15:01:12 RRSAgent has joined #rdf-wg 15:01:12 logging to http://www.w3.org/2012/03/21-rdf-wg-irc 15:01:13 Guus has joined #rdf-wg 15:01:14 RRSAgent, make logs world 15:01:14 Zakim has joined #rdf-wg 15:01:16 Zakim, this will be 73394 15:01:16 ok, trackbot; I see SW_RDFWG()11:00AM scheduled to start now 15:01:17 Meeting: RDF Working Group Teleconference 15:01:17 Date: 21 March 2012 15:01:26 cgreer has joined #rdf-wg 15:01:34 Zakim, this is 73394 15:01:36 ok, gavinc; that matches SW_RDFWG()11:00AM 15:01:49 Zakim, who is on the phone? 15:01:52 +Guus_Schreiber 15:01:55 SteveH has joined #rdf-wg 15:01:56 On the phone I see gavinc, bhyland, sandro, +1.603.897.aaaa, Arnaud, pfps, Guus_Schreiber 15:02:03 Zakim, bhyland is me 15:02:08 +EricP 15:02:14 +davidwood; got it 15:02:20 AndyS1 has joined #rdf-wg 15:02:27 +cgreer 15:02:42 +??P25 15:02:46 zakim, who is here? 15:02:48 zakim, dial ivan-voip 15:02:48 On the phone I see gavinc, davidwood, sandro, +1.603.897.aaaa, Arnaud, pfps, Guus_Schreiber, EricP, cgreer, ??P25 15:02:49 zakim, ??P25 is me 15:02:56 ok, ivan; the call is being made 15:02:57 +AndyS1; got it 15:02:59 +Ivan 15:03:02 On IRC I see AndyS1, SteveH, cgreer, Zakim, Guus, RRSAgent, Souri, Arnaud, gavinc, mischat, pfps, MacTed, LeeF, AndyS, ivan, mdmdm, yvesr, manu1, davidwood, NickH, trackbot, manu, 15:03:05 AlexHall has joined #rdf-wg 15:03:06 ... sandro, ericP 15:03:13 + +1.443.212.aabb 15:03:17 zakim, aabb is me 15:03:18 +AlexHall; got it 15:03:24 zwu2 has joined #rdf-wg 15:03:26 +OpenLink_Software 15:03:32 Zakim, OpenLink_Software is temporarily me 15:03:32 +MacTed; got it 15:03:33 Zakim, mute me 15:03:34 zakim, code? 15:03:36 MacTed should now be muted 15:03:37 zakim, who is here? 15:03:38 the conference code is 73394 (tel:+1.617.761.6200 sip:zakim@voip.w3.org), zwu2 15:03:59 On the phone I see gavinc, davidwood, sandro, +1.603.897.aaaa, Arnaud, pfps, Guus_Schreiber, EricP, cgreer, AndyS1, Ivan, AlexHall, MacTed (muted) 15:04:17 http://www.random.org/coins/ 15:04:19 On IRC I see zwu2, AlexHall, AndyS1, SteveH, cgreer, Zakim, Guus, RRSAgent, Souri, Arnaud, gavinc, mischat, pfps, MacTed, LeeF, AndyS, ivan, mdmdm, yvesr, manu1, davidwood, NickH, 15:04:24 ... trackbot, manu, sandro, ericP 15:04:34 +zwu2 15:04:50 any volunteer for scribing? 15:05:26 +??P36 15:05:26 +LeeF 15:05:33 Zakim, ??P36 is me 15:05:33 +SteveH; got it 15:05:52 You flipped 1 coin of type Polish 5 Złoty: 15:06:33 +??P38 15:06:41 -pfps 15:06:43 scribenick: ericP 15:06:56 Zakim, who's noisy? 15:06:58 Zakim, ??P38 is me 15:06:58 +NickH; got it 15:07:03 Zakim, mute me 15:07:03 NickH should now be muted 15:07:07 MacTed, listening for 10 seconds I heard sound from the following: gavinc (7%), ??P38 (15%), EricP (52%) 15:07:29 Zakim, unmute me 15:07:29 MacTed should no longer be muted 15:07:40 Zakim, mute me 15:07:40 MacTed should now be muted 15:07:41 zakim, mute me 15:07:41 zwu2 should now be muted 15:07:57 Zakim, unmute me 15:07:57 MacTed should no longer be muted 15:08:08 the click seems to be zakim? 15:08:23 Zakim, mute me 15:08:23 MacTed should now be muted 15:08:25 -AndyS1 15:08:28 Zakim, mute ericP 15:08:28 EricP should now be muted 15:08:29 :-) 15:08:36 weak! 15:08:41 -EricP 15:08:53 +EricP 15:09:05 click, click, click 15:09:11 +[IPcaller] 15:09:21 zakim, IPCaller is me 15:09:21 +AndyS1; got it 15:09:24 Ah, that must be AndyS in a snowstorm 15:10:37 PROPOSED to accept the minutes of the 14 Mar telecon: 15:10:50 RESOLVED 15:11:16 topic: action item review 15:11:34 Guus: Turtle actions will be taken up in Turtle agendum 15:11:36 -AndyS1 15:11:46 ... Sandro and I owe actions 15:12:18 topic: getting Turtle to LC 15:12:29 I claim victory on ACTION-154! 15:12:37 ACTION-150? 