IRC log of rdf-wg on 2011-12-14

Timestamps are in UTC.

15:55:20 [RRSAgent]
RRSAgent has joined #rdf-wg
15:55:20 [RRSAgent]
logging to
15:55:22 [trackbot]
RRSAgent, make logs world
15:55:22 [Zakim]
Zakim has joined #rdf-wg
15:55:24 [trackbot]
Zakim, this will be 73394
15:55:24 [Zakim]
ok, trackbot; I see SW_RDFWG()11:00AM scheduled to start in 5 minutes
15:55:25 [trackbot]
Meeting: RDF Working Group Teleconference
15:55:25 [trackbot]
Date: 14 December 2011
15:56:13 [Scott_Bauer]
Scott_Bauer has joined #rdf-wg
15:57:04 [Zakim]
SW_RDFWG()11:00AM has now started
15:57:11 [Zakim]
15:58:36 [Zakim]
15:58:38 [Zakim]
15:58:39 [Zakim]
15:58:52 [AndyS]
zakim, IPCaller is me
15:58:52 [Zakim]
+AndyS; got it
15:59:04 [Guus]
[to Andy: I hope you can scribe, as I haven't seen Olivier for some time]
15:59:38 [Zakim]
15:59:40 [Arnaud1]
Arnaud1 has joined #rdf-wg
15:59:52 [Zakim]
16:00:14 [CGI585]
CGI585 has joined #rdf-wg
16:00:31 [Zakim]
16:00:50 [Zakim]
16:00:54 [AZ]
AZ has joined #rdf-wg
16:01:10 [CGI585]
zakim, CGI585 is me
16:01:11 [Scott_Bauer]
Zakim, BobF is me
16:01:16 [Zakim]
sorry, CGI585, I do not recognize a party named 'CGI585'
16:01:20 [Zakim]
16:01:22 [sandro]
zakim, who is talking?
16:01:24 [Zakim]
+Scott_Bauer; got it
16:01:32 [Zakim]
sandro, listening for 10 seconds I heard sound from the following: Guus (36%), AndyS (59%)
16:01:38 [Zakim]
16:01:40 [MacTed]
Zakim, who's here?
16:01:46 [MacTed]
Zakim, OpenLink_Software is temporarily me
16:01:46 [Zakim]
On the phone I see Guus, AndyS, sandro, charlesgreer, Arnaud_LeHors, Scott_Bauer, LeeF, OpenLink_Software
16:01:48 [MacTed]
Zakim, mute me
16:01:51 [AndyS]
scribe: Andy
16:01:54 [Zakim]
On IRC I see AZ, CGI585, Arnaud1, Scott_Bauer, Zakim, RRSAgent, FabGandon, Guus, LeeF, MacTed, mischat, SteveH, AndyS, mdmdm, davidwood, danbri, manu1, trackbot, NickH, manu,
16:01:59 [AndyS]
scribenick: AndyS
16:02:01 [Zakim]
... sandro, ericP
16:02:02 [CGI585]
zakim, charlesgreer is me
16:02:02 [cygri]
cygri has joined #rdf-wg
16:02:05 [Zakim]
+MacTed; got it
16:02:09 [Zakim]
MacTed should now be muted
16:02:13 [Zakim]
16:02:19 [Zakim]
16:02:21 [Zakim]
+CGI585; got it
16:02:25 [davidwood]
zakim, BernHyland is me
16:02:31 [Guus]
zakim, who is here?
