- From: Eric Prud'hommeaux <eric@w3.org>
- Date: Fri, 12 Jul 2013 23:28:21 -0400
- To: public-rdf-wg@w3.org
Location: https://www.w3.org/2013/meeting/rdf-wg/2013-07-10
RDF Working Group Teleconference
Minutes of 10 July 2013
Seen
Antoine Zimmermann, Arnaud Le Hors, David Wood, Eric Prud'hommeaux, Gavin Carothers, Gregg Kellogg, Markus Lanthaler, Peter Patel-Schneider, Sandro Hawke, Souripriya Das, Ted Thibodeau, Zhe Wu
Scribe
Eric Prud'hommeaux, Sandro Hawke
IRC Log
Original
Resolutions
1. The WG will pursue publication of RDF/JSON as a Note link
2. Resolve ISSUE-137 using at risk text proposed by sandro http://lists.w3.org/Archives/Public/public-rdf-wg/2013Jul/0138.html and a grammar based on ericP's changes http://lists.w3.org/Archives/Public/public-rdf-wg/2013Jul/
0099.html link
Topics
1. describe relationship between RDF Concepts and RDF Semantics in Concepts
2. RDF/JSON
3. NTriples and NQuads
4. TriG
RRSAgent IRC Bot: logging to http://www.w3.org/2013/07/10-rdf-wg-irc
Trackbot IRC Bot: RRSAgent, make logs world
Trackbot IRC Bot: Zakim, this will be 73394
Zakim IRC Bot: ok, trackbot; I see SW_RDFWG()11:00AM scheduled to start in 5 minutes
14:55:47 <trackbot> Date: 10 July 2013
Zakim IRC Bot: SW_RDFWG()11:00AM has now started
Zakim IRC Bot: +EricP
Zakim IRC Bot: +??P0
Zakim IRC Bot: -EricP
Zakim IRC Bot: +EricP
Gregg Kellogg: zakim, I am ??P0
Zakim IRC Bot: +gkellogg; got it
Zakim IRC Bot: +[GVoice]
Eric Prud'hommeaux: http://w3.org/brief/MzM4
Zakim IRC Bot: +davidwood
David Wood: Zakim, who is barking?
Zakim IRC Bot: I don't understand your question, davidwood.
David Wood: Zakim, who is here?
Zakim IRC Bot: On the phone I see EricP, gkellogg, [GVoice], davidwood
Zakim IRC Bot: On IRC I see Arnaud1, pfps, Zakim, RRSAgent, gkellogg, TallTed, trackbot, davidwood, manu, yvesr, mischat, ericP, sandro
Zakim IRC Bot: +OpenLink_Software
Ted Thibodeau: Zakim, OpenLink_Software is temporarily me
Zakim IRC Bot: +TallTed; got it
Ted Thibodeau: Zakim, mute me
Zakim IRC Bot: TallTed should now be muted
Ted Thibodeau: TallTed has changed the topic to: RDF-WG -- http://www.w3.org/2011/rdf-wg/ -- agenda http://www.w3.org/2011/rdf-wg/wiki/Meetings:Telecon2013.07.10
Arnaud Le Hors: is anyone successfully calling Zakim via SIP?
Arnaud Le Hors: it isn't working for me
Zakim IRC Bot: + +081165aaaa
Antoine Zimmermann: Zakim, aaaa is me
Zakim IRC Bot: +AZ; got it
Zakim IRC Bot: +Sandro
(Scribe set to Eric Prud'hommeaux)
Zakim IRC Bot: + +1.707.861.aabb
David Wood: Zakim, who is here?
