F2F10 Minutes

From RIF
Jump to: navigation, search

DRAFT -- Currently Under Review

See F2F10.

See also: IRC log

Day 1

Mike Dean: hello





Sandro Hawke: Hey Mike, the room isn't dialed in yet.
Mike Dean: ok
Axel Polleres: What about that one? http://cgi.w3.org/member-bin/irc/irc.cgi



Missing (guest): ChrisWelty [Scribe assist by Sandro Hawke]
Remote (guest): MikeDean [Scribe assist by Sandro Hawke]

(No activity for 10 minutes)

Sandro Hawke: http://www.w3.org/2005/10/Process-20051014/tr.html#last-call


(Scribe changed to Michael Kifer)

Axel Polleres: if anybody needs anything printed: send me a pdf per mail!

Sandro Hawke: last call means we are done

Sandro Hawke: after DTB is at last call, we can go for CR (candidate implementation). After that, we are not supposed to make any substantive changes to BLD.

At this point, we call for implementation

Christian de Sainte Marie: "Substantive" changes are not allowed to be made after a call for implementations.

Sandro Hawke: "A substantive change (whether deletion, inclusion, or other modification) is one where someone could reasonably expect that making the change would invalidate an individual's review or implementation experience. Other changes (e.g., clarifications, bug fixes, editorial repairs, and minor error corrections) are minor changes."
Sandro Hawke: q?

Sandro/Jos/csma: Example: changing the presentation syntax is a substantive change.

Sandro Hawke: http://lists.w3.org/Archives/Public/public-rif-comments/2008May/

(No activity for 7 minutes)

ACTION: jos add explanatory text to SWC and reply to http://lists.w3.org/Archives/Public/public-rif-comments/2008May/0000.html asking if it explains the matter sufficiently.


ACTION: jdebruij2add explanatory text to SWC and reply to http://lists.w3.org/Archives/Public/public-rif-comments/2008May/0000.html asking if it explains the matter sufficiently.


ACTION: jdebruij2 add explanatory text to SWC and reply to http://lists.w3.org/Archives/Public/public-rif-comments/2008May/0000.html asking if it explains the matter sufficiently. due wednesday

trackbot-ng: Created Action 480 - Add explanatory text to SWC and reply to http://lists.w3.org/Archives/Public/public-rif-comments/2008May/0000.html asking if it explains the matter sufficiently. due wednesday [on Jos de Bruijn - due 2008-06-02].
Adrian Paschke: As a side note: In the W3C HCLS group we are working on a URI note as a guide for a URI-based Naming System
Adrian Paschke: http://sw.neurocommons.org/2007/uri-note/
Axel Polleres: harold... just realized that the "Casts" section seems to have got accidently lost in DTB... will reinsert immediately...

discussion of Dan C's comments about use/mention of IRIs in RIF. Jos will respond to him.

Harold Boley: Axel, so, this discussion was quite helpful.
Harold Boley: About Dan's comment: Couldn't we distinguish two 'aritiy-specialized names' ABC/2 vs. ABC/3, here using Prolog-like name/arity notation?

discussion of Dan C's comment on the fact that BLD does not allow the same symbol to be used in different contexts. This will be a problem for merging of rules (one set may have a predicate as a 2-ary thing and in another as a 3-ary thing, for example).

Decided (guest): to keep discussing this issue.

Axel Polleres: harold, reinserted the CASTS subsection in DTB... as for casts from and to rif:iri, see http://lists.w3.org/Archives/Public/public-rif-wg/2008Mar/0011.html and the thread following that mail. not yet entirely resolved in my opinion, but I think jos had a reasonable proposal within that thread, I will dig it out, so far in the DTB draft it is only marked with a green comment.
Axel Polleres: DCC Chat (91.168.62.122)

(No activity for 9 minutes)


Jos de Bruijn: http://www.w3.org/2005/rules/wiki/Response_to_JC2


Hassan Aït-Kaci: yes
Hassan Aït-Kaci: it is sort of choppy though
Axel Polleres: Hassan, unfortunately, we miss additional microphones for the speaker phone, sorry. hope you can mostly hear us, speak up if no.
Hassan Aït-Kaci: thanks Axel ... we'll do our best ...

(No activity for 10 minutes)

Hassan Aït-Kaci: BTW - I have a question on using &rif;iri vs. rif:iri ... Is it timely to q for it?
Hassan Aït-Kaci: ok ...

ACTION: Jos to add section to SWC which guides people familiar with W3C Semantic Web specs about the surprising differences, eg rif:iri - due Wednesday


Hassan Aït-Kaci: thanks

ACTION: jdebruij2 to add section to SWC which guides people familiar with W3C Semantic Web specs about the surprising differences, eg rif:iri - due Wednesday

trackbot-ng: Created Action 481 - add section to SWC which guides people familiar with W3C Semantic Web specs about the surprising differences, eg rif:iri [on Jos de Bruijn - due 2008-05-28].


Harold Boley: Axel, great, you should bring that up in the DTB session later today.

Discussion of Jeremy's comments about rif:text, rif:iri. Decided: will add clarifications.

(No activity for 13 minutes)

Jos de Bruijn: http://www.w3.org/2005/rules/wiki/Response_to_JC3

(No activity for 9 minutes)

Sandro Hawke: http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0178.html

discussion of Dave R's comments. Most are editorial, which weren't discussed. Others will be discussed elsewhere (eg, in the FLD/BLD parts of the agenda).

Hassan Aït-Kaci: when do you reconvene?
Sandro Hawke: in 30 minutes.
Hassan Aït-Kaci: thanks


(No activity for 31 minutes)


Gary Hallmark: http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0126.html

(No activity for 6 minutes)

(Scribe changed to Harold Boley)

Implementations

Gary Hallmark: ORACLE UCR Use Case 1

Conclusion of rule uses a frame.

Christian de Sainte Marie: Modeled as a frame only because a little easier?

Adrian Paschke: In UCR I gave different formalization of this rules using relations, slots, frames and production rules

Gary Hallmark: OBR's best mapping of BLD goes to a frame because of its bean notion.

Adrian Paschke: http://www.w3.org/2005/rules/wiki/UCR#Negotiating_eBusiness_Contracts_Across_Rule_Platforms

Harold Boley: Use a frame as a conjunction?

Gary Hallmark: Can also use Group to attach conclusion results.

... splitting the And in the then part into two rules.

Jos de Bruijn: For only constants as slot fillers, frames could be used.

Christian de Sainte Marie: Why nesting of <And> <formula> <And>?

Gary Hallmark: Because of the principles of our general XML generator. Could be optimized away for the special case of BLD.

Sandro Hawke: What about the other direction?

Gary Hallmark: Yes. But will need to be worked out.

... We avoid showing durations to end users.

... For ex., days-between in OBR is not exposed to users.

Christian de Sainte Marie: Implementations should have access to such built-ins.

Michael Kifer: rif:new should really be rif:local (because rif:new is not unique).

... If it's a generator, it should be local.

... We need a standard way for handling skolems (via rif:local).

... Problem: The same symbol cannot be used in different contexts.

... Could change the definition of well-formedness. Not allow these skolems in equalities.

Gary Hallmark: Round-tripping with OBR would be a problem.

Sandro Hawke: Sent email about this last week.

Christian de Sainte Marie: 'Hidden conjunction" forbidden?

... multiple slots of a frame in the head.

... normally there is no conjunction in the head.

Michael Kifer: Right, maybe we should allow conjunction in the head, like disjunction in the body.

Christian de Sainte Marie: Can we change the document?

Adrian Paschke: Lloyd-Topor transformation can apply on the conjunctions in rule heads

Michael Kifer: Cannot guarantee change will not introduce errors.

PROPOSED: BLD will include Conjunction in the rule head

PROPOSED: BLD will include Conjunction in the rule head (the "then" part)

Sandro Hawke: +1
Mike Dean: +1
Axel Polleres: +1
Igor Mozetic: +1
Gary Hallmark: +1
Adrian Paschke: 0

0

Jos de Bruijn: 0
Michael Kifer: 0 [Scribe assist by Sandro Hawke]
Adrian Paschke: It is also a question of adoption of BLD, if it is too complex to implement BLD than people will not use it.
Adrian Paschke: of course conjunction can be transformed, but it adds some extra implementation effort
Jos de Bruijn: disjunction in head is a lot harder to implement
Jos de Bruijn: s/head/body/ (failed)

RESOLVED: BLD will include Conjunction in the rule head (the "then" part)

Adrian Paschke: that is true disjunction and also full logical equality are much much more complex

Harold Boley: May increase the burdon on BLD implementers to claim they have a direct (without rewriting) implementation of BLD.

Adrian Paschke: another question: mixing of frames, slots and relations allowed or not?

DTB review & publication

Axel Polleres: Symbol Spaces

Sandro Hawke: Maybe first motivate them before going into the mat.

s/ mat./ math./ (failed)

Sandro Hawke: The doc now doesn't use Curies?

Axel Polleres: It does (now).

Christian de Sainte Marie: q?

Sandro Hawke: Keep Fcts and Preds in the same namespace?

Axel Polleres: builtin-functions vs. builtin-predicates.

... different semantics.

Jos de Bruijn: But user does not see this.

PROPOSED: func: and pred: collapse to one, http:/www.w3.org/2007/rif-builtin#

Axel Polleres: Right.

Christian de Sainte Marie: What's the drawback of having two namespaces?

Jos de Bruijn: It's just more (unnecessary) namespaces.

Sandro Hawke: You have to remember both.

Axel Polleres: If we added a type Boolean, then we could consider collapsing the Fct and Pred namespaces.

Sandro Hawke: so predicate "less-than" and boolean-function "less-than" would need different URIs? [Scribe assist by Sandro Hawke]

Jos de Bruijn: Did we not decide that we will not have to versions (Fct and Pred) of the same builtin?

PROPOSED: keep the two namespaces, pred: and func: ....

Igor Mozetic: Whenever we tried to 'simplify' things in this way, it turned out there were later some complications. So we should keep the separation.

Sandro Hawke: two -- three (axel, igor, harold)
Sandro Hawke: one for all bultings -- two (adrian, jos)
objections (guest): ... none [Scribe assist by Sandro Hawke]
Sandro Hawke: make namespace names singular please [Scribe assist by Sandro Hawke]
Sandro Hawke: informally-resolved -- leave as is, with two namespaces.


Sandro Hawke: Hassan, we're just about to break for lunch  :)
Hassan Aït-Kaci: I am just back from mine ... ;-)

PROPOSED: remove language about all the subtypes of xsf:string

PROPOSED: remove language about all the subtypes of xsd:string being required

RESOLVED: remove language about all the subtypes of xsd:string being required

PROPOSED: instead of all-subtypes-of-decimal, have just Decimal, Integer, Long, .....

Sandro Hawke: RECESS FOR LUNCH.
Axel Polleres: RIF requires that all dialects include the following symbol spaces. Rule sets that are exchanged through RIF can use additional symbol spaces.
Axel Polleres: * xsd:string (http://www.w3.org/2001/XMLSchema#string)
Axel Polleres: * xsd:decimal (http://www.w3.org/2001/XMLSchema#decimal)
Axel Polleres: * xsd:integer (http://www.w3.org/2001/XMLSchema#integer)
Axel Polleres: * xsd:long (http://www.w3.org/2001/XMLSchema#long)
Axel Polleres: * xsd:time (http://www.w3.org/2001/XMLSchema#time).
Axel Polleres: * xsd:date (http://www.w3.org/2001/XMLSchema#date).
Axel Polleres: sorry.
Axel Polleres: :-)


Hassan Aït-Kaci: Enjoy your lunch! ...


(No activity for 20 minutes)



(No activity for 33 minutes)


Sandro Hawke: CONFERENCE CODE HAS CHANGED, SORRY



(Scribe changed to Adrian Paschke)

Continue DTB discussion

Axel Polleres: Goal for outsourcing into a seperate document to define the supported datatypes

Axel Polleres: Current DTB says it defines the supported BLD datatypes

Chris Welty: My understanding was that it is common to all RIF dialects

Sandro Hawke: Each dialect will have a subset of the DTB datatypes and builtins

Michael Kifer: or a superset

Sandro Hawke: DTB has to grow if new datatypes are needed in standardized dialects

Axel Polleres: it was moved / copied from FLD

Michael Kifer: yes, it has been copied from FLD to DTB

Chris Welty: It is a maintenance problem to add all new standard datatypes and builtins to DTB?

Sandro Hawke: only datatypes and builtins from standardize RIF dialects

Chris Welty: we agreed if a standard dialect needs a datatype or builtin function it needs to be in DTB

Chris Welty: If a standard dialect requires a datatype, then that datatype will be in DTB. DTB is extended as necessary. [Scribe assist by Sandro Hawke]
Michael Kifer: No, DTB is the minimal set -- all of DTB must be supported by all dialects. [Scribe assist by Sandro Hawke]

Michael Kifer: dialects may support additional datatypes

Jos de Bruijn: we only talk about standardized dialects

Jos de Bruijn: Every datatype that is required by a dialect specified by the RIF WG will be defined in an updated DTB. [Scribe assist by Sandro Hawke]

Michael Kifer: we would have mandatory and optional datatypes

Michael Kifer: so we'll have "Mandatory" and "Optional" datatypes in DTB. [Scribe assist by Sandro Hawke]

Chris Welty: put it in core

Chris Welty: we could do Mandatory by listing them in Core. [Scribe assist by Sandro Hawke]
Jos de Bruijn: when we say "dialect supports foo" we mean "the dialect REQUIRES all implementors to have their code support foo" [Scribe assist by Sandro Hawke]

Harold Boley: DTP is in some sense parallel to FLD

Michael Kifer: Organize built-ins in groups in DTB

Michael Kifer: let's organize them in groups and link to them from BLD

Sandro Hawke: organize them in levels, e.g. numeric built-ins level 1, level 2

PROPOSED: DTB will provide the menu of datatypes and builtins which diallect dialections can use, by reference, when they state which datatypes and builtins must be supported by implementations.

Michael Kifer: change to ... built-in predicated required by the RIF BLD

PROPOSED: DTB will provide the menu of datatypes and builtins which diallects can use, by reference, when they state which datatypes and builtins must be supported by implementations.

PROPOSED: DTB will provide the menu of datatypes and builtins which dialects can use, by reference, when they state which datatypes and builtins must be supported by implementations.

Sandro Hawke: +1
Jos de Bruijn: +1
Gary Hallmark: +1

+1

Harold Boley: +1

RESOLVED: DTB will provide the menu of datatypes and builtins which dialects can use, by reference, when they state which datatypes and builtins must be supported by implementations.

Igor Mozetic: +1

ACTION: axel to edit DTB to reflect its changed role with regard to DTB (eg in the Abstract)

trackbot-ng: Created Action 482 - Edit DTB to reflect its changed role with regard to DTB (eg in the Abstract) [on Axel Polleres - due 2008-06-02].

(No activity for 6 minutes)

Adrian Paschke: We need durations to implement some of the use cases

PROPOSED: add year-month and day-time duration, but NOT duration, to BLD.

Gary Hallmark: +1
Jos de Bruijn: http://www.w3.org/TR/xquery-operators/#dt-yearMonthDuration
Gary Hallmark: also need numeric-not-equal, numeric-less-than-or-equal, numeric-greater-than-or-equal

PROPOSED: add year-month and day-time duration, but NOT duration, to BLD, as in http://www.w3.org/TR/xquery-operators/#dt-yearMonthDuration

PROPOSED: add year-month and day-time duration, but NOT duration, to BLD (and of course DTB), as in http://www.w3.org/TR/xquery-operators/#dt-yearMonthDuration

PROPOSED: add xs:dayTimeDuration and xs:yearMonthDuration, but NOT duration, to BLD (and of course DTB), as in http://www.w3.org/TR/xquery-operators/#dt-yearMonthDuration

PROPOSED: add xs:dayTimeDuration and xs:yearMonthDuration, but NOT duration, to those required in BLD (and of course DTB), as in http://www.w3.org/TR/xquery-operators/#dt-yearMonthDuration

Sandro Hawke: +1
Axel Polleres: +1
Jos de Bruijn: +1
Igor Mozetic: 0
Sandro Hawke: DaveReynolds, normal number, code 26632.
Axel Polleres: and we should add to DTB: add the functions and operators from http://www.w3.org/TR/xpath-functions/

RESOLVED: add xs:dayTimeDuration and xs:yearMonthDuration, but NOT duration, to those required in BLD (and of course DTB), as in http://www.w3.org/TR/xquery-operators/#dt-yearMonthDuration

Axel Polleres: add the functions and operators from http://www.w3.org/TR/xpath-functions/

ACTION: axel to add the new duration subtypes to DTB

trackbot-ng: Created Action 483 - Add the new duration subtypes to DTB [on Axel Polleres - due 2008-06-02].


ACTION: kifer to make sure BLD includes the appropriate normative reference to DTB to reflect the inclusion of the duration subtypes

trackbot-ng: Created Action 484 - Make sure BLD includes the appropriate normative reference to DTB to reflect the inclusion of the duration subtypes [on Michael Kifer - due 2008-06-02].

Chris Welty: Gary's numeric remark

Gary Hallmark: less_than_or_equal ...

Axel Polleres: you typically have general builtins such as > < <= ...

Gary Hallmark: I only need numeric_greater_than_or_equal ... less_than_or_equal

Chris Welty: short cuts such as >= <= !=

PROPOSED: add builtin predicates to BLD and DTB: <= >= and != for numeric values (they amount to shortcuts, to avoid disjunction).

+1

PROPOSED: add builtin predicates to BLD and DTB: pred:numeric-less-or-equal, pred:numberic-greater-or-equal, pred:numberic-not-equal (they amount to shortcuts, to avoid disjunction).

Jos de Bruijn: +1
Sandro Hawke: +1
Gary Hallmark: +1
Igor Mozetic: +1

+1

RESOLVED: add builtin predicates to BLD and DTB: pred:numeric-less-or-equal, pred:numberic-greater-or-equal, pred:numberic-not-equal (they amount to shortcuts, to avoid disjunction).

Harold Boley: The names for '<=', '>=', etc. should be taken from http://www.w3.org/Submission/SWRL/#8

Sandro Hawke: have an editor from OWL and RIF to discuss about rif:text or using a owl datatype

Christian de Sainte Marie: BLD requires rif:text, so we have a dependency to a document which does not exist if me move it to another document

Chris Welty: keep in DTB

Harold Boley: add an editors note about it

Sandro Hawke: Plan -- DTB keeps rif:text for now, with editor's note, with Feature At Risk in BLD.