16:02:48 [Zakim]
+davidwood; got it
16:02:59 [Zakim]
On the phone I see Guus, AndyS, sandro, CGI585, Arnaud_LeHors, Scott_Bauer, LeeF, MacTed (muted), davidwood, cygri
16:03:08 [gavinc]
gavinc has joined #rdf-wg
16:03:15 [Zakim]
+ +
16:03:22 [Zakim]
16:03:27 [Zakim]
16:03:35 [Zakim]
On IRC I see cygri, AZ, CGI585, Arnaud1, Scott_Bauer, Zakim, RRSAgent, FabGandon, Guus, LeeF, MacTed, mischat, SteveH, AndyS, mdmdm, davidwood, danbri, manu1, trackbot, NickH,
16:03:35 [AZ]
zakim, + is me
16:03:37 [CGI585]
zakim, CGI585 is me
16:03:42 [Zakim]
... manu, sandro, ericP
16:03:42 [FabGandon]
Zakim, + is me
16:03:43 [MacTed]
Zakim, unmute me
16:04:01 [MacTed]
Zakim, mute me
16:04:12 [Zakim]
+AZ; got it
16:04:14 [Zakim]
+CGI585; got it
16:04:18 [Zakim]
sorry, FabGandon, I do not recognize a party named '+'
16:04:20 [Zakim]
MacTed should no longer be muted
16:04:20 [MacTed]
Zakim, CGI585 is charlesgreer
16:04:28 [Zakim]
MacTed should now be muted
16:04:28 [FabGandon]
Zakim, aaaa is me
16:04:35 [Zakim]
16:04:36 [Zakim]
+charlesgreer; got it
16:04:42 [Zakim]
sorry, FabGandon, I do not recognize a party named 'aaaa'
16:04:49 [pfps]
pfps has joined #rdf-wg
16:04:53 [sandro]
16:04:56 [AndyS]
Guus: welcome
16:05:28 [PatH]
PatH has joined #rdf-wg
16:05:41 [JeremyCarroll]
JeremyCarroll has joined #rdf-wg
16:05:42 [FabGandon]
zakim, who is on the phone?
16:05:42 [Zakim]
On the phone I see Guus, AndyS, sandro, charlesgreer, Arnaud_LeHors, Scott_Bauer, LeeF, MacTed (muted), davidwood, cygri, AZ, Peter_Patel-Schneider, [Sophia], gavinc
16:05:43 [pfps]
minutes look OK to me
16:05:55 [FabGandon]
zakim, [Sophia] is me
16:05:55 [Zakim]
+FabGandon; got it
16:06:20 [AndyS]
Accept minutes of telecon of 2011-11-30
16:06:28 [Zakim]
+ +1.443.212.aabb
16:06:38 [AndyS]
Topic: review action items
16:06:45 [AlexHall]
AlexHall has joined #rdf-wg
16:06:51 [AndyS]
(action 119 is done as well)
16:07:20 [AndyS]
Actions for review - accepted and closed
16:07:27 [AndyS]
CLOSE action-122
16:07:28 [trackbot]
ACTION-122 Add list of XSD types to Concepts (assuming no objection from PatH or pfps) closed
16:07:31 [AndyS]
CLOSE action-123
16:07:31 [trackbot]
ACTION-123 Summarize rdf:XMLLiteral options on the list closed
16:07:48 [Zakim]
16:08:13 [Zakim]
16:08:34 [swh]
swh has joined #rdf-wg
16:08:47 [Zakim]
16:09:00 [swh]
Zakim, ??P39 is me
16:09:00 [Zakim]
+swh; got it
16:09:08 [PatH]
zakim, mute me
16:09:08 [Zakim]
PatH should now be muted
16:10:32 [PatH]
zakim, unmute me
16:10:32 [Zakim]
PatH should no longer be muted
16:11:23 [PatH]
zakim, mute me
16:11:23 [Zakim]
PatH should now be muted
16:12:49 [zwu2]
zwu2 has joined #rdf-wg
16:12:50 [PatH]
zakim, unmute me
16:12:50 [Zakim]
PatH should no longer be muted
16:13:30 [PatH]
zakim, mute me
16:13:30 [Zakim]
PatH should now be muted
16:13:40 [PatH]
will do.
16:15:18 [Souri]
Souri has joined #rdf-wg
16:15:27 [pfps]
good schedule
16:15:33 [Zakim]
16:15:45 [AndyS]
next telecons - 21/Dec; skip 28th and then 4/Jan
16:15:48 [PatH]
16:15:57 [AndyS]
Sandro: at risk for 21/Dec
16:16:04 [swh]
regrets for 21st
16:16:08 [AndyS]
... but go ahead
16:16:17 [Arnaud]
I don't think I'll be able to make it next week either
16:16:21 [Souri]
regrets for 21-Dec and 28-Dec
16:16:55 [sandro]
16:17:14 [sandro]
q+ to suggest a little discussion of LDEP workshop
16:18:08 [AndyS]
Topic: Named Graphs
16:18:19 [sandro]
16:18:37 [sandro]
16:18:55 [sandro]
andy: graph literals in RDF graph, like N3
16:19:30 [sandro]
andy: this isn't how they are used much, in the wild. and graph literals are quite big.
