scribenick pfps

<gavinc> trackbot, start meeting

<trackbot> Date: 28 September 2011

<scribe> scribenick: pfps

<scribe> scribenick: pfps

<Arnaud> fyi, I will have to drop at the hour

Admin

<PatH> I will have to drop out in 30 mins for a short while.

PROPOSED: accept minutes of 21 Sep http://www.w3.org/2011/rdf-wg/meeting/2011-09-21

<PatH> +1

RESOLUTION: accept minutes of 21 Sep http://www.w3.org/2011/rdf-wg/meeting/2011-09-21

guus: action items - pending

<yvesr_> done :)

guus: 77 done, 89 [DONE], 90 done, 91 done
... open actions
... action 82 - Sandro?

sandro: no action yet - will try this week

guus: action 86 - Sandro?

<sandro> action-86?

<trackbot> ACTION-86 -- Sandro Hawke to provide example for how code is simpler with language-tag options 3 and 4 vs 1 and 2. -- due 2011-09-14 -- OPEN

<trackbot> http://www.w3.org/2011/rdf-wg/track/actions/86

sandro: 86 is either done or obsolete

guus: everything is to continue

<sandro> it's both!

F2F planning

guus: anything to be done

<sandro> is it disjunction and/or conjunction? :-)

davidwood: note that times have changed (and are now aligned)

<ericP> sandro, do you know the foo to make sure the Stata doors are unlocked?

<ericP> if not, i'll dig into it

<sandro> please do, eric.

guus: agenda for F2F2 is on the wiki, need to make sure that some open time remains
... no need to change more right now

Turtle in HTML

guus: see message from Gavin

gavinc: document needs to be brought up before HTML data task force

<scribe> ACTION: gavinc to liase with HTML data task force to bring up Turtle in HTML document, etc. [recorded in http://www.w3.org/2011/09/28-rdf-wg-minutes.html#action01]

<trackbot> Created ACTION-92 - Liase with HTML data task force to bring up Turtle in HTML document, etc. [on Gavin Carothers - due 2011-10-05].

Liaison with Provenance WG

guus: trust use case from Sandro http://lists.w3.org/Archives/Public/public-rdf-prov/2011Sep/0003.html
... this appears to be a good, driving example

sandro: I can't speak for the PWG, even though I tried to provoke discussion of it there
... Richard had some comments, notably a set of actions

guus: how about RDF to show example(s)?

<PatH> this example seems very ambitious to me. ZDoes the Prov. grp consider it in their scope?

sandro: one issue is when you need a "snapshot", e.g., "That's not right", but then the "that" is fixed
... Richard's proposal does not appear to speak to this

guus: can we come up with some worked examples, showing potential RDF content?

sandro: I'll try to work this into the conversation with Richard

<sandro> will post to public-rdf-prov@w3.org

sandro: on the joint mailing list public-rdf-prov@w3.org

guus: can something be done on this at the F2F?
... we need to prepare something on this for F2F

sandro: can others do other examples?

guus: how about responding to the first example with potential solutions?

<sandro> http://www.w3.org/2011/rdf-wg/wiki/TF-Graphs-UC

the above link is the relevant page for this WG to edit on provenance

guus: volunteers to do something?

gavinc: there is an example - i'll work on it

<yvesr_> just got a nice use-case from Denni V - I asked him to write it up

<yvesr_> about data wikis

<PatH> whatever you decide, it only works if everyone does it.

<gavinc> Exactly.

<sandro> sandro: maybe simplest solution is all script/turtle elements are the default graph, unless they have id tags. This gives you a dataset serialization format (although only when the graph ids have the same base)

guus: please come up with interesting suggestions for proposals

<sandro> Jan

<Guus> Jan Wielemaker

<AndyS> (implicit) base URIs considered harmful.

<Guus> will ask him to look at examples

guus: i'll ask Jan Wielemaker to contribute

<davidwood> AndyS++

<PatH> "My favorite quadstore", a contemporary love story.

<gavinc> www.w3.org/2011/rdf-wg/wiki/Graphs-In-Turtle

guus: provenance will be a main topic at the F2F

gavinc: work on serialization of all this
... looking at what it would take to standarize something like TRIG or n-QUADS and how it could be compatible with Turtle
... proposal from me (GRAPH keyword); another like TRIG (see strawman page)

Status RDF Dataset proposal

gavinc: problem with TRIG is that it isn't n-quads (because TRIG has to have all the graph in one spot)

Strawman proposal Richard: http://dvcs.w3.org/hg/rdf/raw-file/default/rdf-concepts/index.html#section-multigraph

Sandro: I don't like stateful syntaxes

<sandro> ericp

ericP: Turtle already has state - in [], prefixes, etc., so we wouldn't be too much worse

<ericP> <G1> { <G2> { <s> <p> <o> } }

<ericP> <G1> {

<AndyS> Turtle is only concat as whole files. (bNode labels excepted) is that enough?