ACTION: Harold to add AT RISK editor's note in BLD explaining that the IRI identifying rif:text might change.

trackbot-ng: Created Action 485 - Add AT RISK editor's note in BLD explaining that the IRI identifying rif:text might change. [on Harold Boley - due 2008-06-02].
Chris Welty: "rif:text is an AT RISK feature. We expect a joint effort with ... " [Scribe assist by Sandro Hawke]
Sandro Hawke: (agreed)

Axel Polleres: Casts from rif:iri as defined in XML Schema

ACTION: Chris to open issue on casts to/from rif:iri

trackbot-ng: Created Action 486 - Open issue on casts to/from rif:iri [on Christopher Welty - due 2008-06-02].

Jos de Bruijn: cast functions are not defined, yet

Axel Polleres: see section 4.3

ACTION: Axel to change editor's note on casting rif:iri to normal open-issue style, link to new issue on it.

trackbot-ng: Created Action 487 - Change editor's note on casting rif:iri to normal open-issue style, link to new issue on it. [on Axel Polleres - due 2008-06-02].

ACTION: Axel to convert text about concat2, etc, into an editor's note about how the handling of arities is a strawman proposal not yet agreed upon by WG.

trackbot-ng: Created Action 488 - Convert text about concat2, etc, into an editor's note about how the handling of arities is a strawman proposal not yet agreed upon by WG. [on Axel Polleres - due 2008-06-02].

ACTION: axel comment out DTB 4.7 or otherwise make sure it doesn't end up in BLD

trackbot-ng: Created Action 489 - Comment out DTB 4.7 or otherwise make sure it doesn't end up in BLD [on Axel Polleres - due 2008-06-02].

(No activity for 5 minutes)

PROPOSED: Publish DTB as a FPWD once changes decided so far today are made (and reviewed by ...someone...)

Michael Kifer: I want a chance to go over it again, to look for clashes with BLD and FLD. [Scribe assist by Sandro Hawke]

PROPOSED: Publish DTB as a FPWD once changes decided so far today are made (and reviewed by ...someone...)

PROPOSED: Publish DTB as a FPWD once changes decided so far today are made (and reviewed by Chris)

Sandro Hawke: +1 (W3C)
Harold Boley: +1 (NRC)

+1 (REWERSE)

Gary Hallmark: +1 (Oracle)
Axel Polleres: +1 (DERI)
Jos de Bruijn: -0 (FUB)
Dave Reynolds: 0 (HP)
Chris Welty: +1 (IBM)


Chris Welty: +1 (ILOG)

RESOLVED: Publish DTB as a FPWD once changes decided so far today are made (and reviewed by Chris)


Chris Welty: ---break---

(No activity for 10 minutes)

Sandro Hawke: Dinner at McSwiggans, in city center.
Sandro Hawke: at 7:30
Chris Welty: please have your changes finished by then! [Scribe assist by Sandro Hawke]

(Scribe changed to Igor Mozetic)

striping, typed-tagged XML aka Rigid RDF

Sandro Hawke: -> http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0099.html details e-mail

we have nearly full-striped syntax, why not move further?

pro: small step, gives RDF compatibility

Harold Boley: We already have totally fully striped since the LAST f2f, after which I made this transition:
Harold Boley: http://lists.w3.org/Archives/Public/public-rif-wg/2008Mar/0108.html

pro: fallback mechanism is RIF

pro: more implementers can support

Adrian Paschke: It really blows up the syntax

con: even more verbose

Christian de Sainte Marie: main objective is not fully-stiped, but RDF compatible synta

Adrian Paschke: <Const><rdf:value rdf:datatype="&xsd:string">abc</rdf:value></Const> instead of <Const>abc</Const> will significantly blow up RIF documents
Harold Boley: It's not a small step: Sandro brought up:
Harold Boley: ...disadvantages:
Harold Boley: * it makes the XML syntax even more verbose
Harold Boley: * the elements from the RDF namespace can be confusing, even
Harold Boley: off-putting (especially to people who are allergic to RDF)
Harold Boley: * it prevents us from making arbitrary (non-striped) XML constructs
Harold Boley: that might be useful and elegant.
Harold Boley: * it's a course change, late in the day

con: "marketing" issue (some don't want RDF)

pro: "marketing" issue (if there is no dependency, then RDF compatibility is a plus)

Christian de Sainte Marie: for Alex - changing rigid RDF to typed-tagged-XML would solve his opposition

Various (guest): What does RDF Dependency mean? csma: "You can build an implementation without knowing anything about RDF" [Scribe assist by Sandro Hawke]
Christian de Sainte Marie: if that is not guaranteed, then the question is re-open. [Scribe assist by Sandro Hawke]

dependency on RDF namespace is _not_ RDF dependency

Sandro pro: it is self describing

Sandro Hawke: http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0099.html


Sandro Hawke: self-describing = deserialization back into "objects"

Sandro Hawke: Amusing notion: use "rif2" namespace prefix instead of "rdf" namespace prefix.

proposed changes by Sandro:

... add rdf:parseType=Collection to args

Christian de Sainte Marie: this proposal breaks full striping but we need RDF compatibility

... 2) add name under Var

... change from not-striped to striped

... 3) add value role under Const

Michael Kifer: issue is with symbolspace that are not datatypes

Michael Kifer: RDF friendliness brings semantics into syntax

Jos de Bruijn: we need to specify order in the syntax anyway

... 4) rif:iri inside Const is serialized differently

... add 4) uses native RDF support

... add 4) rif:iri and rif:text would disapear from XML serialization

Sandro Hawke: the only change is in the serialization, not in PS

Adrian Paschke: I was wondering what happens if you translate the schema of typed tagged XML into a object-oriented representation in Java using e.g. JAXB

Christian de Sainte Marie: an RDF parser will get RDF triples, not rules

Michael Kifer: is affraid that RDF semantics will be assumed

Christian de Sainte Marie: if this change is zero-cost it is advantageous for the community

... since people will get RDF triples for free and do whatever they want

... Sandro: 5) role tags may have to be in alphabetical order

Harold Boley: People who dont know that RDF is used only as syntax will need to use RDF semantics for that syntax, and will get the RIF semantics wrong!
Dave Reynolds: Harold - that makes no sense to me, how could a bunch of triples presenting the structure of a RIF document have anything to do with RIF semantics?
Harold Boley: Dave, not the structure of a RIF document, but the content of RIF rules.
Dave Reynolds: Harold - no this is just a serialization, no an embedding of RIF semantics in RDF
Axel Polleres: I made a proposal for RDF compatibility which I think was thought through more than what is discussed here now. It was independent from the XML syntax under the assumption that we wanted to keep the XML syntax separate from XML:
Axel Polleres: http://www.w3.org/2005/rules/wg/wiki/AbstractModel

Sandro Hawke: one can load RIF doc into triple store and extract rules by querying it

Axel Polleres: It even has a translation table from XML syntax to RDF.
Axel Polleres: Don't think we should do another unclean attempt to squeeze the XML syntax into RDF *directly*.
Harold Boley: Dave, but receivers of RDF documents by default will need to use Pat's semantics for RDF.

changes would require the following in BLD document:

translation tables, examples, XML schema

Axel Polleres: (BTW: the proposal with the abstract model was taken off the discussion... so, if we reopen the RDF discussion, then I would also want to reopen the discussion of the abstract model, which is more suitable as an RDF serialization of RIF, IMO.)
Harold Boley: We have a still (orally) undiscussed proposal on Identification and Metadata at Any Level:
Harold Boley: http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0036.html ff

(Scribe changed to Gary Hallmark)

Conformance

Christian de Sainte Marie: Conformance clause: Systems which produce or consume RIF documents in the BLD dialect MUST faithfully and completely follow the entire syntax and semantics of BLD as defined in this document.

Michael Kifer: must preserve entailments

Christian de Sainte Marie: but editors don't entail anything

Chris Welty: semantic v. syntax

Michael Kifer: editor doesn't actually DO anything

Sandro Hawke: e.g. biz rule editor imports rif into GUI

Jos de Bruijn: syntax checkers need to validate syntax

Michael Kifer: bijective mapping between 2 languages

... entailments include datatype conformance

Sandro Hawke: what if we encounter a datatype not in BLD?

... that would be an extension of BLD

... syntax check may pass

Jos de Bruijn: rule processor has set of datatypes, and can talk about differences w.r.t. DTB

Sandro Hawke: how to include xs:int in a rif document

... should it convert to xs:decimal, or ignore, or warn?

... should warn or reject

Jos de Bruijn: if conversion is lossless, then it is supported

Dave Reynolds: Sandro - suggest using example of xsd:double which isn't in RIF BLD either and is more of an issue

Michael Kifer: must be in DTB to have a semantics

in the above, "datatype" also includes associated builtins


Michael Kifer: could separate syntactic and semantic conformance

... syntactic subset can be larger


Sandro Hawke: hopes entire languages are translated to RIF, implying many extensions


... allows roundtripping


Christian de Sainte Marie: must be able to add (non-std) extensions to make RIF usable

Michael Kifer: don't require all datatypes are listed in DTB to have a valid RIF document

Christian de Sainte Marie: any other notions of conformance?

Jos de Bruijn: basic notion should be entailment

... first, a RIF consistency check (has a model)

Christian de Sainte Marie: but I can have an inconsistent RIF document



Jos de Bruijn: datatype conformance should be like OWL. Must support DTB, reject others

... for semantic conformance

Michael Kifer: distinguish producers from consumers

Christian de Sainte Marie: given agreement on datatypes (even if not in DTB), entailments are preserved

Michael Kifer: what does it mean to agree?


Jos de Bruijn: must decide what to do about datatypes not in DTB

Chris Welty: conformance defined for a single RIF processor, not for a communicating pair

Michael Kifer: for producer, must be mapping into RIF, and have right entailments

... for consumer, must reject datatypes you do not understand (which must not be in DTB)

Sandro Hawke: From MK's definition, it follows that (1) you MAY include other datatypes in your BLD, and (2) you MUST reject documents which include datatypes you don't (correctly) implement. [Scribe assist by Sandro Hawke]
Michael Kifer: A processor is "BLD Consuming" iff there is a mapping from BLD to the language of the processor that preserves all entailments. [Scribe assist by Sandro Hawke]
Dave Reynolds: If you go with this will there be any conformant implementation other than Flora-2?
Sandro Hawke: Or, any commercial implementation, DaveReynolds? It wouldn't be hard to make academic implementations.
Sandro Hawke: (I think a prolog implementation would be pretty easy, actually.)
Dave Reynolds: Sandro, with full equality ??
Dave Reynolds: Sandro, I don't think it is anything like that easy.
Sandro Hawke: Full equality is easy to implement by brute force, isn't it?
Sandro Hawke: I suspect an FOL prover (implementating paramodulation) could do pretty well, if it has a plug-in architecture for the built-ins. I think otter does, at least.

(No activity for 8 minutes)

much discussion formalizing the conformance statement...

Chris Welty: Given a set set of datatypes D and externals E, a RIF processor is a conformant BLD+D,E consumer if there is a mapping from BLD to the language of the processor that, given a BLD+D,E document, preserves all entailments as specified in the BLD semantics.
Chris Welty: Given a set set of datatypes D and externals E, a RIF processor is a conformant BLD+D,E producer if there is a mapping from the language of the processor to BLD that, given a document in the processor language, preserves all entailments as specified by that processor.


Dave Reynolds: must a consumer be a complete BLD implementation?

Sandro Hawke: yes

Dave Reynolds: concerned about equality

Christian de Sainte Marie: equality in head may be "at risk"

Dave Reynolds: used to have a notion of BLD being a superset

... a conformant impl could be a subset

Christian de Sainte Marie: is this the notion of profiles?

Christian de Sainte Marie: maybe we'll need profiles since we no longer have CORE

Christian de Sainte Marie: maybe solution is to remove equality from BLD

Sandro Hawke: if we need a dialect between CORE and BLD, we'll create one (but I hope not)

... e.g. mini-BLD

Christian de Sainte Marie: if we don't get implementations, we'll have to revisit

Christian de Sainte Marie: does marking some hard things as "at risk" satisfy Dave?

Dave Reynolds: yes

Chris Welty: maybe the conformance statement is "at risk"

Michael Kifer: could have levels of conformance, and lower the bar as needed

Harold Boley: Chris, Yes, let's look at OWL 1 (and 2) Conformance (Normative): http://www.w3.org/TR/owl-test/#conformance

Sandro Hawke: let's talk about syntax extensibility

... must reject syntax you don't understand

Sandro Hawke: http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0143.html

... is somewhat counter to the XML philosophy

... but can't ignore NOT, e.g.

... and we don't require "fallback" processing

Naming

Harold Boley: We can have a notion of (D,E,T)-conformance, where (D,E) is as before and T is the set of (XTAN-) transformation rules used.

Sandro Hawke: doesn't like upper/lower

Christian de Sainte Marie: in PRD, object/class and subclass/class


Christian de Sainte Marie: wants different tags for different content


... makes PS->XML mapping easier

Christian de Sainte Marie: I want each role tag to always contain the same thing -- it would make the XML syntax table much easier to understand,. [Scribe assist by Sandro Hawke]

... because it is not context dependent

Sandro Hawke: "instance" and "class", "subclass" and "superclass"
Sandro Hawke: "sub" and "super". [Scribe assist by Sandro Hawke]

Sandro Hawke: sub/super for SubClass

Sandro Hawke: [ agreed ]
Hassan Aït-Kaci: why not sub/sup ?

Harold Boley: instance/class for Member

Sandro Hawke: you're being facetious, Hassan?
Hassan Aït-Kaci: But I do not really care ... ;-)

Hassan, I though you don't like abbrevs

Hassan Aït-Kaci: It WAS a joke...

ACTION: Harold update BLD to change lower/upper to instance/class and sub/super.

trackbot-ng: Created Action 490 - Update BLD to change lower/upper to instance/class and sub/super. [on Harold Boley - due 2008-06-02].

Chris Welty: doesn't like '->' in frames

... too much like implication

... also -> used for named args

Christian de Sainte Marie: but PS is tomorrow

Christian de Sainte Marie: prefer attribute instead of key

Hassan Aït-Kaci: An attribute is the pair Key->Value

... for the frame case

Hassan Aït-Kaci: There is Frame Attribute (TERM -> TERM) and a Term Attribute (ID -> TERM)




Harold Boley: principle of context-independent role names not maintainable in FLD

Hassan Aït-Kaci: XML tags should be independent of BNF

... distinguish grammar from language

... former is not unique

Harold Boley: We can easily change invisible names such a TERM (not shown in XML instances), but not visible names such as Const, Var, and Expr (shown in XML instances).
Harold Boley: In FLD, we can have both <key> <Const> ... </Const> </key> and <key> <Expr> ... </Expr> </key>.
Hassan Aït-Kaci: when I implemented, I needed a BNF, not EBNF, so I had to modify the grammar, and then the nice grammar properties are gone. [Scribe assist by Sandro Hawke]

... in prototype, had to modify EBNF for jacc, and this ripples through the XML mapping if they are linked

Harold Boley: These two -- <key> <Const> ... </Const> </key> and <key> <Expr> ... </Expr> </key> -- can arise from a single <key> <Var> ... </Var> </key>.
Harold Boley: (Depending on binding the Var to a Const or an Expr.)
Harold Boley: http://www.w3.org/2005/rules/wiki/BLD#Translation_of_the_RIF-BLD_Condition_Language
Hassan Aït-Kaci: q+

Harold Boley: many XML tags have a different PS token, e.g. ?/Var, ^^/Const

Hassan Aït-Kaci: q-
Chris Welty: adjourn
Hassan Aït-Kaci: +1 to adjourn
Hassan Aït-Kaci: bye ...



Day 2

See also: IRC log




Jos de Bruijn: josb has joined #rif


Andreas Harth: aharth has joined #rif
Axel Polleres: AxelPolleres has joined #rif
Chris Welty: ChrisW has joined #rif


Meeting (guest): RIF F2F10 (Day 2), Galway, Ireland, May 27, 2008 [Scribe assist by Chris Welty]



Chris Welty: is anyone waiting to join?
Hassan Aït-Kaci: I am on 26631 - is this ok?
Sandro Hawke: ChrisW, I already starts conf1with a longer lifetime. maybe it'll still work.
Chris Welty: ok
Jos de Bruijn: it's 26632
Hassan Aït-Kaci: ok I'll redial
Sandro Hawke: no!
Chris Welty: no
Harold Boley: Harold has joined #rif
Chris Welty: hassan


Chris Welty: stay on
Mike Dean: i'm also on 26631
Chris Welty: stay there, we will join in a minute
Jos de Bruijn: my mistake
Chris Welty: hassan, rejoin at 26631
Dave Reynolds: DaveReynolds has joined #rif
Hassan Aït-Kaci: ok



John Hall: johnhall has joined #rif
Chris Welty: ChrisW has changed the topic to: RIF F2F10 Code 26631


(Scribe changed to Andreas Harth)

Agenda review

this session: presentation syntax, especially shortcuts

Christian de Sainte Marie: Issue 56

Gary Hallmark: GaryHallmark has joined #rif

Jos de Bruijn: which shortcuts to define

Axel Polleres: sent out proposal yesterday about grammar

Axel Polleres: http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0191.html
Adrian Paschke: AdrianP has joined #rif

Presentation syntax

Igor Mozetic: IgorMozetic has joined #rif

Axel Polleres: do we want to have string lang tag? his proposal is to use grammar from sparql spec

Adrian Paschke: Shortcuts e.g.:
Adrian Paschke: "foo:bar"^^rif:iri <foo:bar> IETF's angular bracket notation
Adrian Paschke: "purchase"^^rif:local purchase locality by default

Jos de Bruijn: just use absolute IRIs not relative IRIRefs

Adrian Paschke: "a b c"^^xsd:string "a b c" Full: quotes are part of ^ syntax
Adrian Paschke: "10"^^xsd:integer 10 as in programming languages
Igor Mozetic: A table with tentative shortcuts is in the following minutes

Jos de Bruijn: in the definition, all w3c standards have full IRIs in their spec

Igor Mozetic: http://lists.w3.org/Archives/Public/public-rif-wg/2008Apr/att-0033/08-rif-minutes.html

Chris Welty: rif doesn't need to know about curis

... do preprocessing to expand iris

Axel Polleres: basically just added the lang tag, ok to use iri, ok to use prefix def in presentation syntax

... still open are string and escaping, we should use the sparql syntax here as well

Axel Polleres: syntax allows to only write the prefix, which is just resolved to the iri of the prefix

Christian de Sainte Marie: decision to make: does it make sense, whether to use iri or iriref, how to address escaping

... any objections?