16:20:18 [sandro]
andy: case 2 is what I though NGs were originally about. like the first one, but it's talking about the literals.
16:20:46 [PatH]
zakim, unmute me
16:20:46 [Zakim]
PatH should no longer be muted
16:20:52 [sandro]
andy: the IRI :graphLiteral1 is naming the literal. maybe :IRI_for_Graph_Literal_1
16:21:04 [sandro]
pat: literal or value?
16:21:14 [sandro]
andy: Really IRI of value, I think.
16:21:35 [PatH]
zakim, mute me
16:21:36 [Zakim]
PatH should now be muted
16:21:52 [sandro]
andy: Case 3a. Putting the Web in, referencing things, doing gets on them....
16:22:14 [sandro]
... Using graph value, then talking about the GET operation that caused you to get it.
16:22:49 [sandro]
... there's a URI, but it's different from where you go the data
16:23:22 [sandro]
... Case 3b, Sandro's Rolling Snapshot, similar, but it's the PUBLISHER who does it.
16:24:08 [sandro]
sandro: not necessarily the original publisher, but yes, the snapshot is a new gbox on the web.
16:24:22 [sandro]
andy: Case 4: my understanidng of jjc's web cache example
16:24:43 [sandro]
... lossy, you've forgotten when you looked at the gbox. A commonly used pattern.
16:24:57 [sandro]
... Case 5 -- primary Topic.
16:25:01 [PatH]
hi danbri. Andy is reviewing
16:25:37 [gavinc]
... can't case 4 become case 3b if you care about the details that get "lost" ?
16:26:01 [Guus]
16:26:04 [Guus]
16:26:11 [sandro]
sandro: CambridgeSemantics has a pattern of bundling data in graphs by a common subject
16:26:13 [gavinc]
16:26:20 [sandro]
(aka Case 5, sort of)
16:26:21 [Guus]
ack Guus
16:26:28 [Guus]
ack gavinc
16:26:29 [LeeF]
That's right, Andy's case 5 is pretty similar to how our tools often work by default
16:26:32 [Zakim]
+ +44.117.230.aacc
16:26:38 [sandro]
Gavin: combing 3b and case 4 -- that works pretty well
16:27:07 [sandro]
gavin: If you apply 3b to 4, you're not losing info.
16:27:21 [sandro]
andy: 3b assumes some other machinery.
16:27:23 [sandro]
16:27:29 [sandro]
16:27:45 [Guus]
ack sandro
16:28:23 [AndyS]
Sandro: insight: I argued that case 5 was bad practice -- I now prefer "RESTful" description.
16:28:27 [gavinc]
sandro, case 5 -can- be restful.
16:28:33 [LeeF]
our case-5-quasi way CAN be used with the URIs dereferenceable to get the graph
16:28:37 [Zakim]
16:28:41 [gavinc]
+1 LeeF
16:29:06 [danbri]
q+ to sandro
16:29:23 [LeeF]
the "naughty" bits of case 5 - as best i can tell - is that you are typing things as both :Graphs and :SomethingElse, when many people think those ought to be disjoint
16:29:28 [pfps]
I worry about the rather underspecified REST forming an underpinning (of anything).
16:29:52 [PatH]
for the record, I am completely confused by Andy's message and his explanation.
16:30:09 [AndyS]
danbri: RESTful sometimes one primary version. Q: what about multiple representations?
16:30:49 [danbri]
danbri: when I hear 'restful' I worry a bit, as people often conflate in a notion that the resource has essentially one primary representation (maybe + content neg) and don't also expect other complexities. SO for example, authenticated access to a personalised resource can still be restful.
16:31:19 [danbri]
...-> which means in practice, for any 'restful' url, you still need to keep track of the state of the deref transaction (user, http headers, date/time etc)
16:31:32 [PatH]
does the machinery change the meanings of anything?
