Admin

resolved: f2f7 minutes accepted

Liason

<PaulVincent> No PRR news

josb: who is owl liason?

sandro: 1st telecon tomorrow
... for new OWL WG

<Doug> I expect to

sandro is OWL liason

csma: sandro, any f2f8 updates?

action on sandro to send reminder to register for f2f8

BLD WD2 issues

csma: split BLD?

josb: defer until after WD2

csma: docs may not be complete enough to split

sandro: is OWL/RDF separable (and ignorable)?

josb: they are completely separable
... no dependencies from main doc

harold: also concerned about separability

sandro: split may not take much effort

csma: editor's note could describe possible future doc split
... editor's notes to call out sections that are not required for specific purposes

josb: confident of semantic independence of owl/rdf section

sandro: good idea to split now

who is in favor of splitting?

<PaulaP> +1

<PaulVincent> +1

<Doug> +1

<Hassan> +1

<sandro> +1 splitting

<MichaelKifer> +1

<LeoraMorgenstern> 1

<LeoraMorgenstern> +1

<IgorMozetic> +1

<AdrianP> -1

<josb> 0

<Harold> 0

<csma> 0

<DaveReynolds> 0

<AllenGinsberg> 0

csma: clear preference to split

<ChrisW> i didn't hear jos' argument

<ChrisW> but I think I am opposed to spitting

<sandro> worth being clear -- in this case "-1" did NOT mean "object". It was already set that no one was objecting.

<csma> PROPOSED: to split RDF compatibility from BLD and publish 2 drafts simultaneously

<ChrisW> wait ten minutes for this vote please

<csma> ok

<sandro> shortname "rif-rdf-owl" not "bld-rdf-owl"

<Harold> I gather we are talking only about RIF-RDF Compatibility

<Harold> not about RIF-OWL Compatibility, for now.

<josb> Harold, the document will also contain OWL compatibility (not in working draft 2, but it should in working draft 3)

csma: prefer rif-bld rather than superscript notation

<MichaelKifer> i dont care. rif-bld is fine

csma: any objections to RIF Basic Logic Dialect and RIF-BLD

<PaulVincent> +1 for RIF-BLD assuming no logicians will be injured...

<csma> PROPOSED: the name of the dialect is RIF basic logic dialect and the short notation is RIF BLD

<Harold> (Sorry for being late with this: Jos, I see. So what would be the short name of the working draft 2 (actually *its* working draft 1) of RIF-RDF Compatibility?)

<ChrisW> how about RiFbLd

<sandro> roflbld

<ChrisW> R-FB*D

csma's proposal is resolved

<Harold> rif:text

<Harold> Michael, RIF-BLD

<josb> It could be called "RIF Semantic Web Compatibility" or "RDF and OWL compatibility of RIF"

csma's resolution ammended to RIF-BLD

csma: do we keep OWL compatibility in WD2?

josb: should be out, because there is no substance yet

csma: subclass and membership in BLD?
... keep and mark as open issues

<Harold> Subclass and membership: Keep it and mark it.

no objections to keep and mark

<josb> there should be a link to the issues tracker

dave: rif:text issues
... <rif:text> vs. string@lang

josb: lexical space is unicode strings, not pairs (string, tag)

<sandro> this is very weird.....

dave: withdraws objection

<sandro> +1 DaveReynolds, I have a worry.... but it's not specific enough to object.

<sandro> Jos: This is a result of putting language tags into the datatype mechanism

<ChrisW> issue for the issues list?

sandro: let others raise the issue if they are worried

<sandro> (very very weakly.)

hassan: issue should be raised

<sandro> ACTION: Dave to raise an issue about xml:lang in rif:text [recorded in http://www.w3.org/2007/10/09-rif-minutes.html#action01]

<rifbot> Created ACTION-356 - Raise an issue about xml:lang in rif:text [on Dave Reynolds - due 2007-10-16].

csma: any objection to rif constants being scoped to ruleset?
... rif local constants

<Harold> Constants rif:local should be scoped to an entire ruleset, otherwise you could not express even simple Prolog textbook examples.

<PaulVincent> +1 to Hassans concerns

no objections to ruleset scope

<josb> one can also use URIs for textbook examples

hassan: but other scopes could be reasonable as well

<csma> c =/= c()

<Harold> likes(john,X) :- likes(X,wine).

<Harold> likes(mary,wine).

<Harold> wine must be scoped to the entire ruleset,

should a constant be distinct from 0-arity function with same name?

<Harold> current-time() would an (active) built-in!

<Harold> current-time would be just a (passive) constant.

harold: they are different in current signature formalism

<ChrisW> this is sounding like the "treatment of builtins" issue

hassan: but logically they are the same (both constants)

michael: should not have 2 different things that are really the same

<IgorMozetic> leave both, a and a()

<Harold> Prolog dosn't have nullary because it's relational. Functional languages do have such nullary functions, e.g. for builtins.

chrisw: distinguish between builtins and functions rather than between constants and functions
... (rhetorically)

<Harold> Also for mapping operator arguments to lists, it's necessary to keep zero arguments corresponding to the empty list [].

<Harold> f(...) can be mapped to unary f([...]).

