Chatlog 2011-03-09

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.

15:57:19 <RRSAgent> RRSAgent has joined #rdf-wg
15:57:19 <RRSAgent> logging to
15:57:28 <cmatheus> rrsagent, make records public
15:57:50 <cmatheus> Scribe: cmatheus
15:58:28 <NickH> NickH has joined #rdf-wg
15:59:50 <AndyS> no zakim?
16:00:04 <NickH> UK phone number seems to be working much better than it was
16:00:20 <JFB> Hi, is the conference code really 73394 (it's not accepted)?
16:00:37 <NickH> JFB: 73394 worked for me
16:00:38 <AZ> JFB, yes it is
16:00:41 <AlexHall> AlexHall has joined #rdf-wg
16:00:50 <JFB> Still trying....
16:00:54 <davidwood> Chair: Guus Schreiber
16:00:58 <AndyS> Code worked for me but no zakim annoucements in IRC.
16:01:12 <cygri> cygri has joined #rdf-wg
16:01:27 <sandro> zakim, this is rdf-wg
16:01:37 <Zakim> Zakim has joined #rdf-wg
16:01:40 <Guus> zakim, who is here?
16:01:40 <Zakim> sorry, Guus, I don't know what conference this is
16:01:42 <pchampin> pchampin has joined #rdf-wg
16:01:42 <Zakim> On IRC I see cygri, AlexHall, NickH, RRSAgent, AZ, pfps, dfensel6, gavin, cmatheus, hsbauer, Guus, JFB, LeeF, webr3, SteveH, AndyS, ivan, davidwood, yvesr, manu, manu1, sandro,
16:01:44 <zwu2> zwu2 has joined #rdf-wg
16:01:45 <Zakim> ... trackbot
16:01:57 <sandro> zakim, this is rdf-wg
16:01:57 <Zakim> sorry, sandro, I do not see a conference named 'rdf-wg' in progress or scheduled at this time
16:02:00 <sandro> zakim, this is rdfwg
16:02:03 <Zakim> ok, sandro; that matches SW_RDFWG()11:00AM
16:02:10 <sandro> zakim, who is on the call?
16:02:10 <Zakim> On the phone I see Scott_Bauer, Dieter, ??P18, Guus, Azimmerm, ??P36, [IPcaller], AlexHall, LeeF, Sandro, zwu2, mhausenblas
16:02:18 <cygri> zakim, mhausenblas is me
16:02:18 <Zakim> +cygri; got it
16:02:21 <zwu2> zakim, mute me
16:02:21 <Zakim> zwu2 should now be muted
16:02:26 <davidwood> I'll be a few minutes late :(
16:02:29 <AndyS> zakim, IPCaller is me
16:02:32 <Zakim> +AndyS; got it
16:02:34 <Zakim> +Peter_Patel-Schneider
16:02:39 <ivan> zakim, dial ivan-voip
16:02:47 <Zakim> ok, ivan; the call is being made
16:02:49 <Zakim> +Ivan
16:02:51 <AxelPolleres> AxelPolleres has joined #rdf-wg
16:02:54 <mbrunati> mbrunati has joined #rdf-wg
16:02:59 <Zakim> +gavinc
16:03:09 <Zakim> +JeanFrancois
16:03:15 <Zakim> +[IPcaller]
16:03:18 <sandro> sandro has changed the topic to: RDF WG Meeting 2011-03-09
16:03:19 <webr3> zakim, i am IPcaller
16:03:19 <Zakim> +Luca
16:03:24 <manu> zakim, code?
16:03:29 <cmatheus> Topic: Admin
16:03:30 <sandro> Agenda:
16:03:31 <Zakim> ok, webr3, I now associate you with [IPcaller]
16:03:37 <Zakim> the conference code is 73394 (tel:+1.617.761.6200 tel:+ tel:+44.203.318.0479), manu
16:03:40 <ivan> +1
16:03:42 <tomayac> tomayac has joined #rdf-wg
16:03:42 <webr3> +1
16:03:45 <Zakim> +AxelPolleres
16:03:46 <cmatheus> Guus: Approve agenda
16:03:46 <JFB> +1
16:03:53 <Zakim> +davidwood
16:03:56 <sandro> PROPOSED: accept previous minutes
16:04:00 <cmatheus> ... minutes accepted
16:04:04 <sandro> RESOLVED: accept previous minutes
16:04:09 <Zakim> +??P55
16:04:11 <Zakim> +[IPcaller.a]
16:04:15 <manu> zakim, I am ??P55
16:04:15 <Zakim> +manu; got it
16:04:19 <cmatheus> ... day light savings time
16:04:25 <Zakim> +[IPcaller.aa]
16:04:36 <mischat> mischat has joined #rdf-wg
16:04:45 <mbrunati> zakim, 
16:04:45 <Zakim> I don't understand '', mbrunati
16:04:56 <cmatheus> ... custom: follow american time
16:05:03 <Zakim> + +1.404.978.aaaa - is perhaps Dzung_Tran?