16:31:54 [pfps]
In particular for RDF, how do we reconcile the difference between the graph-centered RDF and object-centered REST and also open/closed difference and single-/multi-valued difference.
16:32:06 [sandro]
16:32:10 [danbri]
16:32:34 [PatH]
zakim, unmute me
16:32:34 [Zakim]
PatH should no longer be muted
16:33:42 [sandro]
andy: in 3b there isn't a trig document
16:33:48 [pfps]
+1 to PatH
16:33:57 [sandro]
PatH: I can't make semantic sense of this, and will work on it offline.
16:34:43 [AndyS]
Guus: possible strawpoll ; and discussions that have arisen
16:34:43 [PatH]
zakim, mute me
16:34:43 [Zakim]
PatH should now be muted
16:35:39 [sandro]
cygri: Let's resolve the entire GRAPHS discussion by adopting RDF Dataset as in RDF Concepts. And do nothing on the semantics.
16:35:47 [PatH]
zakim, unmute me.
16:35:47 [Zakim]
PatH should no longer be muted
16:35:52 [PatH]
16:36:08 [sandro]
cygri: Some issues about Default Graph, "no name", MAY have name, something. Datamodel from SPARQL, and nothing in Semantics.
16:36:16 [gavinc]
16:36:19 [sandro]
guus: Graph name has to be an IRI?
16:36:23 [sandro]
cygri: Yes.
16:36:24 [AndyS]
q+ to ask that TBL emails get to the list
16:36:26 [Arnaud]
16:37:04 [cygri]
PROPOSAL: Close all graph model+semantics issues by accepting the RDF Datasets design [1] as the data model, and by adding no new semantics.
16:37:08 [sandro]
AndyS: SPARQL docs currently say graph label is IRI, although machinery doesnt really care.
16:37:36 [Guus]
ack PatH
16:37:36 [sandro]
PatH: In my email response to Richard, is that it involves Named Graphs. That seems to modify the semantics.
16:37:51 [AndyS]
PatH: puzzled because it talks about named graph.
16:37:52 [davidwood]
Abstract Syntax for Working with Multiple Graphs:
16:38:00 [sandro]
cygri: Leave the Semantics as it is, leave working in RDF Concepts ED, and close all related open issue.
16:38:09 [AndyS]
cygri: close issue with the current concept doc.
16:38:10 [JeremyCarroll]
16:38:37 [sandro]
PatH: So the Concepts Document would refer to named graphs, and Semantic would be silent on it, leacing a muddle.
16:38:42 [PatH]
zakim, mute me
16:38:42 [Zakim]
PatH should now be muted
16:38:48 [sandro]
cygri: Yes. But who would be opposed to that.
16:39:14 [danbri]
(can't we just make the list writeable to anyone who can post to w3c lists?)
16:39:48 [sandro]
Guys forward TimBLs email to list.
16:40:06 [sandro]
JeremyCarroll: Named Graphs paper suggested a pragmatic treatment of named graphs.
16:40:23 [sandro]
... because we couldnt change the formal semantics
16:40:27 [sandro]
16:40:36 [sandro]
ack JeremyCarroll
16:40:42 [gavinc]
ack AndyS
16:40:42 [Zakim]
AndyS, you wanted to ask that TBL emails get to the list
16:40:50 [PatH]
actually that paper did give a formal semantics. its not hard to do.
16:41:15 [PatH]
masybe jeremy didnt notice it :-)
16:41:20 [JeremyCarroll]
16:41:22 [JeremyCarroll]
16:41:23 [sandro]
STRAWPOLL: Let's close all the GRAPHS issues, with just the currenty text in RDF Concepts
16:41:34 [AZ]
16:41:37 [sandro]
-1 I think we need to try harder, for a while, sadly
16:41:38 [PatH]
16:41:39 [LeeF]
16:41:40 [Arnaud]
16:41:43 [cygri]
16:41:45 [MacTed]
16:41:46 [pfps]
Richard's proposal has the consequence that RDF datasets are not handled in the RDF semantics - which is solely concerned with how RDF graphs work.
