Warning:
This wiki has been archived and is now read-only.

Chatlog 2012-11-28

From RDF Working Group Wiki
Jump to: navigation, search

See panel, original RRSAgent log or preview nicely formatted version.

Please justify/explain non-obvious edits to this page, in your "edit summary" text.

15:51:29 <RRSAgent> RRSAgent has joined #rdf-wg
15:51:29 <RRSAgent> logging to http://www.w3.org/2012/11/28-rdf-wg-irc
15:51:31 <trackbot> RRSAgent, make logs world
15:51:31 <Zakim> Zakim has joined #rdf-wg
15:51:33 <trackbot> Zakim, this will be 73394
15:51:34 <trackbot> Meeting: RDF Working Group Teleconference
15:51:34 <trackbot> Date: 28 November 2012
15:51:35 <Zakim> ok, trackbot; I see SW_RDFWG()11:00AM scheduled to start in 9 minutes
15:58:34 <tbaker> tbaker has joined #rdf-wg
15:58:48 <cygri> cygri has joined #rdf-wg
15:59:01 <Zakim> SW_RDFWG()11:00AM has now started
15:59:10 <pchampin> pchampin has joined #rdf-wg
15:59:22 <Zakim> +MHausenblas
15:59:24 <cygri> zakim, mhausenblas is temporarily me
15:59:25 <Zakim> +cygri; got it
15:59:31 <Zakim> +??P12
15:59:40 <pchampin> zakim, ??P121is me
15:59:40 <Zakim> I don't understand '??P121is me', pchampin
15:59:43 <pchampin> zakim, ??P121 is me
15:59:43 <Zakim> sorry, pchampin, I do not recognize a party named '??P121'
15:59:51 <pchampin> zakim, ??P12 is me
15:59:51 <Zakim> +pchampin; got it
16:00:14 <Arnaud> Arnaud has joined #rdf-wg
16:00:28 <Zakim> +davidwood
16:00:29 <cygri> agenda: http://www.w3.org/2011/rdf-wg/wiki/Meetings:Telecon2012.11.28
16:00:37 <Zakim> +OpenLink_Software
16:00:48 <MacTed> Zakim, OpenLink_Software is temporarily me
16:00:48 <Zakim> +MacTed; got it
16:00:49 <AZ> AZ has joined #rdf-wg
16:00:49 <MacTed> Zakim, mute me
16:00:49 <Zakim> MacTed should now be muted
16:00:54 <Zakim> +??P18
16:00:56 <davidwood1> Chair: David Wood
16:00:58 <gkellogg> zakim, I am ??P18
16:00:58 <Zakim> +gkellogg; got it
16:00:59 <Zakim> + +1.408.992.aaaa
16:01:02 <davidwood1> Zakim, who is here?
16:01:02 <Zakim> On the phone I see cygri, pchampin, davidwood, MacTed (muted), gkellogg, +1.408.992.aaaa
16:01:04 <Zakim> On IRC I see AZ, Arnaud, pchampin, cygri, tbaker, Zakim, RRSAgent, gkellogg, MacTed, gavinc, davidwood1, yvesr, manu, manu1, mischat, trackbot, sandro, ericP
16:01:11 <MacTed> MacTed has changed the topic to: RDF-WG -- http://www.w3.org/2011/rdf-wg/ -- current agenda http://www.w3.org/2011/rdf-wg/wiki/Meetings:Telecon2012.11.28
16:01:14 <Zakim> +Arnaud
16:01:32 <Zakim> +Sandro
16:01:57 <Zakim> +GavinC
16:02:15 <Zakim> +AZ
16:02:34 <gkellogg> scribenick: gkellogg
16:03:00 <davidwood> PROPOSED to accept the minutes of the 21 Nov telecon:
16:03:00 <davidwood> 
16:03:00 <davidwood>    http://www.w3.org/2011/rdf-wg/meeting/2012-11-21
16:03:06 <Zakim> +??P26
16:03:15 <Zakim> +Tom_Baker (was ??P26)
16:03:26 <davidwood> RESOLVED to accept the minutes of the 21 Nov telecon: http://www.w3.org/2011/rdf-wg/meeting/2012-11-21
16:03:45 <Arnaud> zakim, mute me
16:03:45 <Zakim> Arnaud should now be muted
16:03:53 <MacTed> Zakim, who's noisy?
