14:57:40 RRSAgent has joined #rdf-wg 14:57:40 logging to http://www.w3.org/2012/10/24-rdf-wg-irc 14:57:42 RRSAgent, make logs world 14:57:42 Zakim has joined #rdf-wg 14:57:44 Zakim, this will be 73394 14:57:44 ok, trackbot; I see SW_RDFWG()11:00AM scheduled to start in 3 minutes 14:57:45 Meeting: RDF Working Group Teleconference 14:57:45 Date: 24 October 2012 14:58:24 zakim, this is 73394 14:58:24 ok, AndyS; that matches SW_RDFWG()11:00AM 14:58:33 zakim, who is on the phone? 14:58:33 On the phone I see +31.20.598.aaaa, ??P1, ??P9 14:58:43 zakim, ??P9 is me 14:58:44 +AndyS; got it 14:58:51 Zakim, ??P1 is me 14:58:51 +yvesr_; got it 14:58:56 zakim, +31.20 is me 14:58:56 +Guus; got it 14:58:58 Zakim, ??P1 is me 14:58:58 I already had ??P1 as yvesr_, yvesr 14:59:09 zakim, this if rdf 14:59:09 I don't understand 'this if rdf', Guus 14:59:14 +??P3 14:59:18 zakim, this is rdf 14:59:18 Guus, this was already SW_RDFWG()11:00AM 14:59:20 ok, Guus; that matches SW_RDFWG()11:00AM 14:59:24 zakim, ??P3 is me 14:59:24 +gkellogg; got it 14:59:34 +GavinC 15:00:39 cygri has joined #rdf-wg 15:01:08 +??P15 15:01:13 zakim, I am ??P15 15:01:13 +manu; got it 15:01:42 cygri has joined #rdf-wg 15:01:46 AZ has joined #rdf-wg 15:01:51 +Ivan 15:02:00 +MHausenblas 15:02:02 zakim, mhausenblas is temporarily me 15:02:02 +cygri; got it 15:02:11 + +1.540.898.aabb 15:02:21 Zakim, aabb is me 15:02:21 +davidwood; got it 15:02:24 +??P16 15:02:32 zakim, I am ??P16 15:02:32 +mlnt; got it 15:03:23 +??P13 15:03:35 Zakim, ??P13 is me 15:03:35 +AZ; got it 15:04:06 zakim, pick a scribe 15:04:06 Not knowing who is chairing or who scribed recently, I propose gkellogg 15:04:20 +Sandro 15:04:26 scribe: cygri 15:04:33 chair: Guus 15:05:19 +EricP 15:05:20 topic: Admin 15:05:32 zakim, who is on the call? 15:05:32 On the phone I see Guus, yvesr_, AndyS, gkellogg, GavinC, manu, Ivan, cygri, davidwood, mlnt, AZ, Sandro, EricP 15:05:33 PROPOSED to accept the minutes of the 17 Oct telecon 15:05:39 http://www.w3.org/2011/rdf-wg/meeting/2012-10-17 15:05:52 RESOLUTION: Accept the minutes of the 17 Oct telecon 15:05:58 subtopic: Review of action items 15:06:24 guus: I sent the long overdue response to PROV-WG 15:07:08 ... response from PROV-WG: request for clarification of our schedule for finishing the multigraph syntax 15:07:27 ivan: I'll be on both WG's F2F meetings, so can pass on that information 15:07:34 + +1.617.553.aacc 15:07:41 guus: gavinc claims victory on ACTION-189 15:07:43 ... close it 15:07:53 ACTION-188? 15:07:53 ACTION-188 -- Eric Prud'hommeaux to respond to i18n for issue 183 -- due 2012-10-17 -- PENDINGREVIEW 15:07:53 http://www.w3.org/2011/rdf-wg/track/actions/188 15:08:09 +OpenLink_Software 15:08:17 Zakim, OpenLink_Software is temporarily me 15:08:17 +MacTed; got it 15:08:17 ericP: I have done it, but cygri disagreed with my response 15:08:19 Zakim, mute me 15:08:19 MacTed should now be muted 15:08:29 guus: The action is done anyway 15:09:05 ACTION-184? 15:09:05 ACTION-184 -- Richard Cyganiak to review http://www.w3.org/blog/SW/2012/09/12/last-call-constraints-of-the-provenance-data-model/ by 2 Oct -- due 2012-09-26 -- OPEN 15:09:05 http://www.w3.org/2011/rdf-wg/track/actions/184 15:09:10 ACTION-185? 15:09:10 ACTION-185 -- Eric Prud'hommeaux to review http://www.w3.org/blog/SW/2012/09/12/last-call-constraints-of-the-provenance-data-model/ by 2 Oct -- due 2012-09-26 -- OPEN 15:09:10 http://www.w3.org/2011/rdf-wg/track/actions/185 15:09:20 zwu2 has joined #rdf-wg 15:09:34 ericP: I'm not sure if it's still useful to review PROV-C 15:09:44 ivan: I believe it's overcome by events 15:09:51 +PatH 15:10:00 ... you can still of course sent in personal comments 15:10:24 for the record, I sent my personal comments on PROV-CONSTRAINTS, but I don't see anything to say on behalf of RDF WG 15:10:38 cygri: same for my action re PROV-C 15:10:47 patH has joined #rdf-wg 15:10:56 subtopic: Next meeting 15:11:07 guus: F2F on Monday+Tuesday 15:11:28 ... plans for meeting on Sunday evening? 