16:41:46 [AndyS]
+0.5 (want to have dataset patterns)
16:41:47 [pfps]
16:41:50 [charlesgreer]
16:41:51 [AlexHall]
16:41:52 [gavinc]
16:41:52 [zwu2]
+1 (starts with)
16:41:56 [davidwood]
-1, I'd love to end this, but won't in the face of Pat's -1
16:41:57 [AZ]
16:42:10 [swh]
16:42:14 [Souri]
16:42:26 [pfps]
I strongly agree with Sandro, that there needs to be a proposal to move forward.
16:42:29 [swh]
maybe +0.5
16:42:35 [pfps]
... if there is any forward to move.
16:42:39 [pfps]
16:42:49 [AndyS]
q+ to ask about starting points
16:42:59 [AZ]
q+ to say that there is a formal semantic in the current proposal
16:43:05 [davidwood]
The same poll could yield different results in 12 months :)
16:43:07 [PatH]
i dont feel we have rally started on this yet. I will try to get a prposal by next week.
16:43:08 [AndyS]
Guus: minimal needed to add?
16:43:25 [Guus]
ack AZ
16:43:25 [Zakim]
AZ, you wanted to say that there is a formal semantic in the current proposal
16:43:26 [PatH]
16:43:56 [AndyS]
AZ: there is a semantic proposal already
16:44:02 [sandro]
az: In the proposal about datasets, there is already a formal semantics. But it has never been discussed in telecons. It's been there since March. It's minimals. If you have several graphs, then you have several RDF interpretations. What follows from a dataset is what entails from each graph in it.
16:44:51 [PatH]
Im afraid I dont think that idea makes sense. But we can discuss it in email.
16:44:57 [pfps]
Can we have a pointer to this semantics?
16:45:01 [AndyS]
From memory - it seemed OK.
16:45:18 [sandro]
zakim, who is talking?
16:45:29 [Zakim]
sandro, listening for 10 seconds I heard sound from the following: Guus (40%), AndyS (4%), davidwood (4%), AZ (15%), Peter_Patel-Schneider (37%)
16:45:50 [sandro]
pfps: Pat brought up negative consequences that resonated with people. What?
16:46:11 [PatH]
zakim, unmute me
16:46:11 [Zakim]
PatH should no longer be muted
16:46:29 [sandro]
davidwood: Pat said disconnect between RDF Concepts and RDF Semantics.
16:46:30 [sandro]
16:46:38 [PatH]
16:46:42 [AndyS]
ack pfpf
16:46:49 [AndyS]
ack pfps
16:46:55 [Guus]
ack pfps
16:47:02 [gavinc]
Enough so that someone coming to RDF has some CLUE what's going on with named graphs. Right now or with the minimal statement you have to go and read all sorts of other documents to figure out what Named Graphs are used for. And then you find disagreement in those other documents, and then you end up editing TriG
16:47:21 [sandro]
pat: There are many entailments up for grabs. Which ones will be in Semantics?
16:47:36 [sandro]
Guus: If this is only about gboxes, then there should be no change to semantics.
16:47:44 [cygri]
16:48:04 [JeremyCarroll]
16:48:12 [sandro]
pat: When you say a URI names something, that statements has a lot of Semantic Weight. As we've demonstrated, there's a lot of potential misunderstanding over that usage.
16:48:26 [sandro]
[ We need test cases! ]
16:48:28 [cygri]
""^^xsd:anyURI is an IRI and doesn't denote anything in the formal semantics
16:48:39 [AndyS]
ack AndyS
16:48:39 [Zakim]
AndyS, you wanted to ask about starting points
16:48:49 [PatH]
16:49:13 [sandro]
AndyS: Starting points? Are we targetting a design that does not invalidate what people are already doing? Or are we designing a clean sheet solution?
16:49:13 [pfps]
Why not just say that a named graph is a pair consisting of a URI and an RDF graph?
16:49:15 [JeremyCarroll]
@cygri The formal semantics gives the value of "" to that on e
16:49:18 [JeremyCarroll]
it is specified
16:49:26 [sandro]
guus: Charter says don't invalidate existing deployments.
16:49:33 [sandro]
16:50:11 [sandro]
someone: Different people are using the semantics differently
16:50:12 [JeremyCarroll]
q+ to give test case
16:50:30 [Guus]
ack sandro
16:50:45 [AndyS]
Andy: current uses look like several different semantics / patterns of usage
16:50:49 [sandro]
Brad Allen of Elsevier: Named Graph == A document containing RDF that has a presence on the Web, as a container for that graph.
