[06:55:15] <sjoerddebruin>	 Good morning everyone!
[08:53:26] <Jonas_WMDE>	 Tobi_WMDE_SW:  https://gerrit.wikimedia.org/r/#/c/312001/
[08:53:41] <Tobi_WMDE_SW>	 thx Jonas_WMDE 
[08:56:31] <Tobi_WMDE_SW>	 got it Jonas_WMDE 
[08:56:42] <Tobi_WMDE_SW>	 it's because the changes are not deployed yet on test.wikidata.org!
[08:56:55] <Jonas_WMDE>	 möp
[08:57:07] <Tobi_WMDE_SW>	 the issue is, we're always running the latest version of the tests there..
[08:57:34] <Tobi_WMDE_SW>	 probably worth looking into only running the tests from the branch that is actually deployed there
[09:09:50] <Tobi_WMDE_SW>	 Jonas_WMDE: ok, actually jenkins IS always trying to check out the deployed version!
[09:10:00] <Tobi_WMDE_SW>	 but Wikidata is special because we do not deploy regularly..
[09:10:32] <Tobi_WMDE_SW>	 the script uses https://test.wikidata.org/w/api.php?action=query&meta=siteinfo&siprop=general to get the current deployed version of mediawiki and then tries to checkout that version of the extension
[09:10:59] <Tobi_WMDE_SW>	 but wikidata does not have a wmf/1.28.0-wmf.20 branch yet! is is still on .19!
[09:11:01] <Tobi_WMDE_SW>	 https://gerrit.wikimedia.org/r/#/admin/projects/mediawiki/extensions/Wikibase,branches?skip=50
[09:11:05] <Tobi_WMDE_SW>	 Jonas_WMDE: ^^
[09:11:11] <Tobi_WMDE_SW>	 that's the cause of the problem
[09:25:09] <Alphos>	 com.bigdata.rdf.internal.NotMaterializedException: Vocab(2):XSDUnsignedInt(1029222) /// darnit :-(
[09:26:38] <Alphos>	 hm, it stopped doing it, this is normal :-/
[09:26:50] * Alphos won't even comment on the matter :D
[10:52:13] <hashar>	 Hallo. I have updated the CI jobs for Wikidata/Wikibase to disable Zend garbage collection. That apparently is a source of some segfault https://phabricator.wikimedia.org/T142158  
[10:52:20] <hashar>	 all patches courtesy of Paladox :D
[10:56:25] <DanielK_WMDE>	 hashar: thank you!
[11:59:30] <sjoerddebruin>	 Good afternoon, WikidataFacts
[11:59:47] <WikidataFacts>	 good morning ;)
[12:00:21] <Incabell>	 one could think that you live in two very different time zones :D
[12:00:29] <sjoerddebruin>	 What timezone are you? :O
[12:01:12] <WikidataFacts>	 CET, I just got up late :D
[12:01:29] <sjoerddebruin>	 Better start working then. :P
[12:05:53] <WikidataFacts>	 well, I could try to get somewhere with the uppercased Esperanto person labels bot…
[12:06:10] <WikidataFacts>	 I could just run it from PAWS, and ignore that it takes a long time (just let it run)
[12:06:16] <WikidataFacts>	 would that be frowned upon?
[12:06:37] <sjoerddebruin>	 No objections from my side.
[12:51:39] <jon[m]>	 Hello friendly people in this round.
[12:51:40] <jon[m]>	 Does anybody have experiences with deploying a wikidata-query-service together with Blazegraph? We'd want to attach this stack to a custom Wikibase instance.
[12:51:50] <jon[m]>	 It lives at https://base.transformap.co
[12:57:33] <Incabell>	 Lydia_WMDE do you know or who would know?
[12:57:39] <Incabell>	 ^
[12:58:18] <Lydia_WMDE>	 jon[m]: https://www.mediawiki.org/wiki/Wikidata_query_service/User_Manual has infos
[12:58:28] <jon[m]>	 Oh, say hi to Lydia_WMDE (IRC) , we're friends from Maxime. I think we once walked to Diefenbach together?
[12:58:42] <Lydia_WMDE>	 hehe
[12:58:44] <Lydia_WMDE>	 hi!