15:11:32 + +1.650.265.aadd 15:11:49 ivan: The area around TPAC is pretty desolate at night 15:11:50 zakim, +1.650.265.aadd is me 15:11:50 +zwu2; got it 15:12:13 but public transport brings you quickly to the center 15:12:14 davidwood: we could meet in the Hilton lobby 15:12:28 Arnaud has joined #rdf-wg 15:13:07 guus: Next week, summer time stops in Europe but not yet in the U.S. 15:13:19 ... so, difference between the time zones is one hour less 15:13:35 Reminder on Timezones: http://www.timeanddate.com/worldclock/fixedtime.html?msg=RDF+WG+F2F&iso=20121029T0915&p1=333&ah=8 15:14:40 topic: JSON-LD 15:15:11 q+ to suggest a different order 15:15:23 guus: let's start with the discussion of the phrase Linked Data and how it relates to different formats 15:15:32 ... especially Dan's message was interesting 15:15:45 ... in my view, while interesting, this is not on the critical path of this WG 15:15:54 ... I think we should not spend time on it 15:16:09 ericP: Did it come up because JSON-LD references LD? 15:16:10 q+ 15:16:34 guus: it also had to do with the question whether the link to RDF is mandatory for something to be linked data 15:16:57 ... Dan's message was that it's a good slogan 15:17:11 +q to wonder why that community isn't here, and if perhaps that's part of the issue 15:17:24 manu: The JSON-LD spec used to say normatively what is and isn't LD, but this was removed 15:17:25 ack manu 15:17:25 manu, you wanted to suggest a different order 15:17:40 ... so now it just talks informatively and generally about linked data (in the introduction) 15:17:42 +1 to permathreads 15:17:56 ... this approach has largely been agreed in the JSON-LD group and also here in this group 15:18:06 ... we want to avoid these permathreads 15:18:15 ... we have more important things to discuss 15:18:19 q? 15:18:27 ... and what we have in the JSON-LD spec doesn't require this discussion at all 15:18:29 ack cygri 15:18:33 cygri: +1 to what Manu said. 15:18:39 cygri: plus one 2 manu 15:18:42 ack gavinc 15:18:42 gavinc, you wanted to wonder why that community isn't here, and if perhaps that's part of the issue 15:19:02 gavinc: Every time JSON-LD comes up, it is said that the people in this WG are not the target community 15:19:13 q+ about "wrong community" 15:19:20 ack about 15:19:21 ... I am somewhat worried about doing this work in RDF-WG if it's the wrong community 15:19:26 ack "wrong, community" 15:19:33 ack 15:19:45 I see section 3.1 is still there. 15:19:52 q+ 15:20:04 guus: next topics are: what's the relationship between RDF and JSON-LD? and how should that be expressed in the documents? 15:20:19 ... there were messages from Peter 15:20:44 ... gkellogg said: "Is JSON-LD a serialization syntax for all RDF graphs? Yes" 15:20:50 +Sandro.a 15:20:57 -Sandro 15:21:53 manu: Let's identify exactly what we are talking about. There have been strong comments from mhausenblas and pfps saying that the relationship between RDF and JSON-LD has to be normatively specified 15:22:06 q+ 15:22:21 guus: Let's first clarify what the relationship is, between talking about how to express it in the docs 15:22:40 ... Are we in agreement regarding: "Is JSON-LD a serialization syntax for all RDF graphs?" 15:23:00 ivan: The answer is: clearly true 15:23:10 manu: Yes, this was a design requirement 15:23:17 q? 15:23:26 ack 15:23:36 q? 15:23:39 ack manu 15:23:41 guus: So JSON-LD is a serialization syntax for all RDF graphs 15:23:46 ack "wrong, community" 15:24:09 ack manu 15:24:10 ack 'wrong, community' 15:24:11 ack cygri 15:24:25 ack wrong, community 15:24:25 q- 15:24:29 ack "wrong 15:24:31 ack \"wrong, community\" 15:24:31 q- "wrong community" 15:24:39 zakim, help 15:24:39 Please refer to http://www.w3.org/2001/12/zakim-irc-bot for more detailed help. 15:24:41 queue= 15:24:41 Some of the commands I know are: 