16:51:01 [PatH]
a g-box
16:51:09 [LeeF]
thats not what i think a named graph is
16:51:18 [LeeF]
tag:mygraph { ... some triples ... }
16:51:19 [LeeF]
is a named graph
16:51:21 [LeeF]
not on the Web
16:51:26 [danbri]
I disagree
16:51:47 [davidwood]
Are we going to argue now about what a "document" is? I suspect that "document" has now been used in at least as many ways as "graph".
16:51:48 [danbri]
Lots of RDF people understand it as a kind of muddle between SPARQL datasets, and 'where/how I got it...'
16:51:52 [sandro]
a gbox on the web.
16:52:05 [Guus]
16:52:22 [danbri]
( re what-is-a-doc, )
16:52:30 [davidwood]
A "document" cannot equal a g-box in my world. A g-snap, surely.
16:52:30 [PatH]
clearly the world needs restful-named g-boxes.
16:52:37 [AndyS]
Seems very like FROM NAMED in SPARQL -- a usage of the SPARQL RDF datasets
16:52:41 [sandro]
AndyS: I think that's what Dan Connolly had in mind for SPARQL datasets.
16:52:45 [PatH]
I suggest it also needs named graphs.
16:53:22 [PatH]
so we need to distinguish them.
16:53:33 [AndyS]
cygri: world expects that case, we can give the world something like that (maybe more)
16:53:47 [sandro]
cygri:Sandro, Datasets are a step in the right direction to this, "what the world wants"
16:54:43 [AndyS]
cygri: non-denoting in "http://example/resource"^^xsd:anyURI
16:55:02 [LeeF]
BTW, i'd be prety happy with what pfps wrote above: named graph is a pair of (URI, RDF graph) -- since as I understand things, that's exactly how I've always used them
16:55:05 [AndyS]
paht: if we say "uri names X" then we need to say what "names" means
16:55:13 [sandro]
pat: If our spec says "named graphs", we need to be specific about what we mean by that phrase.
16:55:19 [gavinc]
"A graph statement pairs an IRI with a RDF Graph. It is intended that triple statements made using that IRI are being made about the graph."
16:55:31 [danbri]
PatH, don't just say consequences for entailments in RDF; say 'for what people do with RDF data...'
16:55:42 [danbri]
(for those whose eyes glaze over at mention of fancy logic stuff)
16:55:47 [sandro]
q+ to ask if Pat can write down some test cases
16:55:57 [AndyS]
+1 to LeeF -- it's the minimal framework (but the looseness can lead to corner cases later)
16:55:57 [LeeF]
yeah test cases!
16:56:26 [AndyS]
cygri: It's a label.
16:57:16 [sandro]
+1 JJC
16:57:30 [AndyS]
Jeremy: can say different things about same URI.
16:57:41 [AndyS]
q+ to say "patterns"
16:57:50 [sandro]
+1 JJC we need simple test cases to show things like an inconsistency when people dsay different things about the same graph.
16:57:50 [pfps]
But why is it the job of RDF to do what Jeremy wants? For example, RDF doesn't make a=b and a/=b a contradiction.
16:58:12 [LeeF]
16:58:16 [LeeF]
+1 even
16:58:27 [AndyS]
"patterns" are an external semantics -- in-doc semantics mean too many people have to change
16:58:28 [MacTed]
Zakim, who's noisy?
16:58:30 [sandro]
zakim, who is talking?
16:58:39 [Zakim]
MacTed, listening for 10 seconds I heard sound from the following: Guus (67%), Scott_Bauer (51%), cygri (5%), JeremyCarroll (76%)
16:58:48 [Zakim]
16:58:48 [Guus]
16:58:48 [sandro]
16:58:55 [PatH]
pfps, graphs are central to RDF, and no other authority is going to make it clear for us.
