13:47:21 RRSAgent has joined #gld 13:47:21 logging to http://www.w3.org/2012/07/26-gld-irc 13:47:23 RRSAgent, make logs world 13:47:23 Zakim has joined #gld 13:47:25 Zakim, this will be GLD 13:47:25 ok, trackbot; I see T&S_GLDWG()10:00AM scheduled to start in 13 minutes 13:47:26 Meeting: Government Linked Data Working Group Teleconference 13:47:26 Date: 26 July 2012 13:47:35 George has joined #gld 13:50:05 trackbot, start telecon 13:50:08 RRSAgent, make logs world 13:50:10 Zakim, this will be GLD 13:50:10 ok, trackbot; I see T&S_GLDWG()10:00AM scheduled to start in 10 minutes 13:50:11 Meeting: Government Linked Data Working Group Teleconference 13:50:11 Date: 26 July 2012 13:56:32 HadleyBeeman has joined #gld 13:56:59 mhausenblas has joined #gld 13:57:09 PhilA has joined #gld 13:57:36 zakim, this will be GLD 13:57:36 ok, PhilA; I see T&S_GLDWG()10:00AM scheduled to start in 3 minutes 13:57:48 Hi, PhilA! 13:57:56 hi Hadley. 13:58:04 I'm with Rufus & co this aftrernoon 13:58:10 How are things? I feel like I haven't seen you for ages. 13:58:17 Oh, hello to Rufus and co as well! 13:58:21 DaveReynolds has joined #gld 13:58:29 It's been a while. Things OK - looking forward to hols v soon 13:58:37 BenediktKaempgen has joined #gld 13:58:46 T&S_GLDWG()10:00AM has now started 13:58:53 +??P5 13:58:54 +[IPcaller] 13:59:03 Zakim, ??P5 is me 13:59:03 +BartvanLeeuwen; got it 13:59:03 zakim, code? 13:59:04 the conference code is 45394 (tel:+1.617.761.6200 sip:zakim@voip.w3.org), PhilA 13:59:06 zakim, ?p5 is me. 13:59:06 sorry, olyerickson, I do not recognize a party named '?p5' 13:59:08 +GeraldSteeman 13:59:13 +fadmaa 13:59:24 zakim, ??p5 is me. 13:59:24 I already had ??P5 as BartvanLeeuwen, olyerickson 13:59:29 fadmaa has joined #gld 13:59:31 +George_Thomas 13:59:44 Zakim, who is here 13:59:44 fadmaa, you need to end that query with '?' 13:59:45 +BenediktKaempgen 13:59:50 Zakim, who is here? 13:59:50 On the phone I see BartvanLeeuwen, [IPcaller], GeraldSteeman, fadmaa, George_Thomas, BenediktKaempgen 13:59:50 zakim, who is on the phone 13:59:52 On IRC I see fadmaa, BenediktKaempgen, DaveReynolds, PhilA, mhausenblas, HadleyBeeman, George, Zakim, RRSAgent, MacTed, BartvanLeeuwen, martinAlvarez, olyerickson, danbri, 13:59:52 ... trackbot, sandro 13:59:52 I don't understand 'who is on the phone', olyerickson 14:00:07 zakim, who is on the phone? 14:00:07 On the phone I see BartvanLeeuwen, [IPcaller], GeraldSteeman, fadmaa, George_Thomas, BenediktKaempgen 14:00:15 Zakim, mute me 14:00:15 fadmaa should now be muted 14:00:18 zakim, I am ipcaller 14:00:18 ok, olyerickson, I now associate you with [IPcaller] 14:00:25 +??P13 14:00:34 +??P11 14:00:42 +HadleyBeeman 14:00:50 GeraldSteeman has joined #GLD 14:00:52 zakim, who's here? 14:00:52 On the phone I see BartvanLeeuwen, [IPcaller], GeraldSteeman, fadmaa (muted), George_Thomas, BenediktKaempgen, ??P13, ??P11, HadleyBeeman 14:00:54 On IRC I see GeraldSteeman, fadmaa, BenediktKaempgen, DaveReynolds, PhilA, mhausenblas, HadleyBeeman, George, Zakim, RRSAgent, MacTed, BartvanLeeuwen, martinAlvarez, olyerickson, 14:00:54 ... danbri, trackbot, sandro 14:00:57 zakim, ??p13 is me 14:00:59 +martinAlvarez; got it 14:01:07 +??P16 14:01:12 scribe: olyerickson 14:01:14 zakim, mute me 14:01:17 martinAlvarez should now be muted 14:01:30 +mhausenblas 14:01:40 topic: attendance 14:01:40 cygri has joined #gld 14:01:49 Zakim, cygri is with me 14:01:49 +cygri; got it 14:01:51 +sandro 14:02:00 PhilA is with Rufus et.al. from OKFN 14:02:03 zakim, who am I? 14:02:03 I don't understand your question, PhilA. 