16:04:04 <Zakim> MacTed, listening for 10 seconds I heard sound from the following: davidwood (20%), Tom_Baker (28%)
16:04:08 <tbaker> zakim, please mute me
16:04:09 <Zakim> sorry, tbaker, I do not know which phone connection belongs to you
16:04:12 <MacTed> Zakim, mute Tom_Baker
16:04:12 <Zakim> Tom_Baker should now be muted
16:04:24 <MacTed> already fixed...
16:04:29 <tbaker> zakim, I am tbaker
16:04:29 <Zakim> sorry, tbaker, I do not see a party named 'tbaker'
16:04:34 <davidwood> Review of action items
16:04:34 <davidwood> http://www.w3.org/2011/rdf-wg/track/actions/pendingreview
16:04:34 <davidwood> http://www.w3.org/2011/rdf-wg/track/actions/open
16:04:39 <MacTed> Zakim, Tom_Baker is tbaker
16:04:39 <Zakim> +tbaker; got it
16:04:42 <cgreer> cgreer has joined #rdf-wg
16:04:52 <gkellogg> topic: Action Items
16:05:10 <davidwood> CLOSE ACTION-208
16:05:11 <trackbot> ACTION-208 Provide draft of Concepts for WG review closed
16:05:20 <AlexHall> AlexHall has joined #rdf-wg
16:05:24 <AZ> q+
16:05:30 <Zakim> + +1.707.874.aabb
16:05:39 <cgreer> zakim, aabb is me
16:05:39 <Zakim> +cgreer; got it
16:05:41 <gkellogg> CLOSE ACTION-213
16:05:41 <trackbot> ACTION-213 Contact dajobe about Turtle and test suites closed
16:05:41 <davidwood> CLOSE ACTION-213
16:05:41 <trackbot> ACTION-213 Contact dajobe about Turtle and test suites closed
16:05:52 <PatH> PatH has joined #rdf-wg
16:06:05 <davidwood> ack AZ
16:06:07 <gkellogg> I contacted dajobe about Turtle, but haven't heard back yet.
16:06:30 <gkellogg> az: I performed my review of RDF Concepts.
16:06:44 <davidwood> CLOSE ACTION-209
16:06:44 <trackbot> ACTION-209 Write a WG Note on RDF dataset semantics closed
16:06:45 <pfps> pfps has joined #rdf-wg
16:06:57 <AZ> it's 211
16:07:01 <gavinc> OPEN ACTION-209
16:07:12 <davidwood> CLOSE ACTION-211
16:07:12 <trackbot> ACTION-211 Review RDF-Concepts ED http://dvcs.w3.org/hg/rdf/raw-file/default/rdf-concepts/index.html closed
16:07:15 <gkellogg> CLOSE ACTION-211
16:07:15 <trackbot> ACTION-211 Review RDF-Concepts ED http://dvcs.w3.org/hg/rdf/raw-file/default/rdf-concepts/index.html closed
16:07:19 <Guus> Guus has joined #rdf-wg
16:07:42 <gkellogg> topic: Turtle status
16:07:56 <gkellogg> gavinc: haven't heard back from people we asked from.
16:08:10 <Zakim> +PatH
16:08:12 <gkellogg> sandro: how long have we given internationalization working group?
16:08:33 <Zakim> +[IPcaller]
16:08:43 <gkellogg> gavinc: timbl's had a couple of months, everyone else about a month.
16:08:54 <Guus> zakim, [IPcaller] is me
16:08:54 <Zakim> +Guus; got it
16:09:04 <gkellogg> david: we should send a friendly email suggesting that the come back in a week or two.
16:09:13 <gkellogg> … gavinc, please re-ping.
16:09:40 <gkellogg> sandro: we should ping all outstanding commenters. (TimBL)
16:10:11 <gkellogg> gavinc: still need to figure out escaping section for Turtle in HTML.
