W3C

- DRAFT -

RDF Working Group Teleconference

07 Sep 2011

Attendees

Present
yvesr, Guus, +1.707.861.aabb, gavinc, Scott_Bauer, davidwood, Ivan, AZ, Sandro, SteveH, mischat, +1.617.324.aadd, pchampin, ericP, +1.443.212.aaee, LeeF, AlexHall, Thomas, zwu2, MacTed, AndyS, mbrunati, +1.603.897.aagg, Souri, NickH
Regrets
Chair
SV_MEETING_CHAIR
Scribe
alexhall

Contents


<mischat> any head with making http://www.w3.org/2011/08/31-rdf-wg-irc public would be great :)

<mischat> i will bug swh when i see him, he might know ...

<MacTed> mischat - this is the page that should definitely be public -- http://www.w3.org/2011/rdf-wg/meeting/2011-08-31

<MacTed> you get there thru the editable version of the IRC log -- http://www.w3.org/2011/rdf-wg/wiki/Chatlog_2011-08-31

<MacTed> by clicking the "preview nicely formatted version" link in the page head

<MacTed> they all *appear* to be public, having tested with another browser where I'm not logged in

<ivan> trackbot, start telcon

<trackbot> Date: 07 September 2011

<yvesr> anyone using ekiga here? i keep getting 'this passcode is not valid'

<yvesr> found - DTMF needs to be set as RFC2833

<Guus> zhe, can you scribe?

<Guus> you're on the list :-)

<zwu2> Guus, I am not feeling well today, can I scribe next week?

<scribe> scribe: alexhall

<zwu2> Thanks Alex!

Admin

guus: thanks to mischa for fixing minutes, apologies for their lateness

<tomayac> +1 to accept

guus: correct link to minutes is in agenda

<mischat> these are the minutes http://www.w3.org/2011/rdf-wg/meeting/2011-08-31

PROPOSED: to accept the minutes from 31 Aug telecon

RESOLUTION: to accept the minutes from 31 Aug telecon

guus: no pending action items
... open items, danbri is not here

sandro: started conversation, when it's considered stable we can proceed with registering it
... not sure what stable means, from w3c perspective could mean final call
... could probably proceed before then, probably doesn't matter

guus: would prefer to proceed now, hearing no objections

sandro: would like resolution from wg
... prepare template, put to vote for a resolution
... does anybody have reason to think it's not stable?

<sandro> ACTION: sandro to draft well-known URI template and propose WG resolution that it is "stable" enough for IETF. [recorded in http://www.w3.org/2011/09/07-rdf-wg-minutes.html#action01]

<trackbot> Created ACTION-82 - Draft well-known URI template and propose WG resolution that it is "stable" enough for IETF. [on Sandro Hawke - due 2011-09-14].

<sandro> close action-52

<trackbot> ACTION-52 Start conversation on reserving our well-known string (genid) closed

guus: next item, review PA's comments on SPARQL update protocol
... think it's done

<danbri> Hi. I'm in IRC, but I'm not dialing into telecons until I get my Skype dialout fixed (was a payments system problem; allgedly it'll work from tommorrow)

<mischat> no i don't think it was discussed either

guus: was not discussed last week, and PA is not here so we need to keep it open
... need to change the date on action 73, don't need to track it every week

<sandro> action-73?

<trackbot> ACTION-73 -- Fabien Gandon to implement http://www.w3.org/2011/rdf-wg/meeting/2011-08-03#resolution_2 -- due 2011-08-24 -- OPEN

<trackbot> http://www.w3.org/2011/rdf-wg/track/actions/73

guus: action 77 is still open, propose to discuss as part of today's RDF datasets agenda item
... next telecon is 14 Sep, back on normal schedule
... need to agree on policy for who is responsible for responding to public comments

<gavinc> Timely too :\

guus: traditionally editor is responsible for responding
... at the very least, each comment must be politely acknowledged
... suggestion is to think of how each comment impacts the text, point out specific passages or ask for specific changes

ivan: number of commenters were unhappy with current policy of different mailing lists for comments vs. working group
... personally don't have a problem with this, but should we change it?

guus: administratively, it must be done this way.

ivan: rdfa working group has only one mailing list
... supposing it's doable, do we want to make that change?

andy: how does rdfa formally track comments if they're all on the same list?
... one advantage of separate lists is that it's more convenient for tracking comments

<davidwood> +1 to Andy. The separate comments list seems to make it easier to track comments.

ivan: rdfa wg manages by having a very devoted chair who spends a lot of time
... understand the advantage

andy: can't recall exact setup of sparql lists, but wg list is for more conversation and chitchat while public list is for formal comments

guus: happy to leave it as it is as long as there are no strong objections from wg.