14:02:15 -BenediktKaempgen 14:02:17 http://www.w3.org/2011/gld/meeting/2012-06-28 14:02:20 zakim, ??P16 is me 14:02:20 +PhilA; got it 14:02:21 +1 14:02:26 topic: minutes from previous meeting (28 June) 14:02:27 zakim, mute me 14:02:27 PhilA should now be muted 14:03:00 george: minutes accepted 14:03:04 +BenediktKaempgen 14:03:11 topic: new attendees? 14:03:12 zakim, I have Rufus Pollock 14:03:12 I don't understand 'I have Rufus Pollock', PhilA 14:03:14 http://blog.schema.org/2012/07/describing-datasets-with-schemaorg.html is a positive step, indeed - thanks danbri 14:03:25 * no apparent new attendees 14:03:28 +gatemezin 14:03:34 topic: announcements 14:03:42 zakim, PhilA has Rufus 14:03:42 +Rufus; got it 14:03:42 1. PROV in last call 14:04:03 present+ Rufus Pollock 14:04:07 * george as had a look 14:04:12 +1 to going through the provenance work 14:04:17 * would be good for GLD to talk in future 14:04:17 atemezin has joined #gld 14:04:17 zakim, unmute me 14:04:17 PhilA should no longer be muted 14:04:24 topic: DCAT 14:04:41 * thanks to Rufus for feedback/comments 14:04:41 tinagheen has joined #gld 14:04:57 * thanks to Ed Summers (Ed, are you on call)? 14:05:06 +[LC] 14:05:14 Zakim, [LC] is me 14:05:14 +tinagheen; got it 14:05:48 I told Zakim... 14:05:59 ;) 14:06:00 zakim, who is here? 14:06:00 On the phone I see BartvanLeeuwen, [IPcaller], GeraldSteeman, fadmaa (muted), George_Thomas, martinAlvarez (muted), DaveReynolds, HadleyBeeman, PhilA, mhausenblas, sandro, 14:06:03 ... BenediktKaempgen, gatemezin, tinagheen 14:06:03 PhilA has Rufus 14:06:03 mhausenblas has mhausenblas, cygri 14:06:03 On IRC I see tinagheen, atemezin, cygri, GeraldSteeman, fadmaa, BenediktKaempgen, DaveReynolds, PhilA, mhausenblas, HadleyBeeman, George, Zakim, RRSAgent, MacTed, BartvanLeeuwen, 14:06:06 ... martinAlvarez, olyerickson, danbri, trackbot, sandro 14:06:21 rufus & PhilA DCAT 14:07:00 +rsingh2 14:07:06 -> http://spec.datacatalogs.org/ OKFN spec 14:07:19 rufus: "hkh dskjfh *(*)( kjdhfkjd..." 14:08:12 http://lists.w3.org/Archives/Public/public-gld-comments/2012Jul/0016.html 14:08:34 rufus: ??? spec is for harvesting, etc 14:09:14 * two components: spec for catalogs to conform to, protocol for harvesting 14:09:38 * has spoken with Fadi and PhilA 14:09:44 Michael: Just FYI, I've added an data catalogs use case the Linked Data Platform WG Wiki (both ADMS-based and what Rufus is talking about), see -> http://www.w3.org/2012/ldp/wiki/Use_Cases_And_Requirements#Data_catalogs 14:09:54 * has offered suggestions w.r.t. DCAT 14:10:51 * reminder that CKAN is about aggregation, that has been the focus for CKAN spec (and protocol) 14:11:23 PhilA: hopes we can hear (barely...) 14:11:34 * to keep conversation short... 14:11:56 * Rufus & PhilA have thought about difference in purpose between specs 14:12:27 * Particular desire with CKAN spec to specify "required" vs "optional" 14:13:20 Actually, is it common usage to specify "optional" in vocabs design? 