Michael Kifer: not use anglebrackets

Mike Dean: q+

Axel Polleres: we addressed that in the last telecon, just added last line

Michael Kifer: problem is in one place it's really an iri, in other place it's just a marker

Mike Dean: should have ways to use curis and iris

Christian de Sainte Marie: issue is string^^<IRI>

... for const

Mike Dean: s/iris/absolute iris/ (failed)

Michael Kifer: here, it's just a symbol that looks like a iri but it's just a constant

Christian de Sainte Marie: it's mentioning the iri but not using it

Michael Kifer: symbol space is not identified by iri, but just a symbol

Christian de Sainte Marie: you want a different syntax for the different roles of an iri?

Axel Polleres: do we open a can of worms here? for sake of readablilty, i'd buy the conceptual ambiguity

Hassan Aït-Kaci: The point of this is not prettyness but to help Axel have less pain ... ;-)

Axel Polleres: with current proposal we're compatible with n3 syntax

Michael Kifer: the aliases are not required to be iris

Christian de Sainte Marie: need to change it either in bld or here

Christian de Sainte Marie: must the aliases be iris or not?

Dave Reynolds: It's very hard to follow this - where is this alias proposal?

Michael Kifer: curis could be still too long

Hassan Aït-Kaci: Something like ${VAR} in yacc unix etc...
Sandro Hawke: sandro has joined #rif
Axel Polleres: http://www.w3.org/2005/rules/wiki/DTB#Symbol_Spaces

PROPOSED: remove aliases for datatypes in BLD

PROPOSED: remove aliases for symbol space identifiers in BLD

PROPOSED: remove aliases for symbol space identifiers in RIF

Sandro Hawke: does that leave us with this syntax? still possible that say two datatype iris are equal?

Jos de Bruijn: You could have two datatypes that are exactly the same, except for their identifier. [Scribe assist by Sandro Hawke]
Michael Kifer: right. [Scribe assist by Sandro Hawke]
Sandro Hawke: (nods) [Scribe assist by Sandro Hawke]

Christian de Sainte Marie: any more questions?

RESOLVED: remove aliases for symbol space identifiers in RIF

Sandro Hawke: "chat"@en
Sandro Hawke: as short for "chat@en"^^rif:text

Axel Polleres: next: string with lang tag

Christian de Sainte Marie: Const ::= .... | STRING LANGTAG

Christian de Sainte Marie: lang tag in rif:test is mandatory

Axel Polleres: Proposed chane in DTB here: http://www.w3.org/2005/rules/wiki/index.php?title=DTB&diff=1746&oldid=1738

Chris Welty: no shortcut for non-lang-tagged strings?

Sandro Hawke: there should be

PROPOSED: add Const ::= STRING LANGTAG (allowing "chat"@en as short for "chat@en"^^rif:text) and "Const ::= STRING" (allowing "chat" as short for "chat"^^xs:string).

Michael Kifer: MichaelKifer has joined #rif

Michael Kifer: when we lateron introduce modules @ is the right symbol for it

Michael Kifer: how often we use that bit in the presentation syntax?

Jos de Bruijn: for the examples in the doc

Gary Hallmark: we have shortcuts for obscure features, but not for strings and integers

PROPOSED: modify Presentation Syntax to incliude "Const ::= STRING" (allowing "chat" as short for "chat"^^xs:string).

PROPOSED: modify Presentation Syntax to include "Const ::= STRING" (allowing "chat" as short for "chat"^^xs:string).

Sandro Hawke: +1

RESOLVED: modify Presentation Syntax to include "Const ::= STRING" (allowing "chat" as short for "chat"^^xs:string).

Christian de Sainte Marie: any objection?

Gary Hallmark: +1
Igor Mozetic: http://lists.w3.org/Archives/Public/public-rif-wg/2008Apr/att-0033/08-rif-minutes.html
Harold Boley: http://lists.w3.org/Archives/Public/public-rif-wg/2008Apr/att-0033/08-rif-minutes.html
Axel Polleres: NumericLiteral, cf. http://www.w3.org/TR/rdf-sparql-query/#rNumericLiteral
Axel Polleres: (would be the other shortcut we could allow)

topic now: abridged presentation syntax

Christian de Sainte Marie: in minutes there's a table for abridged syntax

Sandro Hawke: how to distinguish between integer and long?

Axel Polleres: for numerical literals the sparql spec could serve as example

Axel Polleres: double not in symbol spaces

PROPOSED: add xsl:double as a required symbol space

Axel Polleres: Can I paste the proposed grammar in the chat?
Axel Polleres: s/xsl/xsd/ (failed)

Jos de Bruijn: why add double?

Sandro Hawke: note that it is not in the list in the charter. [Scribe assist by Sandro Hawke]

Gary Hallmark: important for lot of engineering applications

Sandro Hawke: +1
Jos de Bruijn: 0
Hassan Aït-Kaci: 0
Chris Welty: 0
Dave Reynolds: 0
Gary Hallmark: +1
Mike Dean: +1
Axel Polleres: +.5
Harold Boley: +1
Adrian Paschke: +1

RESOLVED: add xsd:double as a required symbol space

Axel Polleres: ANGLEBRACKIRI  ::= '<' IRIRef '>'
Axel Polleres: STRING  ::= '"' UNICODESTRINGWITHOUTQUOTES '"'
Axel Polleres: CURIE  ::= PNAME_LN | PNAME_NS
Axel Polleres: Const  ::= ANGLEBRACKIRI
Axel Polleres: | CURIE
Axel Polleres: | STRING '^^'ANGLEBRACKIRI
Axel Polleres: | STRING '^^' CURIE
Gary Hallmark: +1.0e0
Axel Polleres: | STRING LANGTAG
Axel Polleres: | STRING
Axel Polleres: | NumericLiteral
Axel Polleres: PNAME_LN, cf. http://www.w3.org/TR/rdf-sparql-query/#rPNAME_LN
Axel Polleres: PNAME_NS, cf. http://www.w3.org/TR/rdf-sparql-query/#rPNAME_NS
Axel Polleres: LANGTAG, cf. http://www.w3.org/TR/rdf-sparql-query/#rLANGTAG

PROPOSED: add shortcut e notation for double

Hassan Aït-Kaci: short ... cat? ;-)

sorry :)

s/shortcat/shortcut (succeeded, 3 lines ago)

discussion about grammar

Sandro Hawke: why do we need long in presentation syntax and datatypes

Chris Welty: get positive negative integer and decimals

Axel Polleres: we can add hooks to link into the sparql grammar

Michael Kifer: prefered to be self-contained, what if sparql changes?

PROPOSED: import NumericLiteral from SPARQL http://www.w3.org/TR/rdf-sparql-query/#rNumericLiteral giving us INTEGER, DECIMAL, and DOUBLE.

PROPOSED: import NumericLiteral from SPARQL http://www.w3.org/TR/rdf-sparql-query/#rNumericLiteral giving us INTEGER, DECIMAL, and DOUBLE to the Presentation Syntax

Adrian Paschke: +1
Gary Hallmark: +1
Axel Polleres: +1
Sandro Hawke: +1
Dave Reynolds: +1
Hassan Aït-Kaci: +1
Chris Welty: +1
Igor Mozetic: +1

PROPOSED: reuse NumericLiteral from SPARQL http://www.w3.org/TR/rdf-sparql-query/#rNumericLiteral giving us INTEGER, DECIMAL, and DOUBLE to the Presentation Syntax

Harold Boley: +1

RESOLVED: reuse NumericLiteral from SPARQL http://www.w3.org/TR/rdf-sparql-query/#rNumericLiteral giving us INTEGER, DECIMAL, and DOUBLE to the Presentation Syntax

Mike Dean: +1


Christian de Sainte Marie: open is rif:local by default?

Axel Polleres: Can we reuse http://www.w3.org/TR/rdf-sparql-query/#rPN_CHARS_BASE ???

Michael Kifer: starts with letter or underscore

PROPOSED: modify presentation syntax so that identifiers (as in C or Java - starting with letter or underscore, allowing digits later), are shortcut for rif:local

Michael Kifer: followed by alphanumeric

Axel Polleres: http://www.w3.org/TR/2006/REC-xml-names11-20060816/#NT-LocalPart
Chris Welty: This is just a shortcut -- no UNICODE -- if you want that, use the ^^rif:local form. [Scribe assist by Sandro Hawke]

Sandro Hawke: most languages set aside a set of keywords for this

Dave Reynolds: How about having locals start with _ to avoid that?

Gary Hallmark: is not code, but presentation syntax

Axel Polleres: Locals shortcut might conflict with keywords... e.g. External( ... )

Michael Kifer: keywords start with dollar sign?

Michael Kifer: keywords like import and external could start with a $ [Scribe assist by Sandro Hawke]

Sandro Hawke: leave out rif:local shortcuts?

PROPOSED: modify presentation syntax so that alphanumeric identifiers starting with "_" are shortcut for rif:local (so _foo is short for "foo"^^rif:local)

Hassan Aït-Kaci: +1
Sandro Hawke: +1

Harold Boley: other character than underscore, use dot like in linux

Sandro Hawke: code convention in java and c for local variables

Adrian Paschke: +1
Dave Reynolds: +1
Harold Boley: +1
Igor Mozetic: +1
Sandro Hawke: ... or local (private) terms, not variables
Jos de Bruijn: 0
Axel Polleres: +1
Gary Hallmark: +0
Mike Dean: +0

Michael Kifer: appreciate the point that "_" prefix represents local things, but in programming languages it's part of the name

... maybe single quotes?

Adrian Paschke: "purchase"^^rif:local becomes _purchase

RESOLVED: modify presentation syntax so that alphanumeric identifiers starting with "_" are shortcut for rif:local (so _foo is short for "foo"^^rif:local)

Axel Polleres: current status of the grammar:
Axel Polleres: Const  ::= STRING '^^'ANGLEBRACKIRI
Axel Polleres: | STRING '^^' CURIE
Axel Polleres: | ANGLEBRACKIRI -> shortcut for rif:iri
Axel Polleres: | CURIE -> shortcut for rif:iri
Axel Polleres: | STRING LANGTAG -> shortcut for rif:text
Axel Polleres: | STRING -> shortcut for xsd:string
Axel Polleres: | NumericLiteral -> shortcut for xsd:integer, xsd:decimal, xsd:double
Axel Polleres: | '_' LocalName -> shortcut for rif:local
Axel Polleres: PNAME_LN, cf. http://www.w3.org/TR/rdf-sparql-query/#rPNAME_LN
Axel Polleres: PNAME_NS, cf. http://www.w3.org/TR/rdf-sparql-query/#rPNAME_NS
Axel Polleres: LANGTAG, cf. http://www.w3.org/TR/rdf-sparql-query/#rLANGTAG
Axel Polleres: NumericLiteral, cf. http://www.w3.org/TR/rdf-sparql-query/#rNumericLiteral
Axel Polleres: LocalName, cf. http://www.w3.org/TR/2006/REC-xml-names11-20060816/#NT-LocalPart

Christian de Sainte Marie: next: lang tag and string

... proposal is to have lang tag seperated from string

PROPOSED: modify Presentaton Syntax adding Const ::= STRING LANGTAG (allowing "chat"@en as short for "chat@en"^^rif:text)

Michael Kifer: concern that @ symbol could be used for modules

Sandro Hawke: MichaelKifer wants to keep @ operator reserved for modules
Hassan Aït-Kaci: q+ to ask a couple of questions on namespaces and entities (when topic gets to that)
Dave Reynolds: That would be confusing, can you use something else for modules?
Mike Dean: q-
Sandro Hawke: the lexer can distinguish normal @ from @-preceded-by-double-quoted-string [Scribe assist by Sandro Hawke]
Axel Polleres: why not have riftext:en, riftext:de, etc. as separate symspaces? :-)

Hassan Aït-Kaci: sometimes we use rif:id, sometimes &rif;id

Dave Reynolds: q+
Axel Polleres: q+
Dave Reynolds: q-
Hassan Aït-Kaci: q-

Christian de Sainte Marie: in this session we talk about presentation syntax

Hassan Aït-Kaci: ok I will ask later
Sandro Hawke: "chat"^^riftext:en
Axel Polleres: "xyz@en"^^rif:text --> "xyz"^^riftext:en

Jos de Bruijn: problem here is we would define iris for languages which is in an rfc and can change

Dave Reynolds: Hassan - in the XML there would only be real IRIs, not Curies. To make XML more readable it is quite common to define entities but the processing of that is done by the XML parser, the translator code would only see the exapansion. See http://www.ibm.com/developerworks/xml/library/x-think9.html as an example. This is one reason I asked that the XML examples in the doc be complete, not fragments.

Sandro Hawke: ok, we could define it openly as a pattern

Christian de Sainte Marie: rif:text might disappear into a document common for rif and owl

Christian de Sainte Marie: we don't discuss that further now

Sandro Hawke: tabled.
Hassan Aït-Kaci: Hassan -> DaveReynolds - Thanks.
Christian de Sainte Marie: So we'll have rif:text marked AT RISK, and the shortcut syntax for it is still considered fluid. [Scribe assist by Sandro Hawke]
Axel Polleres: we could have the shortcut stable, .... [Scribe assist by Sandro Hawke]

Christian de Sainte Marie: next topic irirefs vs iris

... irirefs or absolute iris

Jos de Bruijn: issue with relative iris: you need a base iri to resolve relative iris to absolute iris

... need mechanism to specify base iris if we stay with irirefs

... benefit of irirefs is that

... they're shorter

Dave Reynolds: It is an issue in the XML syntax - should make sure that is included in the discussion.

Axel Polleres: relatively simple, could have a base iri in the preamble

Dave Reynolds: Consider the IRI for identifying a rule to which metadata will be attached. Shouldn't that be relative?
Harold Boley: XML Base Recommendation: http://www.w3.org/TR/xmlbase/#resolution

PROPOSED: IRIs in the XML syntax can be relartive-IRIs (they need not be absolute)

Mike Dean: s/relartive/relative/ (failed)

PROPOSED: IRIs in the XML syntax can be relative-IRIs (they need not be absolute)

Sandro Hawke: +1
Dave Reynolds: q+
Hassan Aït-Kaci: +1
Sandro Hawke: q- AxelPolleres
Axel Polleres: q-
Sandro Hawke: ack DaveReynolds

Dave Reynolds: where exactly use relatvie iris in the XML?

Sandro Hawke: thought about const but maybe there's other places too

Christian de Sainte Marie: table that queston for xml syntax, but focus on presentation syntax

Hassan Aït-Kaci: But I thought the proposed statement mentions that it is for XML syntax ?
Hassan Aït-Kaci: ok

PROPOSED: In Presentation Syntax, the IRIs in rif:iri Consts can be relative. A "base" directive will be added to the preamble.

Dave Reynolds: q+
Sandro Hawke: ack DaveReynolds
Axel Polleres: +1

Dave Reynolds: why does that matter when describing examples

... n3 does not have base directive?

Sandro Hawke: <./foo/bar.x>
Chairs (guest): discussion tabled -- not resolved [Scribe assist by Sandro Hawke]
Christian de Sainte Marie: It is left ambiguous in the presentation syntax. [Scribe assist by Sandro Hawke]
Dave Reynolds: (on irc) Sorry, I was wrong N3 does have an @base FWIW [Scribe assist by Sandro Hawke]

Axel Polleres: propose to adopt syntax proposal from sparql - use backlash to escape quotes inside strings

Mike Dean: s/backlash/backslash/ (failed)

PROPOSED: Adopt SPARQL convention for using backslash to allow quotes within quoted strings.

PROPOSED: Adopt SPARQL convention for using backslash to allow quotes (and cr, lf, tab, etc) within quoted strings.

Hassan Aït-Kaci: +1

PROPOSED: Adopt SPARQL convention for using backslash to allow quotes (and cr, lf, tab, etc) within quoted strings (in Presentation Syntax).

Sandro Hawke: +1
Igor Mozetic: +1
Mike Dean: +1
Axel Polleres: +1
Jos de Bruijn: +1
Chris Welty: +1
Dave Reynolds: +1

RESOLVED: Adopt SPARQL convention for using backslash to allow quotes (and cr, lf, tab, etc) within quoted strings (in Presentation Syntax).

Michael Kifer: where we should elaborte on the escaping?

Chris Welty: --- break ---


(No activity for 14 minutes)

Sandro Hawke: Back from break

(Scribe changed to John Hall)

PROPOSED: Close Issue 56 as addressed by the resolutions this morning.

Sandro Hawke: http://www.w3.org/2005/rules/wg/track/issues/56
Christian de Sainte Marie: csma has joined #rif

Axel Polleres: sent out email summarizing presentation syntax

Sandro Hawke: +1

Chris Welty: proposal - move on

Dave Reynolds: +1
Jos de Bruijn: +++++1

Christian de Sainte Marie: did not close resolution

Igor Mozetic: +1

RESOLVED: Close Issue 56 as addressed by the resolutions this morning.

Axel Polleres: -0.2
Gary Hallmark: GaryHallmark has joined #rif

Chris Welty: closed Issue 56


Christian de Sainte Marie: issue - Jeremy C - not consider RDF entailment

Christian de Sainte Marie: comment 5 in JC email 2

Dave Reynolds: URL for the email?
Chris Welty: http://lists.w3.org/Archives/Public/public-rif-comments/2008May/0004.html
Jos de Bruijn: http://www.w3.org/2005/rules/wiki/Response_to_JC3
Jos de Bruijn: comment B

Christian de Sainte Marie: no discussion today - move on, no change

Christian de Sainte Marie: JC email 1

Jos de Bruijn: http://www.w3.org/2005/rules/wiki/Response_to_JC

Christian de Sainte Marie: coment 15

Chris Welty: point of confusion - is IP a subeset of IR?

Mike Dean: s/subeset/subset/ (failed)

Jos de Bruijn: making it a little easier doe not justify changing

Chris Welty: just changing our view

Chris Welty: leave it

Mike Dean: s/doe/does/ (failed)
Dave Reynolds: Agreed leave it, can't see how we can change RDF, and I would oppose losing simple entailment

Jos de Bruijn: leave things as they are - Jos already has an action

Jos de Bruijn: http://www.w3.org/2005/rules/wiki/Response_to_JC2

Christian de Sainte Marie: JC email 2, comment 19

Jos de Bruijn: comment 19

Christian de Sainte Marie: DL safeness at risk

Jos de Bruijn: JC was referring to something else

... is OK

... not sure that DL safeness restrictions is what people need

Chris Welty: not what this comment is about

Jos de Bruijn: what we discussed yesterday

Chris Welty: ref to Jos comments in document

Chris Welty: csma: marj DL safeness at risk?

s/marj/mark/ (failed)

Chris Welty: action Jos to mark 3.1.1

ACTION: Jos to mark 3.1.1 at risk


Christian de Sainte Marie: drop section 6.2

... from Dave

ACTION: jdebruij2 to mark SWC section 3.1.1 as "AT RISK", with explanation.

trackbot-ng: Created Action 491 - Mark SWC section 3.1.1 as \"AT RISK\", with explanation. [on Jos de Bruijn - due 2008-06-03].