[12:58:49] <jon[m]>	 Yes, we're going to put this into a Docker deployment.
[12:59:13] <jon[m]>	 Let's see how it will work out. Currently investigating triple stores. Any other recommendations, despite Blazegraph?
[12:59:22] <harmonia>	 hello Lydia_WMDE !
[12:59:38] <Lydia_WMDE>	 jon[m]: for use with wikibase? blazegraph :D
[12:59:42] <Lydia_WMDE>	 harmonia: hey! :)
[12:59:52] <jon[m]>	 oké
[13:00:27] <harmonia>	 Lydia_WMDE: I passed the million edits monday :p
[13:00:50] <Lydia_WMDE>	 harmonia: wohooooooooooooo! :D
[13:01:06] <sjoerddebruin>	 <3
[13:01:07] <harmonia>	 ^^
[13:01:17] <Lydia_WMDE>	 jon[m]: if you are happy with your wikibase install and if things work out i'd be happy to have you as one of the users of wikibase on wikiba.se
[13:01:17] <harmonia>	 ^/me is happy
[13:01:24] <Lydia_WMDE>	 you should be! :D
[13:01:30] <Lydia_WMDE>	 cupcakes?
[13:01:38] <Incabell>	 kittens!
[13:01:41] <Lydia_WMDE>	 hehe
[13:01:42] <harmonia>	 cupcakes and tea!
[13:01:51] * Lydia_WMDE serves some for the channel
[13:02:04] * harmonia give pu-ehr tea to all the chan
[13:02:07] <harmonia>	 gives*
[13:02:29] <sjoerddebruin>	 Kitten-shaped cupcakes!
[13:02:33] <Lydia_WMDE>	 lol
[13:03:04] <harmonia>	 with matcha icecream?
[13:03:09] <harmonia>	 :p
[13:03:56] <jon[m]>	 Anybody who ever put this beast (Wikidata Query Service + Blazegraph) into Docker containers? Any preliminary work would right now be very helpful.
[13:03:56] <jon[m]>	 We're starting to read the documentation from above at and may use https://github.com/lyrasis/docker-blazegraph . https://www.mediawiki.org/wiki/Wikidata_query_service/User_Manual#Standalone_service
[13:13:02] <Lydia_WMDE>	 jon[m]: gehel and SMalyshev might be able to tell
[13:15:25] <jon[m]>	 Thanks.
[13:15:42] <jon[m]>	 Let's wait if they are able to respond to the mention.
[13:43:21] * jon[m] are working in https://github.com/TransforMap/docker-wikibase-query-service
[14:06:20] <jon[m]>	 We managed to push some commits which get Blazegraph running in a Docker container.
[14:06:21] <jon[m]>	 Now we are figuring out how to create an RDF export of our Wikibase instance. Any pointers, again?
[14:08:34] <jon[m]>	 Found https://www.mediawiki.org/wiki/Wikidata_Toolkit which had been used to generate http://korrekt.org/page/Introducing_Wikidata_to_the_Linked_Data_Web
[15:03:14] <Jonas_WMDE>	 sjoerddebruin I think I fixed T146318 could you please verify?
[15:03:15] <stashbot>	 T146318: DuplicateReferences doesn't support adding a second copied reference to the same statement - https://phabricator.wikimedia.org/T146318
[15:05:10] <sjoerddebruin>	 Maybe we shouldn't test on the same item. :P
[15:07:03] <sjoerddebruin>	 It seems to save, but it doesn't update the current view. If you know what I mean, Jonas_WMDE.
[15:07:22] <Jonas_WMDE>	 not sure
[15:07:48] <sjoerddebruin>	 It just stays on saving here.
[15:08:36] <Jonas_WMDE>	 argh
[15:10:18] <Jonas_WMDE>	 do you get any js errors?
[15:11:21] <sjoerddebruin>	 TypeError: null is not an object (evaluating 'this._referencesListview.option')
[15:23:23] <Jonas_WMDE>	 sjoerddebruin so when I use the gadget I can reproduce the error
[15:23:31] <Jonas_WMDE>	 when I use the my script it works https://www.wikidata.org/wiki/User:Jonas_Kress_(WMDE)/duply.js
[15:23:48] <sjoerddebruin>	 Note that someone did some edits to the gadget one.