16:05:09 <mbrunati> zakim, [IPcaller.aa] is me
16:05:09 <Zakim> +mbrunati; got it
16:05:19 <sandro> Guus: For European folks, the next two telecons will be an hour early.
16:05:36 <SteveH> SteveH has joined #rdf-wg
16:05:38 <davidwood> We need an international standard for DST…
16:05:42 <Zakim> +Souri
16:05:48 <cmatheus> Guus: next telecoms - note time differences
16:05:55 <cmatheus> ... action item review
16:05:59 <gavin> davidwood, yes, to banish it forever and never speak of it again
16:06:01 <cmatheus> Topic: Action Items
16:06:01 <NickH> Zakim, ??P36 is me
16:06:01 <Zakim> +NickH; got it
16:06:02 <sandro> davidwood, that could never work for north-vs-southern hemisphere, at least.  :-)
16:06:04 <mbrunati> q-
16:06:07 <Souri> Souri has joined #rdf-wg
16:06:11 <Guus> q?
16:06:23 <Zakim> +??P49
16:06:28 <Guus> ack Luca
16:06:29 <SteveH> Zakim, who's on the phone?
16:06:30 <Zakim> On the phone I see Scott_Bauer, Dieter, ??P18, Guus, Azimmerm, NickH, AndyS, AlexHall, LeeF, Sandro, zwu2 (muted), cygri, Peter_Patel-Schneider, Ivan, gavinc, JeanFrancois,
16:06:33 <sandro> zakim, who is talking?
16:06:33 <Zakim> ... [IPcaller], pchampin, AxelPolleres, davidwood, manu, [IPcaller.a], mbrunati, Dzung_Tran?, Souri, ??P49
16:06:35 <mischat> zakim, ??P49 is me 
16:06:35 <Zakim> +mischat; got it
16:06:41 <mischat> zakim, mute me
16:06:41 <Zakim> mischat should now be muted
16:06:45 <Zakim> sandro, listening for 10 seconds I heard sound from the following: Guus (13%)
16:07:04 <SteveH> Zakim, [IPcaller] is me
16:07:04 <Zakim> +SteveH; got it
16:07:07 <sandro> q?
16:07:13 <cmatheus> ... Action 7
16:07:23 <webr3> ->
16:07:25 <cygri> ACTION-7?
16:07:25 <trackbot> ACTION-7 -- Nathan Rixham to write a wiki page clarifying the different "graph" concepts -- due 2011-03-02 -- PENDINGREVIEW
16:07:25 <trackbot>
16:07:26 <cmatheus> ... Nathan wiki page on graph concepts
16:07:32 <cmatheus> Nathan: done
16:07:50 <sandro> close ACTION-7
16:07:50 <trackbot> ACTION-7 Write a wiki page clarifying the different "graph" concepts closed
16:07:51 <AxelPolleres> close ACTION-7
16:07:51 <trackbot> ACTION-7 Write a wiki page clarifying the different "graph" concepts closed
16:07:56 <cmatheus> ... Sandro, can you mark action 7 as closed
16:08:04 <cmatheus> Resolved: Action 7 closed
16:08:08 <sandro> ACTION-12?
16:08:08 <trackbot> ACTION-12 -- Guus Schreiber to talk to paul groth to get a provenance use case for graphs -- due 2011-03-02 -- PENDINGREVIEW
16:08:08 <trackbot>
16:08:16 <cmatheus> ... action 12
16:08:21 <sandro> close ACTION-12
16:08:21 <trackbot> ACTION-12 Talk to paul groth to get a provenance use case for graphs closed
16:08:22 <cmatheus> ... closed
16:08:32 <sandro> ACTION-15?
16:08:32 <trackbot> ACTION-15 -- Guus Schreiber to make hotel suggestions for FTF1 -- due 2011-03-09 -- PENDINGREVIEW
16:08:32 <trackbot>
16:08:45 <cmatheus> ... action 15 - hotel suggestions - done.
16:08:55 <sandro> close ACTION-15
16:08:55 <trackbot> ACTION-15 Make hotel suggestions for FTF1 closed
16:09:00 <cmatheus> Resolved: Action 12 closed
16:09:19 <cmatheus> Resolved: Action 15 closed
16:09:23 <sandro> guus: No hotels at CWI, but three groups close by. not a big city.    at most 35 minutes.
16:09:28 <manu> I have created a wiki page attempting to collect design requirements for RDF in JSON here:
16:09:37 <manu> (That's related to ACTION-16)
16:09:41 <sandro> cmatheus, please don't use "Resolved" for actions.
