Warning:
This wiki has been archived and is now read-only.
Chatlog 2009-10-20
From SPARQL Working Group
See original RRSAgent log and preview nicely formatted version.
Please justify/explain all edits to this page, in your "edit summary" text.
13:57:44 <RRSAgent> RRSAgent has joined #sparql 13:57:44 <RRSAgent> logging to http://www.w3.org/2009/10/20-sparql-irc 13:58:03 <Zakim> Zakim has joined #sparql 13:58:11 <AxelPolleres> zakim, this will be sparql 13:58:11 <Zakim> ok, AxelPolleres; I see SW_(SPARQL)10:00AM scheduled to start in 2 minutes 13:58:32 <Zakim> SW_(SPARQL)10:00AM has now started 13:58:37 <Zakim> + +1.312.863.aaaa 13:58:40 <Zakim> + +1.518.276.aabb 13:58:44 <AxelPolleres> scribe: Axel Polleres 13:59:08 <Zakim> + +3539149aacc 13:59:13 <pgearon> Zakim, +1.312.863.aaaa is me 13:59:13 <Zakim> +pgearon; got it 13:59:20 <AxelPolleres> Zakim, aacc is me 13:59:20 <Zakim> +AxelPolleres; got it 13:59:21 <AndyS> AndyS has joined #sparql 13:59:24 <Zakim> + +0186528aadd 13:59:27 <kasei> Zakim, aabb is me 13:59:27 <Zakim> +kasei; got it 13:59:31 <kasei> Zakim, mute me 13:59:31 <Zakim> kasei should now be muted 13:59:34 <bglimm> Zakim, who am I 13:59:37 <Zakim> I don't understand 'who am I', bglimm 13:59:51 <bglimm> Zakim, who am I? 13:59:51 <Zakim> I don't understand your question, bglimm. 14:00:15 <SimonS> SimonS has joined #sparql 14:00:18 <Zakim> - +0186528aadd 14:00:33 <Zakim> + +7.950.aaee 14:00:35 <kasei> Zakim, who is on the phone? 14:00:35 <Zakim> On the phone I see kasei (muted), pgearon, AxelPolleres, +7.950.aaee 14:00:41 <AxelPolleres> Agenda is at http://www.w3.org/2009/sparql/wiki/Agenda-2009-10-20 14:00:47 <AxelPolleres> AxelPolleres has changed the topic to: http://www.w3.org/2009/sparql/wiki/Agenda-2009-10-20 14:00:58 <Zakim> +??P13 14:01:06 <AndyS> zakim, ??P13 is me 14:01:06 <Zakim> +AndyS; got it 14:01:15 <Zakim> +??P14 14:01:19 <AlexPassant> Zakim, ??p14 is me 14:01:19 <Zakim> +AlexPassant; got it 14:01:19 <SteveH_> SteveH_ has joined #sparql 14:01:25 <AndyS> (probably - zakim is slow) 14:01:35 <SteveH_> I dialed in about 1 min ago, anyone see my ID? 14:01:37 <Zakim> + +0186528aaff 14:01:47 <bglimm> Zakim, +0186528aaff is me 14:01:47 <AndyS> zakim, please mute me 14:01:52 <Zakim> +bglimm; got it 14:01:55 <Zakim> AndyS should now be muted 14:01:56 <Zakim> + +049261287aagg 14:01:59 <bglimm> Zakim, mute me 14:02:00 <Zakim> bglimm should now be muted 14:02:08 <SimonS> Zakim, aagg is me 14:02:11 <Zakim> +SimonS; got it 14:02:21 <AndyS> zakim, please unmute me 14:02:21 <Zakim> AndyS should no longer be muted 14:02:35 <AxelPolleres> still waiting for lee for a minute or two, otherwise we might just start. 14:03:21 <Zakim> + +49.238.aahh 14:03:34 <LukeWM> Zakim, who is on the phone? 14:03:36 <Zakim> On the phone I see kasei (muted), pgearon, AxelPolleres, +7.950.aaee, AndyS, AlexPassant, bglimm (muted), SimonS, +49.238.aahh 14:04:18 <SteveH> Zakim, aaee is [Garlik] 14:04:18 <Zakim> +[Garlik]; got it 14:04:27 <SteveH> Zakim, [Garlik] has SteveH and LukeWM 14:04:27 <Zakim> +SteveH, LukeWM; got it 14:04:39 <bglimm> I can scribe 14:04:43 <bglimm> Zakim, unmute me 14:04:43 <Zakim> bglimm should no longer be muted 14:04:48 <OlivierCorby> Zakim, +49.238.aahh is me 14:04:48 <Zakim> +OlivierCorby; got it 14:04:56 <OlivierCorby> Zakim, who is on the phone? 