15:12:37 ACTION-150 -- Gavin Carothers to fail to update Turtle/N-Triples grammar to not use \u escaping in the local part of prefix names -- due 2012-03-21 -- OPEN 15:12:37 http://www.w3.org/2011/rdf-wg/track/actions/150 15:12:50 +??P25 15:13:06 zakim, ??P25 is me 15:13:06 +AndyS1; got it 15:14:17 cygri has joined #rdf-wg 15:14:30 Guus: do we need to discussion the ":"s in local names issue? 15:14:37 zakim, code? 15:14:37 the conference code is 73394 (tel:+1.617.761.6200 sip:zakim@voip.w3.org), cygri 15:14:44 gavinc: since i didn't file the issue in time, i suspect we'll do nothing about it 15:14:50 ... i'm mostly okay with that 15:15:03 +mhausenblas 15:15:04 ... creates some probs for a poor vocab proposed by facebook 15:15:07 zakim, mhausenblas is temporarily me 15:15:07 +cygri; got it 15:15:42 colons? wacky 15:15:43 ... Facebook uses ":"s in local part of propery names 15:15:50 ivan: already widely used? 15:15:54 gavinc: yup 15:15:59 ack me 15:16:28 CONTINUE ACTION-152 15:16:39 yes 15:16:47 ACTION-153 DONE 15:16:57 example - og:video:height 15:17:06 http://developers.facebook.com/docs/opengraphprotocol/ 15:17:36 ACTION-154 15:17:50 gavinc: we keep the case-sensitivity 15:17:54 The "og:" is effectively a prefix part, the "video:height" is effectively a local part 15:17:58 ... i'll remove the change note 15:18:29 ACTION-154 DONE 15:18:48 PatH has joined #rdf-wg 15:18:57 Guus: note that next week's telecon returns to CT's 5pm 15:18:57 Sorry Im late. 15:19:59 +PatH 15:20:10 zakim, unmute me 15:20:10 zwu2 should no longer be muted 15:20:12 q+ 15:20:19 gavinc: issues were relationship between texp/plain and application/ntriples 15:20:29 ... and @@1 15:20:46 +pfps 15:21:03 zwu2: we'd like to see clarification of consequences of using new encoding on old tool 15:21:26 zakim, mute me. 15:21:26 PatH should now be muted 15:21:29 gavinc: can we say "there may be some issues" instead of "you should continue to serve ntriples as US-ASCII"? 15:21:37 zwu2: agree to a large extent 15:21:52 ... but for maximum backward-compatibility, one should 15:22:10 ... is it acceptable to include that qualifier [in the spec]? 15:22:14 q+ 15:22:21 gavinc: i'm not ok with that but willing to defer to the WG 15:22:29 ... many tools today already accept UTF8 15:22:38 which existing tools actually look at the Content-Type (and charset)? 15:22:44 zwu2: there are many tools which do not accept the new encoding 15:23:01 davidwood: that doesn't mean we should prefer ASCII for the next 10 years 15:23:21 We should *allow*, but not *prefer* ASCII 15:23:50 zwu2: for this new encoding to work with tools, shouldn't we document the backward-compatibility path? 15:23:52 q? 15:24:18 davidwood: as long as ASCII is allowed, how does it hurt backward-compatibility even if it's not preferred? 15:24:43 The issue I think is that to express any kind of preference for ASCII might (we hope) be archaic some years from now, even if it is accurate now. 15:24:44 zwu2: supposed you create an ntriples file in two years, and then i feed it to rapper 15:25:04 ... it will fail. should we have a warning indicating these consequences 15:25:20 note that this is *forward* compatibility 15:25:37 new data on old tools 15:25:49 q? 15:25:52 it is a delicate matter of wording and nuance. 15:26:31 gavinc: i propose to propose some text over the next week. good enough? 15:26:46 Guus: let's leave that to editorial discretion 15:26:52 ack zwu 15:26:59 ack cygri 15:27:12 cygri: it may not be necessary to say what people should do in what case, instead just the facts: 15:27:31 ... utf8 and us-ascii are both valid encodings for turtle 15:27:43 +1 to richard 15:27:43 ... some legacy systems do not handle utf8 15:27:53 +1 richard 15:28:05 +1, describe the technical issues and let people decide 15:28:06 gavinc: yes [going in this direction] 15:28:27 cygri: it's reasonable to point out that conventional tools don't handle utf-8 15:28:29 could also say, at the time of writing.... 