16:59:01 [Zakim]
sandro, listening for 10 seconds I heard sound from the following: Guus (57%), AZ (4%), JeremyCarroll (16%)
16:59:15 [cygri]
ack me
16:59:15 [sandro]
ack sandro
16:59:17 [Guus]
ack cygri
16:59:27 [PatH]
zakim, mute me
16:59:39 [MacTed]
ack JeremyCarroll
16:59:39 [Zakim]
sandro, you wanted to ask if Pat can write down some test cases
16:59:42 [Guus]
ack JeremyCarroll
16:59:57 [Zakim]
PatH should now be muted
16:59:58 [PatH]
will do.
17:00:05 [PatH]
in email soon.
17:00:12 [MacTed]
17:00:14 [Zakim]
JeremyCarroll, you wanted to give test case
17:00:15 [MacTed]
Zakim, unmute me
17:00:29 [sandro]
17:01:11 [sandro]
17:01:12 [Zakim]
MacTed should no longer be muted
17:01:36 [sandro]
andy: Patterns, an attempt to overlay semantics on a TriG document.
17:01:57 [sandro]
andy: and then label documents as using particular patterns.
17:01:59 [PatH]
test cases can also be petterns :-)
17:02:12 [PatH]
17:02:37 [gavinc]
q+ to complain about not being able to work on TriG
17:03:43 [PatH]
zakim, unmute me
17:03:47 [AndyS]
ack me
17:03:47 [PatH]
17:04:01 [sandro]
q+ to suggest what a test case might look like (Entailments/Consistency of comibinations of TriG and RDF Graphs)
17:04:04 [Zakim]
PatH should no longer be muted
17:04:12 [Zakim]
AndyS, you wanted to say "patterns"
17:04:41 [davidwood]
+10 to MacTed
17:04:45 [danbri]
17:04:59 [PatH]
is that trerminology written down? pointer?
17:05:18 [davidwood]
17:05:20 [davidwood]
^ Pat
17:05:23 [PatH]
17:05:43 [sandro]
MacTed: We need to stick with g-box, etc, terms instead of the way we keep using confusing terms.
17:05:48 [pfps]
I'm trying to figure out what overloading is happening with respect to "named graphs".
17:06:11 [PatH]
my bad for using 'g-box'.
17:06:26 [sandro]
pfps, is a named graph a pair of (name, graph) as in the RDF Concepts, or is it "A document containing RDF that has a presence on the Web, as a container for that graph".
17:06:41 [PatH]
i wasnt at the f2f :(
17:06:59 [PatH]
17:07:01 [sandro]
17:07:08 [PatH]
17:07:09 [MacTed]
Zakim, mute me
17:07:09 [Zakim]
MacTed should now be muted
17:07:16 [Guus]
ack MacTed
17:07:24 [davidwood]
<graph container> owl:sameAs "g-box", but the range of sameAs can't be a literal so we end up having the same conversation about denotation...
17:07:39 [PatH]
17:07:59 [sandro]
gavin: First draft of TriG was posted a week and ahalf after F2F, and I've asked at every WG meeting since then about some language in it, and I keep getting different answers.
17:08:10 [Guus]
ack gavinc
17:08:10 [Zakim]
gavinc, you wanted to complain about not being able to work on TriG
17:08:23 [PatH]
we need a group assignment to read and follow a bunch of documents.
17:08:47 [cygri]
gavinc, you can't standardize the serialization when the model is still in flux. no surprise there
17:08:51 [AndyS]
We are quoting from outside WG so safer to use g-* to understand those views.
17:08:51 [Guus]
ack sandro
17:08:51 [Zakim]
sandro, you wanted to suggest what a test case might look like (Entailments/Consistency of comibinations of TriG and RDF Graphs)
17:08:53 [sandro]
TriG document + RDF Graphs ===> Entailments / contradictions
17:09:26 [pfps]
actually I don't see any problem with writing the serialization spec, it is just the "surround" that is difficult
17:09:35 [gavinc]
"A graph statement pairs an IRI with a RDF Graph. It is intended that triple statements made using that IRI are being made about the graph."
17:09:42 [pfps]
... without any consensus on how named graphs are supposed to work
17:09:54 [PatH]
we might need a new format for webbish test cases involving containers.