14:13:21 * wrt protocol/API, need to "park" discussion 14:13:46 george: likes "member submission protocol" idea 14:14:00 * not sure about "conformance" question 14:14:25 * other issues that Rufus/PhilA want to jump into? 14:14:46 Rufus: bullet 4: distribution 14:15:14 the current diagram http://dvcs.w3.org/hg/gld/raw-file/default/dcat/dcat-model.jpg 14:15:14 * superclass vs subclass 14:16:12 PhilA: API is a way to access...but many datasets aren't available for access that way 14:16:42 https://www.w3.org/2011/gld/track/issues/9 14:16:44 * but makes sense to have "distribution" as a resource that can be elaborated upon 14:17:32 * PhilA: in favor of having a resource, and having distribution as a sub-class (specialization) of that 14:17:45 GEorge: Fadi? cygri? 14:17:58 cygri: its complicated...many issues to consider 14:18:08 * forget for a moment history 14:18:15 * forget about current use 14:18:21 * focus on what really makes sense 14:18:37 * agrees that "distribution" is unfortunate 14:18:58 * concept exists on many data catalogs but isn't usually named as such 14:19:26 * origins of name: simply due to the hisoty of CKAN, might not make much sense otherwise 14:19:44 Resource or DatasetResource 14:19:50 (from Rufus) 14:19:53 * w.r.t. "resource," not too happy about the term, too overloaded 14:20:07 * In RDF, resource has a specific meaning 14:20:15 * in webarch, a slightly different maning 14:20:19 Rufus: dcat:DatasetResource ? 14:20:29 * in DCAT, a possible third different maning 14:20:46 * raises the question: how are dataset and resource different? 14:21:10 * w.r.t. webarch, dataset is *already* a resource 14:21:28 * how then are things "under" dataset resources? 14:21:33 Rufus: +1 on API, Download/File etc but we need superclass ... 14:21:38 * May need more specific terms 14:22:38 rufus: this sort of repetition already happens 14:23:01 * remove ambiguity by defining the hierarchy 14:23:12 cygri: how about doing it without the superclass? 14:23:36 fine - what then is the name for the property :-) e.g. dcat:resource or dcat:distribution 14:23:52 we relocate the question to properties from classes ... 14:24:14 cygri: (thinking out loud) only have specific sub-classes, no superclass 14:24:20 * already have shared property 14:24:43 * even if we don't have the superclass, we still have the commonality 14:24:43 but that property should be renamed (sorry not clearer about that ...) 14:24:51 if dcat:Distribution renamed so would property 14:24:52 * makes documentation a bit confusing 14:25:08 * still need to explain/provide guidance 14:25:13 q? 14:25:25 cygri: ditch superclass, rename property to something like dcat:associatedResource? 14:25:27 dcat:distribution -> dcat:resource (or dcat:datasetResource) 14:25:38 cygri: understood - that sounds good ... 14:25:43 (Rufus) 14:26:04 s/cygri:/cygri/ 14:26:50 Rufus: +1 on dcat:associatedResource 14:28:22 Zakim, unmute me 14:28:22 fadmaa should no longer be muted 14:28:32 please repeat last point olyerickson -missed it 14:28:36 did any of that make sense? 14:29:00 fadmaa: one more option 14:29:02 olyerickson: i'm opposed to making any reference to "required" parts of the spec 14:29:16 * accessURL and... 14:29:20 * downloadURL 14:29:41 -rsingh2 14:29:55 * downloadURL used only if provider is sure it is downloadable 14:30:13 rufus: problem with locating addition information 14:31:01 cygri: What is the proposal on the table now? 14:31:16 rufus: leave classes as they are 14:31:36 (sorry that was fadmaa) 14:31:52 * downloadURL points to downloadable dataset file 14:31:53 zakim, who is noisy? 