14:04:56 <Zakim> On the phone I see kasei (muted), pgearon, AxelPolleres, [Garlik], AndyS, AlexPassant, bglimm, SimonS, OlivierCorby 14:04:58 <Zakim> [Garlik] has SteveH, LukeWM 14:04:59 <bglimm> Scribe: bglimm 14:05:00 <Zakim> +EricP 14:05:10 <bglimm> Topic: Adiministrative 14:05:22 <LeeF> LeeF has joined #sparql 14:05:27 <AxelPolleres> PROPOSED: approve http://www.w3.org/2009/sparql/meeting/2009-10-13 14:05:28 <bglimm> s/Adiministrative/Administrative/ 14:05:42 <AxelPolleres> RESOLVED: approved http://www.w3.org/2009/sparql/meeting/2009-10-13 14:05:42 <bglimm> AxelPolleres: Suggest to approve last week's minutes 14:05:59 <Zakim> +??P34 14:06:01 <Zakim> -??P34 14:06:05 <bglimm> ... next week might be difficult for a meeting because of ISWC 14:06:12 <chimezie> chimezie has joined #sparql 14:06:15 <AndyS> Regrets for next week. 14:06:20 <bglimm> ... people will try to participate from there 14:06:36 <Zakim> +??P34 14:06:43 <KjetilK> Zakim, ??P34 is me 14:06:43 <Zakim> +KjetilK; got it 14:06:48 <KjetilK> Zakim, mute me 14:06:48 <Zakim> KjetilK should now be muted 14:06:52 <bglimm> LeeF: We should still have a call. It is the lastbefore F2F. 14:07:05 <bglimm> Topic: F2F 14:07:06 <Zakim> + +1.216.636.aaii 14:07:16 <Zakim> +SteveB 14:07:17 <chimezie> Zakim 636.aaii is me 14:07:28 <LukeWM> q+ 14:07:30 <bglimm> AxelPolleres: seems we are a small group, but some people wish to participate remotely 14:07:34 <chimezie> Zakim, +1.216.636.aaii is me 14:07:34 <Zakim> +chimezie; got it 14:07:57 <chimezie> Zakim, mute me 14:07:57 <Zakim> chimezie should now be muted 14:07:59 <Zakim> -EricP 14:08:00 <bglimm> LeeF: we will only have normal telecon facilities 14:08:11 <bglimm> s/LeeF/EricP/ 14:08:41 <LeeF> zakim, who's here? 14:08:41 <Zakim> On the phone I see kasei (muted), pgearon, AxelPolleres, [Garlik], AndyS, AlexPassant, bglimm, SimonS, OlivierCorby, KjetilK (muted), chimezie (muted), SteveB 14:08:44 <Zakim> [Garlik] has SteveH, LukeWM 14:08:45 <Zakim> On IRC I see chimezie, LeeF, SteveH, SimonS, AndyS, Zakim, RRSAgent, pgearon, AxelPolleres, OlivierCorby, ivan, LukeWM, bglimm, KjetilK, kjetil, AlexPassant, kasei, karl, iv_an_ru, 14:08:47 <Zakim> ... ericP, trackbot 14:08:50 <LeeF> zakim, SteveB has LeeF, ericP 14:08:50 <Zakim> +LeeF, ericP; got it 14:08:51 <bglimm> AxelPolleres: I hope everybody has rejoined the WG by now. 14:09:04 <bglimm> .... if not, please do so. 14:09:05 <sandro> sandro has joined #sparql 14:09:29 <bglimm> ... There is news for the team contacts 14:09:32 <AndyS> q? 14:09:46 <bglimm> EricP: Sandro will take over for me. Ivan will be continuing 14:09:53 <SimonS> q+ 14:10:09 <sandro> zakim, what is the code? 14:10:09 <Zakim> the conference code is 77277 (tel:+1.617.761.6200 tel:+33.4.89.06.34.99 tel:+44.117.370.6152), sandro 14:10:12 <LeeF> Regarding F2F agenda - please see and reply to http://lists.w3.org/Archives/Public/public-rdf-dawg/2009OctDec/0201.html 14:10:16 <Zakim> +Sandro 14:10:39 <bglimm> AxelPolleres: Thanks from all of us to EricP for the work he has done. 14:10:42 <AndyS> Thanks to EricP (DAWG and SPARQL) 14:10:51 <bglimm> ... Sandro also joinded the call. 14:11:27 <bglimm> Sandro: I should be at the second dy of the F2F 14:11:46 <LeeF> ack LukeWM 14:11:51 <AxelPolleres> birte, i can take over scribing 14:11:57 <bglimm> Luke: I was wondering what remotely means for the F2F 14:12:04 <AxelPolleres> scribe: Axel Polleres 14:12:15 <bglimm> Scribe: AxelPolleres 14:12:24 <bglimm> Zakim, unmute me 14:12:24 <Zakim> bglimm was not muted, bglimm 14:12:31 <bglimm> Zakim, mute me 14:12:31 <Zakim> bglimm should now be muted 14:12:43 <AndyS> US is still daylight time? 14:12:46 <AndyS> EU is not. 14:12:49 <AxelPolleres> Lee: dialing into west coast might be challenging. we will see how we can arrange the schedule. 14:13:25 <AxelPolleres> ... we should gather as many issues as possible now, to put the important things in the morning slot, to enable some overlap. 