FTF planning

<mischat> http://www.w3.org/2011/rdf-wg/wiki/F2F2

guus: 8 or 9 at Boston location, several more remotely at bbc
... please add your name to one of these lists

<gavinc> Yay for a 2nd Europen F2F ;)

guus: BBC might have more attendees than Boston, might affect scheduling issues (earlier start time in Boston)

<gavinc> ... ... ... Uh. Right, that sucks

sandro: looks like more people at BBC location, might make more sense to make it there instead

<NickH> haha, if there are morepeople in London, then Boston is the 'second site'?

<yvesr> sandro, although we can't accomodate a lot more people at the bbc location (the room we have is relatively small)

david: can't personally make it to european location

<Scott_Bauer> I've also made travel plans for

<Scott_Bauer> Boston

andy: i know it's harder having a split site, but we get more people this way

<LeeF> Echo the concerns about a two-site meeting without video

gavin: noticed that we don't seem to have video conferencing setup, is that correct?

yves: that's correct, we don't currently have videoconf at BBC, can investigate
... also, cannot accommodate many more people

guus: seems we will end up with a split meeting, need to come up with an agenda to accomodate both sides

<LeeF> oops

<LeeF> i hung up, but intended to say that i'd be happy to do 7am

<LeeF> "happy"

guus: 5 hour time difference, is 8am start in Boston OK?

eric: we could possibly open the door early

guus: breakout groups early in UK, afternoon in US
... common meeting in morning (US)/afternoon (UK)

david: might be hard to organize breakouts by geography instead of interest
... could we overlap the breakouts?

<sandro> So, I've just re-researved MIT's H.320/H.323 video conferencing room, hoping a matching BBC facility will be available.

yves: can keep the building open, only possible problem is that lunch is already arranged

guus: proposed schedule: UK breakouts 10am-noon, lunch noon-1pm, meeting 1pm-??
... US meeting 8am-2pm, lunch, then breakouts

sandro: had previously released video conferencing room, will re-reserve

<scribe> ACTION: guus and david to come up with agenda which works for US and UK locatinos [recorded in http://www.w3.org/2011/09/07-rdf-wg-minutes.html#action02]

<trackbot> Created ACTION-83 - And david to come up with agenda which works for US and UK locatinos [on Guus Schreiber - due 2011-09-14].

<sandro> Hmmmm. Ekiga claims to support H.323, so yeah, people could do it on desktops.

guus: discussed in august what would be the main meeting objectives
... making substantial progress on multi-graph issues is the major objective

<mischat> does that include all of the graph terminology stuff ?

guus: given that other issues are making reasonable progress
... yes, this includes the terminology, sandro is waiting for progress on multigraph before moving on with terminology

Liason with Provenance WG

guus: in august we were in contact with Luc, one of the chairs of the provenance wg
... trying to set up a common time
... their telecon is the same as ours but on thursday
... we will meet after their telecon on 15 Sep (at 12:15pm eastern)
... who in this group can or will attend?

<AZ> I want to attend

<ivan> probably

<Scott_Bauer> probably, need to rearrange a meeting.

<MacTed> I plan to join

<AndyS> Interested ... no specific issues ... want to understand deeply

<davidwood> I will be there

<gavinc> I plan to join

<sandro> I'm inclined to attend.

<SteveH> Would like to be there, but have clashes

<pchampin> not sure yet, I'll have to check

guus: 5 or 6 volunteers in IRC, would also like richard to attend

<mischat> the provenance people only really seem to care about named graphs

<scribe> ACTION: guus to contact Richard to ask to attend provenance WG liason call [recorded in http://www.w3.org/2011/09/07-rdf-wg-minutes.html#action03]

<trackbot> Created ACTION-84 - Contact Richard to ask to attend provenance WG liason call [on Guus Schreiber - due 2011-09-14].