16:10:15 <sandro> (with a deadline of one week)
16:10:34 <gkellogg> chair: guus
16:10:38 <davidwood> Chair: Guus Schreiber
16:10:39 <cygri> zakim, who is on the phone?
16:10:39 <Zakim> On the phone I see cygri, pchampin, davidwood, MacTed (muted), gkellogg, +1.408.992.aaaa, Arnaud (muted), Sandro, GavinC, AZ, tbaker (muted), cgreer, PatH, Guus
16:10:40 <markus> markus has joined #rdf-wg
16:10:54 <gkellogg> guus: richard has a new proposal for bnode definition.
16:11:15 <gkellogg> … message in the agenda, and there were followups
16:11:19 <AZ> q+
16:11:25 <cygri> http://www.w3.org/2011/rdf-wg/wiki/User:Rcygania2/B-Scopes
16:11:33 <gkellogg> cygri: there have been 2 1/2 different proposals so far.
16:11:48 <gkellogg> … scope as bijection, classic, and previous-draft (outdated)
16:12:01 <zwu2> zwu2 has joined #rdf-wg
16:12:29 <gkellogg> … each have proponents and detractors
16:12:36 <Zakim> +??P33
16:12:50 <Zakim> + +1.650.265.aacc
16:13:04 <zwu2> zakim, +1.650.265.aacc is me
16:13:04 <Zakim> +zwu2; got it
16:13:09 <gkellogg> … the classic was my first attempt at phrasing things for the spec, but I changed a couple of ideas from the original version.
16:13:10 <zwu2> zakim, mute me
16:13:10 <Zakim> zwu2 should now be muted
16:13:33 <gkellogg> … Based an AZ's (negative) feedback, I worked on the bijection version, but that has also received negative comments.
16:13:46 <FabGandon> FabGandon has joined #rdf-wg
16:13:53 <gkellogg> … The ball is in my (cygri's) court to resolve now.
16:14:30 <Guus> q?
16:14:45 <Guus> ack AZ
16:14:50 <PatH> +q
16:14:52 <gkellogg> AZ: I think the ball is actually on my side; I haven't yet responded to your last email.
16:14:54 <Zakim> +FabGandon
16:15:06 <gkellogg> … your last email clarified a lot of things, and is going in the right direction.
16:15:23 <Guus> ack PatH
16:16:04 <gkellogg> PatH: speaking as semantics editor, the original version, has all the advantages.
16:16:20 <gkellogg> … it's simplest and mathematically cleanest, and easiest to attach semantics to.
16:16:32 <AZ> q+
16:16:38 <gkellogg> … I think it's way superior to bijections and other recent attempts.
16:17:21 <gkellogg> cygri: the first one was written to describe a possible direction, not as spec text.
16:17:48 <gkellogg> … it would need to be written a little bit differently, to define terms and distribute among sub-sections.
16:18:16 <gkellogg> … as far as the technical design works well.
16:18:45 <gkellogg> … I changed it based on feedback from AZ and AndyS
16:19:04 <gkellogg> david: to be clear, the one marked as outdated.
16:19:12 <gkellogg> … The semantics need to be aligned.
16:19:24 <davidwood> s/david/guus/
16:19:30 <Guus> zakim, who is here?
16:19:30 <Zakim> On the phone I see cygri, pchampin, davidwood, MacTed (muted), gkellogg, +1.408.992.aaaa, Arnaud (muted), Sandro, GavinC, AZ, tbaker (muted), cgreer, PatH, Guus, markus, zwu2
16:19:30 <gkellogg> AZ: I wanted to answer PatH's comments.
16:19:34 <Zakim> ... (muted), FabGandon
16:19:34 <Zakim> On IRC I see FabGandon, zwu2, markus, Guus, pfps, PatH, AlexHall, cgreer, AZ, Arnaud, pchampin, cygri, tbaker, Zakim, RRSAgent, gkellogg, MacTed, gavinc, davidwood, yvesr, manu,
16:19:34 <Zakim> ... manu1, mischat, trackbot, sandro, ericP
16:19:48 <gkellogg> … I think scopes shouldn't be part of the abstract syntax.
16:20:02 <gkellogg> … The scope should be defined as the context in which the variable appears.