<Harold> f() can be mapped to unary f([]).

<Harold> PROPOSED: to split RIF-RDF Compatibility from BLD and publish 2 drafts simultaneously: RIF-BLD (WD 2) and RIF-RDF (WD 1).

chrisw: either 2nd doc is part of BLD or a new dialect
... what does 2nd doc describe?

harold: its a bridge for interoperability, not a dialect
... also applies to dialects other than BLD

josb: 2nd doc describes how 2 languages work together
... also don't have to read RDF doc if you don't care to interoperate with RDF
... not a dialect, but a bridge

<Harold> It starts like this:

<Harold> This section defines combinations of RIF rules with RDF graphs, taking into account the various (normative) entailment regimes defined by RDF. A typical case where RIF rules and RDF graphs are combined is when an RIF rule set refers to one or more RDF data sets or RDFS ontologies (which are also RDF graphs).

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

<PaulVincent> ?? Why is RDF more important to RIF-BLD versus say XML - not sure I get the rationale for embedding RDF into BLD doc...

chrisw: do I need a bridge doc for FLORA, or for XML Schema?

sandro: need such a doc to roundtrip FLORA <-> RIF

<Harold> We are a chartered to produce docs for RIF-RDF Compatibility and RIF-OWL Compatibility. Not *chartered* for any other language.

chrisw: proliferation of "bridge docs" seems antithetical to RIF
... RDF doc creates questionable precedent

<Harold> Chris, this may be only since RDF, OWL and RIF would be the enhanced *uniform* basis for the semweb.

chrisw: ok to be part of BLD due to special semantic web status of RDF

<Harold> (Other languages would not be part of that enhanced W3C SemWeb 'triad'.)

chrisw: if not a dialect, then it creates new category of "bridge" documents

<Harold> Chris, soon we may need another 'bridge' doc:

<Harold> PRD and BLD.

<Hassan> not exactly chris: we do have a charter to treat RDF and OWL specially

<PaulVincent> +1 to Chris' concerns - this is why this should be out of the base BLD doc ...

josb: no matter what you call it, it has to be specified
... i.e. how will RIF work with data language "X"

<Harold> RIF Core = BLD --bridge-- PRD

<Doug> Yes, that was the general consensus

Dave: back to the question of working with external data models
... we support XML schema and RDF
... not unbounded
... FLORA is "out of bounds"

chrisw: combine RDF interop doc with Architecture doc?

josb: not sure

dave: need more work on Arch doc

<Harold> PROPOSED: to split RIF-RDF Compatibility from BLD and publish 2 drafts simultaneously: RIF-BLD (WD 2) and RIF-RDF (WD 1).

<Harold> PROPOSED: to split RIF Compatibility from BLD and publish 2 drafts simultaneously: RIF-BLD (WD 2) and RIF-RDF (WD 1).

<Harold> PROPOSED: to split RIF Compatibility from BLD and publish 2 drafts simultaneously: RIF-BLD (WD 2) and RIF (WD 1).

<sandro> chris: "rif-comp"

<Harold> PROPOSED: to split RIF Compatibility from BLD and publish 2 drafts simultaneously: RIF-BLD (WD 2) and RIF (WD 1).

<Harold> <sandro>

<Harold> PROPOSED: to split RIF Compatibility from BLD and publish 2 drafts simultaneously: RIF-BLD (WD 2) and RIF-COMP (WD 1).

<sandro> RESOLVED: to split RIF Compatibility from BLD and publish 2 drafts simultaneously: RIF-BLD (WD 2) and RIF-COMP (WD 1).

<Harold> Sandro, WIKI-TR currently doesn't process BLD: >>Warning: SGML2PL(sgml): []:2: Element "p" not allowed here<< etc. (cf. my Friday email)

<sandro> Yes, Harold, I'm working on that. (and hating prolog. :-)

Several people have noted issues with IE browser rendering special symbols in our docs

Arch: naming conventions

sandro: not required for WD2

This is scribe.perl Revision: 1.128  of Date: 2007/02/23 21:38:13  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/250/350/
Succeeded: s/graph/draft/
Succeeded: s/objects/objection/
Succeeded: s/and RIF/and RDF/
Succeeded: s/,/:/
Found Scribe: Gary Hallmark
Found ScribeNick: GaryHallmark
Default Present: csma, Sandro, +1.703.453.aaaa, me;, josb, PaulaP, DougL, AllenGinsberg, Doug, Hassan, Gary_Hallmark, Harold, DaveReynolds, AdrianP, PaulVincent, LeoraMorgenstern, IgorMozetic, MichaelKifer, ChrisW
Present: csma Sandro +1.703.453.aaaa me; josb PaulaP DougL AllenGinsberg Doug Hassan Gary_Hallmark Harold DaveReynolds AdrianP PaulVincent LeoraMorgenstern IgorMozetic MichaelKifer ChrisW
Regrets: MohamedZergaoui
Agenda: http://lists.w3.org/Archives/Public/public-rif-wg/2007Oct/0026.html
Got date from IRC log name: 9 Oct 2007
Guessing minutes URL: http://www.w3.org/2007/10/09-rif-minutes.html
People with action items: dave