Jos de Bruijn: not strongly in favour of keeping it

Dave Reynolds: I don't feel strongly about this one either way

... or against

Christian de Sainte Marie: what is status of DLP in OWL-R?

Sandro Hawke: pretty good

Christian de Sainte Marie: mark is as at risk?

Jos de Bruijn: can't do that

Christian de Sainte Marie: couild remove it later

s/couild/could/ (failed)

Jos de Bruijn: if OWL-R is well-designed, could be embedded in RIF

Sandro Hawke: not sure

.. keep 6.2

Chris Welty: no change

Christian de Sainte Marie: Dave 7.1, profiels for imports

s/profiels/profiles/ (failed)

Christian de Sainte Marie: http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0174.html

Jos de Bruijn: section 4.2

... more general isue is that you can specify different profiles but have to pick one

... we now pick the highest

... could have one for the entire document

... if profiles for RDFS and OWL-FULL, OWL-FULL takes precedence, RDFS is not valid

Dave Reynolds: if importing under two profiles, find the lowest one that is higher than both

... if not, abort

Jos de Bruijn: agree

... updated to adopt proposal - see Wiki version

Jos de Bruijn: http://www.w3.org/2005/rules/wiki/SWC#Interpretation_of_Profiles

Dave Reynolds: OK with rewording

Igor Mozetic: some built-in predicates

Jos de Bruijn: need to be updated

... and need to write down the proofs

Chris Welty: what built-in predicates?

Jos de Bruijn: detecting il--typed literals

s/il/ill/ (failed)

Christian de Sainte Marie: is informative - can be changed after last call

Christian de Sainte Marie: show stopper for last call?

Jos de Bruijn: no

Chris Welty: editor's note 2.1.2

Jos de Bruijn: resolved

Chris Welty: end of 3.1.1 - remove and mark as 'at risk'

Jos de Bruijn: 6.1.7 - section to be removed

Jos de Bruijn: syggest removing note in 6.2.3.1

.. 6.2.3.2 remove note?

... will add text

Christian de Sainte Marie: any other features at risk?

Jos de Bruijn: OWL DL annotation entailment 3.2.2.3

Chris Welty: what is problem?

... keep it

Christian de Sainte Marie: marking at risk means we can remove it

... if there is a risk that some implementer will complain, mark as at risk

Jos de Bruijn: leave it - we don't require them to implement

ACTION: csma to review changes

trackbot-ng: Created Action 492 - Review changes [on Christian de Sainte Marie - due 2008-06-03].

PROPOSED: Public SWC as LAST CALL Working Draft, after changes agreed upon this session are made (and checked by CSMA)

PROPOSED: Publish SWC as LAST CALL Working Draft, after changes agreed upon this session are made (and checked by CSMA)

PROPOSED: Publish SWC as LAST CALL Working Draft, after changes agreed upon this session and yesterday are made (and checked by CSMA)

Jos de Bruijn: also changes from yesterday

Sandro Hawke: +1 (W3C)
Harold Boley: +1 (NRC)
Dave Reynolds: +1 (HP)
Jos de Bruijn: +1 (FUB)
Igor Mozetic: +1 (JSI)
Mike Dean: +1 (SRI)
Axel Polleres: +1 (DERI)
Axel Polleres: +1 (IBM)
Gary Hallmark: +1 (Oracle)

+1 (OMG

Michael Kifer: MichaelKifer has joined #rif
Michael Kifer: +1
Adrian Paschke: +1 (REWERSE)


Christian de Sainte Marie: csma has joined #rif
Christian de Sainte Marie: +1 (ILOG)


RESOLVED: Publish SWC as LAST CALL Working Draft, after changes agreed upon this session and yesterday are made (and checked by CSMA)


NAMING CONVENTIONS

BLD document

Chris Welty: agreed only to change upper and lower

... discussing for named arguments and frame slots - different content

Harold Boley: leave as is

Chris Welty: ... easy to handle in XSD

Sandro Hawke: declare - has variable

... quantified variable?

Harold Boley: has class name inside

Chris Welty: declares?

Sandro Hawke: never mind

Gary Hallmark: not happy with Expr

Sandro Hawke: <arg>
Sandro Hawke: <Expr>
Sandro Hawke: <op><Const type="rif:iri">cpt:book</Const></op>
Sandro Hawke: <arg>Author</arg>
Sandro Hawke: <arg><Const type="rif:iri">bks:LeRif</Const></arg>
Sandro Hawke: </Expr>
Sandro Hawke: </arg>

Gary Hallmark: Function for Expr, Predicate for Atom

... Atom is jargon

Sandro Hawke: Equal roles should be left and right

Sandro Hawke: <Equal> should have <left> and <right> not side and side. [Scribe assist by Sandro Hawke]

Christian de Sainte Marie: is symmetric

Harold Boley: prefer not to go back to left and right

Dave Reynolds: +1 on left/right being (slightly) better

Sandro Hawke: do not want to get your rules back from RIF with equalities flipped

Chris Welty: discussion was that equality is symmetric, and we didn't want to force people to choose left and right

PROPOSED: shall we switch from Equal/side/side to Equal/left/right ?

Sandro Hawke: leave it as is: three
Sandro Hawke: changing it: Jos, Sandro, (dave reynolds)
Sandro Hawke: I'm willing to drop it on the graounds that it's a lot of work to change. [Scribe assist by Sandro Hawke]

leave as is

Christian de Sainte Marie: http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0205.html

Hassan Aït-Kaci: email on XML tagging

Sandro Hawke: not about naming

Chris Welty: we just have to be sure that we get it right - we use it when it will work

Harold Boley: have to declare twice - as namespace and entity

Dave Reynolds: Harold - generally people use the entity declaration within the namespace declaration which helps slightly.

Chris Welty: not relevant to this session's topic

Christian de Sainte Marie: also need to look at section 4.2

Christian de Sainte Marie: change name of 'implies' for less-overloaded name

... is not an implication (in logical sense) is some roles

... something like 'conditional'?

Gary Hallmark: could they all be nouns?

s/is some/in some/ (failed)

Christian de Sainte Marie: Change 'manner' to 'profile'

... change 'implies' to 'conditional'

Sandro Hawke: 'payload' to 'content'

'manner' to 'profile' unanimous

Dave Reynolds: RuleBody?
Dave Reynolds: Nah, I withdraw that :-)
Mike Dean: keep implies

ACTION: Harold to change "manner" to "profile"

trackbot-ng: Created Action 493 - Change \"manner\" to \"profile\" [on Harold Boley - due 2008-06-03].
Mike Dean: +1 payload to content

'implies' to 'rule' not decided

Dave Reynolds: +1 payload to content

'payload' to 'content' majority against

Sandro Hawke: "payload" -> "content" fails.

'address' to 'location' majority for

ACTION: Harold to change "address" to "location" for Imports, in BLD.

trackbot-ng: Created Action 494 - Change \"address\" to \"location\" for Imports, in BLD. [on Harold Boley - due 2008-06-03].

Christian de Sainte Marie: small objection to 'rule' for 'implies' - if rule has name, it will be far away from tag 'rule'

Mike Dean: +1 keep implies
Mike Dean: +1 leave as is
Sandro Hawke: implies vs rule: 4 for rule, 5 for implies

Christian de Sainte Marie: did we address all parts of Issue 49?

PROPOSED: Close Issue 49 with decisions made so far today

Sandro Hawke: +1
Igor Mozetic: +1
Jos de Bruijn: -0
Gary Hallmark: +1

RESOLVED: Close Issue 49 with decisions made so far today

RESOLVED: close Issue 49

PROPOSED: close Issue 54

PROPOSED: Close Issue 54 with at-risk label as decided this morning.

Jos de Bruijn: +1

RESOLVED: Close Issue 54 with at-risk label as decided this morning.

RESOLVED: close Issue 54

PROPOSED: Close Issue 60 as decided this morning -- if they are incomparable it's an error

Axel Polleres: AxelPolleres has joined #rif
Jos de Bruijn: +1
Axel Polleres: +1

RESOLVED: Close Issue 60 as decided this morning -- if they are incomparable it's an error

PROPOSED: Go to lunch

Sandro Hawke: BREAK-for-Lunch



Mike Dean: 1 hour break


Attendees (guest): MichaelKifer, Harold, AdrianP,IgorMozetic,sandro,aharth,GaryHallmark,josb,johnhall,csma,ChrisW,AxelPolleres [Scribe assist by Chris Welty]




(No activity for 63 minutes)

Adrian Paschke: AdrianP has joined #rif

(No activity for 11 minutes)



Axel Polleres: AxelPolleres has joined #rif


BLD review

Sandro Hawke: sandro has changed the topic to: RIF F2F10 Code 26633



(Scribe changed to Axel Polleres)



Sandro Hawke: mdean, Hassan, DaveReynolds -- CONFERENCE CODE is 26633

BLD review

Dan's comments on.

Christian de Sainte Marie: What about Dan's comment on arity of predicates?

Sandro Hawke: Problematic on merging rulesets where one uses p with arity n and the other uses p with arity m.

Harold Boley: Harold has joined #rif

Christian de Sainte Marie: if it is an IRI it should have the same arity, if it is a local name, then it is in fact different names.

... answer to dan: this is not a problem, i.e. conflicts on using the same iri with different arities is intended.

Michael Kifer: not sure.

PROPOSED: Answer to Dan is that in BLD, pred and funcs have one arity, and it is correct that the restriction holds even across multiple documents. So the requirement is met -- rulesets can be merged -- but if one ruleset is in error, then the merged version will be able to detect the error.

... for example in PROLOG it is quite common to use the same predicates.

Chris Welty: yes, but we disallow that.

Jos de Bruijn: +1

Christian de Sainte Marie: What do we do on rif:locals on merging? General problem.

Sandro Hawke: +1
Dave Reynolds: Isn't Dan's comment asking for a rephrasing?

Jos de Bruijn: This is - for the import mechanism - well-defined.

Michael Kifer: We defined import, but nor merging.

Jos de Bruijn: imports is not exempted. Doc1: P(..), Q(...) Doc2: P=Q import/merge is a problem. [Scribe assist by Sandro Hawke]

RESOLVED: Answer to Dan is that in BLD, pred and funcs have one arity, and it is correct that the restriction holds even across multiple documents. So the requirement is met -- rulesets can be merged -- but if one ruleset is in error, then the merged version will be able to detect the error.

Chris Welty: who is responding to Dan?

... I will start the wiki page for the response right now.

Adrian Paschke: propose to wait until tomorrow and will respond then, together with UCR responses.

ACTION: AdrianP to respond to Dan2 (about well-formedness)


ACTION: Adrian to respond to Dan2 (about well-formedness)


ACTION: apaschke to respond to Dan2 (about well-formedness)

trackbot-ng: Created Action 495 - Respond to Dan2 (about well-formedness) [on Adrian Paschke - due 2008-06-03].

Christian de Sainte Marie: 3 comments from Jeremy on rif:iri. OWL is unconvinced by rif:iri and rif:text.

Sandro Hawke: I think this is satisfied by our presentation syntax resolutions.

... but we need to respond.

Jos de Bruijn: I will write these responses.

Dave Reynolds: What was the outcome of that one?

Christian de Sainte Marie: next, dave has a comment on equality terms appearing in externals.

... the answer to the question is yes: it is deliberate and legal.

Christian de Sainte Marie: Answer -- it's deliberate, it's okay, it's legal --- [Scribe assist by Sandro Hawke]

Dave Reynolds: but why then disallow External in the head?

Michael Kifer: External as a head, as a predicate, would mean you were re-defining the predicate. [Scribe assist by Sandro Hawke]
Dave Reynolds: Surely the same is true of external functions. [Scribe assist by Sandro Hawke]
Jos de Bruijn: this is the same as moving it to the body and have X=Y in the head. [Scribe assist by Sandro Hawke]
Jos de Bruijn: whereas if you have an external predicate in the head, that requires constraint reasoning, which is different from std rule reasoning. [Scribe assist by Sandro Hawke]

Christian de Sainte Marie: So, shall we allow any External in the head or diallow any Externals in the head?

Dave Reynolds: OK
Jos de Bruijn: So, don't change anything. [Scribe assist by Sandro Hawke]

Michael/josb: that would be a void restriction, because it can't be amulated.

s/can't/can/ (failed)

Christian de Sainte Marie: Dave's comment on BLD XML.

Jos de Bruijn: http://lists.w3.org/Archives/Public/public-rif-wg/2007Nov/0019.html
Jos de Bruijn: nov 27: RESOLVED: We'll refer to XSD 1.0 instead of XSD 1.1 in our document for now, including a clear note that it our intention to change to XSD 1.1 when it becomes available, so that people can use XML 1.1.

... this about the XSD version.

... so, only about datatypes.

Jos de Bruijn: s/27/27, 2007/ (failed)

Dave Reynolds: my comment is about the *XML* version.

Christian de Sainte Marie: objections against saying that we refer to XML 1.0?

PROPOSED: We'll use XML 1.0 (not XML 1.1)

Jos de Bruijn: isn't there a possibility to allow people to use their preferred XML version?

Chris Welty: We are not gonna try to solve that problem, if people can make it work with XML1.1, then it is fine.

PROPOSED: We'll use XML 1.0 (not XML 1.1) for the XML syntax for BLD.

Jos de Bruijn: what is the difference?

Sandro Hawke: fixed reference to unicode in XML 1.0.

... 1.1 more open to speak "different languages".

Gary Hallmark: there's a recomendation to use 1.0 unless features of 1.1 really needed.

Sandro Hawke: let's get back to that later, I will gtry to get an answer within the hour.

Christian de Sainte Marie: comment from Dave on compact IRIs in the XML syntax.

... compact IRIs are not approproate in the XML, because there they are real QNames (?)

Harold Boley: Once we have presentation syntax with prefixes, and entities in the XML, that should be fine.

Dave Reynolds: There's not dependency there!

Michael Kifer: prefix definition will be in BLD, Consts will be in DTB.

... (the pres. syntax)

ACTION: Harold to update all examples for Presentation Syntax and XML syntax for curies and entities. Also add Prefix to presentation syntax.

trackbot-ng: Created Action 496 - Update all examples for Presentation Syntax and XML syntax for curies and entities. Also add Prefix to presentation syntax. [on Harold Boley - due 2008-06-03].

Chris Welty: The XML syntax should be valid XML... full stop.

Christian de Sainte Marie: Dave wants a full XML document as an example.

Harold Boley: that will be a byproduct of my action.


Mike Dean: I've got another telecon. I'll be back in about 1.5 hours.

... I always use the official W3C validators for XML in the examples.

Dave Reynolds: Axell - for the record. My comment was that the use of prefixes in the presentation syntax is orthogonal to the XML. In the XML all IRIs are simply IRs, not CURIEs. Whether we choose to use XML Entities to make the examples more readable is simply presentation style.
Dave Reynolds: s/Axell/Axel/ (failed)

Christian de Sainte Marie: more comments form Dave on the schema.

... 1) rif:type should be used rather than just type.

... 2) rif:type should be resticted to anyURI rather than xsd:string.

Dave Reynolds: Talking to some XML people here (1) is probably not a great idea, I'll withdraw it unless anyone wants it.

Christian de Sainte Marie: Any drawback in qalifying "type"?

Chris Welty: is this the only attribute?

Christian de Sainte Marie: Who's in favor f qualifying attributes?

Sandro Hawke: makes XML more readable... attributes don't need a def namespace.

Dave Reynolds: 0
Sandro Hawke: I favor no namespace for rif:type, so we can have a rif document with no "rif:" anywhere in it, using default namespace. [Scribe assist by Sandro Hawke]

who in favor of qualifying?

Dave Reynolds: OK

0 yes, 4 against, 7 undecided.

PROPOSED: in the RIF XML syntax (as long as we stick with this non-RDF style), attributes will have no namespace (so that we can avoid "rif:" in documents)

meetingroom (guest): meetingroom has joined #rif
Adrian Paschke: +1

PROPOSED: in the RIF XML syntax (as long as we stick with this non-RDF style), attributes will have no namespace (be unqualified) (so that we can avoid "rif:" in documents)

Jos de Bruijn: +1
Sandro Hawke: +1
Dave Reynolds: 0
Harold Boley: +1
Adrian Paschke: +1
Gary Hallmark: +1
Igor Mozetic: +1

RESOLVED: in the RIF XML syntax (as long as we stick with this non-RDF style), attributes will have no namespace (be unqualified) (so that we can avoid "rif:" in documents)

Andreas Harth: RDF or XSLT use that differently... there seems not to be an agreed treatment.

Harold Boley: Dave's 2n bullet refers to this XSD snippet:
Harold Boley: <xs:element name="Const">
Harold Boley: <xs:complexType mixed="true">
Harold Boley: <xs:sequence/>
Harold Boley: <xs:attribute name="type" type="xs:string" use="required"/>
Harold Boley: </xs:complexType>
Harold Boley: </xs:element>

Christian de Sainte Marie: next one. Dave suggests 2) rif:type should be resticted to anyURI rather than xsd:string.

... content of the type cannot be a number, must be a IRI.

Sandro Hawke: slight hesitation for anyURI vs. IRI.

... but that could just be a bugfix.

ACTION: Harold to change type of "type" attribute to xs:anyURI (from xs:string)

trackbot-ng: Created Action 497 - Change type of \"type\" attribute to xs:anyURI (from xs:string) [on Harold Boley - due 2008-06-03].

Jos de Bruijn: in XML Schema datatypes 1.1 anyURI is also used for IRIs.

Christian de Sainte Marie: my own comments on BLD.

Michael Kifer: Right, it doesn't make sense to have these things (Equal, Subclass, etc) be external. [Scribe assist by Sandro Hawke]

... Equal, Member, Subclass should not be allowed to be External.

...also Frame. Suggestion: limit External to ATOMIC.

... discussing External terms in property position in Frames.

Michael Kifer: We can disallow some things in External.

Discussion about the respective parts of the BLD grammar.

Christian de Sainte Marie: currently Externals allowed in slotname position in Frames in BLD.

Chris Welty: Should this now be restrictd?

s/td/ted/ (failed)

Christian de Sainte Marie: Still usure about External(Frame)

Michael Kifer: The semantics is precisely defined.

Christian and Michael trying to clarify what an External Frame actualy means.

... we agreed to disallow external Equal, Member, Subclass.

PROPOSED: Replace Exterman(ATOMIC) with External(ATOM_BASE or FRAME) ... ?

... External Frame still under discussion.

Sandro Hawke: doubts about External used as extension mechanism

Axel Polleres: I thought the set of external schemas is FIXED per dialect.

Michael Kifer: no.

... that is an extension mechanism.

Axel Polleres: RIF FLD says "RIF dialects are always associated with sets of coherent signatures"... I am confused now.