14:32:02 zakim, mute me 14:32:02 PhilA should now be muted 14:32:04 cygri, listening for 10 seconds I heard sound from the following: [IPcaller] (14%), George_Thomas (13%), PhilA (57%), gatemezin (4%) 14:32:23 q+ 14:32:29 * accessURL merely...um...access point 14:32:35 cygri: Doesn't see it 14:32:41 * can't see the distinction 14:33:08 * if we know it's a download, don't see what we would need separate "access" URL 14:33:21 * doesn't think we would want to provide two 14:33:30 * thus why make distinction? 14:33:42 +1 not to making the difference btw "download" and "access"-URL 14:33:48 * seems simpler to have a flag e.g. accessType 14:34:05 * target URL and a type 14:34:14 Rufus is nodding his heada in violent agreement 14:34:19 fadmaa: now that cygri says it that way, not so convinced... 14:34:26 Zakim, mute me 14:34:26 fadmaa should now be muted 14:34:29 ack me 14:34:40 George: Are we back to "associated resource?" 14:35:09 PhilA: +1 on "associated resource" (or is that Rufus?) 14:35:31 Rufus + PhilA are happy 14:35:36 * Rufus and PhilA are happy (together...) 14:35:38 Zakim, unmute me 14:35:38 fadmaa should no longer be muted 14:35:53 zakim, mute me 14:35:53 PhilA should now be muted 14:35:53 ISSUE-9? 14:35:53 ISSUE-9 -- dcat:Distribution and its subclasses are unnecessary -- raised 14:35:53 http://www.w3.org/2011/gld/track/issues/9 14:35:56 George: How does this happiness reflect on our status? 14:36:37 Rufus happy with Richard proposal to abolish superclass and use property 14:36:47 but also happy to keep and rename ... 14:37:09 https://www.w3.org/2011/gld/track/issues/7 14:37:13 cygri: thinks Ed's point under original issue was "access url" could be (directly) the URI of the distribution resource 14:37:22 This is me - cygri is singing my tune :-) 14:37:24 ISSUE-7 14:37:29 ISSUE-7? 14:37:29 ISSUE-7 -- Drop dcat:accessUrl, use the URI of the dcat:Download resource instead -- raised 14:37:29 http://www.w3.org/2011/gld/track/issues/7 14:37:58 fadmaa: confirms: Ed's issue was, why not just use the URI 14:38:01 q? 14:38:10 q+ 14:38:17 ack me 14:39:12 PhilA: Hates conflation of access URI/distribution URI (please explain) 14:40:17 * if URIs are forever, one doesn't need an access URL because URI will suffice 14:40:28 this is Ed's original post: http://www.w3.org/egov/IG/track/issues/38 14:40:32 zakim, mute me 14:40:32 PhilA should now be muted 14:41:29 I have accessURL but it seems to be practical and necessary for a lot of people so, reluctantly, I have to come down on the side of keeping it. Ref. ADMS, chat woth cygri and Rufus today 14:41:36 s/have/hate/ 14:42:03 I hate accessURL but it seems to be practical and necessary for a lot of people so, reluctantly, I have to come down on the side of keeping it. Ref. ADMS, chat woth cygri and Rufus today 14:42:21 zakim, unmute me 14:42:21 PhilA should no longer be muted 14:42:38 zakim, mute me 14:42:38 PhilA should now be muted 14:42:52 Looks like PhilA restated it (and it's how I understood him) 14:43:28 cygri: What do we need a subclass that says "this is a download," etc 14:43:35 q+ 14:43:41 * why can we just describe what kind of a thing it is via properties? 