16:09:54 <cmatheus> sandro, got it.
16:10:10 <sandro> action-18?
16:10:10 <trackbot> ACTION-18 -- Ivan Herman to establish a wiki page for the FTF1 agenda and list initial content -- due 2011-03-09 -- PENDINGREVIEW
16:10:10 <trackbot>
16:10:15 <sandro> close action-18
16:10:15 <trackbot> ACTION-18 Establish a wiki page for the FTF1 agenda and list initial content closed
16:10:22 <cmatheus> ... Action 18 - closed.
16:10:48 <cmatheus> ... Dan Brickley action on named graph. not on call.
16:10:54 <sandro> action-5?
16:10:54 <trackbot> ACTION-5 -- Yves Raimond to draft a use case for named graphs from BBC work -- due 2011-03-02 -- OPEN
16:10:54 <trackbot>
16:11:03 <cmatheus> ... Action 5 remains open
16:11:35 <cmatheus> ... Action for Pat -- regrets on being out for ten days.
16:11:38 <mischat> zakim, unmute me
16:11:38 <Zakim> mischat should no longer be muted
16:11:47 <cmatheus> ... Mishat to provided ...
16:12:05 <sandro> close action-11
16:12:06 <trackbot> ACTION-11 Provide Garlik pov re: use-cases with SteveH closed
16:12:09 <cmatheus> Mischat: turned action into pending reviews.
16:12:12 <mischat> zakim, mute me 
16:12:12 <Zakim> mischat should now be muted
16:12:18 <sandro> action-16?
16:12:18 <trackbot> ACTION-16 -- Manu Sporny to summarize positions that folks have taken via the mailing list onto the wiki in an attempt to figure out which document should be used as a starting point for the RDF in JSON work. -- due 2011-03-09 -- OPEN
16:12:18 <trackbot>
16:12:27 <cmatheus> Guus: action closed.
16:12:59 <cmatheus> Manu: JSon - summarization of positions.  still under discussion.
16:13:06 <sandro> close action-16
16:13:06 <trackbot> ACTION-16 Summarize positions that folks have taken via the mailing list onto the wiki in an attempt to figure out which document should be used as a starting point for the RDF in JSON work. closed
16:13:18 <cmatheus> Guus: you did your action item and it can be closed.
16:13:22 <sandro> action-17?
16:13:22 <trackbot> ACTION-17 -- Gavin Carothers to try and produce a digram based on the g-box, g-snap, g-text model from Sandro's email and this conversation -- due 2011-03-16 -- OPEN
16:13:22 <trackbot>
16:13:37 <cmatheus> ... action 17: g-box diagram
16:14:00 <cygri> it was gavin speaking
16:14:02 <cmatheus> Gavin: haven had time to get to it this week.
16:14:03 <sandro> gavin: I'll do it soon and put it on the wiki page of graph concepts
16:14:21 <cmatheus> Guus:  keep open, hopefully close next week.
16:14:36 <cmatheus> ... ends discussionon actionitems
16:15:01 <cmatheus> Topic: F2F1
16:15:05 <Guus> q?
16:15:22 <mischat>
16:15:30 <cmatheus> Guus: please indicate whether you are attending on wiki page.
16:15:45 <ivan> q+
16:15:48 <cmatheus> ... need a page to track regrets.
16:16:16 <ivan> q-
16:16:58 <cmatheus> Ivan: needs to give list of all participants.  if you don't provide a name you may not be permitted in.
16:17:00 <davidwood> Regrets section added to
16:17:18 <cmatheus> ... will have two meeting rooms plus my office.
16:17:39 <cmatheus> Guus: moving on to the agenda for F2F
16:18:20 <cmatheus> ... proposal: use day one to discussion items on table for task forces and identify what's need to be done for first draft.
16:18:46 <cmatheus> ... structure: reporting, discussion, planning
16:19:09 <cmatheus> ... dinner in the evening by university. lunch offered by Talis.
16:19:31 <cmatheus> ... need to be more specific on topics.  can we live with this layout?
16:19:37 <cygri> +1
16:19:41 <JFB> +1
16:19:42 <manu> +1 for F2F1 Agenda
16:19:58 <mbrunati> +1
16:20:04 <davidwood> +1
16:20:05 <cmatheus> ... F2F1 agenda approved.
16:20:07 <pfps> not much more can be done at this time to figure out agenda for F2F1
16:20:14 <tomayac> +1
16:20:26 <cmatheus> ... next two weeks we will fill in the details.
16:20:35 <cmatheus> Topic: Task Forces
16:21:15 <cmatheus> Guus: general remark: very happy to see so many threads.
