IRC log of sparql on 2011-03-22

Timestamps are in UTC.

13:55:06 [RRSAgent]
RRSAgent has joined #sparql
13:55:06 [RRSAgent]
logging to
13:55:20 [Zakim]
Zakim has joined #sparql
13:55:34 [AxelPolleres]
Zakim, this will be sparql
13:55:34 [Zakim]
ok, AxelPolleres; I see SW_(SPARQL)10:00AM scheduled to start in 5 minutes
13:55:57 [AxelPolleres]
13:56:03 [AxelPolleres]
AxelPolleres has changed the topic to: agenda: (AxelPolleres)
13:56:08 [AxelPolleres]
chair: Axel Polleres
13:56:14 [AxelPolleres]
regrets: Chime
13:59:16 [Zakim]
SW_(SPARQL)10:00AM has now started
13:59:16 [Zakim]
13:59:24 [Zakim]
13:59:36 [SteveH_]
SteveH_ has joined #sparql
13:59:39 [Zakim]
13:59:40 [Zakim]
13:59:40 [Zakim]
13:59:47 [MattPerry]
MattPerry has joined #sparql
13:59:51 [cbuilara]
zakim, ??P5 is me
13:59:51 [Zakim]
+cbuilara; got it
13:59:56 [Zakim]
14:00:00 [Zakim]
+ +49.911.973.4.aaaa
14:00:04 [AndyS]
zakim, ??P15 is me
14:00:06 [Zakim]
+AndyS; got it
14:00:08 [NicoM]
zakim, +49 is me
14:00:10 [Zakim]
+NicoM; got it
14:00:18 [Zakim]
+ +
14:00:34 [OlivierCorby]
zakim, aabb is me
14:00:34 [Zakim]
+OlivierCorby; got it
14:00:35 [Zakim]
+ +1.603.897.aacc
14:00:44 [Zakim]
14:00:47 [MattPerry]
zakim, aacc is me
14:00:48 [Zakim]
+MattPerry; got it
14:01:02 [SteveH_]
Zakim, ??P19 is me
14:01:06 [Zakim]
+SteveH_; got it
14:02:11 [AxelPolleres]
Zakim, who is on the phone?
14:02:11 [Zakim]
On the phone I see kasei, AxelPolleres, cbuilara, AndyS, NicoM, OlivierCorby, MattPerry, SteveH_
14:02:36 [AndyS]
I'll scribe
14:02:43 [AndyS]
ScribeNick: AndyS
14:02:48 [AxelPolleres]
topic: Admin
14:02:50 [AndyS]
Scribe: Andy Seaborne
14:03:00 [AxelPolleres]
PROPOSED: Approve minutes at
14:03:16 [AndyS]
14:03:26 [AxelPolleres]
RESOLVED: Approve minutes at
14:03:55 [Zakim]
14:04:15 [AxelPolleres]
next week back to normal?
14:04:23 [AndyS]
Next meeting : Europe on summer time. Meeting 15:00 UK.
14:04:26 [AxelPolleres]
Next regular meeting: 2011-03-29 @ 15:00 UK / 10:00 EST (scribe: Birte)
14:04:43 [Zakim]
14:04:57 [AndyS]
ScribeNick: LeeF
14:05:15 [AndyS]
Scribe: Lee Feigenbaum
14:05:17 [LeeF]
topic: Comments
14:05:20 [AxelPolleres]
topic: Comments
14:05:24 [Zakim]
14:05:34 [AxelPolleres]
14:05:35 [SteveH]
Zakim, ??P25 is me
14:05:35 [Zakim]
+SteveH; got it
14:05:46 [AndyS]
I have some drafts ready-to-be-approved.
14:05:55 [LeeF]
AxelPolleres: unassigned comments
14:06:01 [LeeF] old one from Jeremy Carrol
14:06:24 [pgearon]
pgearon has joined #sparql
14:06:30 [LeeF] comment on federation extension
14:06:46 [AxelPolleres]
ACTION: Carlos to dradft an answer for comment WG-1
14:06:46 [trackbot]
Created ACTION-413 - Dradft an answer for comment WG-1 [on Carlos Buil Aranda - due 2011-03-29].