Zakim IRC Bot: On the phone I see EricP, gkellogg, [GVoice], davidwood, TallTed (muted), AZ, Sandro, +1.707.861.aabb
Zakim IRC Bot: On IRC I see gavinc, markus, AZ, Arnaud, pfps, Zakim, RRSAgent, gkellogg, TallTed, trackbot, davidwood, manu, yvesr, mischat, ericP, sandro
Gavin Carothers: aabb is me
Gavin Carothers: Zakim, aabb is me
Zakim IRC Bot: +gavinc; got it
Zakim IRC Bot: +[IPcaller]
David Wood: PROPOSED to accept the minutes of the 3 July telecon:
David Wood:
David Wood: https://www.w3.org/2013/meeting/rdf-wg/2013-07-03
Arnaud Le Hors: Zakim, IPCaller is me
Zakim IRC Bot: +Arnaud; got it
Zakim IRC Bot: +??P15
David Wood: RESOLVED to accept the minutes of the 3 July telecon: https://www.w3.org/2013/meeting/rdf-wg/2013-07-03
Markus Lanthaler: zakim, ??P15 is me
Zakim IRC Bot: +markus; got it
PROPOSED: accept https://www.w3.org/2013/meeting/rdf-wg/2013-07-03 as a record of the last meeting
David Wood: Review of action items
David Wood: http://www.w3.org/2011/rdf-wg/track/actions/pendingreview
David Wood: http://www.w3.org/2011/rdf-wg/track/actions/open
David Wood: ACTION-278?
Trackbot IRC Bot: ACTION-278 -- Eric Prud'hommeaux to create new grammar for TriG that includes optional graph keywords and make {} optional around the default graph -- due 2013-07-10 -- OPEN
Trackbot IRC Bot: http://www.w3.org/2011/rdf-wg/track/actions/278
David Wood: ACTION-279?
Trackbot IRC Bot: ACTION-279 -- Sandro Hawke to propose text for TriG feature at risk for both GRAPH keywords and {} being optional around the default graph -- due 2013-07-10 -- OPEN
Trackbot IRC Bot: http://www.w3.org/2011/rdf-wg/track/actions/279
David Wood: close ACTION-278
Trackbot IRC Bot: Closed ACTION-278 Create new grammar for TriG that includes optional graph keywords and make {} optional around the default graph.
David Wood: close ACTION-279
Trackbot IRC Bot: Closed ACTION-279 Propose text for TriG feature at risk for both GRAPH keywords and {} being optional around the default graph.
David Wood: :Reminder: The next telecon will be Wednesday, 24 July
David Wood: (biweekly)
David Wood: staring bio-weekly schedule so next telecon Wed 24 July
Zakim IRC Bot: +AZ
Peter Patel-Schneider: where would this paragraph go?
1. describe relationship between RDF Concepts and RDF Semantics in Concepts
David Wood: do you think it should be verbose?
Peter Patel-Schneider: not sure where it should go
Peter Patel-Schneider: previous version didn't include this
... propose: do nothing
Peter Patel-Schneider: I'm of the opinion that the right thing is to do nothing.
David Wood: we mention semantics in the intro (as a link)
... then under entailment and consistency
... markus noted that 1.7 seems out of place. i agree
... could be fixed if we clarified the relationship between the two documents
... markus, would clarifying the relationshop between the two docs improve 1.7?
Markus Lanthaler: i'd like to move this all to semantics
... it's not relevent to a newcomer and not complete enough
davidwood, but it does include refs which introduces it to readers of concepts
Markus Lanthaler: my goal is to simplify Concepts to not scare off new readers
David Wood: first read should be the primer
Markus Lanthaler: agreed, but most specs will ref Concepts so folks will read that before reading the Primer
David Wood: pfps, how do you feel about moving 1.7 to Semantics?
Ted Thibodeau: Zakim, unmute me
Zakim IRC Bot: TallTed should no longer be muted
Peter Patel-Schneider: something has to reference Semantics
Arnaud Le Hors: you could put a note in the intro recommending to read the primer first
Arnaud Le Hors: just make sure the primer doesn't send people back to concepts from the get-go :)
q+ to say that I agree with markus that concepts readers should be able to read a data model without the hard stuff
David Wood: [at al,] yes, we should point to the primer but we don't have one
Ted Thibodeau: if there's an intended order of reading, that should be indicated at the beginning
David Wood: q?
