RIF Telecon 24-Feb-2008 -- 24 Feb 2009

Admin

<josb> I don't see many regrets

<josb> http://www.w3.org/2005/rules/wiki/TeleconRegrets

<ChrisW> Scribe: Harold

<ChrisW> scribenick: Harold

<ChrisW> http://lists.w3.org/Archives/Public/public-rif-wg/2009Feb/att-0106/RIF_Telecon_17_Feb_2009.htm

<ChrisW> PROPOSED: accept minutes of 17 Feb telecon

<ChrisW> RESOLVED: accept minutes of 17 Feb telecon

Liason

<ChrisW> Scribe: Axel Polleres

<AxelPolleres> scribe: AxelPolleres

<ChrisW> scribenick: AxelPolleres

chrisw: jos, any news from OWL datatypes?

jos: many people in OWL wg in favor of disjointness. not too much new.

F2F13

axel: SPARQL telecons just started today, will be 1hr before RIF, axel might be late.

chrisw: next rif f2f ind cambridge, MA April 17-18, we need a web page.

<csma> http://www.w3.org/2005/rules/wiki/F2F13

Action Review

chrisw: http://www.w3.org/2005/rules/wiki/F2F13 needs to be filled.

<josb> http://www.w3.org/2005/rules/wiki/SWC#Importing_RDF_and_OWL_in_RIF

ACTION 707 - completed by jos.

<trackbot> Sorry, couldn't find user - 707

josw: action 707 is completed.

axel: action 704 completed.
... action 703 pending review.

<josb> can we have this discussion w/o DaveR?

csma: goes into safeness discussion later on.
... next action goes in Pres Syn discussion next week.

leora: my "whiteboard" action continued.

chrisw: shall we drop paul vincent's action?

csma: paul wanted to keeep it.... will remind him.

chrisw: let's go for actions with pending review.
... action 653 can be closed.

ISSUE-93

<csma> http://www.w3.org/2005/rules/wg/track/issues/93

chrisw: the issue is: should datatype iris map to datatypes themselves?

<ChrisW> If a constant c \in Const is an IRI constant "d"^^rif:iri and d is a

<ChrisW> datatype identifier, i.e., d \in DTS, then I_C(d) is the datatype [2]

<ChrisW> identified by d.

michael: I think this change is bad, because of now unforseeable consequences and lots of implied changes.

jos:

<josb> isLiteralOfType/2

jos: now that we have predicates which refer to datatypes, what datatype should the argument of the second argument be? (definition of semantics now somewhat unelegant)
... literal-equals, is-literal-of-type, etc.

<josb> xsd:int=xsd:string

<josb> isLiteralOfType("a", xsd:int)

jos: a strange consequence is that if you write xsd:int = xsd:string that could have weird consequences, if we equate datatypes and their uris, that would fix the semantics and make that inconsistent.
... michael seemed to hav doubts for symbol spaces being sets of constants not of objects.

michael: "1"^^ xs:string would not become the same as "1"^^xs:int if xs:int=xs:string at the moment.
... because IRIs as datatypes are not IRI constants. Jos' proposal would equate the DT-IRIs with sets of constants, doesn't make sense to me.

jos: this is not by accident. I propose to fix the definition of datatypes.

<ChrisW> isLiteralOfType("a", "xsd:string"^^rif:iri)

<josb> xsd:int=xsd:string

(or isLiteralOfType("a", xsd:string) )

<josb> isLiteralOfType("a", "xsd:int"^^rif:iri)

http://www.w3.org/2005/rules/wiki/DTB#pred:isLiteralOfType

<josb> xsd:int=xsd:string

jos: the second argument should be a datatype, i.e. the iris of datatypes should have a fixed interpreetation.

michael: this is what I wanted to avoid.
... I suggest to have the second argumant of isLiteralOfType being literals, I consider the current def wrong.
... e.g. use xs:string.

<csma> k csma

<josb> so it boils down to who finds what more elegant

csma: what is the definition in XML schema?
... we cannot change the definiition of XML schema datatypes.