15:24:41 xxx is yyy - establish yyy as the name of unknown party xxx 15:24:41 if yyy is 'me' or 'I', your nick is substituted 15:24:42 xxx may be yyy - establish yyy as possibly the name of unknown party xxx 15:24:42 I am xxx - establish your nick as the name of unknown party xxx 15:24:42 xxx holds yyy [, zzz ...] - establish xxx as a group name and yyy, etc. as participants within that group 15:24:42 xxx also holds yyy - add yyy to the list of participants in group xxx 15:24:42 who's here? - lists the participants on the phone 15:24:42 who's muted? - lists the participants who are muted 15:24:42 mute xxx - mutes party xxx (like pressing 61#) 15:24:42 unmute xxx - reverses the effect of "mute" and of 61# 15:24:43 cygri: Let's talk about the second point - is JSON-LD only a serialization syntax for an RDF Graph - not quite, somewhat similar to the situation with RDFa 15:24:45 is xxx here? - reports whether a party named like xxx is present 15:24:45 list conferences - reports the active conferences 15:24:45 this is xxx - associates this channel with conference xxx 15:24:45 excuse us - disconnects from the irc channel 15:24:45 I last learned something new on $Date: 2012/10/14 23:37:24 $ 15:24:51 +Arnaud 15:24:58 Yes, but not exactly = no 15:25:03 sorry for being late 15:25:41 cygri: In RDFa it's clear that you have an HTML document, RDFa says that you can recover some of the HTML document as structured data in RDF. In JSON-LD, it's not quite as extreme - in an HTML document, you have lots of other content, but it is effectively the same thing - you can extract information. 15:26:09 cygri: There are things that you can pick up from the JSON document and translate to RDF. 15:26:14 guus: so this is in line with gkellogg's response 15:26:25 ... there are some things that don't survive transformation to RDF 15:26:33 ... so "Is JSON-LD only a serialization syntax for RDF graphs?" is not strictly true 15:26:46 manu: And not completely false either 15:26:50 Souri has joined #rdf-wg 15:27:00 ... there are limited cases where it's not true 15:27:17 + +1.603.897.aaee 15:27:28 the delta of bnodes as predicates isn't really that small, is it? 15:27:29 guus: So, any RDF graph is JSON-LD, but there is a certain, small, delta of things that can be expressed in JSON-LD but not in RDF 15:27:33 I think it is fine for any non-RDF format to be able to represent stuff outside RDF 15:27:36 bnodes as predicates? 15:27:39 zakim, aaee is Souri 15:27:39 +Souri; got it 15:27:42 zakim, aaee is Souri 15:27:42 sorry, ivan, I do not recognize a party named 'aaee' 15:27:55 +1 to patH 15:28:31 q+ 15:28:46 guus: The concern of mhausenblas was: If JSON-LD provides an RDF serialization, then the relationship with RDF has to be normatively specified 15:29:08 q+ 15:29:14 ... cygri has made some efforts to align both already; I note there's issue 168 15:29:19 ack cygri 15:29:22 https://github.com/json-ld/json-ld.org/issues/168 15:29:34 +1 to mhausenblas 15:30:02 http://json-ld.org/spec/latest/json-ld-syntax/#json-ld-data-model 15:30:05 cygri: I'd like to talk about next steps that we talked about in the JSON-LD call. I'm trying to make the JSON-LD data model very clear. If we look at the JSON-LD spec, there is a section 3.1. 15:30:32 https://github.com/json-ld/json-ld.org/issues/174 15:30:37 cygri: That has a short description of the data model - one point I raised was that it leaves a number of questions open. There are some details that are not spelled out (for whatever reason). I raised an issue against the JSON-LD spec (above). 15:31:18 cygri: I spelled out all of the minor points that are left ambiguous. There is a long discussion there where the designers of JSON-LD responded to that. Based on that, I think we're pretty close to a point where we can say that in all detail, what the JSON-LD data model is. 15:31:56 cygri: Once we have done that, it'll be very easy to say exactly how it maps to the RDF data model. We can point out the small delta between the two data models. I have taken an action to write an appendix to the JSON-LD spec that will spell it out in detail. 