14:06:53 [Zakim]
14:07:07 [LeeF]
AxelPolleres: one comment by Kjetil is being superceded by other mails from Kjetil
14:07:25 [LeeF]
AxelPolleres: there are some draft responses on the mailing list, please send comments / +1s on those draft responses
14:07:28 [AndyS]
Andy's outstanding responses: JB-5 (zero responses so far), and SC-1 (has one yes from Axel)
14:07:48 [LeeF]
AxelPolleres: re: Gregg Reynolds's comment, Lee says that offlist communication indicates that Gregg does not expect a response from this WG
14:09:12 [LeeF]
AxelPolleres: should we reconsider the naming of the dataset protocol document again? alternative would be "graph store protocol"
14:09:23 [AxelPolleres]
14:09:47 [LeeF]
AndyS: where is the vote from last time so we can see where things stood?
14:10:03 [AndyS]
14:10:13 [AxelPolleres]
some people replied that we might change the name to "SPARQL 1.1 Graph Store HTTP Protocol
14:10:47 [LeeF] was the list of choices
14:11:34 [LeeF]
AndyS: RDF WG is likely to draw a distinction between the graph and the container for the graph
14:11:41 [LeeF]
AxelPolleres: I don't think we should adopt their terminology at this point
14:11:52 [LeeF]
AndyS: Right, don't know where the terminology will end up, but let's get the framework right
14:12:44 [LeeF]
LeeF: I'll email the WG list floating the possibility of renaming the dataset protocol yet again
14:12:57 [LeeF]
SteveH: users don't notice the distinction between mutable and non-mutable concepts
14:13:57 [Zakim]
14:14:10 [AxelPolleres]
topic: to last call
14:14:17 [AxelPolleres]
14:14:27 [LeeF]
AxelPolleres: query status?
14:14:54 [LeeF]
AndyS: I processed Birte and Axel's reviews that pertain to my sections; I'm up to date
14:15:05 [LeeF]
SteveH: One note in the document that I need to talk to Andy about.
14:15:41 [Zakim]
+ +34.92.38.aadd
14:15:41 [LeeF]
... one piece of algebra that is wrong per Birte and haven't yet been able to figure out what is right
14:16:44 [LeeF]
AxelPolleres: I also identified a few issues with the aggregate details
14:16:49 [LeeF]
SteveH: Haven't yet gotten to those comments
14:16:52 [ericP]
you folks don't [plan to] have a 1.1 test harness which does graph bnode isomorphism for the tester, do you?
14:16:55 [ericP]
(i ask 'cause RDB2RDF group contemplating such a harness 'cause their implementations don't require isomorphism equivalence other than for testing)
14:17:05 [LeeF]
ericP, no plans for that right now
14:17:08 [ericP]
14:17:09 [LeeF]
AxelPolleres: Lee still has open review
14:17:15 [LeeF]
14:17:24 [AndyS]
SPARQL 1.0 does?
14:17:50 [LeeF]
SteveH: Not likely to be completed this month.
14:18:15 [LeeF]
AndyS: there are still pending reviews from Lee and a bit from Axel
14:18:32 [LeeF]
AxelPolleres: mine is OK as is if I don't get to anything else
14:19:42 [LeeF]
LeeF: I'll punt on my review rather than hold back the schedule. My review is more likely to focus on editorial clarity and consistency then on the algebra details.
14:19:48 [AxelPolleres]
Steve: should be ready by April 8
14:19:52 [LeeF]
AndyS: Offer to help SteveH with algebra changes
14:20:06 [LeeF]
SteveH: Thanks, quite possibly helpful.
14:20:35 [LeeF]
SteveH: Could use help looking at Birte's suggested algebra changes for grouping, hard to figure out precise details
14:20:40 [LeeF]
(agreement to take conversation offline)
14:20:42 [AxelPolleres]
Zakim, who is on the phone?
14:20:42 [Zakim]
On the phone I see kasei, AxelPolleres, cbuilara, AndyS, NicoM, OlivierCorby, MattPerry, LeeF, SteveH, pgearon, Sandro, +34.92.38.aadd
14:20:46 [LeeF]
AxelPolleres: update status?