14:14:08 <AxelPolleres> discussion about daylght saving time... 14:14:12 <pgearon> November 1 14:14:24 <AxelPolleres> Axel: think next week in Europe, one but next week US 14:14:28 <pgearon> Daylight time ends on Sunday, November 1 at 2 a.m. 14:14:40 <LeeF> ack SimonS 14:15:34 <AxelPolleres> SimonS: too many issues lately, axel convinced me to stay until FPWD, but will take a break until spring then. 14:15:48 <AxelPolleres> ... hope to rejoin then, will be at ISWC. 14:16:02 <AxelPolleres> Lee: thanks, paul will continue as update editor. 14:16:22 <pgearon> :-) 14:16:27 <AxelPolleres> Paul: have CVS access now, don't see a problem with that. 14:16:39 <chimezie> Zakim, unmute me 14:16:39 <Zakim> chimezie should no longer be muted 14:16:56 <AxelPolleres> ericP: 3 more days to register for TPAC, costs 50$ per day. 14:17:13 <AxelPolleres> http://www.w3.org/2009/11/TPAC/ 14:17:34 <ivan> ivan has left #sparql 14:17:42 <AxelPolleres> eric: if you are late, more expensive and unsure 14:18:22 <AxelPolleres> topic: liaison 14:18:36 <AxelPolleres> Lee: we'll meet hcls at TPAC 14:19:01 <AxelPolleres> eric: we'll care about federation, path expression 14:19:30 <AxelPolleres> simons: I had some chat with hcls some time ago, can give a summary 14:20:08 <AxelPolleres> ... they are split into subgroups. results are 5-6 different knowledge bases, which need to be connected (by federation) 14:20:09 <pgearon> +q 14:20:16 <AndyS> q+ to ask that this be described in email 14:20:28 <AxelPolleres> ... will probably need more than what we currently have under basic federation. 14:20:42 <AxelPolleres> paul: where are we going with federation? 14:21:06 <AxelPolleres> lee: basic federation is a time-allowed feature, no actual meetings yet. 14:21:14 <LeeF> ack pgearon 14:21:24 <AxelPolleres> paul: this is something I'd want to get involved with. 14:21:58 <SimonS> +1 Paul 14:22:02 <LeeF> ack AndyS 14:22:02 <Zakim> AndyS, you wanted to ask that this be described in email 14:22:08 <AxelPolleres> ... passing bindings between queries (paul pls summarise) 14:22:25 <pgearon> I'd like to summarize in an email, if that's OK 14:22:34 <AxelPolleres> andy: would be good to have the requirements written up. 14:23:33 <AxelPolleres> action: paul to summarize federated querying in HCLS together with simonS in email thread 14:23:34 <trackbot> Created ACTION-124 - Summarize federated querying in HCLS together with simonS in email thread [on Paul Gearon - due 2009-10-27]. 14:23:49 <ericP> -> http://swobjects.svn.sourceforge.net/viewvc/swobjects/trunk/tests/7tm_receptors/flat/q.rq example of HCLS federation (using GRAPH instead of SERVICE but that will change soon) 14:23:54 <AxelPolleres> topic: fpwd 14:24:02 <AxelPolleres> topic: query 14:24:08 <AxelPolleres> lee: query all set 14:24:12 <AxelPolleres> q+ 14:24:17 <AxelPolleres> q- 14:24:44 <AxelPolleres> lee: all documents need to be consistent with the new naming scheme, we do that in the process of applying pubrules with eric. 14:25:10 <LeeF> http://www.w3.org/2009/sparql/docs/update-1.0/ 14:25:11 <AxelPolleres> topic: update 14:25:26 <AxelPolleres> http://www.w3.org/2009/sparql/docs/update-1.0/ 14:25:41 <SimonS> q+ 14:25:45 <AxelPolleres> luke: should we also move the urls in CVS? 14:26:01 <AxelPolleres> Lee: can we move that? 14:26:05 <AxelPolleres> Eric: can be done 14:26:27 <LeeF> ACTION: Lee and EricP to move update editor's draft to -1.1 URL 14:26:27 <trackbot> Created ACTION-125 - And EricP to move update editor's draft to -1.1 URL [on Lee Feigenbaum - due 2009-10-27]. 14:26:28 <AndyS> (Or remove the version completely in case ... :-() 14:26:35 <AxelPolleres> Paul: all updated in CVS 14:26:43 <SimonS> q- 14:27:14 <AxelPolleres> Luke: small comments, but nothing critical. section 4.1 has a note which might be changed into an issue. 