15:28:46 Guus: let's leave this as useful input to the editors 15:28:52 +1 PatH, and include the date ;) 15:28:55 "at the time of writing" sounds reasonable 15:29:23 issue-73? 15:29:23 ISSUE-73 -- IRI_REF vs. IRIref -- open 15:29:23 http://www.w3.org/2011/rdf-wg/track/issues/73 15:29:43 Guus: we agreed that you three work this out 15:29:49 Is done in SPARQL. 15:29:49 gavinc: i think we have 15:30:38 AndyS1: action told us to get our act together off-list 15:30:56 ... i have changed the sparql grammar 15:30:58 ack me 15:31:30 IRI_REF => IRIREF 15:31:31 IRIRef => iri 15:31:33 IRIRefOrFunction => iriOrFunction 15:31:34 VarOrIRIref => VarOrIri 15:31:38 it has consequences throughout the documents though, so get it right, guys. ; 15:32:49 these productions have consequences in the turtle semantics (parsing rules) as well 15:32:59 I was under the naive impression that 15:33:28 PROPOSE Resolve Issue-73 by changing the following productions in Turtle: IRI_REF => IRIREF, IRIRef => iri, IRIRefOrFunction => iriOrFunction, VarOrIRIref => VarOrIri 15:33:29 IRI didnt need to have ref added :-( 15:33:50 ack me 15:34:11 not disputing, just muddled. 15:34:27 iri is prefix name and <...> 15:34:38 IRIREF is the <...> from RFC 15:34:43 +1 15:34:46 +1 15:34:49 +1 15:34:50 +0 15:35:06 RESOLVED 15:36:04 RESOLVED: Resolve Issue-73 by changing the following productions in Turtle: IRI_REF => IRIREF, IRIRef => iri, IRIRefOrFunction => iriOrFunction, VarOrIRIref => VarOrIri 15:36:11 AZ has joined #rdf-wg 15:36:29 issue-74? 15:36:29 ISSUE-74 -- Prefixed names and slashes -- open 15:36:29 http://www.w3.org/2011/rdf-wg/track/issues/74 15:37:02 yep 15:37:36 Guus: are we still on track for publication next week? 15:37:44 +??P3 15:37:53 gavinc: i believe we can have somehting for folks to review next week. 15:37:59 zakim, ??P3 is me 15:37:59 +AZ; got it 15:38:08 ... i don't think adding a week [before review] would help 15:38:24 Guus: we keep to current schedule 15:38:48 ... two reviews expected 9 April, so possible decision to publish 11 April 15:39:06 ... Andy brought up tests. when should we discuss them? 15:39:19 ... we can discuss them during LC 15:39:35 gavinc: i originally volunteered to write test cases for turtle 15:40:05 ... won't get to them before LC 15:40:11 topic: named graph semantics 15:40:12 zakim, unmute me. 15:40:12 PatH should no longer be muted 15:40:15 Zakim, mute me 15:40:15 gavinc should now be muted 15:40:42 Guus: we spent some time last week 15:41:00 PatH: may i have another week to write up the idea discussed last week 15:41:10 Guus: would like concrete actions to work out examples 15:41:27 ... i'm worried about how to explain this to folks outside WG 15:41:40 ... could work that out as draft text in the primer 15:41:52 ... i liked PatH's text which described what happens on import 15:42:07 ... which everyone does 15:42:14 I will try to write it as primer-level text. 15:42:24 PatH: draft text for Primer would be useful 15:42:44 ok, sounds great 15:42:56 ACITON: Guus to draft Primer text on importing named graphs 15:43:04 ACTION: Guus to draft Primer text on importing named graphs 15:43:04 Created ACTION-159 - Draft Primer text on importing named graphs [on Guus Schreiber - due 2012-03-28]. 15:43:29 Guus: would be nice to apply use cases to PatH's proposal 15:43:47 fwiw, the point i am having trouble with is having an 15:44:21 iri denote one thing while also indicating a context. 15:45:01 PatH: the biggest prob is having an iri able to identify both an e.g. person or context 15:45:06 ack me 15:45:16 I'm not sure what the problem is here. 15:45:26 q+ 15:45:50 ... the fourth field could be any uri, e.g. used to identify an e.g. person 15:46:04 +1 to AndyS 15:46:12 ... how important is it for the same uri to identify both a real-world thingy and a context? 