14:21:11 [SteveH]
AxelPolleres, yes, I've not read your review yet
14:21:14 [LeeF]
pgearon: will spend some time this week
14:21:31 [LeeF]
AxelPolleres: Alex says we'll have update on formal semantics
14:22:10 [LeeF]
AndyS: there are some yet-unaddressed comments from an earlier review from last August - we need to make sure those aren't lost
14:22:19 [LeeF]
pgearon: i can look through that list this afternoon
14:22:35 [AxelPolleres]
ACTION: paul to check (andy's) old comments on update
14:22:35 [trackbot]
Created ACTION-414 - Check (andy's) old comments on update [on Paul Gearon - due 2011-03-29].
14:22:55 [AxelPolleres]
ACTION: Axel to check back with Alex on update formal semantics
14:22:55 [trackbot]
Created ACTION-415 - Check back with Alex on update formal semantics [on Axel Polleres - due 2011-03-29].
14:23:39 [AxelPolleres]
moduylo formal semantics issue, should be ready for LC by first week of April
14:23:55 [LeeF]
AndyS: other review?
14:24:01 [LeeF]
kasei: Haven't had a chance to start it yet
14:24:19 [LeeF]
... ping me when I should start?
14:24:36 [AxelPolleres]
ACTION: axel to ping greg for review as soon as Update formal semantics is ready
14:24:37 [trackbot]
Created ACTION-416 - Ping greg for review as soon as Update formal semantics is ready [on Axel Polleres - due 2011-03-29].
14:24:48 [LeeF]
AxelPolleres: protocol?
14:25:21 [LeeF]
14:26:25 [AxelPolleres]
LeeF: section 2.1 has some updates, would be good to have feedback
14:27:53 [AxelPolleres]
so, maybe if specifically the reviewers at least could have a look, comments welcome from everybody.
14:28:03 [LeeF]
AxelPolleres: service description?
14:28:19 [LeeF]
kasei: held up on editorial issues
14:28:24 [LeeF]
... 1 is based on protocol document
14:28:36 [LeeF]
... other is on empty graph issue, waiting on ACTION-406 for Lee to provide some text
14:28:40 [bglimm]
bglimm has joined #sparql
14:28:42 [LeeF]
... minor changes, nothing substantive
14:28:53 [LeeF]
AxelPolleres: on protocol document - schedule?
14:29:26 [AxelPolleres]
protocol schedule schedule not entirely clear
14:29:35 [Zakim]
+ +44.186.528.aaee
14:29:39 [AxelPolleres]
service description essentially ready for LC
14:29:48 [LeeF]
kasei: for protocol dependencies I just need stable links into the document
14:29:51 [bglimm]
Zakim, +44.186.528.aaee is me
14:29:51 [Zakim]
+bglimm; got it
14:29:56 [LeeF]
kasei: ready for LC very very soon then, I hope
14:30:03 [LeeF]
AxelPolleres: dataset protocol?
14:30:29 [LeeF]
AxelPolleres: there are some comments pending
14:31:44 [AxelPolleres]
I made some attempt to summarise our position in the draft response to KK-12 ...
14:31:45 [AndyS]
I think bullet one is done.
14:32:04 [AxelPolleres]
... maybe the name change is an issue.
14:32:25 [AndyS]
... and bullet 2 is a change in fundamentals (?)
14:32:30 [LeeF]
AxelPolleres: entailment, main problem is still waiting for a review
14:32:38 [LeeF]
bglimm: yes, and i might need to add something about property paths
14:32:45 [LeeF]
... because they are different from BGP matching
14:33:00 [LeeF]
... and entailment is not defined for paths that can't be reduced to normal BGPs
14:33:07 [LeeF]
AxelPolleres: can we add that to wiki page?
14:33:09 [LeeF]
bglimm: yes
14:34:35 [LeeF]
AxelPolleres: can we move ahead without other reviews?
14:34:39 [LeeF]
LeeF: no process requirement otherwise
14:34:46 [LeeF]
AxelPolleres: if we have no more internal reviews, I'd be OK to go ahead
14:34:51 [LeeF]
(silence, other than typing)
14:35:19 [AxelPolleres]
ACTION: Birte to add aa paragraph in entailment on property paths
14:35:19 [trackbot]
Created ACTION-417 - Add aa paragraph in entailment on property paths [on Birte Glimm - due 2011-03-29].