<scribe> ACTION: guus to distribute agenda for provenance wg call [recorded in http://www.w3.org/2011/09/07-rdf-wg-minutes.html#action04]

<trackbot> Created ACTION-85 - Distribute agenda for provenance wg call [on Guus Schreiber - due 2011-09-14].

Status of RDF Dataset proposal

<mischat> we used to have this page http://www.w3.org/2011/rdf-wg/wiki/TF-Graphs re: Graph's TF

guus: there was discussion between PA and richard on dataset proposal
... didn't seem to make its way into the wiki page

pchampin: have made progress on understanding each other's motivations and are slowly coming towards consensus
... but have not arrived at agreement yet

guus: can't make progress on this today, will go back on agenda for next week

ISSUE-12 language-tagged literals

guus: message from pat on varieties of tagged literals
... not much discussion on this yet

<mischat> iirc Pat was asked last week to put together a review of the various proposals

ivan: questionnaire is not yet open because some questions about final wording
... think it's mostly OK
... this is a personal questionnaire, not a formal vote on behalf of the companies
... also a public questionnaire
... sandro made note right before the call that it's set up as a radio button form, meaning only one choice
... other alternative is a more open-ended form
... don't want to open it up too much or this could keep dragging on

guus: any preference from wg on whether poll should be single-choice or multi-select?
... personally prefer the preferences form, gives more nuance

sandro: have a first choice and second choice and a bunch that i really don't like, no way to express that in current form

ivan: i proposed this poll last week, because without taking sides, all arguments have been extensively aired and we're going in circles now
... try to get a clear view of whether there's a single solution that's obviously a winner, don't want this to drag on for months

<AndyS> Why are we discussing the poll when the WG can't see the form?

david: lee asked sandro for more details on his mailing list comment, can we discuss now?

<ericP> ivan, worth quickly opening it as an exhibit?

<sandro> http://lists.w3.org/Archives/Public/public-rdf-wg/2011Sep/0019.html Pat's Email

sandro: poll is only visible to staff and chairs, same as pat's email but in radio button form
... my comments are relevant to pat's email
... of all the option 2 styles, seems that if I have "foo" with different tags then i have 2 literals with same lexical and datatypes
... how do i tell the difference between these in a system that only supports RDF 1.0-style literals?

<pchampin> @sandro: well, yes, but they differ by their LANG

sandro: lang appears in sparql, but it doesn't appear in a serialization or api
... thought the goal of this was to simplify things by giving everything a datatype

pchampin: spirit of #2 proposals is that we still have 2 styles of literals: those with tag and those without
... most people didn't like idea of open-ended universe of language datatypes

sandro: what is the advantage of option 2?

pchampin: everything is given a datatype, but still has distinct lexical and language parts

<ericP> so option 2 leaves us with two literal types?

sandro: seems a trivial change, you're just giving a little bit of extra information

<sandro> sandro: I just don't see how Option 2 is an improvement over what we have now.

andy: concerns about giving URIs to language datatypes
... doesn't play well with existing subtype behavior
... discussed on the mailing list, but doesn't seem to be reflected in the poll.

<sandro> sandro: I understand language tags to not have a strict hierarchy, so we shouldn't use datatype hierarchy.

pchampin: subtypes in this context means value space of language datatype is subset of tagged literals

<AndyS> IIRC: rdf:Lang-en owl:sameAs ns:name => looses lang info :: must be fixed URIs -> special.

<pchampin> pchampin: rdf:TaggedLiteral value space would contain all the <txt, tag> pairs

sandro: sounds like you're saying these subtypes are implicit but can't be used directly

<pchampin> ... but its lexical space would be empty

<AndyS> What is DATATYPE("foo"@en) in 3* ?

<pchampin> ... rdf:TaggedLiteral/fr would only contain the <txt, "fr"> pairs

<pchampin> ... its lexical space would contain all the strings txt

<pchampin> ... and its L2V would be txt -> <txt,"fr">

ivan: option 3 introduces lots of new datatypes to the picture, expect that inference engines would not like this

<pchampin> ... Simiar to owl:real having values but no lexical space;

<pchampin> ... xsd:decimal having a subset of owl:real, and a lexical space and L2V for that subset.

ivan: owl-rl e.g. would have problems with axiomatic triples because it needs a triple for each type it knows about

andy: language tags in 1.1 will be treated specially no matter what, option 3 buries it in the URI, just moving where the special casing occurs
... now you have to parse the URIs to extract the language tag information