Discussion is whether defining an own external Schema is a new dialect, i.e. an extension, or no.

Igor Mozetic: External could be a SPARQL query, yes?

Michael Kifer: If you add some datatypes or externals, then you have a bigger dialect than BLD. [Scribe assist by Sandro Hawke]
Sandro Hawke: Right. [Scribe assist by Sandro Hawke]

Michael Kifer: If you add some datatypes or externals, then you have a bigger dialect than BLD.

Sandro Hawke: Right.

Sandro Hawke: It's all about the software to you have to have installed to make use of the document. [Scribe assist by Sandro Hawke]
Christian de Sainte Marie: So External(Frame) may be in BLD, but not be used except in BLD-extended. [Scribe assist by Sandro Hawke]
Sandro Hawke: right.... [Scribe assist by Sandro Hawke]
Michael Kifer: Internal Frame defined by rules or facts; External Frame defined by something like a java program. [Scribe assist by Sandro Hawke]
Gary Hallmark: Read-Only. [Scribe assist by Sandro Hawke]
Michael Kifer: Yes. [Scribe assist by Sandro Hawke]
Adrian Paschke: Compliance with BLD should not require that your implementation supports external frames and other experssive features such as procedural attachments (which are for sure needed in real-world implementation)

Michael Kifer: It is like a blackbox.

... external Equal would be possible to define, but wouldn't make sense to use, actually.

PROPOSED: Change External(ATOMIC) to External(ATOM) or External(Frame).

Sandro Hawke: +0 still a little hazy about defing this without defining how to use it, but okay....

Chris Welty: broken link in BLD to "coherent set of such schemas associated[...]"

PROPOSED: Change External(ATOMIC) to External(Atom) or External(Frame).

Sandro Hawke: +0 still a little hazy about defing this without defining how to use it, but okay....
Jos de Bruijn: -0
Dave Reynolds: -0.5

0.0

Igor Mozetic: +1
Gary Hallmark: +0
Dave Reynolds: : If we're going to have external frames, they need to be much more clearly explained in the document. [Scribe assist by Sandro Hawke]

0.0 (not sure why external frames needed and not just External(Atom) )

Jos de Bruijn: For the record: I think they're useless

PROPOSED: Change External(ATOMIC) to External(Atom).

Jos de Bruijn: +1
Igor Mozetic: -1
Harold Boley: +1
Gary Hallmark: -0.5
Dave Reynolds: I don't object to external frames per se but can't follow how the semantics part of the document supports this.

PROPOSED: Change External(ATOMIC) to External(Atom) or External(Frame) and add text explaining how External frames are supported by the semantics.

Igor Mozetic: +1
Jos de Bruijn: -0
Dave Reynolds: -0
Chris Welty: +1
Sandro Hawke: +0.5
Harold Boley: +1

+1

Adrian Paschke: +1

RESOLVED: Change External(ATOMIC) to External(Atom) or External(Frame) and add text explaining how External frames are supported by the semantics.


ACTION: Kifer to add text explaining external frames

trackbot-ng: Created Action 498 - Add text explaining external frames [on Michael Kifer - due 2008-06-03].

Christian de Sainte Marie: Next. Christian's comment on NAU limitation

s/NAU/named argument uniterms/ (failed)

clarified.

Christian de Sainte Marie: How can we create new symbols when inferring a new frame?

... proposals: rif:new, skolem terms, existentials in the head.

Sandro Hawke: writing a translator form N3 to rif needs existentials in the head.

Christian de Sainte Marie: What was your concrete example?

... email?

http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0158.html

Sandro Hawke: Skolemizing would be a fallback.

Adrian Paschke: not the same.

Igor Mozetic: the only problem with skolemization is round-tripping.

Christian de Sainte Marie: What about having that in a builtin?

Michael Kifer: not possible as a builtin.

... rif:new should be symbol not a constant (not an external) that each time you use it is interpreted differently.

Discussion skolem function vs. new constant ongoing.

Jos de Bruijn: +1 burden on user

Chris Welty: What about making skolem funcs a special datatype.

...?

Michael Kifer: It could be a subsymbolspace of rif:local

Dicussion of whether something like gensym is possible.

Jos de Bruijn: +1 burden on user

+1 to +1 of jos.


Shall we go on? Don't see this being resolved soon. We had more promising discussions being cut off today already

Christian de Sainte Marie: We are starting to run in circles.

Gary Hallmark: std metadata would be fine. [Scribe assist by Sandro Hawke]

... if we decide we want this, it will delay. Is it worth?

many no's.

Sandro Hawke: right. forget about this for now. some std metadata later, maybe. [Scribe assist by Sandro Hawke]

break now.

20min until 4.

(No activity for 15 minutes)

Harold Boley: Harold has joined #rif
Harold Boley: Guizhen Yang and Michael Kifer, Reasoning about Anonymous Resources and Meta Statements on the Semantic Web: http://www.wsmo.org/wsml/papers/presentations/lausen_anonymous_resources.ppt
Harold Boley: http://www.ai.sri.com/~yang/papers/jods2003.pdf

scribe???

still me? :-)

Sandro Hawke: AFTERNOON AGENDA: Conformance, rdf-syntax-compatibility, xml schema-extensibility, metadata, readers guide
Christian de Sainte Marie: then -- identify features at risk, and decide on Last Call for BLD. [Scribe assist by Sandro Hawke]

(Scribe changed to Adrian Paschke)

BLD open issues

Chris Welty: compliance definition for BLD

Michael Kifer: separate document for compliance

Michael Kifer: put it in the overview

Chris Welty: last call document needs conformant statement

Sandro Hawke: agree conceptually it could go into another document

Sandro Hawke: but for now it might be in BLD

Harold Boley: what about FLD

Chris Welty: put it in BLD for now

Harold Boley: The general part in FLD doc, the BLD dialect in BLD doc?
Harold Boley: s/dialect /dialect part/ (failed)
Axel Polleres: What about putting it into UCR? We should have a kinda fair distribution among the docs... ;-)

PROPOSED: add the text on http://www.w3.org/2005/rules/wiki/Conformance (more or less) to BLD, probably near the next....

Dave Reynolds: I suggested "Document Conformance" in http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0201.html

Sandro Hawke: statement about syntactic RIF consumer , RIF producer compliance

Dave Reynolds: q+
Sandro Hawke: ack DaveReynolds



Dave Reynolds: question: is schema validation actually enough to validate conformance?

Sandro Hawke: "A conformant BLD Document is one which conforms to all the constraints of this document, including ones which cannot be checked by XML Schema validator".

Michael Kifer: well-formed vs. semantically correct use

Sandro Hawke: RIF consumers must reject a BLD document if .... constraints are not met

Sandro Hawke: for rule engines as consumers we should say something about when a BLD document needs to be rejected

Jos de Bruijn: 26633
Hassan Aït-Kaci: thanks
Jos de Bruijn: is the code

Sandro Hawke: for example a BLD document which use e.g. a new construct ActionRule; a consumer must throw an error

Sandro Hawke: it can not silently ignore it


Chris Welty: you want this strict dialect conformance?

Sandro Hawke: right

Sandro Hawke: A BLD+foo processor should have a strict-BLD mode so that it can run in BLD-only mode.

Chris Welty: we could label it strict conformant and conformant

Chris Welty: strict conformance is exclusive; conformance is inclusive

Sandro Hawke: we need strict conformance, otherwise people will abuse BLD

Michael Kifer: conformance and loose conformance

Sandro Hawke: A BLD processor, strict or loose, must never "repair" inputs or accept input which does not adhere to a standard or third-party defined dialect or extension.
Michael Kifer: "In additionan a RIF consumer must reject any documebnt which is not a DTE formula". [Scribe assist by Sandro Hawke]
Sandro Hawke: "A conformant BLD Document is one which conforms to all the constraints of this document, including ones which cannot be checked by XML Schema validator".

(No activity for 7 minutes)

Hassan Aït-Kaci: Please paste it in the IRC

PROPOSED: accept the conformance statement on http://www.w3.org/2005/rules/wiki/Conformance for BLD, up to the separator line.

Sandro Hawke: +1 good enough for last call

0

Igor Mozetic: +1
Dave Reynolds: +0
Hassan Aït-Kaci: 0
Harold Boley: +1

RESOLVED: accept the conformance statement on http://www.w3.org/2005/rules/wiki/Conformance for BLD, up to the separator line.

Mike Dean: +0

RDF discussion

PROPOSED: The normative exchange syntax for RIF will be glass etchings.

XML Syntax -- type tagging and RDF/Compatibility

Chris Welty: type-tagging syntax

Jos de Bruijn: -0 on conformance statement

Chris Welty: XSLT transformation from current XML syntax to rigid RDF syntax

Sandro Hawke: some people are allergic using RDF name space

Chris Welty: current syntax is ordered

Chris Welty: so we would need to add parsetype collections to get rid of the order for RDF

Chris Welty: if you translate into RDF and back the order is lost

Sandro Hawke: sure. if you use a triple store the order is lost

Harold Boley: The differences 1.-4. in http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0099.html can already now be handled as XSLT expansions.

Christian de Sainte Marie: but this is your (user) problem

Christian de Sainte Marie: you are not forced to use it

Michael Kifer: What is the problem with the XSLT solution

Christian de Sainte Marie: RDF tools can not directly parse a BLD document

Christian de Sainte Marie: so you loose a litte bit of openness

Sandro Hawke: you still need to add parsetype collection

Michael Kifer: XSLT will do that

Dave Reynolds: Could use GRDDL so that the document references the correct dialect-specific XSLT.
Axel Polleres: +1 to Michael and Dave.... rationale, either *real* RDF or don't bother.
Axel Polleres: ... I find a standard GRDDL an elegant solution.

Christian de Sainte Marie: Would you need a different XSLT if you have user-defined functions etc.

Harold Boley: No. You covered BLD and it can be mirrored by XSLT

Hassan Aït-Kaci: I agree with Sandro

Sandro Hawke: it does not provide anything. It does not scale with dialects

Michael Kifer: We give an example and they can modify the XSLT example

Sandro Hawke: How to find this XSLT?

Michael Kifer: they publish it

Dave Reynolds: Sandro - could require a GRDDL entry in the root element for each document (not sure I like it but is dialect friendly)

Sandro Hawke: does not solve anything

Jos de Bruijn: grddl sounds pretty good

yes, grddl could be a dialect and maybe module solution

Hassan Aït-Kaci: Yes with Sandro !!!

Sandro Hawke: I want to use frame rules

Harold Boley: it is like meta programming

Chris Welty: only positional arguments can not transformed into frames

Gary Hallmark: so let's get rid of positional arguments and name them

Harold Boley: we had a breakout session about this and slides exsist

Hassan Aït-Kaci: +1 with ChrisW

Sandro Hawke: I want frame rules in RIF

Chris Welty: the only problem is the order

Sandro Hawke: the two options are use numbers on the arguments or have some ordered flag

Michael Kifer: numbers are a general solution

Igor Mozetic: follow the principle object-oriented XML

Gary Hallmark: isn't is possible to have flag which says if it ordered or not

Sandro Hawke: two questions: how to implement the ordering and do we use rdf namespace to implement a solution

Gary Hallmark: people really don't want rdf namespace

Sandro Hawke: alternative to RDF ns -- ordered="yes"

Gary Hallmark: so use a flag attribute

Hassan Aït-Kaci: What's wrong with Gary's proposal (attribute flags)?

Sandro Hawke: yes it solves the frame rule problem and makes me happy

Harold Boley: HaroldBoley has joined #rif

Michael Kifer: solves the parsetype problem

Harold Boley: On parsing, at position i, <arg> can be expanded to <arg index="i">.

Sandro Hawke: we still have the problem with RDF datatypes

Sandro Hawke: Instead of:
Sandro Hawke: <Const type="&xsd;dateTime">2007-11-23</Const>
Sandro Hawke: it would be:
Sandro Hawke: <Const>
Sandro Hawke: <rdf:value rdf:datatype="&xsd:dateTime">2007-11-23</rdf:value>
Sandro Hawke: </Const>
Sandro Hawke: ts for these:
Sandro Hawke: Instead of:
Sandro Hawke: <Const type="&rif;iri">&cpt;purchase</Const>
Sandro Hawke: it would be:
Sandro Hawke: <Const><rdf:value rdf:resource="&cpt;purchase" /></Const>

Igor Mozetic: let's handle these two issues separated

Igor Mozetic: if we talk about OO XML XSLT can make it RDF readable

Christian de Sainte Marie: requiring XSLT to make it RDF parsable is exactly the same as if we have no RDF compatibility

Harold Boley: we already have Const and Var

Michael Kifer: It is not clear why we need this rdf:value

Harold Boley: <Const>abc</Const> and xyz are the leaves of fully striped markup.
Harold Boley: They also give Type Tags Const and Var for abc and xyz, resp.

PROPOSED: we'll have an "object-oriented" / "type-tagged" /"self-describing" XML, so that frame-rules can operate on RIF documents. Requires something like numbering arguments or rdf:parsetype="collection" or ordered="yes".

like ordered="yes"

Igor Mozetic: +1 for Adrian
Sandro Hawke: +1

+1

Gary Hallmark: +1
Hassan Aït-Kaci: +1
Jos de Bruijn: 0
Michael Kifer: <op><Const type="rif:iri">curr:USD</Const></op>
Michael Kifer: <arg><Const type="xsd:integer">49</Const></arg>
Dave Reynolds: 0

PROPOSED: we'll have an XML such that frame-rules can operate on RIF documents. Requires something like numbering arguments or rdf:parsetype="collection" or ordered="yes".

I understand it like <Atom ordered="yes">...</Atom>?

Harold Boley: why do we not the oid of a frame, it think named arguments would do it

PROPOSED: we'll have an XML such that RIF can operate on RIF documents at a RIF-syntactic-level instead of a DOM level. Requires something like numbering arguments or rdf:parsetype="collection" or ordered="yes".

s/Atom/Frame/ (failed)

Harold Boley: +1
Igor Mozetic: +1 for ordered="yes"
Jos de Bruijn: 0
Hassan Aït-Kaci: +1
Andreas Harth: +1
Dave Reynolds: 0

+1 for ordered="yes"

Chris Welty: +0
Axel Polleres: 0

RESOLVED: we'll have an XML such that RIF can operate on RIF documents at a RIF-syntactic-level instead of a DOM level. Requires something like numbering arguments or rdf:parsetype="collection" or ordered="yes".

s/Frame/Atom/ (failed)

PROPOSED: use an RDF/XML-compatible syntax for RIF (more-or-less following the suggestions of http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0099.html)

-1

Harold Boley: -1
Igor Mozetic: -1
Gary Hallmark: +1
Dave Reynolds: +0.5
Michael Kifer: -1
Hassan Aït-Kaci: +1
Sandro Hawke: +1
Andreas Harth: +1
Jos de Bruijn: 0
Axel Polleres: 0
Chris Welty: +1

PROPOSED: use an RDF/XML-compatible syntax for RIF (more-or-less following the suggestions of http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0099.html) provided it does not make RIF implementations need to know anything about RDF.

Harold Boley: You could easily transform the stripped version into a version with stripe skipping, e.g. XSLT would remove the slots

Harold Boley: Alex Kozlenkov's email implies -1 (http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0204.html)

Adrian Paschke: you then would have a much more compact representation

(No activity for 8 minutes)

Sandro Hawke: Scribe: sandro
Igor Mozetic: I would not object [Scribe assist by Sandro Hawke]
Adrian Paschke: I don't want BLD to include the RDF namespace at all [Scribe assist by Sandro Hawke]
Sandro Hawke: (I can't follow what Adrian is saying.)
Hassan Aït-Kaci: (me neither - no mike)
Adrian Paschke: My objection is that it blows up the syntax, making it larger, adding the <name/value> stripes. [Scribe assist by Sandro Hawke]
Adrian Paschke: apaschke has joined #rif

(No activity for 5 minutes)

Sandro Hawke: ScribeNick: apaschke
Michael Kifer: I would object [Scribe assist by Sandro Hawke]
Harold Boley: I would object [Scribe assist by Sandro Hawke]

PROPOSED: we will NOT use an RDF/XML-compatible syntax for RIF

Adrian Paschke: <Const>abc</Const> is very compact and is easy to read and learn
Christian de Sainte Marie: it would be stupid, but I'm not going to object. [Scribe assist by Sandro Hawke]
Hassan Aït-Kaci: I do not understand why we need to close a door for a potentially useful compatibility
Chris Welty: I agree, it's pretty stupid to go this way. But, oh well, we do. [Scribe assist by Sandro Hawke]
Hassan Aït-Kaci: (cannot hear well)
Sandro Hawke: straw poll: <arg index="3"> vs <args orderer="yes">
Adrian Paschke: <Atom ordered="yes"> ... </Atom>
Sandro Hawke: <formula>
Sandro Hawke: <Atom>
Sandro Hawke: <op><Const ... /></op>
Sandro Hawke: <args rdf:parseType="Collection">
Sandro Hawke: ...
Sandro Hawke: ...
Sandro Hawke: <Const .../>
Sandro Hawke: </args>
Sandro Hawke: </Atom>
Sandro Hawke: </formula>

Christian de Sainte Marie: it is not fully stripped

Christian de Sainte Marie: no, I retracted to what I said

Harold Boley: We could also put it on <Atom>

Sandro Hawke: <formula>
Sandro Hawke: <Atom>
Sandro Hawke: <op><Const ... /></op>
Sandro Hawke: <args ordered="yes">
Sandro Hawke: ...
Sandro Hawke: ...
Sandro Hawke: <Const .../>
Sandro Hawke: </args>
Sandro Hawke: </Atom>
Sandro Hawke: </formula>
Sandro Hawke: <formula>
Sandro Hawke: <Atom>
Sandro Hawke: <op><Const ... /></op>
Sandro Hawke: <arg index="1"> ... </arg>
Sandro Hawke: <arg index="2"> ... </arg>
Sandro Hawke: <arg index="3"><Const .../></arg>
Sandro Hawke: </Atom>
Sandro Hawke: </formula>
Sandro Hawke: <formula>
Sandro Hawke: <Atom>
Sandro Hawke: <op><Const ... /></op>
Sandro Hawke: <args rif:ordered="yes">
Sandro Hawke: ...
Sandro Hawke: ...
Sandro Hawke: <Const .../>
Sandro Hawke: </args>
Sandro Hawke: </Atom>
Sandro Hawke: </formula>
Chris Welty: straw poll -- unanymous to do it in the ordered way. [Scribe assist by Sandro Hawke]

Chris Welty: put ordered attribute on Atom

Harold Boley: we could have convention that arguments and members of lists are ordered, by default

Hassan Aït-Kaci: Using conventions and expecting people to follow them is NOT a good idea

Sandro Hawke: it is not simpler

Igor Mozetic: +1 Hassan
Sandro Hawke: <slot>
Sandro Hawke: <Prop>
Sandro Hawke: <key><Const type="rif:iri">cpt:buyer</Const></key>
Sandro Hawke: <val>Buyer</val>
Sandro Hawke: </Prop>
Sandro Hawke: </slot>
Sandro Hawke: =>
Sandro Hawke: <slot rif:ordered="yes">
Sandro Hawke: <Const type="rif:iri">cpt:buyer</Const>
Sandro Hawke: Buyer
Sandro Hawke: </slot>
Adrian Paschke: What about this:
Adrian Paschke: <formula>
Adrian Paschke: <Atom ordered=”yes”>
Adrian Paschke: <op><Const ... /></op>
Adrian Paschke: <arg> ... </arg>
Adrian Paschke: ...
Adrian Paschke: </formula>
Sandro Hawke: <slots>
Sandro Hawke: <Const type="rif:iri">cpt:buyer</Const>
Sandro Hawke: Buyer
Sandro Hawke: </slots>

PROPOSED: use an XML attribute rif:ordered="yes" (as exemplified above) which works like rdf:parseType="Collection" (and rif:type attribute gets qualified again.)