15:32:18 q? 15:32:45 q? 15:32:55 q+ 15:33:00 cygri: The plan is that this would be a normative appendix that would spell out these differences. Section 3.1 uses slightly different terminology from RDF Concepts, sometimes a different name or slightly different definition, but there is an appendix that makes these differences really clear and explicit. This will provide a normative distinction between JSON-LD and RDF. That's how we're... 15:33:02 ...addressing the issue. 15:33:23 cygri: With my RDF Concepts editor's hat on, I'd be satisfied if the Appendix is normative and the JSON-LD data model is spelled out in sufficient detail. 15:34:25 manu: Yes, this is what we agreed on the JSON-LD call yesterday. Broad agreement in the group to align closely with RDF, while ensuring that the document is understandable to those without enough time to study RDF deeply 15:34:35 ... we've taken a number of actions to make sure that this alignment is happening 15:34:49 ... we believe that's enough to address mhausenblas' and pfps' concerns 15:35:23 ... I'd like to know if anyone disagrees, and how these concerns could be alleviated 15:35:27 q? 15:35:33 ack manu 15:35:34 guus: mhausenblas and pfps are not here 15:35:51 ivan: I don't have problems with the plan you have outlined 15:36:11 ... is the agreement to use the same term with the same definition wherever possible? 15:36:23 q+ 15:36:23 ... I understand the image issue and that you don't want to mention RDF 15:36:34 ack ivan 15:36:40 ... nevertheless, wherever possible, there should be not just a mapping but the same term should be used 15:36:41 q+ to respond to 'use the same terms' 15:37:15 ... maybe the background and reasons can be given somewhere, why it was necessary to define JSON-LD graphs 15:37:20 ack patH 15:37:22 ... apart from the image issue 15:37:29 +1, I'd like to understand what the issues are, and whether it's somehting that needs to be taken into account in RDF 15:37:41 PatH: Agree that same terminology should be used 15:37:56 ... I can see why you'd like to protect the audience from reading the RDF docs 15:38:07 ... But it's important that they know that they're using RDF 15:38:28 ... When things are not called the same there's a risk that things evolve in different directions 15:38:47 ack mamnu 15:38:50 ... If you could overcome the RDF shock problem, that would give security against such danger 15:38:58 ack manu 15:38:58 manu, you wanted to respond to 'use the same terms' 15:39:10 manu: We try not to duplicate when we don't have to. That's not necessarily an easy thing 15:39:19 ... There would be two extremes, we aim for neither 15:39:30 q+ 15:39:31 ... One: Defer to RDF concepts for everything. Not good for the audience 15:39:43 q+ to suggest the sections just need to be informative 15:39:47 ... Two: Don't mention RDF at all. We don't want that either because people should be able to dig deeper 15:40:01 ... Trying to find a middle ground. That's difficult as we need to delve into the details 15:40:25 RDF Concepts uses property too, in relation to predicate 15:40:26 ... For example, we preferred term “property” over “predicate” as that term is more commonly understood 15:40:41 ... so we try to pick the web developer view 15:40:51 "property" is not completely inconsistent with RDF, I would say 15:40:59 ... try to avoid forcing people to have to go to the RDF Concepts document 15:41:16 ... So we're trying to align terminology, but it doesn't always work out that way 15:41:26 q? 15:41:48 ... We appreciate input from this group regarding terminology. We don't want to define a parallel data model 15:41:53 It's not "some people" use property to mean something in Javascript btw, it's the ECMA Script spec