Session Start: Fri Feb 08 15:04:29 2002 [15:04] *** Now talking in #rdfcore [15:04] *** Topic is 'RDF Core WG Telecon 15:00-16:00 UTC Friday - Agenda http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Feb/0234.html' [15:04] *** Set by dajobe on Fri Feb 08 14:22:58 [15:05] - +1.617.896.aacc [15:06] + +1.617.902.aadd [15:07] +??P10 [15:07] gk is scribe [15:07] *** DanC has joined #rdfcore [15:07] * DanC hunts down ftf registration info... 13 registered [15:07] rollcall... [15:08] *** JosD has joined #rdfcore [15:08] *** gk is now known as gk-scribe [15:09] ftf registrants: Masinter, Manola, Carroll, Brickley, Horner, Miller, Klyne, Beckett, McBride, Dean, Stickler, De Roo, Connolly [15:09] Masinter? [15:09] frankm told me he couldn't make it [15:09] yeah... Masinter? [15:09] stevep returns [15:09] * em wonders how many Masinters there are and if this is Larrt [15:09] err. Larry [15:09] yes, it's Larry [15:10] *** PatHayes has joined #rdfcore [15:10] gk taking minutes offline ... [15:10] fyi: 17 registered for the thu RDF IG meeting [15:11] testing [15:11] hey pat [15:11] it's working, PatH [15:11] Larry is coming, eh? [15:11] evidently, gk [15:11] observer? [15:11] Item 7 -- actions all done OK [15:11] * DanC checks in with other friday telcon... offers tentative regrets... [15:11] amazing, it works. [15:12] hi pat! [15:12] Item 8 - syntax working draft [15:12] hi Pat!!!!!!!!!!!!! [15:12] Dave - want people to review XML attributes [15:13] Dave's attributes test case approved [15:13] ((URL??) [15:14] http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Jan/0243.html [15:14] Waiting for some final issue resolutions [15:14] item 9 - postponed (jang absent) [15:14] Item 9 test cases WD [15:14] Jan's action -- continues [15:14] CVS access for Jan -- continues [15:15] Others still have problems with CVS access; problems with authentication [15:15] Item 10 - primer [15:15] Pointer to latest draft has been circulated [insert URL] [15:16] *** danbri has joined #rdfcore [15:16] new primer announcement - http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Feb/0237.html [15:16] (Not same as pointer on RDF home page) ... [thanks em!] [15:16] * danbri late, dialing in shortly [15:16] * em nominates gk-scribe to be an official contributor to primer [15:17] * gk-scribe thinks: like MT, have current and editors working copy pointers??? [15:17] I took a quick peek at http://www.w3.org/2001/09/rdfprimer/rdf-primer-20020127.html . it's getting there. still a little stilted for my taste, but getting there. [15:17] ACTION Brian McB: Update RDFcore home page to point new primer version [15:17] Item 11 Model theory WD [15:17] danbri: MT WD? [15:18] Still awaiting publication. [15:18] see email :( --- ball thoroughly dropped this week [15:18] ACTION: eric miller, re-allocate action to publish this [15:18] yeah, just checking you didn't want to say more [15:18] (publication was approved last week)# [15:18] PatH: is it a problem if there's a new version in 2 weeks??? [15:19] Nobody would object. [15:19] Item 12, MT for containers [15:19] oh crap; I still haven't written up my idea on MT for containers. [15:19] oh well. [15:19] Suggestion: simply treat containers as ordered, with the type name beinbg a hint to implementers about how to use them. [15:20] There is a problem ... (missed details) [15:21] +danbri [15:21] I was thinking of an asBag() function: asBag([x,x,y]) = asBag([x,y,x]). so _:x a rdf:Bag implies _:x rdfaux:asBag _:b. [15:21] Patrick: wants there to be a common superproperty for all container membership properties to simplify enumeration of contents [15:23] Brian: this is beyond current charter, don't want to take it on now. [15:23] I use the idiom { ?elt ?n ?cont. ?n rdf:type rdfs:ContainerMembershipProperty } implies ... . It would perhaps be nicer to have a superproperty of the container membership property, rather than (or in addition to) a class of them, but I can live with either. [15:23] ACTION: Brian, add issue for common superproperty for rdf:_1, ... etc. [15:24] ... to issues list [15:24] beyond charter? we can change RDFS, no? [15:24] we can change RDFS, indeed. [15:24] DanC, danbri: are you on the call? [15:24] no [15:24] or just irc? [15:25] (we might decide rdfs:StatementOccurance worth defining, for eg.) [15:25] Problem(ette?) with proposal ... bagId cannot specify which is _1, _2, etc. [15:25] just joined call. +danbri [15:25] thanks [15:25] ... hence difficult to construct definitive mapping to N-triples; proposed MT would treat these as different, even though intent is that they're clearly the same [15:26] EricM, feel free to take any of my comments here and read them into the meeting. [15:26] they're being read as you type :) [15:27] DanBri: asks for implementation feedback on this issue. [15:28] s/asks/offers/ I think? [15:28] * em is looking for latest MT doc to publish... http://www.coginst.uwf.edu/users/phayes/w3-rdf-mt-current-draft.html seems to be timing out? [15:28] DanBri: suggests that the schem,a spec can easily accommodate some of these "smaller" changes to the original RDF specs [15:29] em: WFM [15:30] Brian: shouldn't waste time/effort patching up something that's broken, to have it reworked by a subsequent WG [15:31] PatH: thinks this would be easy to fix; Brian fears deatjh by a thousand cuts; Pat promises not to ask for any more ... [15:32] PROPOSED: extend container vocabulary with generic container membership property, superproperty of rdf:_1, :_2, etc. Nothing else. [15:33] I 2nd that. [15:33] keep ContainerMembershipProperty? [15:33] Patricks asks why not rdf:li, Chair says not for discussion now. [15:33] keep that too. [15:34] ACTION: PatH, white up container proposal and put to the group for approval. [15:34] s/white/write/ [15:34] [thanks...] [15:35] ACTION: jjc, post URI of description of bagId problem to the list [15:36] ACTION: Brian, as above if Jeremy cannot [15:37] ... (keyword for problem -- non-deterministic generation of N-triples) [15:37] Item 13. Reification [15:37] No test cases were sent t PatH, so nothing to present. [15:38] *** DanCon has joined #rdfcore [15:38] FrankM: has assembled process proposal from earlier messages. [15:39] Process suggested is basically to assemble possible *sets of* entailments, and decide between the proposals. [15:39] http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2001Oct/0013.html [15:39] +Sergey [15:39] (The bagID ambiguity problem= [15:40] ah, I was looking for "non-deterministic" [15:40] (Thanks -- I'll put that in the minutes) [15:41] FrankM's message http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Feb/0239.html [15:41] 3 cases: (1) two reified statements; property on one entailed to the other [15:41] (2) statement entails existence of its reification? [15:41] (3) reification entails existence of statement? [15:44] Sigh... 26 months of statings/statements discussion. -- http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0068.html Meaning of RDF:Statement From: Dan Brickley (danbri@w3.org) Date: Sun, Dec 12 1999 [15:44] *** JosD has quit IRC (Remote closed the connection) [15:46] [[DanBri: I think we're looking at exactly that issue]] [15:47] [[I think we're very limited for examples without stepping outside core RDF]] [15:47] 2Use the editor's working draft pointer on the WG home page, its [15:47] http://www.coginst.uwf.edu/users/phayes/w3-rdf-mt-current-draft.html [15:48] ok, thanks PatHayes [15:50] RonD: provenance was very much a vital use-case for reification. If the original group didn't get it right then let's get that use-case sorted [15:50] *** Disconnected Session Close: Fri Feb 08 15:50:39 2002 Session Start: Fri Feb 08 15:52:26 2002 [15:52] *** Now talking in #rdfcore [15:52] *** Topic is 'RDF Core WG Telecon 15:00-16:00 UTC Friday - Agenda http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Feb/0234.html' [15:52] *** Set by dajobe on Fri Feb 08 14:22:58 [15:52] *** GKGK is now known as gk-scribe-again [15:52] Patrick: I don't see point of Statement view [15:53] Chaur: lets have straw poll? [15:53] RonD: Is the intent to match the original intent, or just clarify the document?? [15:53] *** JosD has joined #rdfcore [15:54] RonD: do we agree that support for provenance use-case is more important than exact conformance to current spec?? [15:54] *** DanC has quit IRC (Read error: 110 (Connection timed out)) [15:55] RonD: Do we work to support provenance use-case, or formalize existing reification?? [15:56] *** sergey has joined #rdfcore [15:58] The group overwhemingly, unanimously supports that we should focus, in principle, on addressing the provenance use-case. [15:59] Brian: Can we reduce this to a choice between entailment decisions??? [15:59] * DanCon finishes other telcon, tunes in... [15:59] FrankM: the question, then, is how do we do provenance? [16:00] hmm... my technical understanding of provenance goes well beyond anything in the RDF 1.0 spec. I'm nervous. [16:00] PatH: we don't necessarily have to do provenance, so much as arrange RDF that it doesn't frustrate efforts to add in provence. [16:01] We've I think decided that provenance is what we want this stuff to do for us; that doesn't rule out our concluding that (a) M+S got it wrong (b) getting it right can be done later. [16:01] FrankM: need to distinguish between "lierally said" and "said that..." [16:01] SWAD implementation of provenance uses log:semantics (i.e. accessing the state of the web) and log:includes (which perhaps isn't too far out there). [16:02] Brian: don't want to issue licence to go and do a whole bunch of new stuff. [16:03] *** gk-scribe has quit IRC (Read error: 104 (Connection reset by peer)) [16:03] * gk-scribe-again test [16:03] * gk-scribe-again test: ack [16:03] ack [16:03] *** dajobe has quit IRC (Read error: 104 (Connection reset by peer)) [16:03] it would be nice to know which host the logger connects to [16:03] *** dajobe has joined #rdfcore [16:04] Item 14, datatypes [16:05] Patrick/Sergey are working on a synthesized proposal; PatH has made some suggestions. [16:05] * gk-scribe-again I think PatH approach was good, because literals denote strings, yet it seems to address the other issues. [16:06] PatrickS: likes PatH proposal everything but section 3. [16:07] ---- [16:07] ACTION FrankM, to take reification discussion forward based on addressing provenance use-case. [16:07] Meeting closed. [16:07] *** gk-scribe-again is now known as gk [16:08] * dajobe -> #rdfig [16:08] *** dajobe has left #rdfcore [16:09] PatH proposal -- returns to tidy literals denotomng strings; all the other clever stuff is now loaded on rdf:value (and subproperties) [16:09] rdf:value means roughly: has a literal representation as