16:21:27 <cmatheus> ... some worry that we might go outside the charter.
16:21:42 <cmatheus> ... over next few weeks we need to start restricting outselves.
16:21:49 <mischat> s/Guss/Guus/
16:21:54 <cmatheus> ... must need way to manage things over coming year.
16:22:00 <cygri> q+
16:22:07 <cmatheus> ... can't do everything we've talked about over last two weeks.
16:22:14 <cmatheus> ... any comments?
16:22:50 <cmatheus> Richard: good to have broad discussion to get issues on table, but also important to make clear what charter is.  what it allows us to do and what it doesn't.
16:23:12 <cmatheus> ... would be useful for thos ewho wrote the charter to speak up when things out of scope come up.
16:23:14 <sandro> q+ to say writers intent doesn't matter....
16:23:30 <cygri> q-
16:23:55 <sandro> q-
16:24:10 <cmatheus> Guus: agreed, we need to start limiting discussions.  point well taken.
16:25:09 <cmatheus> David: we don't want to let the conversation to get out of control but people should feel free to make proposals and voice opinions.
16:25:13 <AndyS> Are changes that affect other RECs in or out of scope? (by intent - words do no restrict this WG)
16:25:21 <cmatheus> Guus: moving on to task force discussions.
16:25:36 <cmatheus> ... starting with Turtle TF
16:25:38 <mischat>
16:26:05 <cmatheus> ... potential deliverables for Turtle work.
16:26:12 <cygri>
16:26:17 <cmatheus> Richard: on background of the wiki page -- I created it.
16:27:01 <cmatheus> ... some terms in discussions where captured in wiki page.  
16:27:47 <manu> +1 to Qurtle, N-Quads++
16:27:58 <manu> -1 to everything else (current opinion, may change)
16:28:01 <cmatheus> ... summary: make minimal fixes (e.g. aligning with sparql); super turtle would add some additional properties; qurtle would add quad support.
16:28:19 <ivan> q+
16:28:26 <sandro> I suggest the names N-Triples2 and N-Quads2 are more descriptive than "++" since it's not backward compatible.
16:29:02 <cmatheus> ... still under discussion.  n-triple++ could be recognized and turned into a minimal format for exchanging triples. 
16:29:25 <cmatheus> ... n-quads - take current proposal and add option for naming contexts/graphs.
16:29:35 <gavin> Sandro, why would it not parse current n-triples/n-quads?
16:29:36 <cmatheus> ... idea is to turn it into a specification.
16:29:52 <cmatheus> ... has been said that fourth element should be required.
16:30:28 <cmatheus> ... another proposal: rdf-tuples like csv.  more like a serialization of a sparql result set.
16:30:37 <sandro> gavin, I suppose it could be, but with utf-8 I'd exepect the \uXXXXXXXX syntax to be removed.
16:30:54 <cmatheus> Ivan: comment 1: maybe worth adding what the media types are.
16:31:10 <SteveH> sandro, it's useful for non-UTF-8 systems, like some version of emacs
16:31:34 <cmatheus> ... qurtle should be seperated by media types
16:31:47 <tomayac> +1 for having different media types
16:31:55 <SteveH> +1
16:31:56 <AndyS> sandro, \uXXXX and \UXXXXXXXX in Turtle today.
16:31:59 <cmatheus> ... comment 2: on current page, rdf-tuples is not mentioned in the charter.
16:32:23 <cmatheus> ... whole issue around n-triples and their extension is not in the charter either.
16:32:37 <sandro> zakim, who is talking?
16:32:49 <Zakim> sandro, listening for 10 seconds I heard sound from the following: gavinc (70%), Ivan (8%)
16:33:02 <cmatheus> Gavin: charter speaks of revising existing notes and other parts of RDF
16:33:22 <cmatheus> Ivan: yes, it talks of that on the edge. but rdf-tuples is not in the charter.
16:33:35 <cmatheus> ... n-triples could be squeezed a little bit.
16:33:49 <gavin> +q
16:34:07 <Guus> ack ivan
16:34:33 <cmatheus> Ivan: strictly speaking the charter may exclude super-turtle.
16:34:42 <sandro> q+
16:34:54 <Guus> ack gavin
16:35:10 <cmatheus> ... charter says standardize turtle and add graph support.  doesn't include extensions beyond graph support.
16:35:44 <Guus> ack samdro
16:35:51 <cmatheus> Gavin: n-turtle and quads on list because they could use the additional symbol.
16:35:52 <ivan> ack sandro 
16:35:54 <Guus> ack sandro
16:36:00 <pchampin> +1 to standardize tokens
16:36:10 <AndyS> q+
16:36:19 <SteveH> reverse paths are only in the pattern side
16:36:19 <cmatheus> Sandro: don't think standardized turtle excludes things like reverse paths in sparql 1.1.