14:35:32 [AxelPolleres]
ACTION: Axel to review proprtey paths paragraph in Entailment
14:35:33 [trackbot]
Created ACTION-418 - Review proprtey paths paragraph in Entailment [on Axel Polleres - due 2011-03-29].
14:35:41 [LeeF]
AxelPolleres: federated query status?
14:35:58 [LeeF]
AxelPolleres: re: entailment, there is still pending discussion of d-entailment and canonicalization
14:36:06 [LeeF]
... let simmer for a week and see if we get more clarification
14:36:08 [LeeF]
... no traffic
14:36:21 [LeeF]
... we said if we don't get clarification then we might just drop it
14:36:32 [LeeF]
bglimm: not sure we accomplish anything by waiting another week
14:37:03 [LeeF]
bglimm: either remove it or strengthen it to define a minimum set of datatypes that need to be supported
14:37:09 [LeeF]
AxelPolleres: straw poll!
14:37:25 [LeeF]
bglimm: could be an option to leave as is, but seems to be agreement that it's not useful in its current form
14:37:43 [AxelPolleres]
Strawpoll: +1 change D-entailment towards canonicalization -1 drop D-entailment
14:37:51 [SteveH]
14:37:52 [AndyS]
+1 canonicalize
14:37:56 [cbuilara]
14:37:56 [MattPerry]
14:37:56 [bglimm]
14:37:57 [kasei]
14:37:57 [pgearon]
14:37:58 [NicoM]
14:37:58 [LeeF]
-1 (haven't heard of any systems that actually do SPARQL + D-entailment)
14:38:08 [AxelPolleres]
14:38:45 [LeeF]
bglimm: I would need help defining what data types must be supported
14:38:47 [AxelPolleres]
Zakim, who is on the phone?
14:38:47 [Zakim]
On the phone I see kasei, AxelPolleres, cbuilara, AndyS, NicoM, OlivierCorby, MattPerry, LeeF, SteveH, pgearon, Sandro, +34.92.38.aadd, bglimm
14:39:02 [LeeF]
q+ to ask about canonicalization
14:39:07 [LeeF]
AxelPolleres: Matt, could you help?
14:39:10 [LeeF]
MattPerry: sure
14:40:27 [LeeF]
LeeF: I don't think canonicalization (of data being input to a store) is at issue here. This is about stores that do NOT canonicalize data on the way in but enforce D-entailment rules at query time, and we don't know of any systems that do that
14:41:04 [AndyS]
+1 - suggest it is on input; and for datatypes, suggest coverage - not MUST
14:41:40 [AxelPolleres]
if you do and don't do canonicalization, you get different answers... that's an entailment regime, and it's done by systems
14:43:36 [LeeF]
LeeF: I don't understand how canonicalization of data input has anything to do with entailment at query time
14:43:44 [LeeF]
AndyS: the entailment regimes do specify legal graphs for input
14:43:59 [LeeF]
bglimm: I'd rather not put in the work if there's no benefit to it
14:44:16 [LeeF]
AxelPolleres: I read that Oracle does this
14:44:23 [AndyS]
RIOT can canonicalize on input
14:44:35 [LeeF]
MattPerry: is there things in D-entailment that would not be captured by doing canonicalization at input time?
14:44:39 [LeeF]
bglimm: you also need to do RDFS entailment
14:45:15 [LeeF]
MattPerry: I can send out a list of the canonicalization that we do.
14:45:36 [AxelPolleres]
ACTION: Matthew to sync with Birte on Datatypes for canonicalisation
14:45:36 [trackbot]
Created ACTION-419 - Sync with Birte on Datatypes for canonicalisation [on Matthew Perry - due 2011-03-29].
14:45:52 [LeeF]
AxelPolleres: federated query status?
14:45:58 [bglimm]
Zakim, mute me
14:45:58 [Zakim]
bglimm should now be muted
14:46:05 [LeeF]
cbuilara: applied the comments from Axel and Lee
14:46:07 [AxelPolleres]
topic: Fed Query
14:46:10 [LeeF]
... sent email
14:46:19 [LeeF]
... wiki page with the comments of how I handled these comments
14:46:24 [LeeF]
... waiting for answers
14:46:32 [LeeF]
... need to clarify some things
14:46:39 [LeeF]
... from Axel and Lee
14:47:03 [LeeF]
AxelPolleres: link to wiki page?