14:27:38 <AxelPolleres> SimonS: think I changed all into issues this morning. 14:27:40 <LukeWM> http://www.w3.org/2009/sparql/docs/update-1.0/#sec_examples 14:28:29 <AxelPolleres> SimonS: I changed the note to refer to section 2.1, in section 2.1 you have four related issues... 14:28:41 <AxelPolleres> Luke: ok, that's fine for me. 14:29:21 <AxelPolleres> Luke:other thing, mentioned graph graph patterns which weren't there 14:29:34 <AxelPolleres> SimonS: it is actually there , production 25 14:30:25 <AxelPolleres> Lee: the "red stuff" (stylesheet related) will be fixed today 14:31:02 <AndyS> q+ about XSLT 14:31:18 <AxelPolleres> ... which is a result of XSL not being complete. 14:31:37 <LeeF> I'd also like us to end up with a single XSLT, hoping to do that 14:32:03 <LeeF> ack about 14:32:04 <LeeF> ack XSLT 14:32:17 <AxelPolleres> Eric: if you find something wrong in the XSL, let us know, you'll catch us on IRC the next few hours. 14:32:20 <chimezie> Zakim, mute me 14:32:20 <Zakim> chimezie should now be muted 14:33:02 <AxelPolleres> Andy: you change the XSL? 14:33:10 <AxelPolleres> Lee: yes. 14:33:20 <AndyS> Query document: XSLT: <b> <cite> and tables / boxes 14:33:29 <AxelPolleres> Andy: i see some things removed, e.g. tables not having borders, etc. 14:34:03 <bglimm> +1 to Andy 14:34:47 <AxelPolleres> Andy: we should have one style across all documents.... which is actually a good intent in the new W3C styles. 14:35:01 <ericP> those cursed systems folks have raised expectations 14:35:31 <bglimm> I would like to have elements for query and results in examples 14:35:40 <AxelPolleres> Lee: let us know if you need some special items in the XSL 14:36:06 <AxelPolleres> topic: protocol 14:36:17 <LukeWM> q+ 14:36:21 <AxelPolleres> http://www.w3.org/2009/sparql/docs/protocol-1.1/ 14:36:39 <AxelPolleres> Lee: alex/andy pointed some minor things... 14:36:42 <LeeF> ack LukeWM 14:36:57 <AxelPolleres> ... I did some global search on "SOAP" and tackled that. 14:37:08 <AxelPolleres> Luke: some more remark on update. 14:37:16 <LeeF> http://www.w3.org/2009/sparql/track/issues/22 14:37:40 <AxelPolleres> ... issue-22 doesn't belong to update. 14:38:05 <LukeWM> pgearon, it's in section 4.1 14:38:05 <AxelPolleres> Lee: paul, can you remove ISSUE-22 from update. 14:38:33 <AxelPolleres> Lee: before closing that issue... 14:39:05 <AxelPolleres> ... I want to have a call where we rush through issues, once we have FPWD out. 14:39:19 <AxelPolleres> ... I saw that simon/kjetil created issues. 14:39:34 <sandro> (I'm used to anyone can PROPOSE and issue, but only the chairs can declare it OPEN.) 14:39:45 <ericP> ISSUE+ don't abuse issues 14:39:56 <AxelPolleres> ... we might want to have a policy on that, please consider not adding issues, rather let's create/manage issues through chairs. 14:39:57 <chimezie> Zakim, unmute me 14:39:57 <Zakim> chimezie should no longer be muted 14:40:26 <AxelPolleres> topic: http-update 14:40:27 <AxelPolleres> http://www.w3.org/2009/sparql/docs/http-rdf-update/ 14:40:33 <AxelPolleres> chime: no issues here 14:40:33 <kasei> Zakim, unmute me 14:40:33 <Zakim> kasei should no longer be muted 14:40:41 <AxelPolleres> lee: all set, thanks to chime. 14:40:52 <AxelPolleres> topic: service description 14:40:52 <kasei> http://www.w3.org/2009/sparql/docs/service-description-1.1/xmlspec.xml 14:41:36 <AxelPolleres> q+ 14:41:48 <AlexPassant> q+ to ask about http://www.w3.org/2009/sparql/service-description# being online before FPWD 14:42:12 <AxelPolleres> q+ to ask about http://www.w3.org/2009/sparql/track/issues/43 14:42:19 <pgearon> Lee, I've removed issue 22 14:42:27 <LeeF> ack AxelPolleres 14:42:27 <Zakim> AxelPolleres, you wanted to ask about http://www.w3.org/2009/sparql/track/issues/43 14:42:30 <LeeF> thanks, pgearon 14:42:42 <KjetilK> should the document link to ISSUE-44? 