16:36:26 <AndyS> q-
16:36:27 <SteveH> not in CONSTRUCT { } for e.g.
16:36:28 <AxelPolleres> q+
16:36:29 <cmatheus> Ivan: sparql 1.1 doesn't include reverse paths.
16:36:37 <AndyS> Not in INSERT DATA {}
16:36:46 <SteveH> or CONSTRUCT
16:36:52 <ivan> ack AxelPolleres 
16:36:54 <cygri> q+
16:37:02 <cmatheus> Axel: reverse paths are in sparql 1.1.
16:37:18 <cmatheus> Sandro: not an insert or construct.  they are sort of there and not there.
16:37:23 <Guus> ack cygri
16:37:23 <ivan> ack cygri 
16:38:21 <cmatheus> Richard:charter items:syntax to support multiple graphs.  may be sufficient to support quads.  that's how it was introduced into discussion.
16:39:17 <cmatheus> Guus: at F2F need to focus on quads issue.  and what kind of documents can be produced for the discussion.  suggestions for actions to be taken here?
16:39:47 <cmatheus> Daivid: what about a survey on what direction the group wants to take.
16:39:50 <mischat> wonders what the dependency of quad serialisation in the turtle task force is on the graphs task force ?
16:40:11 <cmatheus> Guus:  have already had these discussions.   would make more sense to summarize the ideas.
16:40:16 <mischat> q+
16:40:20 <ivan> s/Daivid/Manu/
16:40:23 <cmatheus> ... or is that not true.
16:40:24 <mischat> zakim, unmute me 
16:40:24 <Zakim> mischat should no longer be muted
16:40:48 <sandro> q?
16:40:50 <cmatheus> Mischat:  there's a massive dependancy on what comes out of graphs task force.
16:40:54 <sandro> q+
16:40:58 <ivan> ack mischat 
16:41:01 <cmatheus> ... how do people fel about that.
16:41:03 <mischat> zakim, mute me
16:41:03 <Zakim> mischat should now be muted
16:41:11 <mischat> ok
16:41:17 <AndyS> "restore" creeps into dataset publishing.
16:41:22 <cmatheus> Guus: possible alternative route: at F2F just focus on turtle.
16:41:23 <ivan> ack sandro 
16:41:51 <cmatheus> Sandro: find it hard to think of grpah issue in isolation.  thinking about in context of turtle (or somtehing) would be useful.
16:41:59 <SteveH> +1 to sandro
16:42:01 <mischat> +! to sandro
16:42:03 <mischat> +1
16:42:05 <ivan> +1 to sandro
16:42:05 <webr3> and that turtle is just turtle?
16:42:08 <cmatheus> ... suggesting that the turtle issue may be handled by graph tf instead.
16:42:09 <webr3> +1 though
16:42:11 <manu> +1 to sandro
16:42:16 <davidwood> +1 cannot handle in isolation
16:42:17 <pchampin> +1
16:42:22 <mbrunati> +1
16:42:37 <AndyS> Start with TriG
16:42:48 <webr3> yes, trig example 3 is good
16:42:48 <cmatheus> Guus: could there be a strawman proposal for what qurtle could look like on table at F2F?
16:43:07 <mischat> do we need both TriG and N-quads ? 
16:43:11 <cmatheus> Sandro: suggests using name trig instead of qurtle.
16:43:27 <gavin> Yes.
16:43:32 <SteveH> people use turtle and n-triples
16:43:33 <NickH> mischat: parsing performance?
16:43:34 <SteveH> so, yes
16:43:35 <cmatheus> Guus:  do we need both TriG and N-Quads?
16:43:52 <mischat> q+
16:43:54 <cmatheus> ... should we follow Sandro's suggestion for using TriG as strawman proposal?
16:43:58 <ivan> ack mischat 
16:44:04 <sandro> (that was Andy's suggestion I was seconding)
16:44:15 <LeeF> I don't see the harm in having both. I haven't seen any significant cost to the community from having both N-triples and turtle
16:44:37 <gavin> And there is significant benifit
16:44:52 <ivan> ACTION: mischat to make a survey on what serializations triple stores use in the wild
16:44:52 <trackbot> Created ACTION-19 - Make a survey on what serializations triple stores use in the wild [on Mischa Tuffield - due 2011-03-16].
16:44:53 <cmatheus> Mischat:  happy to go through existing implementations of data stores and n-quads.
16:45:00 <LeeF> gavin, yup
16:45:22 <Souri> We use N-Triples quite a bit and planning to support N-Quads as well
16:45:24 <AndyS> q+
16:45:37 <cmatheus> Guus: would a document on turtle have an appendix on how to hande graphs? or a separate doc?
