.w3.org/2007/08/14-rif-minutes.html#action01]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.128 (CVS log)
$Date: 2007/08/14 16:32:26 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.128  of Date: 2007/02/23 21:38:13
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Found ScribeNick: PaulVincent
Found Scribe: Paul Vincent
Default Present: ChrisW, +95362aaaa, josb, csma, +1.443.539.aabb, Harold, Hassan_Ait-Kaci, AllenGinsberg, +1.512.773.aacc, PaulVincent, Jeff_Pan, Gary_Hallmark, DougLenat, Leora_Morgenstern, Stella_Mitchell, Sandro, IgorMozetic, Dave_Reynolds, MichaelKifer
Present: ChrisW +95362aaaa josb csma +1.443.539.aabb Harold Hassan_Ait-Kaci AllenGinsberg +1.512.773.aacc PaulVincent Jeff_Pan Gary_Hallmark DougLenat Leora_Morgenstern Stella_Mitchell Sandro IgorMozetic Dave_Reynolds MichaelKifer
Agenda: http://lists.w3.org/Archives/Public/public-rif-wg/2007Aug/0010.html
Got date from IRC log name: 14 Aug 2007
Guessing minutes URL: http://www.w3.org/2007/08/14-rif-minutes.html
People with action items: minutes

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


[End of scribe.perl diagnostic output]