See also: IRC log
Chris, should I be scribing this?
<ChrisW> hang on, leora
<sandro> Doug: How will we prevent a harmful proliferation of dialects?
<ChrisW> Scribe: Leora Morgenstern
<ChrisW> scribenick: lmorgens
<ChrisW> last meeting minutes: http://lists.w3.org/Archives/Public/public-rif-wg/2010Jul/att-0047/RIF_telecon_27_July_2010.htm
<sandro> Sandro: I'm hoping the mailing list (email@example.com) will help people coordinate and make new dialects be interoperable.
<ChrisW> PROPOSED: approve last minutes
<ChrisW> RESOLVED: approve last minutes
No agenda amendments
<sandro> doug: I think it's importat for the WG to say something about this, pushing people toward the mailing list and toward working together on these dialects.
sandro: nothing new in SPARQL; Chime is planning to respond to Jos's email.
sandro: 2 open issues and 1
ought-to-be-open issue about import.
... It appears, from discussion with Dave, that underlying issue of these issues has been resolved; not sure yet.
... 2 issues: whether to use RDF type arcs in encoding
dave: OWL uses type arcs
... it's not just the syntax: OWL has the RDF semantics for these as well. (check)
sandro: differences between Dave's preferences and mine may reflect two different user groups with which we deal.
leora: Are there any specific examples
you can give of how your approaches would differ?
... I am finding this discussion somewhat abstract; others may also.
dave: preference -- use xml syntax, and RDF type arcs to disambiguate
chris: confirming that we're at stage where either is open to other's position.
<sandro> predicate+function instead of just op, univars+exivars instead of something like declare, allTrue+anyTrue instead of formula(s)
chris: confiming that argument against type arcs is that they are more verbose.
<ChrisW> PROPOSED: No type arcs in RDF syntax
chris: <audibly flips coin> tails, therefore no type arcs.
<sandro> chris flips a coin, in comes up as proposed.
<DaveReynolds> +0 (preference for type arcs, as per email, but no fundamental objection)
leora: -1 million
leora: My preference is for type arcs, too.
<DougL> (FOR the type arcs)
leora: Got disconnected; will call back
<mdean> also prefers rdf:type arcs
<ChrisW> PROPOSED: use type arcs in RDF syntax
<sandro> +0 it's fine
<ChrisW> RESOLVED: use type arcs in RDF syntax
<ChrisW> RESOLVED: use type arcs in RDF syntax (Closing issue 101)
<sandro> need plural version of formula and declare.
<sandro> suggesting: formulas, vars
<DaveReynolds> seems OK to me
leora: no preference
<sandro> PROPOSED: close issue 102 with names formulas and vars
leora: Webster's has both formulae and formulas
<DaveReynolds> 0 (I'd normally use formulae but don't care :))
leora: sandro, I'm wrong: Webster's gives formu-ligh and formu-lee as the pronunciations for formulae
<sandro> RESOLVED: close issue-102 with names formulas and vars
leora: but it doesn't matter, because we're using formulas
sandro: discuss issues of linking RIF,
... we need some RDF triple that says go use this RIF rule set. OWL uses a unary import to do this. We need a binary import, because we have to specify a profile.
<sandro> <docURI> rif:importedWithProfile <profileURI>
sandro: still must determine what to call it; Axel calls this useWithProfile
<sandro> <URI-of-imported-document> rif:importedWithProfile <profileURI>
sandro: Does anyone have strong feelings, or can Axel and I just go forward?
chris: Is the intent to say: run these RIF rules over RDF document, or merge RDF syntax of RIF rules with graph that I have?
sandro: the first. Think of it as an assert.
chris: that's not an import.
<DaveReynolds> +1 to Chris, owl really is an import it is an inclusion, this is a processing directive
sandro: Still, that's what OWL calls
... would you be happier if we used Assert?