16:20:21 <gkellogg> … I don't want to see something such as <ID,Scope>, as the scope depends on the context.
16:20:30 <gkellogg> PatH: the scope is the context.
16:20:41 <gkellogg> AZ: the same BNode could exist in separate contexts
16:20:48 <gkellogg> PatH: no, that's what causes the problems.
16:21:03 <gkellogg> guus: as a non-expert, scope as context would be my understanding too.
16:21:20 <gkellogg> AZ: I also have to answer PatH's email.
16:21:21 <cygri> q+
16:21:31 <gkellogg> ack AZ
16:21:46 <gkellogg> guus: i'd like to get text in revised WD.
16:21:59 <gkellogg> … Do we need this text for a revised WD? I think yes.
16:22:12 <gkellogg> … Can we go with the proposal?
16:22:34 <gkellogg> cygri: I was working under the understanding that 107 doesn't need to be resolved to publish.
16:22:59 <gkellogg> … Since there will still be an LC draft, I think it would be fine to publish what we have now with minor editorial changes
16:23:11 <gkellogg> … We can then revisit ISSUE-107 until afterwards.
16:23:36 <AZ> +1 for a new public WD quickly, and solve remaining issues for a Last Call version
16:23:58 <gkellogg> … Regarding ISSUE-107, the crucial thing is that a BNode can only appear in some local scope, and not anywhere else.
16:24:00 <davidwood> +1 to Richard - publish a WD after AZ's editorial comments are in and address ISSUE-107 in the LC.
16:24:27 <gkellogg> … If we can make this work, it would resolve the odd situation which would imply that the same BNode can appear in multiple places.
16:24:44 <PatH> In order to write the semantics of bnodes rationally, I need the notion of the scope/context that the bnodes in a given graph are all in. I would prefer to use the phrasing that "the graph is in the scope".
16:24:56 <gkellogg> … The semantics suggests that if you get the graph from someplace else, you may get the same BNode which would overlap.
16:25:12 <PatH> +1 to wha richard is saying.
16:25:17 <davidwood> +1 to Richard (again). Let's get rid of blank nodes that transit scopes.
16:25:20 <gkellogg> … This is why we need to clarify in the abstract syntax that BNodes exist only in a given scope.
16:25:30 <gkellogg> … This addresses graph union use cases.
16:25:44 <gkellogg> … This is the main way we can improve on the current situation.
16:25:57 <davidwood> q?
16:26:00 <AZ> to PatH, "graph in a scope" is fine, not bnode in a scope
16:26:10 <davidwood> ack cygri
16:26:24 <gkellogg> guus: Let's move to publication of WD.
16:26:42 <gkellogg> … Let's leave inclusion of ISSUE-107 to editor's discretion.
16:27:02 <gkellogg> cygri: fine, I'd like to get the WD out ASAP, and it's not on the critical path.
16:27:21 <PatH> Az, you have not given any reason for your objection, so I can't take it seriously. The nodes are the things that the scopes are needed for.
16:27:48 <gkellogg> pfps: my review in progress.
16:28:10 <gkellogg> … I don't expect anything too big.
16:28:19 <gkellogg> … There are some required wording changes.
16:29:12 <cygri> status of Concepts: http://lists.w3.org/Archives/Public/public-rdf-wg/2012Nov/0317.html
16:29:36 <gkellogg> cygri: the link is for all open issues that can wait until after LC.
16:29:48 <gkellogg> s/after/before/
16:30:21 <gkellogg> guus: I'd like to take a provisional decision to publish depending on pfps' review.
16:30:41 <gkellogg> cygri: i'll address AZ and pfps' comments before publishing.
16:30:59 <AZ> I see no reason, from my side, that there would be any troublesome changes
16:32:11 <gkellogg> PROPOSAL: publish revised WD of RDF Concepts
16:32:27 <gkellogg> … under the condition that editors reach agreement with reviewers.