14:47:31 [AxelPolleres]
Carlos, add missing issues to Last call page on wiki, please
14:47:34 [AndyS]
I'd take the grammar extract out - it's probably out of date. Link to ServiceGraphPattern
14:48:33 [AxelPolleres]
ACTION: Carlos to look into removing grammar extracts from Fed query and direclty link to current grammar
14:48:33 [trackbot]
Created ACTION-420 - Look into removing grammar extracts from Fed query and direclty link to current grammar [on Carlos Buil Aranda - due 2011-03-29].
14:48:34 [Zakim]
14:48:48 [cbuilara]
fed Query review:
14:49:02 [LeeF]
AxelPolleres: JSON results format?
14:49:09 [LeeF]
AndyS: no changes yet, concentrating on query
14:49:15 [LeeF]
... JSON can go on a different schedule if need be
14:49:34 [Zakim]
14:49:39 [bglimm]
Zakim, unmute me
14:49:39 [Zakim]
bglimm should no longer be muted
14:49:46 [NickH]
Zakim, ??P13 is me
14:49:49 [Zakim]
+NickH; got it
14:50:37 [AxelPolleres]
Birte: entailment... ready by end of first week of april
14:50:37 [LeeF]
cbuilara: if comments this week, need 1-2 weeks at most to apply final comments
14:50:46 [bglimm]
Zakim, mute me
14:50:46 [Zakim]
bglimm should now be muted
14:51:10 [AxelPolleres]
Carlos: reaslistic 1st or 2nd week of april (assuming open questions are addressed by Lee and Axel)
14:51:23 [LeeF]
AxelPolleres: overview document?
14:51:27 [LeeF]
... Lee and Axel to discuss offline
14:52:24 [LeeF]
topic: test cases
14:52:33 [LeeF]
AxelPolleres: looking for volunteers to look through test cases and figure out what is ready for approval
14:52:44 [LeeF]
... look through minutes, mailing list, etc.
14:52:55 [MattPerry]
Property Path tests are ready as far as I'm concerned
14:53:01 [AndyS]
I've added some bad syntax tests recently.
14:53:21 [AxelPolleres]
Zakim, pick a victim
14:53:21 [Zakim]
Not knowing who is chairing or who scribed recently, I propose MattPerry
14:53:24 [LeeF]
LeeF: suggest agenda items for next week: property path tests and bad syntax tests
14:53:33 [OlivierCorby]
I can contribute to test cases
14:54:17 [Zakim]
+ +1.540.412.aaff
14:54:28 [AxelPolleres]
ACTION: Olivier to look through test cases and provide a summary by next TC
14:54:28 [trackbot]
Created ACTION-421 - Look through test cases and provide a summary by next TC [on Olivier Corby - due 2011-03-29].
14:54:40 [pgearon]
Zakim, aaff is me
14:54:40 [Zakim]
+pgearon; got it
14:56:19 [LeeF]
topic: action review
14:56:28 [LeeF]
AxelPolleres: (reviews pending actions)
14:56:55 [kasei]
407 is done
14:57:11 [AxelPolleres]
close ACTION-407
14:57:11 [trackbot]
ACTION-407 Change conformance section in SD to refer to including "at least" one triple... closed
14:57:27 [LeeF]
14:57:52 [AndyS]
408 is done (currently pending) will close. review of 407.
14:58:52 [MattPerry]
14:59:05 [AxelPolleres]
close ACTION-394
14:59:42 [AxelPolleres]
close ACTION-393
14:59:43 [trackbot]
ACTION-393 Review query closed
14:59:54 [AxelPolleres]
close ACTION-390
14:59:54 [trackbot]
ACTION-390 Update comments page closed
15:01:31 [MattPerry]
15:01:34 [Zakim]
15:01:35 [Zakim]
15:01:37 [Zakim]
15:01:37 [LeeF]
15:01:39 [Zakim]
15:01:43 [Zakim]
15:01:47 [Zakim]
- +34.92.38.aadd
15:01:49 [Zakim]
15:01:51 [Zakim]
15:01:53 [Zakim]
15:01:58 [AxelPolleres]
ACTIONS upto 397 will be reviewed next time.