Doug: Maybe something like plug-in
<sandro> <URI-of-ruleset-to-assert> rif:assertWithProfile <profileURI>
<sandro> <URI-of-ruleset-to-assert> rif:usedWithProfile <profileURI>
<sandro> Chris: I like "used"
Chris: Anyone read Axel's proposal, besides Sandro? Feedback?
chris: Sandro, sounds like you are saying: you want the RIF rules to run over graph, but don't necessarily want the triples.
sandro: I'll take guidance from this
call (regarding "used"), will write text with Axel, and get
back to you.
... <discussion of feasibility of doing reverse translation>
<sandro> ACTION: sandro add text to RIF-in-RDF about documents being of type RIFCoreDocument, etc. [recorded in http://www.w3.org/2010/08/10-rif-minutes.html#action01]
<trackbot> Created ACTION-1037 - Add text to RIF-in-RDF about documents being of type RIFCoreDocument, etc. [on Sandro Hawke - due 2010-08-17].
chris: Christian and Gary need to be here to discuss RIF/XML doc
<leora and chris to finish up their parts of the Primer this afternoon; leora to send email to reviewers (Sandro and Gary)>
chris: Discussion of Test Cases.
... do we time-stamp the test case repository? leave it open in case anyone wants to add new test cases?
sandro: would like to keep it open
... but that requires someone to maintain it.
... could label pages/test cases as WG approved, community approved (e.g., 3 people say in email they've looked at it and approved it)
... If RIF is somewhat successful, people should be willing to do that.
... Test Case doc should be published as note instead of working draft.
chris: we should go through remaining
proposed test cases, see if they should be approved.
... mostly PRD, one RDF combination.
chris: <proposes going through some non-PRD cases now>
leora: I'm not clear: why shouldn't it import?
<ChrisW> PROPOSED: accept RDF_Combination_Invalid_Constant_1
leora: ah, okay, I see that now.
<sandro> If you wanted to say ex:a ex:p "http://example.org/#b"^^rif:iri you would write it in turtle as ex:a ex:p <http://example.org/#b>
<ChrisW> RESOLVED: accept RDF_Combination_Invalid_Constant_1
<ChrisW> If ui, 1 ≤ i ≤ n, refers to an RDF graph that uses a typed literal of the form "s"^^rif:iri or "s"^^rdf:PlainLiteral, the document must be rejected.
<sandro> PROPOSED: accept RDF_Combination_Invalid_Constant_2
<ChrisW> ACTION: Chris to move approved test cases to approved status [recorded in http://www.w3.org/2010/08/10-rif-minutes.html#action02]
<trackbot> Created ACTION-1038 - Move approved test cases to approved status [on Christopher Welty - due 2010-08-17].
<sandro> RESOLVED: accept RDF_Combination_Invalid_Constant_2
<sandro> Note the TestCase template seems to be broken for second imported documents.
<ChrisW> PROPOSED: approve RDF_Combination_Invalid_Profiles_1
<ChrisW> RESOLVED: approve RDF_Combination_Invalid_Profiles_1
<sandro> Dave: It'd be nice to have a working implementation run this
<sandro> sandro; Indeed.
<ChrisW> PROPOSED: approve UCR_4.1a
<ChrisW> RESOLVED: approve UCR_4.1a
dave: ontology could be more compact; there are some redundancies in the declarations.
sandro: We should approve it anyway. Some noise adds it some flavor.
<sandro> sandro: But I'm fine with trimming it.
<ChrisW> ACTION: leora to send stella recommendations to improve UCR-4.7a test case [recorded in http://www.w3.org/2010/08/10-rif-minutes.html#action03]
<trackbot> Created ACTION-1039 - Send stella recommendations to improve UCR-4.7a test case [on Leora Morgenstern - due 2010-08-17].
chris: action should include renaming the test case, since it no longer points back to ucr.
<DougL> (I am the 512 one I think)
ACTION: Chris to move approved test cases to
approved status [recorded in
[NEW] ACTION: leora to send stella recommendations to improve UCR-4.7a test case [recorded in http://www.w3.org/2010/08/10-rif-minutes.html#action03]
[NEW] ACTION: sandro add text to RIF-in-RDF about documents being of type RIFCoreDocument, etc. [recorded in http://www.w3.org/2010/08/10-rif-minutes.html#action01]
[End of minutes]