16:32:34 <pfps> +1
16:32:35 <gkellogg> +1
16:32:37 <cygri> +1
16:32:38 <davidwood> +1
16:32:39 <pchampin> +1
16:32:40 <FabGandon> +1
16:32:40 <zwu2> +1
16:32:41 <markus> +1
16:32:44 <gavinc> +1
16:32:47 <PatH> +1
16:32:49 <AZ> +1
16:33:02 <gkellogg> RESOLVED: publish revised WD of RDF Concepts under the condition that editors reach agreement with reviewers
16:33:13 <sandro> +1
16:33:21 <gkellogg> guus: thanks to reviewers and editors.
16:33:33 <gkellogg> topic: JSON-LD
16:34:38 <davidwood> Zakim, who is on the phone?
16:34:38 <Zakim> On the phone I see cygri, pchampin, davidwood, MacTed (muted), gkellogg, +1.408.992.aaaa, Arnaud (muted), Sandro, GavinC, AZ, tbaker (muted), cgreer, PatH, Guus, markus, zwu2
16:34:41 <Zakim> ... (muted), FabGandon
16:34:47 <cygri> scribe: cygri
16:35:05 <cygri> gkellogg: we're getting through the issue list
16:35:11 <Guus> q?
16:35:21 <cygri> ... one major thing we're working for is richard's contribution
16:35:35 <cygri> ... another issue: graphs vs datasets
16:35:59 <cygri> ... can you content-negotiate between JSON-LD and turtue, and how does the presence of named graphs affect it?
16:36:05 <cygri> ... this seems related to ISSUE-105
16:36:29 <cygri> ... what an agent should do if treating the data source as a graph, it should ignore named graphs
16:36:37 <cygri> ... for example in SPARQL FROM
16:36:55 <cygri> ... provenance is an important use case
16:37:47 <PatH> Ugh. Don't like anyone saying how others OUGHT to use dataset structure.
16:37:50 <cygri> ... we'd like to put provenance into a named graph and make statements about the default graph
16:37:56 <PatH> +q
16:38:03 <Guus> Guus has joined #rdf-wg
16:38:11 <gkellogg> q?
16:38:17 <davidwood> ack PatH
16:38:40 <cygri> PatH: i object to laying down the law on how people ought to use datasets
16:38:56 <cygri> ... some people will use it one way and some in other ways
16:39:07 <sandro> q?
16:39:26 <cygri> ... if we knew how they use it, we could have defined a semantics, but we couldn't because of disagreement on use
16:39:49 <Guus> +1 to PatH; I was planning to make the same comment
16:40:04 <davidwood> q+
16:40:15 <cygri> ... a better way would be to put a triple into the default graph that directs attention to a named graph
16:40:20 <sandro> q+
16:40:22 <cygri> ... legislating the use of datasets is a bad idea
16:40:24 <markus> markus has joined #rdf-wg
16:40:42 <cygri> gkellogg: what is the required behaviour of an agent that receives a dataset when it's looking for a graph?
16:41:07 <cygri> ... options: use only the default graph; use only the named graph corresponding to the document location; or merge the graphs
16:41:49 <cygri> ... if i poke an endpoint, i would expect to get just triples
16:42:12 <cygri> ... except if the answer is json-ld, then i'd expect to get provenance info as well
16:42:23 <cygri> PatH: that's one way but there are other ways
16:42:26 <Guus> ack davidwood
16:42:40 <PatH> pointer?
16:42:44 <Zakim> +??P1
16:42:48 <cygri> davidwood: LDP-WG had a long thread about "LDP would benefit from being RESTful"
16:43:18 <cygri> ... it was felt that LDP creates something on top of REST and perhaps shouldn't
16:43:35 <cygri> ... situation here is similar. don't define something on top of RDF abstract syntax
16:43:52 <MacTed> MacTed has joined #rdf-wg
16:44:50 <cygri> sandro: i'm sympathetic to dave's and pat's points
16:44:59 <cygri> ... but i don't know how to solve this
16:45:02 <PatH> Join the club :-)
16:45:25 <cygri> ... if the same syntax can carry graphs and datasets, that's a problem
16:45:28 <cygri> q+
16:45:36 <Guus> ack sandro
16:46:18 <cygri> ... simply using the default graph when graphifying a dataset seems least problematic
16:46:28 <gavinc> This is ONLY for JSON-LD
16:46:45 <cygri> (scribe can't keep up with crosstalk)
16:46:50 <gkellogg> no, we need to consider it for any serialization that provides datasets
16:46:53 <gkellogg> q+
16:47:02 <Guus> ack cygri
16:47:34 <gkellogg> cygri: we wanted to make a distinction between dataset syntaxes and graph syntaxes.