itself 15:41:53 ack ivan 15:41:56 q 15:42:05 q+ 15:42:31 ivan: I'm fine with the direction. We can't do anything useful here really until the text from Richard is available and the description of the data model is considered final 15:42:38 ack davidwood 15:42:38 davidwood, you wanted to suggest the sections just need to be informative 15:42:43 ... The direction seems fine and I wish you good luck 15:43:08 davidwood: Some reasonable concerns have been raised that the JSON-LD documents should not redefine what has already been defined elsewhere 15:43:34 ... We have a good mechanism for that. Just mark those sections informative. 15:43:54 ... Then we can simply note that this is the JSON-LD spec speaking to its readership 15:43:58 q+ to say that some parts can't be informative 15:44:05 ... This might be all we need to do here 15:44:23 guus: It might be useful to go through the terminology delta at the F2F 15:44:54 cygri: I think we need to look at the data models and where there is a difference between them... regardless of the words that are used. 15:45:01 q- 15:45:24 cygri: As far as I understand, in a JSON-LD graph, you can have a free-floating node. You can just have an IRI that exists as an independent node. You can't have a node that doesn't at least have one statement. 15:45:40 cygri: That's one of these subtle differences. Is there a reasonable technical reason that we have these differences? 15:45:51 How about ? 15:46:13 Alt view - in RDF all nodes exist always. 15:46:16 cygri: If there is a reason we have it, wouldn't it be more dangerous to lean too heavily on RDF Concepts that could create more confusion than just saying that they're two indepdendent data models. 15:46:36 manu, two points. (1) where there are genuine technical overloadings of meaning then of course you might need to use different terms, and "property" might be one. (2) Richard is rewriting RDF Concepts to keep it as brief and snappy as possible, so i tmight serve to be a barrier against the dreaded model theory for your readers. 15:46:43 cygri: I think it will be useful to go through the differences in the data model, discuss what they are and how the mapping to RDF happens. 15:47:08 guus: I think this would be one of the most useful things to do at the F2F, might solve many of the issues 15:47:33 ack patH 15:48:00 ack manu 15:48:00 manu, you wanted to say that some parts can't be informative 15:48:42 +1 to speaker 15:48:49 manu: I think it's useful to discuss the delta, but am afraid of getting into long debates of the JSON-LD data model. It doesn't help if the RDF-WG talks in detail about the JSON-LD data model as this won't have a clear result 15:49:11 guus: The discussion can be limited. Some discussion might be necessary to get the endorsement of this group 15:49:26 +1 to path 15:49:34 PatH: RDF-WG shouldn't get into minute details of the JSON-LD data model. Waste of time 15:49:35 q+ 15:49:57 ack cygri 15:50:16 manu: Not saying we shouldn't discuss it. Just avoid getting bogged down in details 15:50:40 -Sandro.a 15:51:07 manu: can we have some resolutions to capture the agreement we seem to have? 15:51:41 PROPOSED PROPOSAL: JSON-LD MUST provide a serialization syntax for all RDF graphs. 15:51:43 PROPOSED PROPOSAL: Add a normative section to the JSON-LD Syntax specification outlining how the JSON-LD data model aligns with the RDF data model described in RDF Concepts. 15:51:45 PROPOSED PROPOSAL: If necessary, add precision to the JSON-LD data model such that it is crystal clear as to how the JSON-LD data model maps to the RDF data model. 15:52:21 cygri: We shouldn't do resolutions on statements of fact... we should do it on things we intend to do. 15:53:44 Should JSON-LD provide a way to recognize whether a given serialization is or is not legal RDF (other than translating it into RDF nad back again, that is) ?? NOt sure how to word this. 15:54:10 patH: No, down that road lies insanity 15:54:19 patH, We plan to put a conformance section in the spec... which might do what you want. 15:54:38 PROPOSED PROPOSAL revision: JSON-LD syntax MUST support serialization of (any?) all RDF graphs. 