14:43:58 q- 14:44:15 Rufus +1 14:44:22 * thinks it is helpful to have rich set of subclasses to help client "understand" what is available/what can be done 14:45:03 * if there is info attached that say "this is a download" etc is useful 14:45:31 * but is in favor of accomplishing this clarification with structure of subclasses 14:45:55 * bottom line: do it through subclasses or do it with properties? 14:46:15 ack me 14:46:24 George: Back to Rufus & PhilA 14:47:06 rufus: avoid doing it as a hack 14:47:25 * having distinct subclasses is preferred 14:47:50 * happy for superclass to be there, happy for it to go... 14:48:07 Rufus "subclasses" (distinct classes) is a very good idea 14:48:37 so, drop dcat:Distribution superclass; keep subclasses; rename dcat:distribution property to dcat:associatedResource? 14:48:43 happy for super class to go if we retain discuss classes "dcat:Download" etc 14:48:43 PhilA: conscious of time; will try to send email to come to resolution 14:49:01 Rufus +1 to cygri proposal 14:49:02 suggestion: examples might be good 14:49:16 zakim, mute me 14:49:16 PhilA should now be muted 14:49:30 George: Recap and clarification would be appreciated 14:49:56 I would be happy with enything that is clear and is extensible 14:50:08 action: PhilA to summarise discussion and make a proposal for agreement or not 14:50:08 Created ACTION-73 - Summarise discussion and make a proposal for agreement or not [on Phil Archer - due 2012-08-02]. 14:50:14 ack me 14:50:19 olyerickson's concern is adding confusion for adopters 14:50:32 topic: "What this means" (PhilA) 14:50:51 * these vocabs are on the "rec track" 14:51:09 Rufus (from email) Clarify "conformance". Suggest either removal (is it relevant for 14:51:10 vocabs?) or phrasing as: "Conformance means using terms from this 14:51:10 vocabulary whenever they are relevant" 14:51:12 * does it make sense to have a "conformance statement" 14:51:20 * if so, what is it intended to do? 14:51:26 (apologies for line breaks!) 14:51:38 * mostly we mean, "please do not re-invent these terms" 14:52:07 zakim, mute me 14:52:07 PhilA should now be muted 14:52:15 ack me 14:52:16 * For DCAT, is it appropriate to have conformance statement, and of so what should it say? 14:52:32 Isn't conformance about interop? If every application does it's own profile then there's no interop. 14:52:33 George: wants Sandro to weigh in? 14:52:59 -tinagheen 14:53:00 Sandro: Has been thinking, likes Rufus's comments in email earlier 14:53:27 * conformance clauses are there to keep people on track 14:53:47 * the more precise people are, the more helpful 14:53:53 cygri: agrees with Sandro' 14:54:02 q? 14:54:11 * unclear what it means to conform (unlike many other specs) 14:54:28 Q+ 14:54:55 zakim, I am [ipcaller] 14:54:55 ok, olyerickson, I now associate you with [IPcaller] 14:55:25 * can we add conformance/non-conformance in the spec (reasonably)? 14:55:30 * how to make it work? 14:55:44 * much easier with a protocol than e.g. a vocabulary 14:57:27 q? 14:57:34 ack olyerickson 14:57:40 An idea: "The nature of a vocabulary is that each person's use is likely to be different. However, inclusion of a term in this vocabulary is in indication that the community sees it as being important and/or useful! 