16:45:51 <cmatheus> ... separate doc makes extra doc harder to use.
16:45:56 <ivan> ack AndyS 
16:46:04 <webr3> and if you publish quads, i need a quint store, (recurse up to RDBMS)
16:46:06 <cmatheus> Sandro: if soemone is focused just on Turtle a smaller doc is better.
16:46:16 <cygri> q+
16:46:40 <cmatheus> AndyS: would suggest a single doc.  if TriG doc is free standing there would be a lot that would have to be copied over.
16:46:42 <SteveH> could be 3 docs + a grammar doc
16:47:12 <AndyS> happy for SteveH suggestion as well.  One technical doc.
16:47:18 <webr3> confirm? so more-than-turtle is now part of Graphs-TF, or still Turtle TF (is turtle tf, just for "turtle" as we have it now or)?
16:47:21 <cmatheus> Richard: seconds Andy's statement.  add n-triples to grammar.  additions would probably be quite low.  so vote for a single doc.
16:47:28 <danbri> danbri has joined #rdf-wg
16:47:34 <cmatheus> Guus: let's have this as a dicsussion point at F2F.
16:48:02 <cmatheus> David: is "more than turtle" part of Graph TF?
16:48:16 <cmatheus> Sandro: if it relates to graph additions then yes.
16:48:28 <ivan> s/David/Nathan/
16:48:36 <cmatheus> Guus:  so TriG will be part of Graph TF discussions.
16:48:47 <cmatheus> Topic: JSON TF
16:48:57 <manu>
16:49:02 <cmatheus> Manu: started with list of questions.
16:49:12 <manu>
16:49:13 <cmatheus> ... morphed into a set of preliminary use cases.
16:49:31 <cmatheus> ... if ou can think of a use case that's not here please add it.
16:49:50 <cmatheus> ... change into a set of rdf/json design requirements.
16:50:06 <cmatheus> s/change/changed/
16:50:24 <manu>
16:50:43 <cmatheus> Guus:  the answers to the questions seem to be divided into two groups.
16:51:01 <cmatheus> ... those in favor of user-friendly and whose for machine friendly approach.
16:51:19 <cmatheus> Manu: some confusion due to my statement of questions.
16:51:29 <gavin> I think the an ?s ?p ?o syntax -can- be simpler to explain and use then a complicated "Easy" to use syntax. N-Triples vs. RDF/XML ;)
16:51:32 <cmatheus> ... maybe better to talk about use cases first.
16:51:52 <cmatheus> Guus: for the F2F the main goal will be to decide which of the two approaches should be the major one.
16:52:02 <cmatheus> Manu:  I believe that is the case.
16:52:15 <cmatheus> ... should there be two serializations or just one.
16:52:27 <cmatheus> ... some people pushing for machine readable version.
16:52:28 <cygri> q+
16:52:42 <cmatheus> ... some people argue there's not enough time to do both.
16:53:03 <cmatheus> Guus: fro chairs perspective you will have a strong push to do only one.
16:53:27 <ivan> ack cygri 
16:53:29 <ivan> q+
16:53:32 <Zakim> -mischat
16:53:35 <cmatheus> Richard: assuming we do one syntax, should it be human friendly, or should it be machine friendly, or should it be a compromise?
16:53:38 <gavin> +q
16:53:44 <ivan> ack ivan 
16:53:45 <cmatheus> ... which of those to work on is the main question at the moment.
16:54:22 <cmatheus> Ivan: I asked on the mailing list but didn't get answer the question of what are the communities we are targeting.
16:54:40 <tomayac> there are two strong camps in this wg for one, and for two serializations. 
16:54:44 <cmatheus> ... not of interest to traditional RDF communities.  they are perfectly happy with turtle and won't use JSON.
16:54:47 <manu> +1 to Ivan - RDF in JSON serialization is /not/ for those that are using RDF today w/ RDF/XML and TURTLE
16:54:52 <gavin> -q
16:54:53 <SteveH> +1
16:55:08 <LeeF> Some of us are using RDF today with JSON though
16:55:09 <Guus> [chair hat off]: i disagree with Ivan, we do a lot with JSON in our applications
16:55:13 <NickH> Ivan: I disagree - parsing JSON is super fast in PHP
16:55:13 <cmatheus> ... the other community is not well represented in this group.
16:55:27 <NickH> Ivan: don't need to write a 'new' parser
16:55:35 <cmatheus> Manu: what I intended with the machine-friendly/human-friendly question.
16:55:42 <Guus> q?
16:55:42 <pchampin> q+
16:55:45 <tomayac> i consider myself part of this "other community" ;-)
16:55:47 <webr3> NickH, +1, it's most useful for js environments too (as in node/rhino etc)
16:55:56 <davidwood> [chair hat off] I have come to see use cases for both developers and Web authors as separate requirements.