15:54:45 Avoid "crystal" in a resolution. 15:54:53 guus: not sure about spending telecon time on the exact phrasing 15:55:17 PROPOSAL: JSON-LD syntax MUST support serialization of all RDF graphs. 15:55:19 +1 15:55:20 +1 15:55:21 +1 15:55:21 +1 15:55:21 +1 15:55:21 +1 15:55:23 +1 15:55:24 +1 15:55:24 +1 15:55:26 +1 15:55:26 +1 15:55:26 +1 15:55:28 +1 15:55:36 +1 15:56:05 +1 with fingers in ears around datasets 15:56:07 +1 15:56:26 RESOLUTION: JSON-LD syntax MUST support serialization of all RDF graphs. 15:57:32 RDF Graphs are sets of triples only 15:58:13 gkellogg: the intention is that JSON-LD serializes RDF datasets too, like TriG 15:58:21 guus: let's not try to get resolution on that now 15:58:24 have to leave for another meeting. bye. 15:58:25 Does JSON-LD allow bnodes for graphs? 15:58:30 -zwu2 15:58:35 AndyS: Yes. 15:58:48 PROPOSAL: Add a normative section to the JSON-LD Syntax specification outlining how the JSON-LD data model aligns with the RDF data model described in RDF Concepts. 15:59:06 outlining//stating 15:59:13 Suggest simply ban them - not as a syntax issue. 15:59:14 PROPOSAL: Add a normative section to the JSON-LD Syntax specification stating how the JSON-LD data model aligns with the RDF data model described in RDF Concepts. 15:59:27 +1 15:59:30 +1 15:59:31 +1 15:59:31 why stating and not defining? 15:59:32 +1 15:59:32 +1 15:59:32 +0 15:59:39 +1 15:59:40 +1 15:59:42 +1 15:59:49 +1 15:59:55 +1 15:59:56 ok 16:00:04 +0 prefer removal of 3.1 16:00:05 +1 16:00:12 cygri: a normative statement is a definition, isn't it? 16:00:31 Arnaud, stating requires unambiguous and clear but not the exact form. 16:00:32 3.1 is the "JSON-LD Data Model" section 16:00:45 +1 16:00:48 gavinc: I still think the separate terms are a mistake, but it's editorial discretion 16:01:07 RESOLUTION: Add a normative section to the JSON-LD Syntax specification stating (=defining) how the JSON-LD data model aligns with the RDF data model described in RDF Concepts. 16:01:10 3.1 could be informative 16:01:39 AndyS, the drafts don't sync automatically to w3.org.. the most recent are always at json-ld.org 16:01:48 gkellogg: indeed, and I wouldn't mind a reusable into to RDF modelish that could be used in intros to Turtle, TriG, as well 16:01:54 :-) 16:01:54 intro 16:02:17 topic: Turtle LC 16:02:24 ericP: no replies from i18n group yet 16:02:41 guus: Can you send a reminder? 16:03:46 ... to make a resolution regarding LC at the F2F, we need all responses to the LC comments 16:04:01 guus: I think the other outstanding response is from Tim 16:04:02 David, we are all pedantic *in our own way* 16:04:21 ericP: I talked to him, he grumbled but said “whatever”. Just need to get him to write that down 16:04:28 :-) 16:04:38 ... this was about reverse properties in Turtle 16:05:03 topic: Graphs 16:05:36 guus: We had three proposals last week, and resolved one. Do we want to continue with the other two? 16:06:21 ... I think this captured the outcome of last week's discussion: The WG makes no normative statement on whether implementations that parse and store information from TriG documents MAY or MAY NOT turn the TriG default graph into a named graph with a name chosen in an implementation-dependent way. 16:07:00 ... is this a correct formulation? and is it useful to capture this as a resolution? 