David Wood: ack ericP
Zakim IRC Bot: ericP, you wanted to say that I agree with markus that concepts readers should be able to read a data model without the hard stuff
David Wood: in the beginning of Concepts, we reference these other documents
Ted Thibodeau: Zakim, mute me
Zakim IRC Bot: TallTed should now be muted
... if we move 1.7 to semantics and define entailment and consistency there, Concepts will have few refs to Semantics
Antoine Zimmermann: Zakim, who's speaking?
... in that section, we just define some terms.
Zakim IRC Bot: AZ, listening for 10 seconds I heard sound from the following: [GVoice] (79%), davidwood (68%)
... should a readers of Concepts understand those terms
Peter Patel-Schneider: i'm uncomfortable with moving 1.7 from Concepts and into Semantics
Markus Lanthaler: are these terms used again in Concepts?
... i see one later ref to "entailment" in an example
Gavin Carothers: "Two RDF graphs A and B are equivalent if they make the same claim about the world. A is equivalent to B if and only if A entails B and B entails A." That one's important
... you could argue that it's basic knowledge for RDF, but...
Ted Thibodeau: Zakim, who's here?
Zakim IRC Bot: On the phone I see EricP, gkellogg, [GVoice], davidwood, TallTed (muted), AZ, Sandro, gavinc, Arnaud, markus, AZ
Zakim IRC Bot: On IRC I see Souri, pchampin, zwu2, gavinc, markus, AZ, Arnaud, pfps, Zakim, RRSAgent, gkellogg, TallTed, trackbot, davidwood, manu, yvesr, mischat, ericP, sandro
Markus Lanthaler: https://dvcs.w3.org/hg/rdf/raw-file/default/rdf-concepts/index.html#entailment
Zakim IRC Bot: +Souri
Markus Lanthaler: http://json-ld.org/spec/latest/json-ld/#data-model
David Wood: http://lists.w3.org/Archives/Public/public-rdf-wg/2013Jun/0126.html
Peter Patel-Schneider: If we are making changes to 1.7, I would make some non-controversial editorial changes. I would also remove the stuff on union and merge (which might be somewhat more controversial). I'll send out a message on
this.
Peter Patel-Schneider: Consensus on everything except my worries about JSON numbers.
Markus Lanthaler: http://json-ld.org/spec/latest/json-ld/#relationship-to-rdf
Markus Lanthaler: we discussed an unambiguos mapping from JSON datatypes to RDF datatypes
... that addressed pfps's concearns
... i just updated the spec 10 mins ago. everything should be in place
Peter Patel-Schneider: i'd have been content yesterday
... this is all painful
... JSON is a loose spec which does thousands of things with numbers
... e.g. 32 bit integers
Sandro Hawke: here we're improving JSON by using RDF's [really XML Schema's] precise definitions
Peter Patel-Schneider: implementations on the ground are likely to use "JSON number"
Sandro Hawke: JSON-LD steers you away from "number" if you care about round-tripping
Peter Patel-Schneider: JSON has a notion of a fraction number, e.g. 12.3 has a fractional part of "3"
Peter Patel-Schneider: 1.1E1 is the canonical form of "11"^^xsd:double
15:33:08 <ericP> ... that ".1" is the fractional part
... that ".1" is the fractional part
... if that part's nailed down, i think JSON-LD provides a consistent and coherent view of the world
... no idea what the JSON reception would be
Gavin Carothers: http://tools.ietf.org/html/draft-staykov-hu-json-canonical-form-00#section-2.1.1 btw
Markus Lanthaler: Peter's quote: ... "without, of course, getting bogged down on things like how many Unicode surrogate characters can dance on the head of a JSON string" :-)
Peter Patel-Schneider: we talk about unicode codepoints but JSON talks about unicode characters
David Wood: what will the JSON-LD group do with this?