16:47:46 <gkellogg> … This is why TriG is not an extension of Turtle.
16:48:11 <sandro> actually.......     steve might really have a problem with json-ld
16:48:19 <gkellogg> … It's really a legacy consideration for Turtle, but it doesn't apply to JSON-LD, as there isn't a large set.
16:48:40 <gkellogg> … There is still the semantics issue. Are we messing too much with the architecture.
16:48:59 <sandro> its like con-neg between turtle and trig.
16:49:04 <cygri> gkellogg: there are issues with tying this to the syntax
16:49:07 <PatH> The same point applies to the future. What is a legacy issue facing backwards becomes a usability issue facing forwards. But what the hell, I'm not going to ever uyse JSON in any case.
16:49:09 <Guus> ack gkellogg
16:49:12 <cygri> ... consumers do content negotiation
16:49:29 <cygri> ... and don't know before that they may be dealing specifically with JSON-LD
16:49:42 <cygri> ... there were suggestions to add named graphs to RDFa
16:49:52 <cygri> ... so these issues are not tied to a specific syntax
16:49:58 <Zakim> +??P1
16:50:02 <gavinc> Simple answer, you can't.
16:50:07 <cygri> ... but to the question how datasets ought to be interpreted in a graph context
16:50:09 <cygri> q+
16:50:11 <gavinc> I'm a fan of You Can't.
16:50:14 <PatH> +1 tp last speaker.
16:50:36 <Zakim> -GavinC
16:50:44 <gavinc> hi?
16:50:49 <gavinc> Hey, the power went out ;)
16:50:57 <gkellogg> cygri: Perhaps Sandro's option works: if you can only consume triples, and you're confronted with a dataset, use the default graph and ignore everything else.
16:51:12 <gkellogg> … It seems like the obvious approach.
16:52:02 <gkellogg> guus: what about the option where you have a graph named the same as the document?
16:52:20 <gkellogg> sandro: the usual case is where there are no named graphs, just the default graph.
16:52:41 <gkellogg> … If you convert Turtle to JSON-LD you have a dataset with a default graph and no named graphs.
16:52:55 <Zakim> +GavinC
16:53:20 <gavinc> Turtle is a single graph.
16:53:40 <cygri> sandro: if you convert a dataset to JSON-LD, named graphs are intact
16:53:45 <gkellogg> sandro: if you turn a dataset into JSON-LD, you get a dataset.
16:53:46 <gavinc> JSON-LD can be either a single graph or a dataset
16:53:56 <cygri> davidwood: JSON-LD can serialize datasets. turtle can't.
16:54:16 <cygri> guus: we might have to get back to this issue later
16:54:30 <cygri> ... other issues on JSON-LD?
16:54:34 <cygri> gkellogg: that's what occupied most of the time recently
16:54:40 <Guus> q?
16:54:43 <cygri> ... the other thing is the RDF alignment
16:54:45 <cygri> ack me
16:54:49 <Guus> ack cygri
16:55:01 <cygri> sandro: when will the documents be ready for review?
16:55:32 <cygri> gkellogg: blocked on richard
16:55:51 <cygri> ... other discussion was on renaming the API document to something focusing more on processors and algorithms
16:56:30 <PatH> David, thnx. I should have said trig :-)
16:56:36 <davidwood> davidwood has left #rdf-wg
16:56:38 <cygri> ... other discussion was on moving more normative text from Syntax to API, and make the Syntax doc more primer-ish
16:56:48 <gkellogg> markus: there are two opinions, that there should be two documents and there's no value in a primer, and other see the advantage of a single normative document.
16:56:50 <cygri> markus: no consensus on that
16:57:02 <davidwood> davidwood has joined #rdf-wg
16:57:24 <gkellogg> markus: My opinion is that the syntax should be separate from algorithms, as most people are just interested in the syntax.