16:07:13 :-) 16:07:37 PatH: Did we decide this? 16:07:44 As long as it is not "the same" dataset. i.e. when written out. 16:07:52 guus: I tried to summarize the consensus of last week 16:08:07 q+ 16:08:12 +meh 16:08:38 davidwood: As I recall, we couldn't quite get consensus last week 16:08:39 I don't think it that proposal says anything 16:08:43 so sure, I agree with it 16:08:55 Zakim, unmute me 16:08:55 MacTed should no longer be muted 16:09:02 q? 16:09:07 davidwood: wondering if such a resolution is useful 16:09:12 q+ 16:09:14 gkellogg: I see no value in such a statement 16:09:16 q+ 16:09:22 ack gkellogg 16:09:47 ack MacTed 16:09:49 MacTed: I agree with Gavin, the statement says nothing 16:09:57 ... it's an open issue that we didn't get agreement on yet 16:10:08 I think Guus wants a resolution handy to stop further discussion, Think of it as a Chair's light-saber. 16:10:16 Fundamentally, this is a question of interoperability. No implementor will care whether their own interpretation of the spec is "correct" until they try to interoperate with others. 16:10:59 "MIGHT"? 16:11:18 Test - is it visible in some way? 16:11:21 ack ivan 16:11:23 cygri: might not be worth making a resolution on this 16:11:37 Write the 'may or may not' in lower case. 16:11:48 ivan: One value of having it recorded: It doesn't affect the document, but it affects us 16:11:52 how about a note in the document? 16:12:05 OPTIONALLY OPTIONAL 16:12:15 ... so unless some really new information comes up, chairs can shut down discussion on this particualr issue 16:12:38 PROPOSED: The WG makes no normative statement on whether implementations that parse and store information from TriG documents MAY or MAY NOT turn the TriG default graph into a named graph with a name chosen in an implementation-dependent way. 16:12:38 Zakim, mute me 16:12:38 MacTed should now be muted 16:12:39 q+ 16:12:47 q+ 16:13:01 PatH: suggest decapitalizing the MAY / NOT 16:13:23 linclark has left #rdf-wg 16:13:23 PROPOSED: The WG makes no normative statement on how implementations that parse and store information from TriG documents handle the TriG default graph. 16:14:12 in that case, why do we need to have a default graph in trig? 16:14:19 I like cygri's idea 16:14:24 cygri: maybe phrase it as a positive instead of negative? 16:14:36 ack cygri 16:14:39 ack cygri 16:14:42 ack AndyS 16:14:46 ack AndyS 16:14:58 I didnt hear that properly 16:15:29 cygri's suggestion comes down to "continue the discussion until we figure out whether MAY or MAY NOT is what we'll say" 16:15:35 AndyS: I'm worried that this permits giving a name to the default graph and ...? (scribe fail) 16:16:03 PROPOSED: The RDF WG recognises that implementations may have more than one local interpretation of a TriG default graph and makes no normative statement on how implementations that parse and store information from TriG documents handle the TriG default graph. 16:16:12 how about something like: whether implementations that parse and store information from TriG documents turn the TriG default graph into a named graph is implementation dependent? 16:16:33 Nice try, Guus. 16:16:35 depends if a SPARQL query is changed. 16:16:36 ivan has left #rdf-wg 16:16:41 guus: Let's postpone it 16:16:45 ... thanks for joining! 16:16:54 -LeeF 16:16:56 -manu 16:16:57 -MacTed 16:16:58 ... see some of you next week! 16:16:58 -Souri 16:16:58 -Ivan 16:17:00 -yvesr_ 16:17:00 -AndyS 16:17:02 -Arnaud 16:17:08 -gkellogg 16:17:10 -mlnt 16:17:11 Hey, where are details of how to call in? 16:17:13 -EricP 16:17:15 RRSAgent, make logs public 16:17:27 Indeed, can we get the remote details? 16:17:39 Is it the same call? 16:17:43 -cygri 16:17:59 -AZ 16:18:07 patH, it's on the f2f page: http://www.w3.org/2011/rdf-wg/wiki/FTF3 16:18:26 trackbot, end meeting 16:18:26 Zakim, list attendees 16:18:26 As of this point the attendees have been +31.20.598.aaaa, AndyS, yvesr_, Guus, gkellogg, GavinC, manu, Ivan, cygri, +1.540.898.aabb, davidwood, mlnt, AZ, Sandro, EricP, 16:18:30 ... +1.617.553.aacc, LeeF, MacTed, PatH, zwu2, Arnaud, +1.603.897.aaee, Souri 16:18:34 RRSAgent, please draft minutes 16:18:34 I have made the request to generate http://www.w3.org/2012/10/24-rdf-wg-minutes.html trackbot 16:18:35 RRSAgent, bye 16:18:35 I see no action items