15:01:59 [Zakim]
15:02:04 [AxelPolleres]
15:02:05 [LeeF]
RRSAgent, make logs world
15:02:24 [AxelPolleres]
Zakim, who is on the phone?
15:02:24 [Zakim]
On the phone I see AxelPolleres, NicoM, OlivierCorby, LeeF
15:06:02 [sandro]
Zakim, who is here?
15:06:02 [Zakim]
On the phone I see AxelPolleres, NicoM, OlivierCorby, LeeF
15:06:03 [Zakim]
On IRC I see bglimm, pgearon, SteveH, Zakim, RRSAgent, cbuilara, OlivierCorby, AndyS, LeeF, AxelPolleres, karl, ericP, NickH, kasei, trackbot, sandro
15:06:30 [sandro]
kasei, we decided against jsonp, didn't we? I now think that was a mistake.....
15:06:41 [kasei]
15:06:54 [kasei]
still time to change... :)
15:07:27 [AndyS]
Sandro - could you send an email to the list about it.
15:09:22 [sandro]
well, I've started doing some coding where I need jsonp, and it works great, and it occurs to me that it'll still be a while before practically-all browsers implement it, even if just because it takes years for people to upgrade their browsers.
15:09:27 [sandro]
Will do, AndyS
15:10:15 [AndyS]
The "years" argument seems quite important. Good that semweb now has such problems!
15:11:29 [Zakim]
15:11:30 [Zakim]
15:27:42 [sandro]
AndyS, kasei, I take it back. I didn't realize that CORS had been out there quite so long already. I did the math, and 86% of current users are on a browser that supports CORS. The ones who don't are IE 6 or 7 (9%), Opera (2.5%), and other ancient versions (2%). I don't care too much about support for them, I guess.
15:29:57 [sandro]
(The opera case is truly bizarre. They don't support it, but the Editor of the CORS TR is Anne van Kesteren, and he works for Opera. I'm sure there's an interesting story here.)
15:30:29 [AndyS]
sandro - facts - good. And IE6 or 7 is unlikely to run the apps anyway (users not that sort?, behind firewalls banning them etc etc)
15:31:04 [sandro]
16:47:38 [AndyS]
AndyS has joined #sparql
17:46:47 [SteveH]
SteveH has joined #sparql
18:06:16 [AndyS]
AndyS has joined #sparql
18:19:29 [pgearon]
pgearon has joined #sparql
18:37:40 [pgearon]
pgearon has joined #sparql
20:35:37 [karl]
karl has joined #sparql
20:51:37 [LeeF]
is there a name for rdfs:Resource - rdfs:Literal? is that even a reasonable question?
21:03:56 [kasei]
i wish there was... i don't think (but am not sure) owl:Thing is what you want.
21:04:23 [kasei]
not a bit owl person, but think that can cover literals
21:05:06 [kasei]
i'm worried that it's not a reasonable question for the logics side of things, but that it's a very reasonable thing to want in practice.
21:05:37 [LeeF]
kasei, thanks, that's exactly where i'm at
21:05:45 [LeeF]
owl:thing == rdfs:Resource, so yeah, that's not what i want
21:05:52 [LeeF]
but i want some way to write down metadata about sparql functions
21:05:56 [LeeF]
such as the parameters they expect
21:06:06 [LeeF]
and want to be able to say that a function wants a <URI> type thingy
21:06:43 [LeeF]
In SPARQL 1.0 we just talk about IRIs, without giving them a URI
21:10:51 [kasei]
despite my (possibly wrong) impressions of owl:Thing, seems like you should be able to express this in owl as !rdfs:Literal
21:15:42 [SteveH]
SteveH has joined #sparql
21:15:49 [AndyS]
owl:Thing isa rdfs:Class but does not contain rdfs:Class. Can some be a literal and have a IRI?
21:16:44 [AndyS]
LeeF: "a function wants a <URI> type thingy" ==> That's really about the graph node type - not the thing being named.
21:48:56 [LeeF]
dang, missed AndyS
23:23:10 [bglimm]
bglimm has joined #sparql