14:57:48 ack [IPcaller] 14:57:57 zakim, mute me. 14:57:57 [IPcaller] should now be muted 14:58:08 ack me 14:58:16 Sandro: not much experience w.r.t. "comformance" 14:58:40 DCAT conformance means that a provider uses the entire DCAT vocabulary when publishing catalog and dataset metadata. 14:59:01 Rufus: Please explain what DCAT spec means when it (currently) refers to conformance 14:59:10 sandro: I don't know of any cases in the industry where people have had difficulties around conformance with a vocabulary, so it's hard to know what problems to address. 14:59:33 -BartvanLeeuwen 14:59:43 * DCAT spec currently says, conformance means using it all. 14:59:44 q? 14:59:58 +1 to rufus, the current wording is not good 14:59:59 zakim, mute me 15:00:00 PhilA should now be muted 15:00:08 George: We have a lot to work on w.r.t. "conformance" 15:00:19 BartvanLeeuwen has left #gld 15:00:26 * protocol usage might inform/guide 15:00:36 topic: Wrap Up 15:00:45 q+ 15:00:49 * thanks to olyerickson for typing 15:00:56 Rufus thank-you! 15:01:04 * thanks to Rufus & PhilA for recent work on this 15:01:08 I'll send some follow up mails/issues 15:01:08 -DaveReynolds 15:01:10 zakim, unmute me 15:01:11 Thanks! 15:01:12 martinAlvarez should no longer be muted 15:01:14 * thanks everyone for productive call 15:01:17 +q 15:01:25 martinAlvarez: Remark (informative) 15:01:31 thanks for joining Rufus! 15:01:43 Thanks Rufus for your comments... 15:01:45 * currently in charge of recommendation for Spanish govt 15:01:53 * al levels of govt 15:02:08 * guidance on publishing govt data using DCAT 15:02:19 * would like it in a stable format 15:02:39 martinAlvarez, what's the timeline on this? 15:02:42 q? 15:02:43 George: Encouraged to collaborate on GLD cubmission 15:02:45 zakim, mute me 15:02:45 martinAlvarez should now be muted 15:02:49 ack martinAlvarez 15:02:54 HadleyBeeman: hard to hear 15:02:56 Rufus: would like to work with martinAlvarez on that! 15:03:04 * been thinking about data discivery 15:03:09 Please, Rufus :-) 15:03:16 ack me 15:03:17 * any thoughts on discovery et.al. 15:03:26 ack HadleyBeeman 15:03:35 PhilA: repeat question 15:03:40 HadleyBeeman: as been brainstorming 15:03:47 * what behaviors are assumed 15:04:01 * behaviors w.r.t. interacting with datasets 15:04:32 PhilA: This is pertinent w.r.t. the protocol/API work currently being done at CKAN 15:04:33 I should look to the definition of an API work that Rufus is doing? (to confirm) 15:04:35 HadleyBeeman, there are some thoughts here: http://dvcs.w3.org/hg/gld/raw-file/default/dcat-ucr/index.html 15:05:44 Sounds like we should all talk more. Thanks, Rufus, PhilA and cygri! 15:05:46 zakim, mute me 15:05:46 PhilA should now be muted 15:05:46 thanks all! 15:05:49 George: End with massive +1 15:05:49 Cheers, all! 15:05:55 I don't know how to close this out 15:05:57 bye all 15:06:01 HadleyBeeman has left #gld 15:06:03 can someone help???? 15:06:04 -mhausenblas 15:06:06 -fadmaa 15:06:08 -GeraldSteeman 15:06:09 martinAlvarez has left #gld 15:06:10 Thanks. 15:06:12 -HadleyBeeman 15:06:15 -PhilA 15:06:16 -sandro 15:06:18 -martinAlvarez 15:06:18 rrsagent, bye 15:06:18 I see 1 open action item saved in http://www.w3.org/2012/07/26-gld-actions.rdf : 15:06:18 ACTION: PhilA to summarise discussion and make a proposal for agreement or not [1] 15:06:18 recorded in http://www.w3.org/2012/07/26-gld-irc#T14-50-08