<csma> isLiteralOfType(xsd:string, rif:iri)

<josb> xsd:string="a"

<josb> rif:iri=xsd:string

chrisw: we don't change XML schema DTs, but only talk about rif:iri here.

michael: I can send a proposal per mail (please paste link)
... basically, second argument should be a (string) literal.

What's wrong about keeping it like we have it (side question)?

jos: why not do it in the same way as RDF and OWL (where URIs denote datatypes)?
... we don't want xs:int, etc. to denote anything.

p("1"^^a) :- a=xs:int, p(1)

chrisw: maybe we should move on...

ISSUE-92

<csma> http://www.w3.org/2005/rules/wg/track/issues/92

josb: both michael and I don't like the current definition, I personally would prefer even michael's solution to the current definition.

<josb> http://www.w3.org/2005/rules/wiki/DTB#Functions_and_Predicates_on_Strings

http://www.w3.org/2005/rules/wiki/DTB#Functions_and_Predicates_on_Strings

<josb> we allow the equivalent forms External( func:concat( str1, str2) ) and External( func:concat( str1 str2 str3 ) )

<josb> In my view it does change the bld syntax

<Harold> Polyadic syntax for associative binary functions should be not a problem.

axel, michael, jos: discussing multiple arity of external functions such as func:concat.

jos: originally the requirement for fixed arity and disjoint names came from the same restriction in FO-logic definitions.

chrisw: the quickest way to go is to leave it the way to go is to leave it the way it is and reword DTB just to reflect that this is just for explanatory purposes in DTB.

josb: quite sure that there are no problems if we do it for external preds/funcs, but not sure about logical preds/funcs.

csma: do not remember implementation discussion at this point.

chrisw: is there anyone on the phone objecting arbitrary arity for anything?
... (anything means logical and external), implementations seem to support it perfectly well.
... noboddy objects, something to think about.

Safeness/Termination

axel: agree that we can't leave the description of multiple arity like that.

<josb> http://www.w3.org/2005/rules/wiki/Core#Safeness

csma: we should discuss issue-80 first.

Issue-80

<csma> http://www.w3.org/2005/rules/wg/track/issues/80

last mail on that: http://lists.w3.org/Archives/Public/public-rif-wg/2009Feb/0072.html

chrisw: generic comparison should only work if datatypes are the same.

<josb> my feeling: not have generic comparison

axel: but if not, should they be false or undefined? (currently false)

<josb> if we have them: get rid of specific ones and make sure the generic ones behave exactly like the specific ones do now

csma: do we have an equality predicate that does not do propagation?

chrisw: yes "="

csma: that makes is-literal-equal redundant?

"Editor's Note: pred:literal-equal seems to be partially redundant w.r.t the usual equality in BLD."

jos: our current equality and neither is-literal-equal does not cover "=" in Xpath/XQuery, but the specific ones do

jos, it is not talking about 1.000 = 1.00, is it?

jos: equality is not identity in XML schema.

<ChrisW> ACTION: jos to find an example of equality between datatypes within the same value space that is not identity [recorded in http://www.w3.org/2009/02/24-rif-minutes.html#action01]

<trackbot> Sorry, amibiguous username (more than one match) - jos

<trackbot> Try using a different identifier, such as family name or username (eg. jdebruij2, jderoo)

jos: I will draft an example.

<ChrisW> ACTION: josb to find an example of equality between datatypes within the same value space that is not identity [recorded in http://www.w3.org/2009/02/24-rif-minutes.html#action02]

<trackbot> Created ACTION-710 - Find an example of equality between datatypes within the same value space that is not identity [on Jos de Bruijn - due 2009-03-03].

<josb> yes

chrisw: I have no problem with literal-equal being redundant, any problems with literal-not-equal.

josb: I would rather want to drop them both.

axel: can probably dial in next week earlier, but not all of the time.

- DRAFT -

RIF Telecon 24-Feb-2008

24 Feb 2009

Attendees

Contents

Admin

Liason

F2F13

Action Review

ISSUE-93

ISSUE-92

Safeness/Termination

Issue-80

Summary of Action Items

Scribe.perl diagnostic output