[15:26:57] <Jonas_WMDE>	 yes I saw that, but I can not read the diff unfortunately :(
[15:33:58] <Jonas_WMDE>	 Sorry sjoerddebruin  :(
[15:39:20] <sjoerddebruin>	 :(
[15:57:14] <jon[m]>	 We would still be stuck with generating a Wikibase dump for the Wikidata Toolkit Client to convert to RDF. Any suggestions appreciated.
[16:06:55] <dennyvrandecic>	 does anyone here know about the OSM/wikidata work?
[16:07:16] <dennyvrandecic>	 I wonder why the wikidata tags on OSM are not being autopopulated from the wikipedia tags
[16:07:21] <dennyvrandecic>	 aude: ?
[16:08:15] <jon[m]>	 dennyvrandecic (IRC): A friend who's sitting next to me is also jumping on IRC now. They have had this exact question at the State of the Map in Brussels last weekend.
[16:08:39] <dennyvrandecic>	 ha, very timely :)
[16:08:54] <jon[m]>	 Nobody knows how to create the JSON dumps of Wikibase? https://www.wikidata.org/wiki/Wikidata:Database_download
[16:09:32] <speciesATosm>	 dennyvrandecic: We had the OSM/wikidata discussion on the recent State of the Map conference...
[16:10:16] <dennyvrandecic>	 *nod*
[16:10:35] <speciesATosm>	 Automatic 'imports' into OpenStreetMap are done only in very few cases.
[16:12:00] <speciesATosm>	 The reason the wikidata tag is not set automatically is that often the wikipedia tag does not correspond to the 'real' object... e.g. on a monument the wikipedia tag is set to the subject, e.g. an old king. Setting the wikidata tag therefore automatically would be wrong.
[16:12:29] <dennyvrandecic>	 often?
[16:13:00] <speciesATosm>	 in a lot of cases. The 'policy' for OSM on imports is that you have to check each object manually.
[16:13:04] <dennyvrandecic>	 ah, so the wikipedia tags have been used rather indiscriminately
[16:13:57] <dennyvrandecic>	 we have a lot of cases where a wikipedia tag is used not to encode equality, but rather relatedness
[16:14:00] <dennyvrandecic>	 is that fair to say?
[16:14:03] <speciesATosm>	 Yeah :-/
[16:14:03] <speciesATosm>	 That's partly the fault of OSM users, and we do not want to 'legitimate' this wrong data by adding the wrong WD keys
[16:14:18] <dennyvrandecic>	 understood, makes sense
[16:14:40] <dennyvrandecic>	 is there an effort ongoing to solve that?
[16:14:46] <speciesATosm>	 e.g. for a monument the correct wikidata tag would then be "subject:wikidata"=Q* .
[16:15:08] <speciesATosm>	 There is not a coordinated effort ongoing yet.
[16:15:22] <dennyvrandecic>	 i.e. if someone took the data, analyzed it, and created a dataset that has a higher correctness
[16:15:31] <speciesATosm>	 but it would make a nice "weekly task".
[16:15:47] <dennyvrandecic>	 is there a chance of uploading it, or would it still require manual individual assessment
[16:16:11] <speciesATosm>	 Even if you analyze it, it would only simplify the manual correction. Yes, the current policy is that you still have to check each object manually.
[16:16:31] <speciesATosm>	 We on OSM have a big history of 'failed' imports, so we are rather strict on new imports.
[16:16:54] <dennyvrandecic>	 understood. it is unfortunate though
[16:17:08] <speciesATosm>	 I met one user at SotM, who tried to do that automatically - he was banned rather quickly^^.
[16:17:14] <dennyvrandecic>	 if we have a dataset with 0.99 correctness, it is likely that it will be better than the human evaluation
[16:17:42] <speciesATosm>	 The OSM data working group would still have objections, even if its 99%.
[16:18:02] <dennyvrandecic>	 is there any indication human evaluation is better than 99%? :D
[16:18:07] <dennyvrandecic>	 or even reaches it?