17:10:10 [AndyS]
sandro: IRI labels the graph
17:10:15 [PatH]
+1 pfps
17:10:26 [Zakim]
17:10:27 [AndyS]
... workign out what "labels" can be
17:10:29 [Zakim]
17:11:41 [LeeF]
17:12:06 [AndyS]
pfps: parser/grammar outputs a dataset.
17:13:04 [sandro]
STRAWPOLL: We'll use TriG as our multigraph syntax, given we agree on some little details. TriG is a syntax for serializaing an RDF Dataset.
17:13:04 [cygri]
17:13:12 [MacTed]
TriG as opposed to ... N-Quads? something(s) else?
17:13:13 [davidwood]
17:13:14 [LeeF]
17:13:15 [cygri]
17:13:16 [pfps]
17:13:17 [zwu2]
17:13:20 [PatH]
17:13:24 [sandro]
+0 not sure
17:13:26 [swh]
+1 (but only internally, not published)
17:13:36 [AlexHall]
17:13:41 [danbri]
17:14:00 [sandro]
cygri, prefers in n-quads
17:14:10 [AndyS]
+1 TriG and N-quads (c.f. Turtle and N-triples)
17:14:13 [AZ]
17:14:15 [MacTed]
+0 I generally prefer not to implicitly bless anything, when it's possible to present multiple (so I'm likewise against presenting *only* in N-Quads, if both are viable)
17:14:29 [LeeF]
I'm +1 for N-quads too
17:14:38 [LeeF]
since we're now discussing that :)
17:14:43 [PatH]
I have to leave very soon.
17:14:46 [AndyS]
Not only N-Quads - unreadable at scale (scale = 2+ graphs)
17:15:18 [PatH]
im feeling pentagonal.
17:16:00 [sandro]
np, Andy
17:16:29 [sandro]
guus: Thanks everyone.
17:16:32 [zwu2]
17:16:38 [PatH]
17:16:43 [sandro]
... sounds like we need to pay more attention to our graphs terminology
17:16:57 [sandro]
... and work on what needs to be added for Graphs.
17:17:11 [Zakim]
17:17:12 [Zakim]
17:17:12 [Zakim]
17:17:13 [Zakim]
17:17:13 [Zakim]
17:17:13 [Zakim]
17:17:15 [Zakim]
17:17:17 [Zakim]
17:17:19 [Zakim]
17:17:21 [Zakim]
17:17:23 [Zakim]
17:17:24 [FabGandon]
FabGandon has left #rdf-wg
17:17:26 [Zakim]
- +1.443.212.aabb
17:17:27 [Zakim]
17:17:31 [AlexHall]
AlexHall has left #rdf-wg
17:17:34 [Zakim]
17:17:42 [Zakim]
17:17:45 [Zakim]
17:17:48 [Zakim]
17:17:57 [sandro]
RRSAgent, pointer?
17:17:57 [RRSAgent]
17:18:16 [Zakim]
17:19:22 [Scott_Bauer]
Scott_Bauer has left #rdf-wg
17:21:01 [Zakim]
17:27:28 [Zakim]
17:27:29 [Zakim]
SW_RDFWG()11:00AM has ended
17:27:30 [Zakim]
Attendees were Guus, AndyS, sandro, Arnaud_LeHors, LeeF, Scott_Bauer, MacTed, cygri, davidwood, Peter_Patel-Schneider, AZ, gavinc, charlesgreer, FabGandon, +1.443.212.aabb, zwu2,
17:27:34 [Zakim]
... PatH, swh, Souri, danbri, JeremyCarroll, Arnaud
18:09:35 [Arnaud]
Arnaud has joined #rdf-wg
18:20:14 [Arnaud]
Arnaud has left #rdf-wg
19:19:39 [Zakim]
Zakim has left #rdf-wg
21:41:20 [davidwood1]
davidwood1 has joined #rdf-wg
21:56:24 [mischat]
mischat has joined #rdf-wg
21:58:09 [davidwood]
davidwood has joined #rdf-wg
21:58:53 [davidwood]
davidwood has joined #rdf-wg
21:59:43 [davidwood]
davidwood has joined #rdf-wg
22:36:26 [davidwood]
davidwood has joined #rdf-wg
22:55:00 [trackbot]
trackbot has joined #rdf-wg
23:00:01 [trackbot]
trackbot has joined #rdf-wg