Hassan Aït-Kaci: +1
Igor Mozetic: +1
Gary Hallmark: +1


Dave Reynolds: What does "works like rdf..." mean - will you have a seq/list type in RIF frames?
Hassan Aït-Kaci: no phone...



PROPOSED: use an XML attribute rif:ordered="yes" (as exemplified above) or using an equivalent unique method to specify order, which works like rdf:parseType="Collection" (and rif:type attribute gets qualified again.)

Christian de Sainte Marie: csma has joined #rif


Christian de Sainte Marie: meeting room has been cut off the phone
Hassan Aït-Kaci: yes
Christian de Sainte Marie: trying to call back


Harold Boley: +1
Sandro Hawke: +1
Gary Hallmark: +1
Jos de Bruijn: 0
Adrian Paschke: +1
Hassan Aït-Kaci: +1
Mike Dean: +1
Sandro Hawke: DaveReynolds, I mean as far as the grammar and general purpose/intent.
Hassan Aït-Kaci: Arguing about arguments? A meta-argument...? ;-)

RESOLVED: use an XML attribute rif:ordered="yes" (as exemplified above) or using an equivalent unique method to specify order, which works like rdf:parseType="Collection" (and rif:type attribute gets qualified again.)

Dave Reynolds: Sandro - I guessed that :-) but if the goal is to make RIF manipulatable by RIF you have to close that gap and specify the mapping.
Adrian Paschke: Metadata

Metadata

Christian de Sainte Marie: http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0036.html

Harold Boley: explains proposal for metadata

Jos de Bruijn: Why isn't identifier simply and IRI?

Michael Kifer: yes, it can be an iri

Sandro Hawke: Curries?

Jos de Bruijn: Link between metadata and identifier?

Harold Boley: now it is totally decoupled

Jos de Bruijn: what is the advantage?

Michael Kifer: it gives you more freedom, refer to other pieces to metadata

Hassan Aït-Kaci: I have the same questions as Jos...

Sandro Hawke: in the example is pd identifier for the group?

Jos de Bruijn: it is the identifier of the frame not of the group

Michael Kifer: there is no formal relation

Christian de Sainte Marie: I would like to say that a certain rule is called "cmp" in a group of rules containing only one rule

Sandro Hawke: *i _g1 *m _g2[dc:creator->"csma"] Group ( ... )
Sandro Hawke: *i _g1 *m _g1[dc:creator->"csma"] Group ( ... )
Hassan Aït-Kaci: ???
Sandro Hawke: Hassan, the syntax is from http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0036.html
Hassan Aït-Kaci: I know - I did not understand your example Sandro

Harold Boley: we allow crossreferences between metadata

Jos de Bruijn: I dissagree with the snapshot proposal

Jos de Bruijn: with the new proposal we can identify rules, so it overcomes my issue

Harold Boley: it is open how deep it will go ; could be on var

Sandro Hawke: annotation(......

Sandro Hawke: id and meta roles; optional

Harold Boley: XML syntax is given in the end of document

Christian de Sainte Marie: compatibility with PRD, a rule set will have parameters, how do I distinguish a group with and without parameters

Sandro Hawke: you make different groups

Hassan Aït-Kaci: e.eg priority
Hassan Aït-Kaci: for ruleflow synchrony as well
Hassan Aït-Kaci: s/e.eg/e.g./ (failed)
Hassan Aït-Kaci: Ruleset parameters change the semantics of the ruleset

Michael Kifer: it has nothing to with metadata, currently group has no parameters

Christian de Sainte Marie: Group can be used in other dialects, PRD

Christian de Sainte Marie: Currently in PRD you have ruleset, we could use the same syntax

Harold Boley: Group  ::= 'Group' IRIMETA? '(' (RULE | Group)* ')'

Sandro Hawke: it is orthogonal to metadata

Dave Reynolds: +1 to Sandro (I need to go soon and would like to get to the end of the metadata discussion)
Adrian Paschke: +1 to Sandro and Hassan, let's decide on metadata or postpone it
Harold Boley: Hassan, I think people want to have dinner very soon here. So I guess only a few minutes left.
Dave Reynolds: I object to use of unconstrained formulas for the metadata

Gary Hallmark: Why is it a formula?

Sandro Hawke: Metadata could be in a separate document


Igor Mozetic: +1
Sandro Hawke: two ids the same -- syntax error? [Scribe assist by Sandro Hawke]
Harold Boley: I think so.... [Scribe assist by Sandro Hawke]
Dave Reynolds: I will miss significant parts of tomorrow. For the record I would formally oppose an unconstrained notion of metadata. Conjunctions of frames would be ok.





(No activity for 89 minutes)


(No activity for 88 minutes)

Sandro Hawke: sandro has joined #rif

(No activity for 6 minutes)

Sandro Hawke: sandro has joined #rif

(No activity for 11 minutes)

Gary Hallmark: GaryHallmark has joined #rif


Day 3

See also: IRC log

(Scribe changed to Igor Mozetic)

Chris Welty: ChrisW has joined #rif

PROPOSED: Close Issue 34 as addressed by text currently in BLD

John Hall: johnhall has joined #rif

PROPOSED: Close Issue 34 as addressed by text currently in BLD at http://www.w3.org/2005/rules/wiki/BLD#Conformance_Clauses

Gary Hallmark: +1
Sandro Hawke: +1

+1

Adrian Paschke: +1

RESOLVED: Close Issue 34 as addressed by text currently in BLD at http://www.w3.org/2005/rules/wiki/BLD#Conformance_Clauses

BLD left-overs from yesterday

How to order args in XML syntax (ordered="yes")

Sandro Hawke: http://www.w3.org/2005/rules/wiki/Signalling_Ordering
Christian de Sainte Marie: csma has joined #rif

Harold Boley: multiple children under a role should be ordered by convention

Dave Reynolds: DaveReynolds has joined #rif



Harold Boley: Ordering Criteria for RIF's Object Orient XML: The child elements (roles) of class tags are unordered; the child elements roles are ordered.

Why using convention instead of being explicit?

Harold Boley: (explains convention of auto-detecting when ordering matters) [Scribe assist by Sandro Hawke]
Sandro Hawke: <args>x</args>

Sandro Hawke: is the above ordered or not?

Sandro Hawke: his proposal http://www.w3.org/2005/rules/wiki/Signalling_Ordering

Adrian Paschke: I liked the attribute solutions order="yes".
Adrian Paschke: at best on the level of atom <Atom order="yes">...</Atom>

Sandro Hawke: one needs a flag to indicate order

show of hands: 1 prefers the second option, the rest the first

Harold Boley: Examples:
Harold Boley: <Atom>
Harold Boley: <op><Const type="&rif;iri">&cpt;scheduled</Const></op>
Harold Boley: <arg>item</arg>
Harold Boley: <arg>scheduledate</arg>
Harold Boley: </Atom>
Harold Boley: ==>
Harold Boley: <Atom>
Harold Boley: <op><Const type="&rif;iri">&cpt;scheduled</Const></op>
Harold Boley: <args>
Harold Boley: item
Harold Boley: scheduledate
Harold Boley: </args>
Sandro Hawke: prefers ordered="yes" to <LIst> --- 9-to-1 (only Sandro prefers <List>)
Harold Boley: </Atom>
Harold Boley: <slot>
Harold Boley: <Prop>
Harold Boley: <key><Name>&cpt;title</Name></key>
Harold Boley: <val><Const type="&rif;iri">&bks;LeRif</Const></val>
Harold Boley: </Prop>
Harold Boley: </slot>
Harold Boley: ==>
Harold Boley: <slot>
Harold Boley: <Name>&cpt;title</Name>
Harold Boley: <Const type="&rif;iri">&bks;LeRif</Const>
Harold Boley: </slot>

Gary Hallmark: explicit ordered="yes" affects semantics

Igor Mozetic: what if we have explicit ordered="no" and the rest is assumed ordered?

Sandro Hawke: not sure

Sandro Hawke: maybe attribute ttxml:collection="yes"
Sandro Hawke: maybe attribute ttxml:singleton="yes"
Sandro Hawke: maybe attribute ttxml:collection="yes" (optional if there are two or more child elements)

PROPOSED: RIF adopts an Object Oriented XML with the following Ordering Criteria : The child elements (roles) of class tags are unordered; the child elements roles are ordered. The only roles where this matters are the non-unary ones: args and slot. Editorss Note: This is at-risk and can be replaced by an ordered="yes" attribute.

Adrian Paschke: why not simply use rif_ordered="yes"?

PROPOSED: RIF adopts an Object Oriented XML with the following Ordering Criteria : The child elements (roles) of class tags are unordered; the child elements of roles are ordered. An XML attribute ttxml:collection="yes" is used on for emphasis (optional if there are two or more child elements)

PROPOSED: 1 prefers inexplicit convention, majority explicit order

Adrian Paschke: the implementation of an attribute group for this ordered attribute is rather easy and can be easily added where needed

Gary Hallmark: keep it simple!

PROPOSED: we use "ordered=yes" to indicate ordered arguments in XML

PROPOSED: who prefers rif namespace: 4 yes vs. 4 no

PROPOSED: RIF will use rif:ordered="yes". This item will be marked "at risk", saying the name and XML details on this bit may change.

Sandro Hawke: +1
Chris Welty: +1

no objections

Axel Polleres: +1
Harold Boley: 0 (abstain)

+1

RESOLVED: RIF will use rif:ordered="yes". This item will be marked "at risk", saying the name and XML details on this bit may change.

Sandro Hawke: http://www.w3.org/2005/rules/wiki/Terse_consts


Dave Reynolds: Doesn't that cause problems for schema validation if the set of datatypes is extensible?

Christian de Sainte Marie: concern about datatypes extensibility

Dave Reynolds: That would inhibit people doing extensions if you have to rewrite the schema to just an another data type.

Michael Kifer: vars can have symbol spaces in the future dialects

Axel Polleres: This looks nice to me. How is the similar behavior in RDF/XML's XML Schema handled?
Sandro Hawke: RDF/XML doesn't have an XML schema, Axel.  :-/

Christian de Sainte Marie: agrees with Dave

Axel Polleres: :-)

Axel Polleres: we could have shortcuts in XML

Sandro Hawke: q?
Axel Polleres: I withdraw that, agree, it would raise too much ambiguity in the XML
Harold Boley: Extra attributes inside Const such as <Const type="..." arity="..." > would not be possible.
Dave Reynolds: Stick with current
Sandro Hawke: likes this:5 prefers Consts: 5

strawpoll on http://www.w3.org/2005/rules/wiki/Terse_consts: 5 vs 5

Gary Hallmark: the list of datatypes should be enumerated in the schema, so it can catch the errors. [Scribe assist by Sandro Hawke]
Dave Reynolds: +1 with csma

Christian de Sainte Marie: stick with current

Harold Boley: Let's not put IT at risk!

+1 for Harold

PROPOSED: Close Issue 55, with decisions made so far this meeting.

Gary Hallmark: GaryHallmark has joined #rif

tabled for after break

Sandro Hawke: I'm very hesitant on that -- I'd like to somehow keep the door open on Terse_consts, somehow. [Scribe assist by Sandro Hawke]

metadata

Harold Boley: http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0036.html
Sandro Hawke: http://www.w3.org/2005/rules/wiki/Metadata_examples

meta as frame conjunction, not arbitrary formula

otherwise the same proposal as by Harold

the short form is an alternative proposal


the short form covers the comments

Dave Reynolds: What's the preference over, can't follow the discussion?
Sandro Hawke: these, Dave: http://www.w3.org/2005/rules/wiki/Metadata_examples
Jos de Bruijn: http://www.w3.org/2005/rules/wiki/Metadata_examples

Dave, between the two options above

Dave Reynolds: The second one again fails the xml-schema test unless we have fixed metadata.
Dave Reynolds: ???? how to you distinguish between metadata and extensions?
Dave Reynolds: You could have a shorter syntax than #1 without loosing schema

Sandro Hawke: extensions are just extensions of Classes, not roles

Christian de Sainte Marie: who prefers Harold's proposal with formula restricted to conj of frames: 7 [Scribe assist by Sandro Hawke]
Dave Reynolds: 1 is OK, 2 I'm not so happy with, but a short form syntax like that but schema-friendly seems possible
Christian de Sainte Marie: sandro's proposal -- any unknown roles is metadata : 2 [Scribe assist by Sandro Hawke]
Christian de Sainte Marie: who doesn't like either: 0 [Scribe assist by Sandro Hawke]

PROPOSED: 7 for option 1, 2 for option 2, nobody objects to either

Dave Reynolds: <atom id="...>
Dave Reynolds: <meta>
Christian de Sainte Marie: http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0036.html
Dave Reynolds: Sorry, let me work something out then paste, will take a min or two

PROPOSED: Adopt the XML syntax for metadata in http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0036.html, using conjunction-of-frames instead of all formulas.

Christian de Sainte Marie: return to the AT RISK features

Dave Reynolds: How about:
Dave Reynolds: <Atom id="http://example.com/foo/bar">
Dave Reynolds: <meta order=yes>
Dave Reynolds: <Const type="&rif;iri">&dc;creator</Const>
Dave Reynolds: <Const type="&xs;string">John Smith</Const>
Dave Reynolds: </meta>
Dave Reynolds: <meta order=yes>

(Scribe changed to Gary Hallmark)

Dave Reynolds: <Const type="&rif;iri">&dc;creator</Const>
Dave Reynolds: <Const type="&xs;string">Csma</Const>
Dave Reynolds: </meta>
Dave Reynolds: </Atom>

Dave Reynolds: problem with shortcut is metadata mixed with rule markup

Christian de Sainte Marie: also, shortcut limits metadata to be about container only

Jos de Bruijn: also, how to do structured metadata

Dave Reynolds: Agreed, this means you can't do structured metadata

PROPOSED: Adopt the XML syntax for metadata in http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0036.html, using conjunction-of-frames instead of all formulas.

no supporters for Dave's suggestion

PROPOSED: Adopt the XML syntax for metadata in http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0036.html and given as the first example on http://www.w3.org/2005/rules/wiki/Metadata_examples, using conjunction-of-frames instead of all formulas.

Sandro Hawke: +1
Chris Welty: +1
Dave Reynolds: +1
Gary Hallmark: +1
Adrian Paschke: +1
Jos de Bruijn: +0

RESOLVED: Adopt the XML syntax for metadata in http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0036.html and given as the first example on http://www.w3.org/2005/rules/wiki/Metadata_examples, using conjunction-of-frames instead of all formulas.

Mike Dean: +0
Sandro Hawke: change me to +0
Igor Mozetic: +1
Harold Boley: +1

Christian de Sainte Marie: any objection to put metadata on all Class elements?

Jos de Bruijn: so you can put IDs on Const?

... it may be an IRI already

Sandro Hawke: Const could be Plancks constant (worthy of a comment)

PROPOSED: the <id> and <meta> elements can occur under any Class element (this matter is underspecified in 0036, and previous resolution).

Jos de Bruijn: oid of meta frames can be anything

Sandro Hawke: +1
Dave Reynolds: 0
Gary Hallmark: +1
Jos de Bruijn: 0
Chris Welty: +1
Adrian Paschke: +1
Igor Mozetic: +1

RESOLVED: the <id> and <meta> elements can occur under any Class element (this matter is underspecified in 0036, and previous resolution).

Axel Polleres: +1

PROPOSED: Close Issue 51 (metadata syntax and rule identification) give the decisions made so far this meeting.

Sandro Hawke: +1

Dave Reynolds: IDs can be any Const, maybe should limit to IRI

Sandro Hawke: nice to have locals

... also nice to have existential to avoid having ot invent locals to nest frames

Christian de Sainte Marie: Yes, IDs can be anything --- locals, iris, ... [Scribe assist by Sandro Hawke]
Harold Boley: +1

Dave Reynolds: doesn't like having number as ID of element

Andreas Harth: can I refer to rules from another doc? answer: yes

Chris Welty: ack DaveReynolds
Sandro Hawke: sense of room: no one likes things like numbers as Ids, but it's not worth saying anything about.l
Dave Reynolds: -0

RESOLVED: Close Issue 51 (metadata syntax and rule identification) give the decisions made so far this meeting.

Sandro Hawke: (to Andreas) you can move metadata around, and it really doesn;t change anything (except maybe for editors) ---- nods around room. [Scribe assist by Sandro Hawke]

Christian de Sainte Marie: what about comments?

Sandro Hawke: rdfs:comment

... don't use xml comments, they are ignored

Sandro Hawke: http://www.w3.org/TR/rdf-schema/#ch_comment

Michael Kifer: should list recommended metadata property names

... and include one for comment

Sandro Hawke: Dublin Core 1.1 http://dublincore.org/documents/1999/07/02/dces/

PROPOSED: Close Issue 58 (Comments) by suggesting people use Dublin Core for metadata (http://dublincore.org/documents/1999/07/02/dces/) eg dc:comment

Sandro Hawke: point people at dublin core (also has comment)

Adrian Paschke: AdrianP has joined #rif
Sandro Hawke: (but we're on break)

(No activity for 6 minutes)

Dave Reynolds: Last time we discussed this there was negative feedback about using rdf/rdfs properties and I suggested DC as outlined in http://lists.w3.org/Archives/Public/public-rif-wg/2008Apr/0175.html

(No activity for 5 minutes)

Chris Welty: http://www.w3.org/TR/owl-ref/#Annotations




Chris Welty: owl points people to list of other properties for metadata (annotations)

PROPOSED: Close Issue 58 (Comments) by suggesting people use Dublin Core, RDFS, and OWL for metadata, along the lines of http://www.w3.org/TR/owl-ref/#Annotations

... rdfs:comment, etc.