16:55:59 <gavin> +q
16:56:02 <cmatheus> ... with machine serialization you transform the rdf in to a json structure and you're don.
16:56:16 <pfps> I agree with Guus - the main reason I see for working with JSON here is to allow JSON stuff to feed into RDF (which I am already using)
16:56:25 <AndyS> That was my understanding of machine-serialization. Speaks to : "goal is to provide an RDF serialization as complete as possible"
16:56:36 <cmatheus> with human friednly it's more along line of json-ld.  
16:56:41 <mischat_> mischat_ has joined #rdf-wg
16:56:48 <davidwood> machine-friendly serialization is *easy*.  Human-friendly is not.
16:56:57 <mbrunati> +1 to andy
16:57:08 <NickH> +1 to davidwood
16:57:08 <cmatheus> ... with machine friendly don't care how humans will use it.  with human friendly you expect users to use data like they use it today.
16:57:28 <cmatheus> ... expect things to map to associative arrays.  can use things without a heavy api.
16:57:39 <webr3> q?
16:57:48 <Zakim> +[IPcaller]
16:57:57 <ivan> ack pchampin 
16:58:02 <mischat> zakim, [IPcaller] is me
16:58:06 <Zakim> +mischat; got it
16:58:09 <mischat> zakim, mute me 
16:58:10 <Zakim> mischat should now be muted
16:58:29 <cmatheus> Webr3: probably I have with RDF serialization.  JSON is both human and machine friendly.
16:58:44 <ivan> s/Web3/pchampin/
16:58:47 <Guus> q?
16:58:59 <cmatheus> ... rdf is very different.  the underlining data structure are very different from the documents.
16:59:18 <cygri> maybe rename? "json developer friendly" and "j-triples++"
16:59:30 <cmatheus> ... I like Andy's proposal to let program to handle JSON as usual without having to parse it into a real graph structure.
16:59:35 <webr3> +1 to "rdf goggles"
16:59:41 <ivan> ack gavin 
16:59:46 <Guus> ack gavin
17:00:38 <pfps> +inf to the madness of RDF/XML
17:00:53 <NickH> +1 to gavin!
17:01:05 <webr3> gavin, it was my understanding that human friendly meant simple kv objects that can be used without an api - not "to write easily"
17:01:21 <cmatheus> GavinL making it human friendly make is easier to author.  have a problem with the term "friendly" as applied here -- neither one seems very friendly.
17:01:44 <gavin> Sure, but RDF JSON is "simple" KV objects... that happen to be triples rather then more complicated data struture
17:01:54 <cmatheus> Manu: some people want to translate data in JSON.  some would like data to be immediately usable in a JS program.
17:02:07 <pfps> gavin:  "human friendly" made RDF/XML hostile for both humans and other machines
17:02:19 <AndyS> My worry is the "human friendly" is unclear.  App task seems to influence the friendliness approach.  So TF is quite a lot of work (life of WG), a lot of WG energy.  Happy is doable but I'm doubtful.
17:02:29 <cmatheus> ... if that division exists, we have quite a bit of talking to do to come to consensus. 
17:02:30 <webr3> gavin, something you can do, obj.age - rather than.. well working w/ triples
17:02:34 <AndyS> s/is doable/if doable/
17:02:34 <tomayac> +1 for sandro's view  that there're 2 camps
17:02:45 <ivan> s/sandro/manu/
17:02:54 <pchampin> I agree about the "2 camps" view
17:03:01 <LeeF> I'm in the "interested in serializing triples to JSON" camp, but in fairness I'm also not terribly worried about the need for a standard here.
17:03:19 <SteveH> +1 to LeeF 
17:03:27 <webr3> gavin, well, it wouldn't be RDF.. it would be simple objects w/ a subject - or just some rdf goggles
17:03:32 <cygri> +1 to the concern about ending up with something� like RDF/XML
17:03:42 <gavin> RDFa works great :D
17:03:44 <AndyS> LeeF, module MIME registration.  Good to know format of incoming.
17:03:47 <tomayac> sorry, thanks for the correction, ivan
17:03:55 <gavin> But looks like triples in the end?
17:03:55 <AndyS> LeeF, yes, modulo MIME registration.  Good to know format of incoming.
17:04:05 <cmatheus> Manu: in RDFa would have a tree and set properties from the subject.
17:04:27 <cygri> webr3++
17:04:29 <cmatheus> ... we seem to think that think about triples is easy but the rest of the world thinks about objects.
17:04:39 <Zakim> -mischat
17:04:57 <ivan> q+
17:04:59 <cmatheus> ... objets have key-vaue pairs.  they map to triples but users don't see the mapping.