14:43:28 <LukeWM> which document, KjetilK ? 14:43:38 <KjetilK> LukeWM, the HTTP Update 14:43:45 <LukeWM> ah, OK 14:43:50 <bglimm> Should I add that issue to the entailments doc then? 14:43:55 <AxelPolleres> Lee: if we flag ISSUE-43 now in SD, it might be confusing 14:44:03 <bglimm> Ok 14:44:05 <AxelPolleres> Axel: fair enough 14:44:22 <bglimm> Zakim, unmute me 14:44:22 <Zakim> bglimm should no longer be muted 14:44:29 <AlexPassant> LeeF: i'm on the queue :) 14:44:33 <kasei> Zakim, mute me 14:44:33 <Zakim> kasei should now be muted 14:44:39 <AxelPolleres> Lee: birte, would be good to have the issue in the entailment doc. 14:44:44 <LeeF> ack AlexPassant 14:44:44 <Zakim> AlexPassant, you wanted to ask about http://www.w3.org/2009/sparql/service-description# being online before FPWD 14:45:23 <AlexPassant> either put the vocab online or mention that it's not here *yet* 14:45:25 <AlexPassant> yes LeeF 14:45:39 <kasei> good catch. should also have agreement on the iri (?) 14:45:39 <AxelPolleres> Alex: can we get the namespace http://www.w3.org/2009/sparql/service-description# online before FPWD? 14:45:56 <AxelPolleres> Eric: we could put a placeholder document 14:46:03 <AlexPassant> yes 14:46:06 <kasei> Zakim, unmute me 14:46:06 <Zakim> kasei should no longer be muted 14:46:30 <AxelPolleres> greg: I have just couned that URI, is it ok wirth everyone? 14:47:23 <AxelPolleres> ... is that inline with W3C naming convention? 14:47:23 <sandro> this is a namespace? for XML or RDF? 14:48:00 <sandro> www.w3.org/ns/service# 14:48:11 <sandro> www.w3.org/ns/sparql-service# 14:48:48 <sandro> yeah, I didn;'t mean to leave out "sparql" 14:49:35 <AxelPolleres> andyS: maybe better to not block the namespace now, since it gives it more weight, we might not want that, unless we are sure about the namespace. 14:49:46 <sandro> There MUST be a document at the namespace URI if the namespace is used in a TR. 14:50:14 <sandro> yes, it can say "subject to change, etc" 14:50:24 <AxelPolleres> sandro: pubrules prsecribe that there has to be a document at any namespace mentioned. 14:50:27 <sandro> yes 14:50:54 <sandro> and note in the Spec where the NS is mentions. 14:50:55 <AxelPolleres> Lee: anybody unhappy with www.w3.org/ns/sparql-service# ? 14:51:02 <SteveH> +1 to +description 14:51:04 <LeeF> sparql-service# vs. sparql-service-description# 14:51:08 <SteveH> it's a little confusing otherwise 14:51:13 <AxelPolleres> greg: prefer service description. 14:51:26 <sandro> "description" seems redundant to me, but I'm new. 14:51:34 <AxelPolleres> +1 to include description 14:51:37 <ericP> +1 to apathy 14:51:45 <SteveH> +0 to apathy 14:52:04 <AxelPolleres> Lee: www.w3.org/ns/sparql-service-description# be it 14:52:07 <LeeF> ACTION: Eric and Lee to request .../ns/sparql-service-description# and put a "subject to change" doc there 14:52:07 <trackbot> Created ACTION-126 - And Lee to request .../ns/sparql-service-description# and put a "subject to change" doc there [on Eric Prud'hommeaux - due 2009-10-27]. 14:52:27 <AxelPolleres> eric: HTML for now there. 14:52:49 <AxelPolleres> ... anybody objecting to conneg there? 14:53:01 <sandro> html for now, RDFa and/or conneg later 14:53:06 <kasei> LeeF, will you change that ns in the doc when you are making changes today, or should I? 14:53:07 <AxelPolleres> sandro: we can do RDFa later on 14:53:08 <AndyS> if RDFa, then RDF as well please. 