Markus Lanthaler: we spend a long time on round-tripping
... i think the spec is clear anough about it. some corner cases like "1.0" becomes an integer
Peter Patel-Schneider: the point is that the syntax for JSON numbers uses frac and fraction part for the .1 in 1.1E1 so fractional part needs to be distinguished from that
... but the RDF-to-JSON defaults to using the string representation so you default to clean round-tripping
David Wood: so don't expect many changes between now and REC
Arnaud Le Hors: well, I would still be happy to say something about RDF/JSON
David Wood: Zakim, who is here?
Zakim IRC Bot: On the phone I see EricP, gkellogg, pfps, davidwood, TallTed (muted), AZ, Sandro, gavinc, Arnaud, markus, AZ, Souri
Zakim IRC Bot: On IRC I see Guus_, Souri, zwu2, gavinc, markus, AZ, Arnaud, pfps, Zakim, RRSAgent, gkellogg, TallTed, trackbot, davidwood, manu, yvesr, mischat, ericP, sandro
Arnaud Le Hors: +q to say something about RDF/JSON
David Wood: ack Arnaud
Zakim IRC Bot: Arnaud, you wanted to say something about RDF/JSON
David Wood: q?
Peter Patel-Schneider: By the way, when are Semantics and Concepts going to LC publication?
2. RDF/JSON
Arnaud Le Hors: after reading pierre antoine and andy's comments, i don't think they're serious obstacles
Peter Patel-Schneider: +1 (NB)
... before editing the spec, i want to know whether it's likely to proceed
David Wood: gregg proposed an alternative format to JSON-LD, which i don't think this group has time
Arnaud Le Hors: i discussed this with Gregg at SemTech
... he said that they have a rep which is essentially similar to RDf/JSON
David Wood: putting RDF/JSON out as a note makes it easy for folks to translate
... someone can later make that REC-track
Gregg Kellogg: we'd need a internal step which creates an ID map when flattening
Markus Lanthaler: you can have the same shape, but you need a top-level node
PROPOSED: The WG will publish RDF/JSON as a Note
Arnaud Le Hors: +1
David Wood: +1
PROPOSED: The WG will pursue publication of RDF/JSON as a Note
Arnaud Le Hors: +1
Markus Lanthaler: -0.5
Souripriya Das: +1
Gavin Carothers: +1
+1
David Wood: +1
Zhe Wu: +1
Gregg Kellogg: +0.5
Peter Patel-Schneider: +1
Ted Thibodeau: +0
Sandro Hawke: (to be clear -- this is NOT a decision to publish)
Antoine Zimmermann: +1
Sandro Hawke: +0.5
Markus Lanthaler: i find it confusing that the same WG publishes two competing JSON formats
Gavin Carothers: Hey! I'm WRITING 4 competing formats :P
Sandro Hawke: (agreed, there's some confusion with JSON-LD -- but Arnaud has promised the document will be clear about it)
David Wood: yeah, but in the first year, we had lots of discussion
Arnaud Le Hors: I already added this to the RDF/JSON draft: "On the other hand, the RDF Working Group decided to put JSON-LD on the Recommendation track (see resolution of May 30, 2012). If you have no specific reason to use this
document instead of JSON-LD, you are therefore encouraged to use JSON-LD."
... the use cases for these two serializations were completely separate. (that's why i'm not concearned)
RESOLVED: The WG will pursue publication of RDF/JSON as a Note
Arnaud Le Hors: I hope this addresses some of Markus's concern which I fully understand
Zakim IRC Bot: -pfps
3. NTriples and NQuads
David Wood: propose to move from pursuing a Note to pursuing a Rec
PROPOSED: The WG will pursue N-Triples/N-Quads as a Rec instead of a Note
Sandro Hawke: i understand this doesn't bind us, just advice to editor
Gavin Carothers: +∞
Gregg Kellogg: +1
Gavin Carothers: +1
David Wood: +0.5
Ted Thibodeau: +1
Zakim IRC Bot: +[GVoice]
Peter Patel-Schneider: zakim, gvoice is me
Zakim IRC Bot: +pfps; got it
Antoine Zimmermann: +0.5
Eric Prud'hommeaux: Does this raise the bar on what an RDF Implementation is? [ Scribe Assist by Sandro Hawke ]
Eric Prud'hommeaux: Is there a presumption that every RECOMMENDED syntax is supported by every system? [ Scribe Assist by Sandro Hawke ]
Markus Lanthaler: Arnaud, quick question just out of curiosity: does RDF/JSON serialize all numbers as strings?