... dc:creator, dc:description, dc:date, etc

PROPOSED: Close Issue 58 (Comments) by suggesting people use Dublin Core, RDFS, and OWL for metadata, along the lines of http://www.w3.org/TR/owl-ref/#Annotations -- specifically owl:versionInfo, rdfs:lable, rdfs:comment, rdfs:seeAlso, rdfs:isDefinedBy, dc:creator, dc:description

Jos de Bruijn: http://dublincore.org/documents/dces/

PROPOSED: Close Issue 58 (Comments) by suggesting people use Dublin Core, RDFS, and OWL properties for metadata, along the lines of http://www.w3.org/TR/owl-ref/#Annotations -- specifically owl:versionInfo, rdfs:label, rdfs:comment, rdfs:seeAlso, rdfs:isDefinedBy, dc:creator, dc:description, dc:date, foaf:maker (when creator is an object, not a string).

Andreas Harth: foaf:maker is better than dc:creator because range is IRI not just string

PROPOSED: Close Issue 58 (Comments) by suggesting people use Dublin Core, RDFS, and OWL properties for metadata, along the lines of http://www.w3.org/TR/owl-ref/#Annotations -- specifically owl:versionInfo, rdfs:label, rdfs:comment, rdfs:seeAlso, rdfs:isDefinedBy, dc:creator, dc:description, dc:date, foaf:maker

Chris Welty: these are just suggestions. users can also invent new ones if they like

PROPOSED: Close Issue 58 (Comments) by suggesting people use Dublin Core, RDFS, and OWL properties for metadata, along the lines of http://www.w3.org/TR/owl-ref/#Annotations -- specifically owl:versionInfo, rdfs:label, rdfs:comment, rdfs:seeAlso, rdfs:isDefinedBy, dc:creator, dc:description, dc:date, foaf:maker. This goes in the metadata section of BLD.

Jos de Bruijn: +1

Michael Kifer: where does this info go. we have no metadata section

Sandro Hawke: +1
Igor Mozetic: +1
Axel Polleres: AxelPolleres has joined #rif
Axel Polleres: +1
Adrian Paschke: +1
Harold Boley: +1

PROPOSED: Close Issue 58 (Comments) by suggesting people use Dublin Core, RDFS, and OWL properties for metadata, along the lines of http://www.w3.org/TR/owl-ref/#Annotations -- specifically owl:versionInfo, rdfs:label, rdfs:comment, rdfs:seeAlso, rdfs:isDefinedBy, dc:creator, dc:description, dc:date, foaf:maker. This goes in BLD (near where the metadata syntax goes).

Michael Kifer: +1
Andreas Harth: +1
Chris Welty: +1

RESOLVED: Close Issue 58 (Comments) by suggesting people use Dublin Core, RDFS, and OWL properties for metadata, along the lines of http://www.w3.org/TR/owl-ref/#Annotations -- specifically owl:versionInfo, rdfs:label, rdfs:comment, rdfs:seeAlso, rdfs:isDefinedBy, dc:creator, dc:description, dc:date, foaf:maker. This goes in BLD (near where the metadata syntax goes).

ACTION: Harold to add metadata syntax and commentary to BLD.

trackbot-ng: Created Action 499 - Add metadata syntax and commentary to BLD. [on Harold Boley - due 2008-06-04].
Michael Kifer: <* .... *>
Jos de Bruijn: http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0036.html
Michael Kifer: *i .... *m ...

Michael Kifer: above are alternatives for presentation syntax

Sandro Hawke: one for *i and *m

Sandro Hawke: prefers round or curly braces to pointy ones

Sandro Hawke: (* ... *)
Sandro Hawke: agreed to use (* ... *) for the metadata syntax.

ACTION: kifer to metadata to mathematical description

trackbot-ng: Created Action 500 - Metadata to mathematical description [on Michael Kifer - due 2008-06-04].

PROPOSED: close Issue 55 (striping and xml syntax and rdf/xml syntax compatibility) addressed by decisions made so far this meeting

Sandro Hawke: +0.8
Jos de Bruijn: 0
Dave Reynolds: 0
Chris Welty: +1+1

RESOLVED: close Issue 55 (striping and xml syntax and rdf/xml syntax compatibility) addressed by decisions made so far this meeting

Harold Boley: We could spread the parens around it: (* id meta term *)

Jos de Bruijn: metadata target is ambiguous in presentation syntax, e.g. (* comment *) ?o[s->v] could apply to frame or oid

Harold Boley: could enclose target in the (* *)

Michael Kifer: could add regular parens to disambiguate

Jos de Bruijn: could define precedence for comment "operator"

Jos de Bruijn: by default metadata binds as loosely as possible [Scribe assist by Sandro Hawke]

Sandro Hawke: comment has lowest precedence

Harold Boley: E.g.: (* "http://sample.org"^^rif:iri "pd"^^rif:local[dc:publisher->w3:W3C Group ( rule fact ) *)
Sandro Hawke: Instead: (* id frames *) o[s->v] metadata applies to whole frame

Sandro Hawke: don't have to formalize precedence until/unless we formalize rest of PS

Sandro Hawke: and - ((* id frames *) o)[s->v] metadata applies to const "o"
Harold Boley: Instead: (* id frames o[s->v] *) metadata applies to whole frame
Harold Boley: (* id frames o *) [s->v]
Harold Boley: metadata applies to const "o"
Sandro Hawke: while we're being silly: /* ... */ is shortcut for (* _foo _foo[rdfs:comment->" ..." *)

Michael Kifer: issue with Harold's proposal is most of the time the *) is far away from the (*

Dave Reynolds: Ah I like /* ... */ !

PROPOSED: we don't need to settle things like precedence in the PS for now. we're fine.

Harold Boley: But pretty-printing always helps.
Sandro Hawke: +1
Axel Polleres: Dave, we could use the good old /* */ for real comments in the PS :-)
Axel Polleres: ... wouldn't be too bad actually.
Dave Reynolds: Sure, allow free text but anything parsable as a frame is treated as metadata ;-)
Axel Polleres: My idea is: (* ... *) for metadata in PS, /* */ for commenting out things in PS

PROPOSED: move Issue 57 (xml syntax extensibility) out of critical path

Axel Polleres: which I find nice.
Sandro Hawke: +1
Jos de Bruijn: +1

RESOLVED: move Issue 57 (xml syntax extensibility) out of critical path

Adrian Paschke: +1
Axel Polleres: +1
Harold Boley: Instead: (* id frames function(a1 ... aN) *) metadata applies to whole function vs. (* id frames function *)(a1 ... aN) metadata applies to function only.

Metadata suvivability

Harold Boley: Instead: (* id frames predicate(a1 ... aN) *) metadata applies to whole predicate application vs. (* id frames predicate *)(a1 ... aN) metadata applies to predicate only.
Harold Boley: s/whole function/whole function application/ (failed)

Sandro Hawke: metadata in RIF should be put in comments in target language

Christian de Sainte Marie: comments could be less survivable

PROPOSED: We say metadata (including comments) SHOULD survive the translation from-and-back-to RIF

Chris Welty: but we don't distinguish comments from other metadata

Christian de Sainte Marie: don't care about XML comments

PROPOSED: We say metadata SHOULD survive the translation from-and-back-to RIF

Sandro Hawke: +1
Harold Boley: +1
Dave Reynolds: +1
Mike Dean: +1
Gary Hallmark: +1
Igor Mozetic: +1

RESOLVED: We say metadata SHOULD survive the translation from-and-back-to RIF

Christian de Sainte Marie: should preserve metadata even if you don't understand it

ACTION: Harold to put this in Conformance section of BLD

trackbot-ng: Created Action 501 - Put this in Conformance section of BLD [on Harold Boley - due 2008-06-04].

PROPOSED: closs Issue 59 as discussed in this meeting

Adrian Paschke: +1
Sandro Hawke: +1
Jos de Bruijn: +1
Chris Welty: +1
Axel Polleres: +1
Christian de Sainte Marie: +1

RESOLVED: close Issue 59 as discussed in this meeting

Mike Dean: s/closs/close (succeeded, 1 lines ago)

Relative IRIs

Harold Boley: related to Prefix:

Sandro Hawke: base and prefix are separate issues

... base overrides location of document

Chris Welty: base is for unprefixed IRIs

Axel Polleres: base is not required, but prefix is required

Sandro Hawke: relative IRIs work well for imports

Michael Kifer: don't need both base and prefix

... relative IRIs most useful when no base is specified

Dave Reynolds: The notion of base is well-defined for XML, that's not the issue. The issue is whether to allow relative IRIs or not and if so where.

Michael Kifer: need to define semantics

Chris Welty: its a preprocessing step

Jos de Bruijn: shares michaels concern

Dave Reynolds: presentation syntax should use absolute IRIs with curies. relative IRIs are in XML only

Jos de Bruijn: how to map a relative IRI to domain of interpretation

Dave Reynolds: you don't map relative IRI to domain of intepretation, there are no relative IRIs in the semantics or the abstract structure - this is entirely an XML syntax issue
Chris Welty: We just need some statement that there is a pre-processing step involve base and relative IRIs, turning them into absolute IRIs. [Scribe assist by Sandro Hawke]

we seem to agree that relative to absolute expansion is a preprocessing step

Sandro Hawke: (we all seem agreed on this)
Dave Reynolds: We *could* choose to have relative IRIs in the presentation syntax as well, as Turtle does, but we don't have to and it seems better to not do so right now given how we are treating the presentation syntax as not a concrete syntax
Sandro Hawke: For how you resolve URIs -- what the default Base is -- see for example http://www.w3.org/TeamSubmission/turtle/#sec-uris
Sandro Hawke: or much more verbosely, http://www.w3.org/TR/xmlbase/#resolution
Dave Reynolds: Why do you want relative IRIs in the presentation syntax at all?

Sandro Hawke: wants relative IRIs in PS because of PS->XML mapping

... i.e. roundtrip between XML and PS

Dave Reynolds: +1 just in Xml
Sandro Hawke: two=just in xml, three=in both one=just absolute
Jos de Bruijn: i would object to disallowing relative URIs in the XML. [Scribe assist by Sandro Hawke]


PROPOSED: In Presentation Syntax, the IRIs in rif:iri Consts can be relative. A "base" directive will be added to the preamble.

Dave Reynolds: -0.5

PROPOSED: In Presentation Syntax, the IRIs in rif:iri Consts can be relative.

Sandro Hawke: +1
Andreas Harth: +1
Igor Mozetic: 0
Adrian Paschke: 0
Jos de Bruijn: -0
Christian de Sainte Marie: 0
Harold Boley: +0.25
Gary Hallmark: 0
Axel Polleres: +1
Jos de Bruijn: s/disallow relative/object to disallow relative/ (failed)

RESOLVED: In Presentation Syntax, the IRIs in rif:iri Consts can be relative.

ACTION: kifer to put relative IRI handling, with base directive, in BLD

trackbot-ng: Created Action 502 - Put relative IRI handling, with base directive, in BLD [on Michael Kifer - due 2008-06-04].

ACTION: Harold to update BNF with base directive for relative IRI handling in PS

trackbot-ng: Created Action 503 - Update BNF with base directive for relative IRI handling in PS [on Harold Boley - due 2008-06-04].
Dave Reynolds: what's the syntax in the presentation syntax?
Dave Reynolds: Just in <..>?
Adrian Paschke: a base directive

XML 1.0 or 1.1

Dave Reynolds: Base directive is easy, I mean how to the relative IRIs themselves appear, just in <..>?
Jos de Bruijn: ".."^^rif:iri and <...>
Dave Reynolds: josb - thanks
Harold Boley: Of course, the "base" directive uses an upper case first letter: Base (which is a directive similar to Prefix).

Chris Welty: xml 1.1 may have issues with normalizing strings

PROPOSED: We'll use XML 1.0 (not XML 1.1) for the XML syntax for BLD.

Sandro Hawke: advice from experts is to use 1.0

Sandro Hawke: +1
Igor Mozetic: 0
Adrian Paschke: +1 for 1.0
Dave Reynolds: +1

... use 1.0 "as amended" to pick up new unicode chars

Sandro Hawke: (advice is also to say "as amended", and to use the infoset if possible.) [Scribe assist by Sandro Hawke]

PROPOSED: We'll use XML 1.0 as amended (not XML 1.1) for the XML syntax for BLD.

Jos de Bruijn: http://www.w3.org/TR/xml/
Sandro Hawke: link to http://www.w3.org/TR/REC-xml/
Sandro Hawke: http://www.w3.org/TR/xml/
Sandro Hawke: is right
Jos de Bruijn: http://www.w3.org/TR/xml
Harold Boley: The mapping of the PS level of the above directives to XML is not entirely trivial: it will not just be a simple table row for both Base and Prefix.

PROPOSED: We'll use XML 1.0 as amended http://www.w3.org/TR/xml (not XML 1.1) for the XML syntax for BLD.

Adrian Paschke: +1
Mike Dean: +1
Sandro Hawke: +1 again
Dave Reynolds: +1
Axel Polleres: +1
Jos de Bruijn: ?0
Gary Hallmark: +1
Igor Mozetic: 0

RESOLVED: We'll use XML 1.0 as amended http://www.w3.org/TR/xml (not XML 1.1) for the XML syntax for BLD.

Dave Reynolds: Harold - there should be *no* such mapping

Michael Kifer: back to metadata disambiguation - propose to attach metadata no lower than FORMULA

ACTION: Harold to add XML 1.0 statement to BLD

trackbot-ng: Created Action 504 - Add XML 1.0 statement to BLD [on Harold Boley - due 2008-06-04].
Dave Reynolds: Harold - just as the prefixes should be indenpendent, there should be *no* requirement that the entities in the XML having anything to do with how you choose to define prefixes in the presentation syntax, really

at risk

ACTION: Sandro come up with style for "At Risk" comments in document.

trackbot-ng: Created Action 505 - Come up with style for \"At Risk\" comments in document. [on Sandro Hawke - due 2008-06-04].

Christian de Sainte Marie: rif:text, rif:ordered at risk

Chris Welty: equality?

Christian de Sainte Marie: at risk in head at least

Harold Boley: Dave, did you look at OWL 2's Functional Syntax, which uses uniform Prefix declarations, which will then need to split into two kinds for the XML level?

Michael Kifer: in body its just identity (easy)

Harold Boley: This is one advantage of a Presentation Syntax.
Dave Reynolds: Harold - no, but it's true you would have to do that. The point is the there should be *no* round tripping expectations between presentation syntax and the XML

PROPOSED: make equality-in-the-head (that is, equality that is more than syntactic sugar) a feature-at-risk.

Michael Kifer: equating two built-in-functions can't be re-wrtiten without equality-test. [Scribe assist by Sandro Hawke]

PROPOSED: make equality-in-the-head a feature-at-risk.

Sandro Hawke: +1
Igor Mozetic: +1
Andreas Harth: +1
Christian de Sainte Marie: +1 [Scribe assist by Sandro Hawke]
Axel Polleres: +1
Christian de Sainte Marie: +1
Adrian Paschke: 0
Gary Hallmark: +1
Jos de Bruijn: 0

RESOLVED: make equality-in-the-head a feature-at-risk.

Dave Reynolds: 0 (i'd rather the whole of equality was at risk)

Christian de Sainte Marie: conformance clause at risk?

Sandro Hawke: note discussion of equality-in-body -- cannot be rewritten away in the case of testing for the same result from two built-in functions.
Sandro Hawke: (can just use builti-in tests for equality.)


Christian de Sainte Marie: at risk because of "strict mode"

PROPOSED: Mark "at risk" the strictness part of the conformance clause

Sandro Hawke: +1
Jos de Bruijn: +1
Adrian Paschke: +1
Igor Mozetic: 0
Christian de Sainte Marie: +1

RESOLVED: Mark "at risk" the strictness part of the conformance clause

Harold Boley: +1

Chris Welty: reminder that we can't make substantive changes to non-at-risk parts of the doc after last call

Sandro Hawke: a redo of last call costs at least 4 weeks extra review time

PROPOSED: Advance BLD to Last Call, pending satisfactory completion of the editors decided at this meeting.

PROPOSED: Advance BLD to Last Call, pending satisfactory completion of the edit decided at this meeting.

are we ready for Last Call?

PROPOSED: Advance BLD to Last Call, pending satisfactory completion of the edits decided at this meeting.

we scroll thru the BLD doc, making sure all editors notes are taken care of

ACTION: kifer to make editorial changes associated with DTB links in BLD and remove editor's notes

trackbot-ng: Created Action 506 - Make editorial changes associated with DTB links in BLD and remove editor's notes [on Michael Kifer - due 2008-06-04].

Christian de Sainte Marie: the xml syntax translation table needs editorial work, but not a problem for last call

PROPOSED: Advance BLD to Last Call, pending satisfactory completion of the edits decided at this meeting.

Christian de Sainte Marie: who is the reviewer

Sandro Hawke: we all can be

Sandro Hawke: +1 (W3C)
Igor Mozetic: +1 (JSI)
Gary Hallmark: +1 (Oracle)
Adrian Paschke: +1 (REWERSE)
John Hall: +1 (OMG)
Mike Dean: +1 (SRI)
Jos de Bruijn: 0 (FUB)
Michael Kifer: +1 (self)
Axel Polleres: +1 (DERI)
Harold Boley: +1 (NRC)
Chris Welty: +1
Chris Welty: +1 (IBM)
Christian de Sainte Marie: +1 (ILOG)

RESOLVED: Advance BLD to Last Call, pending satisfactory completion of the edits decided at this meeting.



(Scribe changed to Andreas Harth) agenda planning


prd

Christian de Sainte Marie: changes since last time

... tightened up overview, removed diagrams in syntax section

... added running example

Sandro Hawke: i'd benefit from a five-ten minute summary what prd is

Christian de Sainte Marie: not an attempt to improve pr language, but on common serialisation format

... tried to use the same syntax as bld, but not the case everywhere (some could be expressed in bld, some cannot)

... operational semantics is being described in terms of states and transitions

Jos de Bruijn: josb has joined #rif

... proposed table to convert xml syntax to presentation syntax

... i think we can do without presentation syntax in prd

... semantics specified using plodkin-style system

... next are rule instantiation, conflict resolution,and halting test

... useful to describe the semantics in three-step system rather than just one function

... method to resolve conflicts differs from system to system, other methods are shared between systems

... conflict resolution requires some discussion

... halting test also differs across systems

Harold Boley: question about three-step approach - is it applied to single rule or the entire process?

Christian de Sainte Marie: describing the transition from one step to another step describes the semantics

Harold Boley: could finite state machines or finite automata also used?