14:53:16 <LeeF> ACTION: Lee to update NS in serv. description document 14:53:16 <trackbot> Created ACTION-127 - Update NS in serv. description document [on Lee Feigenbaum - due 2009-10-27]. 14:53:17 <sandro> plus 6 grddl transforms. :-) 14:53:32 <kasei> Zakim, mute me 14:53:32 <Zakim> kasei should now be muted 14:53:37 <AxelPolleres> topic: entailment document 14:53:42 <AndyS> OK if server side ;-) 14:53:52 <AxelPolleres> birte: all in CVS, looks alright for me. 14:54:03 <AxelPolleres> ... not sure how it looks for everyone. 14:54:06 <LeeF> to set access go to http://URI/URI/,access 14:54:23 <AxelPolleres> eric: you should be able to set .access 14:54:29 <bglimm> http://www.w3.org/2009/sparql/docs/entailment/xmlspec.xml 14:55:26 <sandro> I see it in CVS. 14:55:29 <AxelPolleres> birte: might be a commit issue. 14:55:41 <AxelPolleres> eric: let's pick that up after the call on irc. 14:56:02 <AndyS> I see it in CVS. last update Oct 20 15:42 14:56:11 <AxelPolleres> lee: did you address all the issues we talked about? 14:56:50 <AxelPolleres> birte: yes, but might need more reviewers, removed entailment regimes we haven't really worked on, just left RDF/RDFS 14:57:03 <AxelPolleres> lee: should be ok. 14:57:07 <sandro> 7 years bad luck 14:57:32 <AxelPolleres> topic: overall fpwd issues 14:58:20 <AxelPolleres> eric: we will turn around a lot at grammar, how many docs have grammar in them? 14:58:43 <sandro> weird. the mirrors say http://www.w3.org/2009/sparql/docs/entailment doesn't exist, but when I did a an update -d on 2009/sparql it was created for me, with contents. 14:58:46 <LeeF> Regrets: IvanH, Orri 14:58:53 <bglimm> Zakim, mute me 14:58:53 <Zakim> bglimm should now be muted 14:59:12 <AxelPolleres> andy/eric discussing grammar... 15:00:10 <AxelPolleres> Andy: query isn't xmlspec'ed yet. is in HMTL, should be ok. 15:00:22 <AxelPolleres> ... more important things than styling so far. 15:01:05 <SteveH> +1 to aggregates 15:01:10 <AndyS> +1 15:01:12 <AxelPolleres> Lee: 60mins up, would like to continue shortly about the aggregate issues. 15:01:15 <AxelPolleres> +1 15:01:31 <AxelPolleres> andy: we may not agree, but let's start the issues. 15:01:50 <SteveH> not really , no 15:01:50 <AxelPolleres> Lee: almost all include we should include count. 15:01:52 <Zakim> -Sandro 15:02:01 <AxelPolleres> ... do we have a starting point for that? 15:02:24 <AxelPolleres> Andy: shall we start with the relational one? 15:02:25 <AndyS> COUNT, MIN, MAX, SUM 15:02:33 <LukeWM> AVG? 15:02:56 <LeeF> GROUP_CONCAT 15:03:09 <LeeF> SAMPLE 15:03:09 <AxelPolleres> basically any XPATH/XQuery funct that accepts sequences could be allowed? 15:03:10 <AndyS> StsveH - can you define it in email? Still hazy on details (prob happy with it) 15:03:30 <pgearon> first? 15:03:33 <SteveH> AndyS, sure 15:03:34 <AndyS> XP/XQ has issues in the details 15:03:37 <AxelPolleres> e.g. string-join 15:03:41 <AndyS> SteveH, ta 15:03:46 <LeeF> STDDEV 15:04:01 <AxelPolleres> lee: statistics, e.g. STDDEV 15:04:17 <ericP> my pref would have been COUNT, MIN, MAX, SUM, AVG, but i think a system that has debugging from the start is a very good developer draw 15:04:28 <ericP> so also like SAMPLE 15:04:29 <pgearon> we'd be a laughing stock if we didn't include COUNT 15:04:30 <LukeWM> I'd like AVG to be called something different. 15:04:48 <AxelPolleres> COUNT seems to be agreed. 15:05:13 <pgearon> +1 for MIN/MAX (mixed types aside) 15:05:41 <AxelPolleres> axel: MIN/MAX don't add expressivity, but very useful 15:06:04 <AxelPolleres> andyS: actually they do add: usable in GROUP BY 15:06:42 <AxelPolleres> q+ 15:07:21 <LeeF> ack AxelPolleres 15:07:34 <AxelPolleres> SUM also sounds like it has support 15:07:49 <AndyS> q+ 15:07:52 <AxelPolleres> AVG might need to be named different. 