[16:19:10] <dennyvrandecic>	 we're talking about 700k items here :) I am not sure many humans will maintain a high correctness on such a dataset
[16:19:48] <speciesATosm>	 No, there is not. But at OSM, we do not thrust external datasets in general.
[16:19:49] <speciesATosm>	 Then it may takes time until every OSM wp key has its corresponding WD key. We have time :-)
[16:20:32] <dennyvrandecic>	 sure :) I won't fight policies of other collaborative projects. I know how these are made on the wikimedia projects and how useful it is to critizise them as an outsider
[16:20:34] <speciesATosm>	 I do not say it is impossible to do an automated import here - it is just very improbable that the OSM community would agree to do it automaticly
[16:21:04] <dennyvrandecic>	 would you have a link to the process?
[16:21:32] <jon[m]>	 Let's believe the WikiData case could be rather different of any other import, given the collaborative alignment and importance of both datasets to the Public Domain.
[16:21:42] <speciesATosm>	 http://wiki.openstreetmap.org/wiki/Automated_Edits_code_of_conduct
[16:23:45] <dennyvrandecic>	 thank you for the link, i will take a look
[16:24:24] <dennyvrandecic>	 I assume a tool where an OSM contributor can oauth in and can then click yes or no or skip would be OK?
[16:24:55] <jon[m]>	 Oh that sounds lovely. I have heard of those Wikidata games only lately.
[16:24:58] <speciesATosm>	 Yeah, I guess it would do
[16:25:04] <dennyvrandecic>	 great thanks
[16:25:08] <dennyvrandecic>	 and speciesATosm, since I have you here
[16:25:15] <dennyvrandecic>	 I understand relation IDs are not persistent
[16:25:26] <dennyvrandecic>	 which is why it is discouraged to link from Wikidata to OSM
[16:25:27] <speciesATosm>	 Also an option to "fixme" flat an entry would be nice.
[16:25:36] <dennyvrandecic>	 speciesATosm: good idea re fixme
[16:26:01] <speciesATosm>	 oh, relation IDs are pretty stable.
[16:26:01] <speciesATosm>	 only nodes and ways are not, because they are often migrated to relations.
[16:26:10] <dennyvrandecic>	 ah
[16:26:11] <dennyvrandecic>	 OK
[16:26:13] <dennyvrandecic>	 good to know
[16:26:22] <dennyvrandecic>	 so relations are fine, nodes and ways rather not
[16:26:24] <dennyvrandecic>	 (noting)
[16:26:38] <speciesATosm>	 thats why there is a OSM relation id property on WD, but not a node/rel property.
[16:26:47] <dennyvrandecic>	 one more question - even if they are not stable, would you ever reuse a previous id for a different object?
[16:27:00] <dennyvrandecic>	 speciesATosm: makes sense!
[16:27:26] <speciesATosm>	 not by chance - you have to manually talk to the API to do that.
[16:27:37] <dennyvrandecic>	 ah, great
[16:27:52] <dennyvrandecic>	 but there are no "redirects" kept from old ids to the new one?
[16:28:13] <speciesATosm>	 unfortunately not.
[16:28:13] <speciesATosm>	 there are some workarounds, but very few users actually use that.
[16:28:27] <dennyvrandecic>	 ok, thanks, just trying to understand and learn
[16:28:44] <dennyvrandecic>	 i am not criticising :) just learning what the restrictions are
[16:28:52] <speciesATosm>	 But there are plany to use the "wikidata" tag on an object to have some sort of "stable" identifier for an object.
[16:29:52] <dennyvrandecic>	 that sounds nice
[16:30:30] <dennyvrandecic>	 is there a datadump extract with such metadata, keys, labels, etc. but without the actual geo-data?
[16:30:53] <speciesATosm>	 because it acts as some sort of UUID for an object. And an OSM "thing" can be a 'collection' of more than one items, e.g. the different segments of a street.
[16:31:14] <speciesATosm>	 You can easily create such a datadump with tools like osmconvert.
[16:31:24] <dennyvrandecic>	 ok, great
[16:31:34] <dennyvrandecic>	 ah, different segments makes sense, indeed
[16:32:22] <speciesATosm>	 osmconvert can generate csv-files and statistics from a planet dump.