16:58:19 <cygri> gkellogg: some issues have to be resolved with RDF-WG. document renaming would require new TR shortnames for instance
16:58:34 <cygri> guus: when is next version expected for review?
16:58:36 <cygri> q+
16:58:46 <cygri> gkellogg: mostly waiting for richard
16:58:58 <PatH> In years to come its a lot easier to find what you want in a document rather than find the right document. (See the OWL specs for example)
16:59:01 <Guus> ack cygri
16:59:17 <gkellogg> cygri: Ive' had this issue for a number of weeks; I've made progress, but it's not complete.
16:59:41 <gkellogg> … It's difficult to commit to a specific date, as I've been consumed by RDF concepts for the last couple of weeks, and there's still work to do.
17:00:44 <zwu2> have to go to another meeting. bye!
17:00:53 <Zakim> -zwu2
17:00:57 <cygri> gkellogg: richard is in a unique position as he has looked at the specs in detail and knows concepts
17:01:29 <gkellogg> cygri: it seems unlikely that I'll be able to finish in the next week.
17:02:22 <gkellogg> sandro: To write this, someone needs to know both the RDF model and the JSON-LD model, so I would hope that it could be phrased in those terms easily.
17:02:54 <gkellogg> cygri: from the work I've done, the difficulty has been that the JSON-LD data model isn't explicit enough; It's a simplification that doesn't spell out the details enough.
17:03:27 <gkellogg> … It hasn't been designed by defining a data mode and then describe how the data model is expressed in JSON; it needs to be reverse engineered from the algorithms.
17:04:04 <gkellogg> … The main issue isn't knowledge of concepts, but to know what the data model is, you need to study the algorithms to see what parts survive and what doesn't survive.
17:04:45 <gkellogg> … what I've been trying is to do through the algorithms and see what survives when you flatten and go back.
17:05:26 <gkellogg> markus: cygri, do you have a draft available?
17:05:34 <gkellogg> cygri: there's a wiki page I started.
17:05:55 <cygri> http://www.w3.org/2011/rdf-wg/wiki/JSON-LD_Data_Model
17:06:34 <gkellogg> guus: I suggest that gregg and markus can use this to write up the data model themselves.
17:08:28 <cygri> gkellogg: i won't get to this in the next week. might have time to work on it over christmas.
17:08:44 <gkellogg> markus: I'll try to work on this.
17:08:46 <cygri> markus: i will try to come up with something till next week. can't promise though
17:09:08 <cygri> sandro: we said at the F2F that we don't want JSON-LD to have a different data model from RDF
17:09:09 <gkellogg> sandro: I want to remind the group that at the F2F we didn't want to have anything that wasn't in the RDF data model.
17:10:01 <gkellogg> guus: a few minutes for RDF Semantics.
17:10:02 <cygri> topic: RDF Semantics
17:11:00 <gkellogg> guus: a nice start would be existing draft marked up with issues.
17:11:12 <gkellogg> PatH: I can do something like this, but it will be a couple of weeks.
17:11:33 <gkellogg> guus: The named graph stuff won't change the semantics much, but there are other small issues.
17:12:02 <gkellogg> ACTION: PatH to provide update to RDF semantics within two weeks.
17:12:02 <trackbot> Created ACTION-215 - Provide update to RDF semantics within two weeks. [on Patrick Hayes - due 2012-12-05].
17:12:55 <gkellogg> guus: it would be nice to point to a draft by the end of january where it's clear where we are
17:13:02 <gkellogg> … Important for extension request.
17:13:31 <gkellogg> … If we can have the draft with issues before christmas, we'll be in good shape.
17:13:50 <gkellogg> sandro: holiday schedule?
17:14:25 <gkellogg> guus: proposal is 19 December, then two week break.
17:14:37 <gkellogg> … no telecom on 26 December or 2 January.
17:15:02 <gkellogg> … This means we need to do some careful planning, so we're ready for the extension request.
# SPECIAL MARKER FOR CHATSYNC.  DO NOT EDIT THIS LINE OR BELOW.  SRCLINESUSED=00000366