15:46:28 SteveH: it doesn't have to be santioned 15:46:51 ack AndyS 15:46:58 ... people will do what they do so it's not critical to make it so flexible 15:47:06 AndyS1: two subcases: 15:47:22 ... .. person, and e.g. LeeF doesn't expect the spec to cover it 15:47:29 ... .. location on the web 15:47:37 That is a correct characterization of my position - we use URIs to denote both real-world things and graphs, and we don't really care if the spec covers it 15:47:41 ... could you do the easy one now, so we can see it? 15:47:51 ... would be nice to compare them side-by-side 15:48:12 PatH: hmm, not as easy as i thought, but can get it out in the next couple days 15:48:22 ack me 15:48:27 ... i'll document the simplest, and report its flaws 15:48:53 Guus: your second case splits where the web resource is static or dynamic 15:49:13 these three seem to come up consistently and cover the requirements 15:49:23 the context comes in the triples involving the URI -- the URI itself is opaque... 15:49:23 :URI a :person vs :URI a :webpage vs :URI a :electron 15:49:24 :URI a :indeterminate-thing has attributes 15:49:24 eventually you have enough attributes to know which it is -- or you know which attributes to discard, because they're not relevant to the thing with which you are concerned... 15:49:27 ... if we go for a solution, these should be the basis 15:50:18 davidwood: i thought that we decided even static was likely to change over time, just a matter of degree 15:50:25 case 1: static graph containers: always the same value // case 2: Changing contents (changing value) [maybe a continuum] 15:50:36 PatH: while that's true philosophically, there are those that we treat as static 15:50:54 ... so when static things change, we expect stuff to break 15:51:13 ... the dynamic stuff is engineered to change 15:51:47 davidwood: are you trying to make this isomorphic with things with things which should be time invariant? 15:52:02 +q 15:52:05 ack me 15:52:21 ack me 15:52:28 ack gavinc 15:52:34 PatH: people should be able to declare stuff invariant 15:52:44 ... e.g. signing, endorsing 15:52:53 eric is better than I am at writing my thoughts, which is scary. 15:52:55 s/attributes to know which it is/attributes to know which class it belongs to/ 15:53:02 s/not relevant to the thing/not relevant to the class/ 15:53:43 gavinc: i don't really agree RDF emulate snapshots for signing as every signing mechanism snapshots into a bytestream 15:54:05 PatH: how does one record a signed document as a web resource for later review? 15:55:14 http://en.wikipedia.org/wiki/Magnet_URI_scheme 15:55:46 gavinc: magnet URIs are designed for crypticographically identifying a resource for later search 15:56:24 Guus: magnet uris are a draft open standard 15:56:45 Zakim, unmute me 15:56:46 MacTed should no longer be muted 15:56:57 http://magnet-uri.sourceforge.net/ 15:57:43 PatH: back to named graphs, maybe this makes a use case irrelevent. will consider 15:58:22 sandro: last i recall was gavinc and i disputed whether you needed crypto in the endorsement use case 15:59:03 sorry I have another meeting coming up. bye. 15:59:11 -zwu2 15:59:19 gavinc: crypto gets you the ability to use magent uris to identify things on the web and get them without dereference 15:59:59 q+ to say we *can* factor crypto out 16:00:19 AndyS1: there are existing tools and techniques using crypto 16:00:36 ... crypto not necessary technically 16:00:55 ack me 16:00:56 ericP, you wanted to say we *can* factor crypto out 16:01:00 ?seem to be 2 cases? 1. controlling who gets to see the resource 2. ensuring that the resource referred to is the right one. I think (?) magnet uris are for the first? 