<gavinc> langmatches is going to be defined really really really funckily for option 3

sandro: people who care about language tags will always have to treat them specially, option 3 allows people who don't care about them to ignore them

andy: option 2 puts it in the syntax, not in the datatype

<SteveH> I think sandro has a point

sandro: don't understand how the mechanism in option 2 is supposed to work, main basis for not liking that
... i see options 3 and 4 simplifying application code for lots of people, not so for option 2

<gavinc> really, the iri rdf:langTag-zh-cmn-a-bbb-a-ccc is an improvement? :\

<sandro> guus: Am I right that option 2 has aestheic advantage, but not much of an advantage for implementors.

guus: have a feeling that the main advantage of option 2 is an aesthetic advantage, is that true?

andy: don't think any of them have a clear technical advantage, everything requires special handling of some sort

sandro: would like to eliminate branching of literal handling in apis

gavin: datatype iris for moderately complex language tags are going to be a pain

ivan: need final decision on what form to make the poll

<zwu2> Ivan, you worry about inference blow up, how many language tags do you think there will be?

<MacTed> +1 example for each! a concrete thing will vary with each proposal ... show the variations?

<gavinc> API (RDF Interfaces) in Python for Literal ;) https://github.com/norcalrdf/pymantic/blob/master/pymantic/primitives.py#L203

<sandro> ACTION: sandro to provide example for how code is simpler with language-tag options 3 and 4 vs 1 and 2. [recorded in http://www.w3.org/2011/09/07-rdf-wg-minutes.html#action05]

<trackbot> Created ACTION-86 - Provide example for how code is simpler with language-tag options 3 and 4 vs 1 and 2. [on Sandro Hawke - due 2011-09-14].

ivan: will need to change poll. is 2 weeks ok?

<zwu2> thanks Alex!

guus: close poll night before telecon 2 weeks from now

<gavinc> avoid last weeks fun ;)

thanks gavin, was trying to look that up

<gavinc> np

<MacTed> trackbot, end meeting

Summary of Action Items

[NEW] ACTION: guus and david to come up with agenda which works for US and UK locatinos [recorded in http://www.w3.org/2011/09/07-rdf-wg-minutes.html#action02]
[NEW] ACTION: guus to contact Richard to ask to attend provenance WG liason call [recorded in http://www.w3.org/2011/09/07-rdf-wg-minutes.html#action03]
[NEW] ACTION: guus to distribute agenda for provenance wg call [recorded in http://www.w3.org/2011/09/07-rdf-wg-minutes.html#action04]
[NEW] ACTION: sandro to draft well-known URI template and propose WG resolution that it is "stable" enough for IETF. [recorded in http://www.w3.org/2011/09/07-rdf-wg-minutes.html#action01]
[NEW] ACTION: sandro to provide example for how code is simpler with language-tag options 3 and 4 vs 1 and 2. [recorded in http://www.w3.org/2011/09/07-rdf-wg-minutes.html#action05]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.136 (CVS log)
$Date: 2011/09/07 18:24:37 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
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/ivan/guus/
Succeeded: s/asked//
Succeeded: s/???/Guus/
Succeeded: s/subtype/subset/
Found Scribe: alexhall
Inferring ScribeNick: AlexHall
Default Present: yvesr, Guus, +1.707.861.aabb, gavinc, Scott_Bauer, davidwood, Ivan, AZ, Sandro, SteveH, mischat, +1.617.324.aadd, pchampin, ericP, +1.443.212.aaee, LeeF, AlexHall, Thomas, zwu2, MacTed, AndyS, mbrunati, +1.603.897.aagg, Souri, NickH
Present: yvesr Guus +1.707.861.aabb gavinc Scott_Bauer davidwood Ivan AZ Sandro SteveH mischat +1.617.324.aadd pchampin ericP +1.443.212.aaee LeeF AlexHall Thomas zwu2 MacTed AndyS mbrunati +1.603.897.aagg Souri NickH

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

Found Date: 07 Sep 2011
Guessing minutes URL: http://www.w3.org/2011/09/07-rdf-wg-minutes.html
People with action items: david guus sandro

WARNING: IRC log location not specified!  (You can ignore this 
warning if you do not want the generated minutes to contain 
a link to the original IRC log.)


[End of scribe.perl diagnostic output]