15:07:53 <chimezie> q+ consensus on general, expected behavior of aggregate functions 15:08:00 <LeeF> ack AndyS 15:08:04 <chimezie> q+ 15:08:07 <LeeF> ack consensus 15:08:07 <Zakim> consensus, you wanted to comment on general, expected behavior of aggregate functions 15:08:17 <AxelPolleres> Lee: different naming was suggested by Luke 15:08:19 <LeeF> q+ chimezie to discuss consensus on general, expected behavior of aggregate functions 15:08:27 <AxelPolleres> Luke: better call it "mean" 15:08:43 <SteveH> can have AVG as an alias for MEAN 15:08:59 <SteveH> a la isURI and isIRI 15:09:01 <AxelPolleres> Lee: AVG would assume same behaviour as in SQL 15:09:16 <pgearon> MEAN implies a numeric operation, while people might try to apply AVG to other things... but what's the average of "foo" and "bar"? :-) 15:09:44 <kasei> pgearon, probably depends on the current text encoding :) 15:09:46 <AxelPolleres> Lee: would lean towards AVG 15:10:04 <pgearon> +1 for aggregates 15:10:18 <AxelPolleres> either both SUM and AVG or no arithmetic at all 15:10:32 <pgearon> I mean, for and "average" aggregate 15:10:50 <LeeF> ISSUE: Should the average/arithmetic mean aggregate be called AVG, MEAN, both, or something else? 15:10:50 <trackbot> Created ISSUE-50 - Should the average/arithmetic mean aggregate be called AVG, MEAN, both, or something else? ; please complete additional details at http://www.w3.org/2009/sparql/track/issues/50/edit . 15:11:27 <LukeWM> no 15:11:33 <kasei> 0 15:11:35 <SteveH> STDDEV is quite useful 15:11:41 <AxelPolleres> Lee: Excel has a core set that includes SDTDEV, does anybody think we should include that? 15:12:12 <AxelPolleres> ... extensibility should enable including these. 15:12:40 <AxelPolleres> Lee: GROUPCONCAT very much like string-join 15:12:53 <AxelPolleres> ... we use it all the time. 15:13:02 <AndyS> Name is horrible! 15:13:09 <kasei> is the "GROUP" in that just to distinguish it as an aggregate function? 15:13:22 <SteveH> kasei, yes 15:13:26 <AxelPolleres> ... no strong feeling whether it should belong to thee core. 15:13:29 <SteveH> the name is horrible, yes 15:13:46 <KjetilK> +1 15:13:48 <SteveH> +1 15:13:55 <AndyS> 0 15:14:02 <LeeF> ack chimezie 15:14:02 <Zakim> chimezie, you wanted to discuss consensus on general, expected behavior of aggregate functions 15:14:02 <AxelPolleres> ... I am not wedded to the name "GROUPCONCAT", aside that would we want to include it? 15:14:10 <kasei> AndyS, seems tricky without the ability to escape the underlying values first 15:14:17 <SteveH> defn, from mysql, FWIW: http://dev.mysql.com/doc/refman/5.0/en/group-by-functions.html#function_group-concat 15:14:34 <AndyS> escaping? 15:14:40 <kasei> for csv 15:14:40 <AxelPolleres> chime: shouldn't we rather think of criteria of what a well-behaved function should look like? 15:14:44 <AndyS> StevbeH - does that escape sep? 15:14:59 <AxelPolleres> ... than discussing the single aggregates? 15:15:05 <SteveH> AndyS, no 15:15:12 <SteveH> not in mysql anyway 15:15:18 <AndyS> NP - add a fucntion call. 15:15:21 <SteveH> yup 15:15:27 <AndyS> q+ 15:15:36 <AxelPolleres> Lee: we aren't even sure yet whether we want aggregates to define single return values. 15:15:38 <LeeF> ack AndyS 15:16:02 <AxelPolleres> AndyS: I already have a definition which ties it to a single value. 15:16:33 <Zakim> -chimezie 15:16:53 <AxelPolleres> Lee: will the issues we have discussed today go into the FPWD? 15:16:57 <AndyS> -99 to definition. 