Eric Prud'hommeaux: Do we have too many syntaxes? [ Scribe Assist by Sandro Hawke ]
Gavin Carothers: Yes, but N-Triples and N-Quads aren't the ones [ Scribe Assist by Sandro Hawke ]
Eric Prud'hommeaux: Yes, but in 2004 N-Triples was specified, but they DIDNT tell the world to expose their data as N-Triples. [ Scribe Assist by Sandro Hawke ]
David Wood: Eric, this is RESOLVED, and Eric please raise an issue about this [ Scribe Assist by Sandro Hawke ]
Markus Lanthaler: +0
Eric Prud'hommeaux: you're saying we have another avenue for guidance about syntaxes, not just "Every Recommendation". [ Scribe Assist by Sandro Hawke ]
David Wood: Now we have all these different syntaxes, so we should make some statement, in Concepts, about what you should comply with. [ Scribe Assist by Sandro Hawke ]
David Wood: I'd say N-Triples and N-Quads and everything else is optional [ Scribe Assist by Sandro Hawke ]
Sandro Hawke: ?!?!?!?!
Sandro Hawke: -1 to that
Sandro Hawke: eric; I'd say w3.org/TR is how you know which syntaxes to implement
David Wood: but practically we have so many [ Scribe Assist by Sandro Hawke ]
Sandro Hawke: LDP says "you have to use Turtle"
... you're saying that every government has to expose as NTriples to guarantee interop"
David Wood: i have to think about it
Gavin Carothers: Simple, name an RDF implementation that DOESN'T implement N-Triples
Zhe Wu: N-TRIPLES is really popular
Souripriya Das: Oracle parses N-Triples and N-Quads
... if we have a half dozen serialization formats, we can't call an impl that doesn't implement them all "non-compliant"
[discussion of proliferation of NTriples and NQuads]
Arnaud Le Hors: the one advantage of having all these formats is that it should make it clear to anyone that RDF isn't RDF/XML :)
Ted Thibodeau: that's actually a significant advantage :-)
ISSUE: guidance to RDF users and developers about which syntaxes to parse and publish
Trackbot IRC Bot: Created ISSUE-138 - Guidance to RDF users and developers about which syntaxes to parse and publish; please complete additional details at <http://www.w3.org/2011/rdf-wg/track/issues/138/edit>.
Antoine Zimmermann: btw OWL/XML is a REC and OWL implementations do *not* have to support it
(Scribe set to Sandro Hawke)
4. TriG
Eric Prud'hommeaux: http://www.w3.org/2005/01/yacker/uploads/trigS_new?lang=perl
Ted Thibodeau: (except for the part where RDF/XML seems to be the only serialization that includes "RDF" [and not just "R"] in its name)
Eric Prud'hommeaux: I created a grammar that takes care of (1) blank nodes as graph identifiers, (2) allow the GRAPH keyword, (3) allow { } to be optional around default graph
Eric Prud'hommeaux: I made an LALR(1) grammar for this
Gavin Carothers: blank nodes as graph identifiers are already in the Grammar in the the current ED
Eric Prud'hommeaux: interesting grammatical points.
Gregg Kellogg: Also LL(1)
Eric Prud'hommeaux: Noticed one could do: [ :a :b ] { .... }
Gregg Kellogg: Also, possibly (1 2) { … }
Eric Prud'hommeaux: Andy pushed back, saying this syntac might be used for something else, and exceeds sparql
Gavin Carothers: Aligning with SPARQL, and then going right past it.....