17:05:09 <ivan> ack ivan
17:05:24 <cmatheus> Ivan: you seem to have jumped to a conclusion too quickly.
17:05:36 <cmatheus> ... the reason for the problem in acceptance is not the fact you have triples.
17:05:37 <mischat_> mischat_ has joined #rdf-wg
17:05:46 <gavin> Triples aren't hard. English majors get them in 15-20 minutes ;)
17:05:51 <davidwood> The "natural" format is the graph, not triples nor quads nor XML nor JSON, etc.  Let's not confuse serialization syntax with the graph.
17:05:52 <tomayac> we should try to think json/objects, and less semantic web
17:05:58 <cmatheus> .. rather the dominating syntax (RDF-XML) made it very difficult to see that there are triples.
17:06:00 <AndyS> I disagree it hasn't worked.  RDF is not a substitution for something else.
17:06:13 <manu> q+ to speak to triples.
17:06:24 <ivan> ack manu
17:06:24 <Zakim> manu, you wanted to speak to triples.
17:06:49 <cmatheus> Manu:  why are we trying to convince people to use triples when they are already comfortable with objects and JSON?
17:06:51 <pchampin> @Ivan: I agree about the problem coming partly from RDF/XML
17:07:05 <pchampin> but my experience is that some people also have difficulties accepting *triples*
17:07:06 <gavin> 'cause objects suck at linking?
17:07:09 <cygri> q+
17:07:18 <cmatheus> ...we shouldn't be trying to retrain the world in how they express their data.
17:07:23 <davidwood> +1 to Manu.  Objects are *most appropriate* for UIs.
17:07:31 <ivan> ack cygri 
17:07:39 <webr3> manu, yes, but that means /not/ changing their current data - so would be more of a data transformation map
17:08:05 <gavin> Maybe we need JSON RDF Syntax and JSON GRDDL?
17:08:06 <cmatheus> Richard: take this with a grain of salt.  you could make the same argument for designing rdf-xml the way it was deisgned.
17:08:17 <cmatheus> ... some people naturally think in trees.
17:08:19 <manu> q+ to discuss HTML+RDFa and why it was successful.
17:08:42 <cmatheus> ... danger if we say let's just treat everything as objects and somehow we'll get out our triples.
17:08:59 <ivan> ack manu 
17:08:59 <Zakim> manu, you wanted to discuss HTML+RDFa and why it was successful.
17:09:01 <cmatheus> ...  why shouldn't I be concerned with this rdf-xml trap?
17:09:25 <tomayac> q+
17:09:29 <cmatheus> Manu:  RDFa was successful because we built it on top of html.
17:09:36 <SteveH> RDF/XML is was very widely used too, it's just not liked
17:09:40 <SteveH> RSS1 for e.g.
17:09:46 <sandro> I don't agree that RDFa was more successful than RDF/XML.
17:09:49 <ivan> q+
17:09:56 <cmatheus> ... was very hard to sell rdf-xml to the world.  rdfa was easy to sell.  not sure why this was the case.
17:09:58 <ivan> ack tomayac 
17:10:08 <cmatheus> Thomas:  triples are not that hard.
17:10:14 <AndyS> Real data is published in N-Triples.  Semi :-)
17:10:25 <LeeF> triples are pretty easy... triples + datatypes + languages + blank nodes + URIs are harder :-)
17:10:45 <cmatheus> ... we at Google are seeing that once people see that the data is just triples they realize that it's not that hard.
17:10:53 <mbrunati> +1 lee and the model stuff as graph
17:10:58 <ivan> q-
17:11:30 <cmatheus> ... my point is that for the rdf-json serialization we can think triples.  let's not limit ourselves and not hide the fact that it is rdf.  it is triples, nothing more nothing less.
17:11:53 <cmatheus> Sandro: I'm wondering if there's candidate syntax that does what you're asking for?
17:12:52 <cmatheus> Thomas: haven't read them all.  elements of some do.  trying to reach a compromise I think we can come up with a bridge between the camps and the development communities.
17:13:18 <cmatheus> Manu:  would it help to go through some of the design requirements?
17:13:46 <Guus> can anybody hear me?
17:13:58 <davidwood> I do hope to make progress on this ftf.  Some things are just easier that way.
17:14:01 <Guus> -
17:14:04 <davidwood> Guus: no :(
17:14:19 <cmatheus> Thomas: this should be something to discuss at F2F.  Some of you have been here for over 10 years some like me have been here for like 1 year or so.  we should get together and try to be objective.
17:14:26 <Zakim> -Guus
17:14:35 <AndyS> I see Turtle as object-ish but linking is first class.  JSON only has strings.
17:14:46 <cygri> excellent point AndyS