16:01:16 ack me 16:01:33 found it --- http://www.w3.org/2011/rdf-wg/wiki/Example_of_Endorsement_Use_Case 16:02:36 ack me 16:04:33 MacTed: you have to hash a gsnap 16:04:49 or i can ask some trusted party to copy it 16:05:09 or you may trust me to nto alter the text, and i just want to talk *about* that graph 16:05:28 sandro, pls talk us through the page you pointed to 16:05:34 +1 16:05:36 http://www.w3.org/2011/rdf-wg/wiki/Example_of_Endorsement_Use_Case 16:06:09 zakim, mute me 16:06:09 PatH should now be muted 16:06:37 [conversation about guuids and crypto] 16:06:47 +1 to whoever just spoke 16:07:04 that was ted 16:07:04 sandro: gavin created this wiki page 16:07:26 ... started with use case "alice want to say that she agrees that 'Bob is named Bob'" 16:07:42 <#me> foaf:name "Bob", 16:07:50 http://lists.w3.org/Archives/Public/public-rdf-wg/2012Jan/0064.html 16:08:08 ... Bob says "<#me> foaf:name "Bob"" 16:08:59 http://www.w3.org/2011/rdf-wg/wiki/Example_of_Endorsement_Use_Case 16:09:07 ... adding crypto isn't required for this use case 16:09:45 ... gavinc's using something cryptographically derived from bytes instead "g775" 16:10:12 gavinc: if someone is externally signing the docs, you don't need the crypto in trig 16:10:35 q+ 16:10:46 q+ to comment against Sandro's message 16:11:02 trust =/= security? 16:11:10 ack me 16:11:43 Zakim, mute me 16:11:43 EricP should now be muted 16:11:47 zakim, unmute me 16:11:47 PatH should no longer be muted 16:12:19 q+ 16:13:18 sandro: is there anything that this group needs to do to make this example work? 16:14:15 This WG does not need to solve this problem - it needs to enable it or just not block it. 16:15:11 sandro: I will happily agree that crypt-URIs *are* useful for when you DONT want to use TRIG. When the data is too long, or something. 16:15:24 q- 16:15:30 ack Guus 16:15:36 ack davidwood 16:15:36 davidwood, you wanted to comment against Sandro's message 16:15:42 davidwood: critique of sandro's mail, alice could have made up "Bob's name is 'Bob'" 16:15:44 david just made mt point. 16:15:59 mt//my 16:17:08 what "GRAPH"? gsnap, gbox, gtext? I think there is slop happening... 16:17:28 PatH: there's not much point is endorsing a mathmatical abstraction 16:18:04 i wonder if annotea was an endorsement system 16:18:25 time rescues us again.... 16:18:32 Guus: if we don't get some examples, we have to go with the minimal solution 16:18:55 -cgreer 16:18:58 sandro: or admit defeat (preferrable to claiming victory on a non-success) 16:18:59 -PatH 16:19:17 -AZ 16:19:22 - +1.603.897.aaaa 16:19:22 We should define TriG syntax -- it's out there now -- ditto nquads 16:19:25 ck me 16:19:25 -pfps 16:19:27 ack me 16:19:29 -Arnaud 16:19:30 -AlexHall 16:19:30 -sandro 16:19:35 -NickH 16:19:37 -Guus_Schreiber 16:19:37 -AndyS1 16:19:41 -gavinc 16:19:44 -Ivan 16:19:46 http://www.w3.org/2011/rdf-wg/wiki/Scribes 16:19:49 :-) 16:19:49 syntax without semantics, AndyS ? Maybe.... 16:19:51 http://www.w3.org/2011/rdf-wg/wiki/Scribes 16:20:11 -MacTed 16:20:34 -EricP 16:20:35 -cygri 16:20:37 -davidwood 16:20:39 If we can't define the semantics we can at least enable data exchange. 16:21:03 AndyS I like it as a first step 16:21:03 +1 16:21:07 yep. true. 16:21:17 It will happen, the only Q is whether we help interoperability or not by fixing details. 16:23:30 -SteveH 16:23:36 AlexHall has left #rdf-wg 16:25:09 -LeeF 16:25:10 SW_RDFWG()11:00AM has ended 16:25:10 Attendees were gavinc, sandro, +1.603.897.aaaa, Arnaud, pfps, Guus_Schreiber, EricP, davidwood, cgreer, AndyS1, Ivan, +1.443.212.aabb, AlexHall, MacTed, zwu2, LeeF, SteveH, NickH, 16:25:10 ... cygri, PatH, AZ 16:59:51 SteveH_ has joined #rdf-wg 17:19:33 davidwood has joined #rdf-wg 17:45:45 mischat has joined #rdf-wg 17:58:38 mischat_ has joined #rdf-wg 18:41:33 mischat_ has joined #rdf-wg 18:46:34 Zakim has left #rdf-wg 20:00:05 AndyS has joined #rdf-wg 20:59:31 mischat has joined #rdf-wg