15:17:26 <AxelPolleres> Steve: already a list of candidates there, which we might extend, no definition at this point. 15:17:41 <AxelPolleres> Lee: let's continue the list ... 15:17:47 <AxelPolleres> ... SAMPLE. 15:17:48 <AndyS> q+ 15:17:56 <SteveH> SAMPLE is neccesary if we don't allow aggregate values to be projected, which had support previously 15:18:18 <AxelPolleres> q+ worried about non-determinism (as usual). 15:18:20 <AndyS> ack me 15:18:23 <LeeF> ack AndyS 15:18:46 <AndyS> ah - "non aggregated variables" ?? 15:18:53 <pgearon> what about just using FIRST? 15:18:58 <AxelPolleres> Lee: axel, yes it is nondeterministic, but it is very explcicitly non-deterministic. 15:19:09 <pgearon> +q 15:19:15 <AndyS> A change to design but NP. 15:19:17 <LeeF> ack pgearon 15:19:35 <AndyS> q+ 15:19:54 <AxelPolleres> Paul: FIRST is fine in conncection with order by. 15:20:03 <SteveH> q+ 15:20:06 <LeeF> ack AndyS 15:20:14 <AxelPolleres> ... without order by SAMPLE is just the same as FIRST. 15:20:25 <SteveH> q- 15:21:08 <AxelPolleres> Andy: seems there are different assumptions here, don't see how you can sort the variabes coming out of a group by. 15:21:21 <SteveH> in SQL yu can do AGGREGATE (?x ORDER BY ...) 15:21:30 <AndyS> MIN or MAX nearly= FIRST? 15:21:31 <AxelPolleres> paul: honestly, it's a while since I chekced that in SQL 15:22:40 <SteveH> action? 15:22:57 <LeeF> PROPOSED: To include aggregates for COUNT, MIN, MAX, SUM, AVG and discuss others down the road 15:23:04 <SteveH> seconded 15:23:15 <bglimm> +1 15:23:17 <AndyS> +1 15:23:18 <pgearon> +1 15:23:29 <LeeF> RESOLVED: To include aggregates for COUNT, MIN, MAX, SUM, AVG and discuss others down the road, no abstentions or objections 15:23:35 <iv_an_ru> +1 15:24:03 <AxelPolleres> +0.95 (guess we can get around the involved issues, so fair enough) 15:24:30 <AxelPolleres> Lee: let's close... some more hint for those at ISWC 15:25:09 <AxelPolleres> ... we have a panel at ISWC, Wendesday, Oct 28, 10:30 15:25:34 <AxelPolleres> AndyS: When will the agenda go out for F2F, was due. 15:25:49 <AxelPolleres> LeeF: aggregate issues, for sure. 15:25:58 <AxelPolleres> ... stil soliciting issues. 15:26:19 <AxelPolleres> Lee, Axel to do the agenda by end of the week. 15:26:31 <AndyS> Tues OK : Mon - not first thing 15:27:15 <LeeF> adjourned 15:27:15 <bglimm> Zakim, unute me 15:27:15 <Zakim> I don't understand 'unute me', bglimm 15:27:19 <bglimm> yes 15:27:21 <Zakim> -KjetilK 15:27:26 <bglimm> Zakim, unmute me 15:27:26 <Zakim> bglimm should no longer be muted 15:27:27 <AxelPolleres> Axel: suggest we collect issues by end of the week and on next call discuss how to schedule them best 15:27:27 <Zakim> -SimonS 15:27:28 <Zakim> -kasei 15:27:30 <Zakim> -AndyS 15:27:32 <AndyS> Thanks all. 15:27:33 <Zakim> -AlexPassant 15:27:40 <AndyS> Was not me @AndyS then! 15:27:44 <Zakim> -OlivierCorby 15:27:44 <AxelPolleres> rrsagent, make records public 15:27:45 <Zakim> -[Garlik] 15:27:53 <AxelPolleres> Zakim, who was on the phone? 15:27:53 <Zakim> I don't understand your question, AxelPolleres. 15:28:00 <AxelPolleres> Zakim, list attendees 15:28:00 <Zakim> As of this point the attendees have been +1.518.276.aabb, +3539149aacc, pgearon, AxelPolleres, +0186528aadd, kasei, +7.950.aaee, AndyS, AlexPassant, bglimm, +049261287aagg, SimonS, 15:28:03 <Zakim> ... SteveH, LukeWM, OlivierCorby, EricP, KjetilK, chimezie, LeeF, Sandro 15:28:19 <pgearon> ericP, can you let me know when the renaming and other updates are done please? I'll update cvs once you let me know 15:29:21 <AxelPolleres> rrsagent, make records public # SPECIAL MARKER FOR CHATSYNC. DO NOT EDIT THIS LINE OR BELOW. SRCLINESUSED=00000498