[16:32:35] <dennyvrandecic>	 great, I'll fetch that and take a look
[16:33:01] <speciesATosm>	 try it with a smaller extract first, e.g. liechtenstein.
[16:33:23] <dennyvrandecic>	 good idea
[16:33:51] <dennyvrandecic>	 osm has no plans to include historic boundaries and data like the spread of certain species, i assume?
[16:34:18] <speciesATosm>	 no, only current data.
[16:34:18] <speciesATosm>	 For historical data, there is the OpenHistoricalMap.
[16:34:25] <dennyvrandecic>	 thanks!
[16:34:34] <speciesATosm>	 For the spread of certain species, I would suggest to use Umap.
[16:35:11] <speciesATosm>	 Or create a fork of OSM, as OHM did would make even more sense.
[16:35:22] <dennyvrandecic>	 openhistoricalmap is down. ah well, there's the wiki page on it on osm
[16:36:13] <speciesATosm>	 oh, sad :-/
[16:36:58] <dennyvrandecic>	 another question
[16:37:12] <dennyvrandecic>	 is the relation between administrative regions somehow encoded in osm?
[16:37:18] <dennyvrandecic>	 i have seen that the level is encoded
[16:37:36] <dennyvrandecic>	 but was wondering if also what county belongs to which state is there, or similar
[16:37:44] <dennyvrandecic>	 and whether this is used for quality checks
[16:37:55] <speciesATosm>	 'level' is used for the 'building level' an item is on.
[16:38:43] <dennyvrandecic>	 sorry, admin_level
[16:39:00] <speciesATosm>	 To derive the 'political' boundaries an item is in, we usually use nominatim, the geocoder. It has a very nice API.
[16:39:27] <speciesATosm>	 But please don't DDOS it with 700k items^^
[16:39:37] <dennyvrandecic>	 :D
[16:39:42] <gehel>	 jon[m]: sorry for the late answer. I'm busy at Ops offsite this week, but WDQS in docker seems like a kool idea! I'd be happy to help, but saddly not before next week
[16:39:43] <dennyvrandecic>	 wasn't the plan :)
[16:40:06] <dennyvrandecic>	 for 700k items I'd probably just download the data completely and try to setup a local install of nominatim
[16:40:29] <jon[m]>	 gehel: Thanks, it seems to be progressing.
[16:41:06] <jon[m]>	 Now solved the often confusing documentation by using our own pad. MediaWiki, Wikidata and Wikibase want to be carefully distinguished domains.
[16:41:09] <speciesATosm>	 Yeah, setting up nominatim would be the way to go.
[16:41:42] <jon[m]>	 gehel (IRC): Make sure to track our repository, which we are going to enrich with our findings.
[16:41:50] <jon[m]>	 https://github.com/TransforMap/docker-wikibase-query-service
[16:42:16] <dennyvrandecic>	 but there's no one checking, e.g. "oh, this county is supposed to be in this state, but the state boundaries do not cover the county"?
[16:42:42] <gehel>	 jon[m]: I'm now watching that repo!
[16:43:30] <gehel>	 It might make sense at some point to provide a data container with a recent set of blazegraph data...
[16:43:39] <gehel>	 But that would be a pretty big one...
[16:44:01] <speciesATosm>	 There are some automatic relation checkers which at least check for unclosed relations, but I don't know of any who check the topology of relations - seems like a good idea to implement something that .... fetches the relation of states etc from Wikidata and check if they are correct in OSM too :-)
[16:45:23] <dennyvrandecic>	 yeah, just thinking loud :)
[16:46:09] <dennyvrandecic>	 ok, thank you so much, speciesATosm for answering all my questions
[16:46:14] <dennyvrandecic>	 and your patience
[16:46:17] <dennyvrandecic>	 this was superhelpful
[16:46:30] <dennyvrandecic>	 jon[m]: thanks for connecting
[16:47:33] <dennyvrandecic>	 speciesATosm: oh, one more thing: regarding something like taxa map, we could also host, say a kml file on commons, and link to that?
[16:47:46] <dennyvrandecic>	 that would also allow to use that as an overlay over an osm generated map easily?