Christian de Sainte Marie: number of states is not necessarily finite

Harold Boley: datalog's not necessarily finite if your alphabet is infinite

Gary Hallmark: here we can define new things

Chris Welty: let's look at the reviews, start with gary's

Michael Kifer: question regarding halting test

... assume you have core, have a test to stop after five applications, you may end up without a model?

Christian de Sainte Marie: yes

... if you have only assert in conlusion and no negation in condition, per default it should halt when you have minimal model

Michael Kifer: is there some generally accepted way to explain that?

Christian de Sainte Marie: given that description of semantics you cannot use fixpoint as a test, halt is when you have no transition any more

Harold Boley: result is a set of facts? you consider only the final configuration but not the actions?

Christian de Sainte Marie: same as when you're looking at brd ruleset, you're only interested in the model

... actions not covered in the semantics yet, we have to discuss that

... i wrote down the common points of pr systems, left out other things

Harold Boley: in the current version only working memory can be changed

Christian de Sainte Marie: nice property of semantics is that it's very compact

... execute might be difficult (e.g. what transition is a print?)

Gary Hallmark: if x != 0 inc(x) and start with x=1 the system will never terminate, if true inc(x) will only fire once in my system

Christian de Sainte Marie: http://lists.w3.org/Archives/Public/public-rif-wg/2008May/0166.html

Christian de Sainte Marie: discussing gary's comment

Chris Welty: what about syntax?

Christian de Sainte Marie: derive presentation syntax from xml (bld is organised the reverse way)

Adrian Paschke: AdrianP has joined #rif

Gary Hallmark: should the organisation of prd follow the bld document?

Christian de Sainte Marie: not necessarily, easier to go from xml to presentation syntax

Chris Welty: issue is presentation not how syntax was derived

... do you use the presentation syntax at all?

Gary Hallmark: 1.3.2 example has different presentation syntax than defined

Christian de Sainte Marie: syntax is not yet defined in the beginning, just use a rule syntax

Chris Welty: let's first identify and collect issues, then resolve

Christian de Sainte Marie: if we use imply in bld we'll use it in prd as well

... forall can be nested in prd

... constraints on variable can be imposed in the forall

Chris Welty: is that necessary?

Harold Boley: http://www.w3.org/2005/rules/wg/draft/rif-prd/#Forall

Christian de Sainte Marie: not really, in pr system the evaluation of clauses must be ordered, but they really have different scope

... could flatten everything and then reconstruct

... but representation of rules should preserve optimisations that use forall ordering

... but community undecided on that

Chris Welty: any points or issues that should be resolved before 1st wd

Gary Hallmark: need some vision of a core which can span all rule languages

... currently bld and prd does not have any intersection

... syntax side-by-side comparison would be a first step towards common core

s/comparision/comparison (succeeded, 1 lines ago)

Christian de Sainte Marie: two goals: i) get pr community interested (those people do not care about bld)

... ii) but also interoperate with other kind of rules

... pr community will not read bld

... proposal is to have table with similarities and overlaps in seperate section

Gary Hallmark: would nice to have the core more explicit

Christian de Sainte Marie: will align syntax

Gary Hallmark: I should be able to have one syntax for rules which can move between PR systems and Prolog systems. [Scribe assist by Sandro Hawke]

Adrian Paschke: core with both syntactic and semantic overlap is desirable

... syntax should be aligned

Christian de Sainte Marie: forall is the only point where there's considerable differences, will update tagnames to current ones

Gary Hallmark: write pattern and tests as conjunction into one formula

Christian de Sainte Marie: i want to target only the pr community

Gary Hallmark: i'd like to use that standard to bridge logic and production rule communities inside oracle

Christian de Sainte Marie: need that for if-then-else

Gary Hallmark: it's a trivial rewrite using a negation

Christian de Sainte Marie: there's no negation in bld

Chris Welty: would an editor's note on nested forall's suffice for now?

Christian de Sainte Marie: there should even be a specific request for comments from reviewers

Igor Mozetic: IgorMozetic has joined #rif

Chris Welty: if we align syntax, table with overlap between prd and bld, and add editor's note to ask community for comments, can we go to first wd?

Gary Hallmark: re order, you could have a policy that rules can only fire once

Christian de Sainte Marie: order related to age of rule could be another policy

ACTION: csma align syntax, table with overlap between prd and bld, and add editor's note to ask community for comments

trackbot-ng: Created Action 507 - Align syntax, table with overlap between prd and bld, and add editor's note to ask community for comments [on Christian de Sainte Marie - due 2008-06-04].

ACTION: csma to work out policies for pick (refraction, recency, priority, sequential)

trackbot-ng: Created Action 508 - Work out policies for pick (refraction, recency, priority, sequential) [on Christian de Sainte Marie - due 2008-06-04].

Gary Hallmark: it's also common to give a limit on cycles

ACTION: gary and adrian to review draft working draft

trackbot-ng: Created Action 509 - And adrian to review draft working draft [on Gary Hallmark - due 2008-06-04].

Adrian Paschke: we have not discussed update and execute

Christian de Sainte Marie: propose to table update

... say to say we have execute but semantics not defined yet

Gary Hallmark: could even remove update

Christian de Sainte Marie: would be nice if we have assign

... with assign you can remove old value and use a new one

Gary Hallmark: assign used to synchronise working memory with external storage

Harold Boley: If <pattern> is only needed for <Member>-like restrictions on <declare>ed variables, then I think -- instead of introducing a new <pattern> role -- the <declare> role with <Member>-like fillers should be kept. (Still, this restricting use could be called a 'pattern'.)

Gary Hallmark: only have assert and remove for now, execute would need to be integrated with external functions

ACTION: csma to remove UPDATE, EXECUTE and ASSIGN from PRD

trackbot-ng: Created Action 510 - Remove UPDATE, EXECUTE and ASSIGN from PRD [on Christian de Sainte Marie - due 2008-06-04].

PROPOSED: Publish PRD as a FPWD, given the editorial changes decided so far this meeting (after confirmation of edits by Gary and Adrian).

Harold Boley: Can we replace 'remove' with the more typically used 'retract'?

Christian de Sainte Marie: the prr group uses remove

ACTION: csma to change REMOVE to RETRACT

trackbot-ng: Created Action 511 - Change REMOVE to RETRACT [on Christian de Sainte Marie - due 2008-06-04].

PROPOSED: Publish PRD as a FPWD, given the editorial changes decided so far this meeting (after confirmation of edits by Gary and Adrian).

Sandro Hawke: +1 (W3C)
Christian de Sainte Marie: +1 (ILOG)
Harold Boley: +1 (NRC)
Gary Hallmark: +1 (Oracle)
Jos de Bruijn: +1 (FUB)
Chris Welty: +1 (IBM)
Adrian Paschke: +1 (REWERSE)
Axel Polleres: AxelPolleres has joined #rif
Igor Mozetic: +1 (JSI)
Axel Polleres: +1 (DERI)
John Hall: +1 (OMG)
Michael Kifer: 0 (self)

RESOLVED: Publish PRD as a FPWD, given the editorial changes decided so far this meeting (after confirmation of edits by Gary and Adrian).

WG Futures

Sandro Hawke: http://www.w3.org/2005/rules/wiki/Extension_Request_2008

administration, future of working group

Sandro Hawke: future meetings, extension request, publication details

Sandro Hawke: ===Publication Details===

Chris Welty: schedule for DTB?

Axel Polleres: end of the week

Axel Polleres: by May 30 [Scribe assist by Sandro Hawke]

Jos de Bruijn: swc by monday

... June 2

Christian de Sainte Marie: swc review by june 6

ACTION: chris review Axel's changes to DTB

trackbot-ng: Created Action 512 - Review Axel's changes to DTB [on Christopher Welty - due 2008-06-04].

Michael Kifer: BLD until June 16

Michael Kifer: by Jun 16 [Scribe assist by Sandro Hawke]
Harold Boley: by June 6 [Scribe assist by Sandro Hawke]

Chris Welty: working group review of BLD by June 23

Michael Kifer: reivew DTB by June 16

s/reivew/review/ (failed)

Sandro Hawke: Let's try to get UCR on telecons. [Scribe assist by Sandro Hawke]

Christian de Sainte Marie: changes to PRD by June 2

Gary Hallmark: review by June 4

Adrian Paschke: review by June 6

Sandro Hawke: do you make changes to FLD as welL?

Michael Kifer: yes

Michael Kifer: FLD by June 16

Chris Welty: reviewing FLD by June 23

Sandro Hawke: publication date of June 23rd-ish?

Sandro Hawke: publication date -- June 24th.

future of WG

Adrian Paschke: UCR similar schedule for publication
Sandro Hawke: === RIF Extension ===
Sandro Hawke: draft proposal http://www.w3.org/2005/rules/wiki/Extension_Request_2008

Chris Welty: need to ask for another extension - how long before everything's moved to rec?

Sandro Hawke: jos, when are you changing level of participation?

Jos de Bruijn: want to bring bld, swc, dtb to rec, not so much involved with future dialects

Sandro Hawke: need to get bld and swc to recommendation, get implementatons and test suite

... want to fld and dtb torec

Jos de Bruijn: dependencies between bld and dtb

Sandro Hawke: certainly need to be stable

Jos de Bruijn: dtb is essentially part of bld

Sandro Hawke: want to consider to all get them to rec


Sandro Hawke: mdean, we're dialing


Mike Dean: thanks
Christian de Sainte Marie: csma has joined #rif



Christian de Sainte Marie: http://www.w3.org/2005/rules/wiki/Extension_Request_2008
Sandro Hawke: http://www.w3.org/2005/rules/wiki/Extension_Request_2008

Chris Welty: no document on fallback mechanism yet, can we take that to rec realistically?

Sandro Hawke: we should try

... UCR could be a note

Adrian Paschke: interested in events, reaction dialect

Christian de Sainte Marie: must make sure the prd can be extended to events

Adrian Paschke: some reaction rules build on logical formalism, others to pr

Chris Welty: i'm nervous about prd a chair being an editor

... need more participation from the pr community

Chris Welty: discussion about more pr involvement during next telecon

Christian de Sainte Marie: want to re-inforce: wrote and rewrote the draft three times because nobody else wanted to do it

Chris Welty: who's doing the work on core?

Christian de Sainte Marie: we should make clear that core specifies the fragment that is implementable in both logical and production systems

Michael Kifer: can't that be an appendix on BLD, we've had that before?

Chris Welty: core should be core document

Axel Polleres: why not just a document to restrict bld?

... many things were outsourced to dtb

... what's the use of bld then?

Harold Boley: Sandro, "Fallback Mechanism" (XTAN, ...) reminds me of KIF's partial specifications.

Michael Kifer: takes a long time to specify a dialect, way to go is to restrict one dialect

Sandro Hawke: why not have bld and prd editors write two pages to restrict their dialect to core?

Sandro Hawke: i'm tempted to go for 18 months instead of one year

Chris Welty: would prefer to scope work to fit into one year

Chris Welty: why not move core and fallback to wd?

Sandro Hawke: but there's no unity to rif without those

Christian de Sainte Marie: why not commit somebody to combine the specialisations into core? adrian would know both camps

Harold Boley: could we move the name fallback mechanism to interoperation mechanism?

Harold Boley: would not mind downgrading core and fallback mechanism to wd

Michael Kifer: other ways to achieve unity via a framework

Chris Welty: how about merging the two into core and interoperability

Gary Hallmark: how about test cases?

Sandro Hawke: implied in the documents

Sandro Hawke: merged core and fallback mechanism documents in wiki

Michael Kifer: we'd need lpd in

Chris Welty: 18 months is a long commitment for a chair

Christian de Sainte Marie: same as


Mike Dean: i've got another telecon. I'll try to call back in about 90 minutes.

Michael Kifer: what's the process (e.g. for adding events) if the group is finished

Christian de Sainte Marie: at some point need to terminate group to re-charter it

Sandro Hawke: there's wg's like css that run for twelve years now, others like owl finished, paused, and were re-established

(Scribe changed to Axel Polleres)

Chris Welty: I would like to opt for one year.

PROPOSED: The WG requests a 1-year request, with the work plan/description http://www.w3.org/2005/rules/wiki/Extension_Request_2008

Chris Welty: What we have looks like a realistic plan.

Harold Boley: with which events should we align?

... RR, business rules, etc.

PROPOSED: The WG requests a 1-year extension, with the work plan/description http://www.w3.org/2005/rules/wiki/Extension_Request_2008

Sandro Hawke: +1 (W3C)

+1 (DERI)

Igor Mozetic: +1 (JSI)
Christian de Sainte Marie: +1 (ILOG)
Harold Boley: +1 (NRC)
Jos de Bruijn: +1 (FUB)
Chris Welty: +1 (IBM)
Michael Kifer: +1 (self)
Gary Hallmark: +1 (Oracle)
Adrian Paschke: +1 (REWERSE)

RESOLVED: The WG requests a 1-year extension, with the work plan/description http://www.w3.org/2005/rules/wiki/Extension_Request_2008


UCR

Chris Welty: igor, you reviewed it?

Sandro Hawke: we also need to talk about f2f schedule then.

Adrian Paschke: completely restructured the draft according to our discussions.

... compacted CSFs removing discussions, pictures, etc.

that was section 3.

s/3/2/ (failed)

now Section 3.

Section three outlines the documents.

Sandro Hawke: Isn't this kind of like the reader's guide we need anyway?

Michael Kifer: Is it still "Use Cases and Requirements"?

Jos de Bruijn: this section is neither UC nor R.

Chris Welty: could go into a separate overview/guide document.

Adrian Paschke: Section 4

Axel Polleres: abridged syntax now in DTB

... should be pointed at.

Discussion about syntax in the examples.

... Escpecially about :- vs. <-

Chris Welty: This needs to be changed to RIF PS.

Christian de Sainte Marie: ad overfull boxes on the drafts, you should use colons for intention, then it works.

Adrian Paschke: Use cases will be updated. next section Requirements.

... distinction between phase 1 and phase 2 changed to "general requirments" and "objectives" which we could address but we do not promise.

Sandro Hawke: change "Objectives" to "Expected Future Requirements" or "Wishlist"

ChrisW/Sandro: Objectives should rather be "future requirments" or "desiderata" or "wishes".

Chris Welty: you think XYZ June is realistic deadline?

Adrian Paschke: yes.

Chris Welty: We need UCR on the Tele conf agenda in two weeks.

ACTION: csma put UCR (esp. requirements) on agenda

trackbot-ng: Created Action 513 - Put UCR (esp. requirements) on agenda [on Christian de Sainte Marie - due 2008-06-04].

s/XYZ/16/ (failed)

Sandro Hawke: requirements should not have been changed.

Adrian Paschke: I only restructured.

Igor Mozetic: We should have an example with blanknodes in the head/skolemization, to indicate limitationes.

Adrian Paschke: this is basically UC1, could be another example there.

Sandro Hawke: we should have a one-paragraph version of each UC aside vidually.

Christian de Sainte Marie: check whether this is already there, we had that request already.

Harold Boley: Maybe move things to an appendix?

Adrian's gonna look at making the UCs shorter along these lines.

Chris Welty: After this version, lets see if we can cut the use cases way down, to a compelling core. [Scribe assist by Sandro Hawke]

(No activity for 12 minutes)







AxelPolleres has joined #rif


(No activity for 12 minutes)

future F2Fs.

Chris Welty: We have 4 possible dates in October.

TPAC, ISWC, RuleML, RR, Orlando or Karlsruhe or Mandelieu.

Christian de Sainte Marie: we have september dificulties typically with people on teaching assignments.

Jos de Bruijn: I am travelling secnd half of Sept.

Chris Welty: at least one person opposing TPAC, Oct 23-24

Sandro Hawke: OCT 23-24 in south of france (at TPAC).... Who likes? 6

Christian de Sainte Marie: We might consider have one more in between.

... last week of august.

Sandro Hawke: I can host any of the august meetings.

Gary Hallmark: I could try in Oregon.

Suggested dates 28-29 August.

... for both those options.

Portland pro: 4, Boston pro: 5.

Sandro Hawke: I have to back off 29th, that is a bad time for MIT.

... will change for alternatives.

s/change/check/ (failed)

Sandro Hawke: back to previous state: 28-29 IS ok for MIT.

FLD

?

ACTION: Gary confirm hosting offer for Aug 28-29

trackbot-ng: Created Action 514 - Confirm hosting offer for Aug 28-29 [on Gary Hallmark - due 2008-06-04].

ACTION: Sandro confirm hosting offer for Aug 28-29

trackbot-ng: Created Action 515 - Confirm hosting offer for Aug 28-29 [on Sandro Hawke - due 2008-06-04].

FLD

Discussion (guest): FLD should not *require* other dialects to be derived from FLD by spezialization but that should be weakened to "expected" or alike.

Sandro Hawke: something like: The WG plans to base future standard dialects on FLD, so any dialects being developed to become a stardard should anticipate fitting in with FLD or justify deviating from (or extending) FLD. [Scribe assist by Sandro Hawke]

Christian de Sainte Marie: it is ambiguous whether the PS normative or not.

I don't know whether this is also in BLD or FLD, but I removed this in DTB: "The compact URI notation is not part of the RIF-BLD syntax." Likewise, I think all things which say that the PS is not normative should be removed.

(that was not a scribecomment but a personal one)

Dave Reynolds: DaveReynolds has joined #rif

Discussion on that the EBNF does not represent the Presentation syntax, since it misses some constraints of it.

Michael Kifer: we will not change this, but add a clarifying text.

Christian de Sainte Marie: before going to WD, do we need more reviews?

Chris Welty: I think we need to have it reviewed once again.

Christian de Sainte Marie: let's ask for reviews and then decide for publication.

Chris Welty: we have only June 16-23 for the reviews.

Jos and ChrisW will review FLD

ACTION: chris to review FLD [june 23]

trackbot-ng: Created Action 516 - Review FLD [june 23] [on Christopher Welty - due 2008-06-04].

ACTION: jos to review FLD [june 23]


ACTION: jdebruij2 to review FLD [june 23]

trackbot-ng: Created Action 517 - Review FLD [june 23] [on Jos de Bruijn - due 2008-06-04].

PROPOSED: conditional on reviews by Jos and Chris, publish FLD as 2cnd WD

Chris Welty: +1 (IBM)
Igor Mozetic: +1 (JSI)
Gary Hallmark: +1 (Oracle)
Jos de Bruijn: +1 (FUB)
Sandro Hawke: +1 (W3C)
Harold Boley: +1 (NRC)
John Hall: +1 (OMG)
Chris Welty: +1 (Ilog)
Michael Kifer: +1

+1 (DERI)

RESOLVED: conditional on reviews by Jos and Chris, publish FLD as 2cnd WD

PROPOSED: BEER!

Igor Mozetic: +2
Chris Welty: +1
Jos de Bruijn: +1
Gary Hallmark: +3

+0.5 (Bulmers)

RESOLVED: BEER!