<ericP> <G2> { <s> <p> <o> }

<ericP> }

<sandro> sandro: so good practice might be to start every turtle doc with "@graph . #go back to the default graph" if you think there might be concats.

ericP: some of this can be addressed by fiddling with embedding

<Zakim> ericP, you wanted to say the semantics might be easier if each graph simply added to the "database" so repeats simply append

ericP: concatenation has some negative consequences
... line-orientation also runs counter to concatenability

<LeeF> We repeat the same graph ID a fair amount

<LeeF> (in our TriG documents)

<LeeF> And it's additive

AndyS: the (?) implementation of TRIG actually allows graph id repetition

gavinc: but the semantics may not be best

<sandro> = vs +=

sandro: N3 has an implicit "=", so concatentation isn't the intended meaning

AndyS: which version of N3

sandro: both

<sandro> sandro: for this week, I'm inclined to go with @graph, since it's more clear to me what it means.

<ericP> with the curly braces approach, we have a short way to write the grammar:

<ericP> [1] trigDoc ::= ( directive | trigGraph )*

<ericP> [2] trigGraph ::= IRIref "="? "{" turtleDoc" "}" "."?

<ericP> and lean on the turtle spec

sandro: @graph form - semantics??

<gavinc> @graph . #go back to the default graph

gavinc: not complete

<sandro> sandro: when you hit the end of the turtle file, do you know exactly which triples are in each named graph?

guus: any relationship to documents

??

sandro: dataset = gbox, gsnap is different

gavinc: need something that shows complete graph contents

sandro: yes

<sandro> sandro: maybe = syntax will be clearer these are gsnaps

<sandro> (uh oh - phone problems - may drop off without warning.)

ericP: i fiddled with making trig lean on Turtle: http://www.w3.org/2010/01/Turtle/Trig

<AndyS> Why not make Trig simply +Turtle (unnamed {} optional) +NQuads (outside {})

sandro: how about splitting the difference?

AndyS: the strawman is more line oriented
... this helps with relationship to n-quads

<sandro> [....] is nesting, yeah.

gavinc: trig introduces nesting to turtle, which doesn't have much

<yvesr_> although only one degree of nesting (in trig)

Strawman proposal Richard: http://dvcs.w3.org/hg/rdf/raw-file/default/rdf-concepts/index.html#section-multigraph

guus: as Richard is not here, let's move on
... ISSUE-12 language-tagged literals
... Results of poll so far: http://www.w3.org/2002/09/wbs/46168/tagged_literals/results
... the proposals with only a few "no" votes are 1 and 2d
... the other proposals are very likely to be problematic
... option 2d has fewer prefer not votes, so it appears to be a more-liked solution

<AZ> +1 to go on with 2d

guus: shall we spend *more* time discussing this, or should we just try for 2d

<gavinc> +1 to go with 2d

<LeeF> I'm happy with 2d

<pchampin> +1 to go on with 2d

<yvesr_> guus++

<ericP> Arnaud, did you say that the current RDF spec says that language tags are normalized to lower case?

davidwood: should we spend some time looking at cleaning up various uses of string literals

<gavinc> ericP, yes

guus: spending scarce F2F time may not be the best way
... can we put forward a proposal (next week?) in preparation for a final decision at the F2F?

<ericP> gavinc, wasnt' that why we gave up on the datatype approach?

davidwood: OK

<ericP> oh well, i'm not going to push anything

guus: i'll propose a resolution for next week (but maybe the initial proposer would be better)

<AndyS> ericP - yes - sec 6.5 of "concepts"

sandro: we need to try to make a final resolution - no backsliding!

AOB

<scribe> ACTION: guus to propose a resolution for next week on language-tagged-literals [recorded in http://www.w3.org/2011/09/28-rdf-wg-minutes.html#action02]

<trackbot> Created ACTION-93 - Propose a resolution for next week on language-tagged-literals [on Guus Schreiber - due 2011-10-05].

guus: adjourned

<yvesr_> bye!

<AZ> bye

<zwu2> bye

This is scribe.perl Revision: 1.136  of Date: 2011/05/12 12:01:43  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/HTML/HTML data/
Succeeded: s/ericP/AndyS/
Found ScribeNick: pfps
Found ScribeNick: pfps
Inferring Scribes: pfps
Default Present: +33.4.77.42.aabb, +1.443.212.aacc, Arnaud, +1.707.861.aadd, Peter_Patel-Schneider, MacTed, gavinc, AlexHall, davidwood, Scott_Bauer, Guus, PatH, AZ, yvesr_, +1.781.899.aaff, Sandro, pchampin, +1.650.265.aagg, zwu2, EricP, +1.603.897.aahh, Souri, AndyS, LeeF
Present: +33.4.77.42.aabb +1.443.212.aacc Arnaud +1.707.861.aadd Peter_Patel-Schneider MacTed gavinc AlexHall davidwood Scott_Bauer Guus PatH AZ yvesr_ +1.781.899.aaff Sandro pchampin +1.650.265.aagg zwu2 EricP +1.603.897.aahh Souri AndyS LeeF

WARNING: No meeting chair found!
You should specify the meeting chair like this:
<dbooth> Chair: dbooth

Found Date: 28 Sep 2011
Guessing minutes URL: http://www.w3.org/2011/09/28-rdf-wg-minutes.html
People with action items: gavinc guus

WARNING: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.