[16:48:33] <speciesATosm>	 You could use Umap to display a map with an external file easily. I don't know if kml is supported, I usually use geojson.
[16:48:47] <dennyvrandecic>	 or geojson
[16:49:03] <dennyvrandecic>	 i don't care so much about the specific format, as long as it is open and easily usable by the tools :)
[16:49:16] <dennyvrandecic>	 yes, i guess i should have said geojson, i am merely old :)
[16:49:17] <hoo>	 DanielK_WMDE: DanielK_WMDE_: here?
[16:51:18] <speciesATosm>	 This is really easy to do with Umap. Create a new Umap, and use an external like as datasource.
[16:51:38] <speciesATosm>	 s/like/link/
[16:51:49] <dennyvrandecic>	 sweet :)
[16:52:28] <dennyvrandecic>	 my dream come true would be to for the wikidata sparql service to transparently be able to draw relations based on a query result
[16:52:34] <dennyvrandecic>	 i think this is already possible
[16:52:49] <dennyvrandecic>	 and then mix that with a umap/custom geojson
[16:53:01] <dennyvrandecic>	 well, sounds like we're pretty close to that :)
[16:54:05] <speciesATosm>	 dennyvrandecic: do you mean osm relations or any WD relations between entities?
[16:54:54] <dennyvrandecic>	 in this case osm relations. I.e. make a query like { ?county inState wd:California } and display the result on a map
[16:55:01] <dennyvrandecic>	 so to quickly see which counties are missing, etc.
[16:55:32] <dennyvrandecic>	 i have to check the work the foundation did recently, I think that covers this already mostly
[16:55:42] <dennyvrandecic>	 wikimedia foundation
[16:55:51] <dennyvrandecic>	 sorry, i should be more specific while talking to you :D
[16:57:05] <speciesATosm>	 For displaying OSM relations, just directly link them: https://www.openstreetmap.org/relation/1155955 - your browser should dynamically load the boundaries and display them.
[16:57:31] <dennyvrandecic>	 yes, that's one - but I want *all* the counties of california
[16:57:57] <speciesATosm>	 Ah - maybe an overpass query would do it...
[16:58:11] <dennyvrandecic>	 (noting down)
[16:58:24] <dennyvrandecic>	 osm has plenty of cool tools in its infrastructure
[17:01:43] <speciesATosm>	 maybe something like that: http://overpass-turbo.eu/s/iDy - if you dig through https://wiki.openstreetmap.org/wiki/Overpass_API/Overpass_QL, there are also functions to query objects inside some boundaries...
[17:02:39] <dennyvrandecic>	 cool
[17:02:44] <dennyvrandecic>	 I'll play around with that
[17:02:57] <dennyvrandecic>	 yay, a proprietary query language! :D
[17:03:57] <speciesATosm>	 Yeah :-/
[17:03:57] <speciesATosm>	 But there was nothing like that available ... maybe sometimes when OSM is available as RDF, some thing like GeoSPARQL would do it.
[17:04:24] <dennyvrandecic>	 I actually have no idea how appropriate something like sparql is for osm
[17:04:25] <dennyvrandecic>	 :)
[17:04:41] <dennyvrandecic>	 osm is a unique resource in its availability
[17:04:57] <dennyvrandecic>	 it might well be that a proprietary language will remain the best way
[17:05:14] <dennyvrandecic>	 but let's see. i like standardization :)
[17:05:25] <speciesATosm>	 me neither ;-)
[17:05:25] <speciesATosm>	 But at least there were some attempts by http://linkedgeodata.org/
[17:05:53] <speciesATosm>	 It was down for some years, and I don't know how active it is maintained.
[17:06:41] <dennyvrandecic>	 yeah, I would prefer not to discuss their approach to these things publicly :D
[17:28:52] <dennyvrandecic>	 ow, speciesATosm, that query on http://overpass-turbo.eu/s/iDy  you sent me know merely gives me error messages
[17:29:05] <dennyvrandecic>	 speciesATosm: it seems I am not a good influence on OSM infra :D
[17:29:07] <dennyvrandecic>	 An error occured during the execution of the overpass query! This is what overpass API returned:
[17:29:07] <dennyvrandecic>	 Error: runtime error: open64: 2 No such file or directory /osm3s_v0.7.52_osm_base Dispatcher_Client::1
[17:55:24] <speciesATosm>	 dennyvrandecic: Maybe try it with a smaller BBOX? I never had this error - Just tested, it worked now.