Eric Prud'hommeaux: ASK { { <s> <p> <o> } GRAPH <x> { ... } }
Eric Prud'hommeaux: Resolved that it's fine to make GRAPH optional and make { ... } optional
Eric Prud'hommeaux: ASK { { <s> <p> <o> { <s2> <p2> <o2> } } GRAPH <x> { ... } }
Sandro Hawke: ewww!
Eric Prud'hommeaux: normal use: ASK { <s> <p> <o> GRAPH <x> { ... } }
Sandro Hawke: and that's what we'd like in TriG
David Wood: (missed)
Eric Prud'hommeaux: { { <s> <p> <o> { <s2> <p2> <o2> } } GRAPH <x> { ... } } turns into { <s> <p> <o>. <s2> <p2> <o2> GRAPH <x> { ... } }
Eric Prud'hommeaux: Resolved that it's fine to make GRAPH optional and make { ... } optional
Eric Prud'hommeaux: is there a mechanism to survey....?
Eric Prud'hommeaux: eg wiki page of implementations
Sandro Hawke: Sure, let's link to the implementation risk in the At Risk text.
Zakim IRC Bot: -Arnaud
Gavin Carothers: I think we already resolved to add these, last meeting. I was going to normalize against existing trig grammar and include it.
Eric Prud'hommeaux: also, optional trailing dot inside curlies
Eric Prud'hommeaux: I borrow from SPARQL not TURTLE.
+1
David Wood: We've gotten through our agenda!
Sandro Hawke: I don't recall us actually resolving 137
http://lists.w3.org/Archives/Public/public-rdf-wg/2013Jul/0138.html
PROPOSED: Resolve ISSUE-137 using at risk text proposed by sandro http://lists.w3.org/Archives/Public/public-rdf-wg/2013Jul/0138.html and a grammar based on ericP's changes http://lists.w3.org/Archives/Public/public-rdf-wg/2013Jul/
0099.html
+1
David Wood: +1
Gavin Carothers: +0.5
Gregg Kellogg: +1
Eric Prud'hommeaux: +1
Ted Thibodeau: +1
Zhe Wu: +1
Souripriya Das: +1
RESOLVED: Resolve ISSUE-137 using at risk text proposed by sandro http://lists.w3.org/Archives/Public/public-rdf-wg/2013Jul/0138.html and a grammar based on ericP's changes http://lists.w3.org/Archives/Public/public-rdf-wg/2013Jul/
0099.html
David Wood: AOB?
Markus Lanthaler: +1
Gavin Carothers: We'll also be adding Andy's test cases to the TriG test suite.
Zhe Wu: bye
Zakim IRC Bot: -AZ
ADJOURN
Zakim IRC Bot: -AZ
Zakim IRC Bot: -pfps
Zakim IRC Bot: -Souri
Markus Lanthaler: bye
Zakim IRC Bot: -markus
https://dvcs.w3.org/hg/rdf/file/0f2466565bc8/trig/tests2/trig-kw-graph-08.trig
https://dvcs.w3.org/hg/rdf/file/0f2466565bc8/trig/tests2
Zakim IRC Bot: -EricP
(No events recorded for 5 minutes)
https://dvcs.w3.org/hg/rdf/file/0f2466565bc8/trig/tests2/trig-kw-graph-01.trig
Gregg Kellogg: https://dvcs.w3.org/hg/rdf/log/0f2466565bc8/trig/tests2/trig-kw-graph-08.trig
Zakim IRC Bot: -gkellogg
Zakim IRC Bot: -Sandro
Zakim IRC Bot: -gavinc
Zakim IRC Bot: -davidwood
Zakim IRC Bot: -TallTed
Zakim IRC Bot: SW_RDFWG()11:00AM has ended
Zakim IRC Bot: Attendees were EricP, gkellogg, davidwood, TallTed, +081165aaaa, AZ, Sandro, +1.707.861.aabb, gavinc, Arnaud, markus, Souri, pfps
--
-ericP
Received on Saturday, 13 July 2013 03:28:52 UTC