[17:55:34] <dennyvrandecic>	 trying
[17:55:44] <dennyvrandecic>	 ah yeah indeed
[17:55:57] <dennyvrandecic>	 yeah, maybe the bounding box was bad
[17:56:07] <dennyvrandecic>	 oh wow, i am surprised how much work is still left in OSM
[17:56:19] <dennyvrandecic>	 i would have assumed that municipalities and counties etc. are pretty complete
[17:56:22] <dennyvrandecic>	 fun!
[17:58:42] * DanielK_WMDE_ is always happy to see dennyvrandecic play with data :)
[17:58:50] <dennyvrandecic>	 :D
[17:58:53] <dennyvrandecic>	 data is awesome
[17:59:38] * Harmonia_Amanda has no idea what they are talking about but she is always good for a little datalove
[18:06:47] <dennyvrandecic>	 speciesATosm: one last question, if you don't mind - when I search for my adress, I can find it in OSM (yay) - but when I am in edit view in the same area, there is nothing, no house, no object to put the adress on
[18:06:55] <dennyvrandecic>	 is there some magic going on with adresses?
[18:10:25] <WikidataFacts>	 sjoerddebruin: bot is running now :)
[18:10:51] <DanielK_WMDE_>	 dennyvrandecic: i'm making slides for a keynote, and would like to give some historical background. do you have some slides handy  about How It All Started in 2011?
[18:10:54] <sjoerddebruin>	 WikidataFacts: :)
[18:11:10] <DanielK_WMDE_>	 dennyvrandecic: i remember you presenting something like this at wikimania a few years ago...
[18:11:23] <DanielK_WMDE_>	 err, how it all started in 2005, i mean :P
[18:14:51] <DanielK_WMDE_>	 haha, holy shit! https://commons.wikimedia.org/wiki/Category:Wikidata_presentations_in_English
[18:15:37] <SMalyshev>	 just 56? That's not too many
[18:16:37] <jon[m]>	 Hi room, speciesATosm (IRC) and me are still evaluating how to create a valid Wikibase RDF dump for importing into Blazegraph for the Query Service.
[18:16:56] <jon[m]>	 Unfotunately the dumpRdf.php from Wikibase doesn't create an output that the munge.sh script likes.
[18:16:57] <dennyvrandecic>	 DanielK_WMDE_: I might have lost my slidesets from back then :(
[18:17:06] <dennyvrandecic>	 or have them only on some disks somwhere
[18:17:19] <DanielK_WMDE_>	 never mind then.
[18:17:22] <dennyvrandecic>	 oh wait
[18:17:25] <dennyvrandecic>	 I should have one thing
[18:17:28] <jon[m]>	 Now we're trying dumpJson.php again, but the wdtk-client does not like our inputfile. What can we do?
[18:17:34] <dennyvrandecic>	 we had a 10 year SMW presentation last year
[18:17:38] <dennyvrandecic>	 i should have a slideset on that
[18:17:44] <dennyvrandecic>	 yay!
[18:17:48] <speciesATosm>	 dennyvrandecic: You are based in the US? Nominatim is also using TIGER data when looking up addresses. Could you send me the link to the query you are using?
[18:17:53] <dennyvrandecic>	 I'll share with you DanielK_WMDE_
[18:18:11] <DanielK_WMDE_>	 dennyvrandecic: oh, i wasn't thinking of the original SMW presentation; rather the slides you had about 2005 when you presented wikidata at wikimania in... 2012, i think.
[18:18:21] <DanielK_WMDE_>	 thanks!
[18:22:29] <dennyvrandecic>	 speciesATosm: https://www.openstreetmap.org/search?query=1430%20Grove%20Street%2C%20Alameda%2C%20CA#map=19/37.76077/-122.23498
[18:27:37] <dennyvrandecic>	 shared two presentations I had here, DanielK_WMDE_
[18:29:15] <speciesATosm>	 dennyvrandecic: yes, the address is only found in the TIGER dataset, not yet in OSM. You may add it :-)
[18:33:24] <DanielK_WMDE_>	 dennyvrandecic: thank you! 
[18:33:32] <DanielK_WMDE_>	 i find myself writing an actual text for this o_O
[18:33:41] <DanielK_WMDE_>	 usually i just make slides, and then wing it...
[18:45:12] <SMalyshev>	 jon[m]: which specific problems do you have with dumpRdf?
[18:51:21] <SMalyshev>	 DanielK_WMDE: do you know what is the procedure for updating wikidata build that is deployed? I've made 1.4.0 for purtle, but I have no idea if it was deployed and where
[18:51:29] <SMalyshev>	 and Special:Version just says 1.0
[20:10:26] <dennyvrandecic>	 DanielK_WMDE: DanielK_WMDE_: still there?
[20:10:47] <dennyvrandecic>	 speciesATosm: thanks! that's good to understand!
[20:14:42] <dennyvrandecic>	 DanielK_WMDE: I did find the slides from wikimania 2011&12 here, we go: http://simia.net/download/wikidata-wikimania-2011.pptx and http://simia.net/download/wikidata-wikimania-2012.pptx
[21:10:26] <jon[m]>	 gehel (IRC) + Lydia_WMDE (IRC) : Our progress of setting up a dockerised wikibase query service is stuck at the second installation step: [loading data](https://github.com/TransforMap/docker-wikibase-query-service/issues/1).
[21:29:31] <gehel>	 jon[m]: still around?
[22:07:45] <DanielK_WMDE__>	 dennyvrandecic: thank you! i'll look at them tomorrow.
[22:15:43] <DanielK_WMDE__>	 SMalyshev: can you give these folks a hint? i suspect they are missing the right params to dumpRdf.php https://tree.taiga.io/project/transformap/task/353
[22:16:35] <DanielK_WMDE__>	 jon[m]: ah, you are still here. talk to SMalyshev.
[22:22:34] <SMalyshev>	 DanielK_WMDE: I have hard time understanding what's going on... it just says "not accepted" but I have no idea what that means
[22:22:46] <SMalyshev>	 jon[m]: have any details on that?
[22:23:58] <SMalyshev>	 ah ok I found the log
[22:24:05] <DanielK_WMDE__>	 SMalyshev: does the munger eat ttl? or does it want triples?
[22:24:06] <SMalyshev>	 looks like the base URL is not set
[22:24:32] * DanielK_WMDE__ didn't find the log
[22:24:41] <SMalyshev>	 if your entities are not wikidata, you need to set base url
[22:26:40] <DanielK_WMDE__>	 SMalyshev: i see: Error munging null org.wikidata.query.rdf.tool.exception.ContainedException: Didn't get a revision id for [(_:genid-d83776a395a04eafa83b5e9cccbeb6de-genid1, http://www.w3.org/1999/02/22-rdf-syntax-ns#type, http://www.w3.org/2002/07/owl#Restriction
[22:26:51] <SMalyshev>	 jon[m]: try munge.sh .... -- -w base.transformap.co
[22:26:54] <DanielK_WMDE__>	 is that caused by a missing base URL?
[22:27:26] <SMalyshev>	 DanielK_WMDE: hmm that's something else
[22:27:41] <DanielK_WMDE__>	 SMalyshev: you can log in using your github id to leave a comment, it seems
[22:28:06] <DanielK_WMDE__>	 .oO(what's tiga.io anyway?)
[22:29:40] <SMalyshev>	 DanielK_WMDE: not allowed to comment
[22:29:47] <DanielK_WMDE__>	 hmpf
[22:29:55] <DanielK_WMDE__>	 .oO(what's a ContainedException?)
[22:30:11] <SMalyshev>	 it's a kind of exception in WDQS
[22:30:22] <DanielK_WMDE__>	 SMalyshev: well, let's hope jon[m] reads his backlog.
[22:30:52] <DanielK_WMDE__>	 SMalyshev: yea, i gathered that much :D
[22:31:55] <SMalyshev>	 DanielK_WMDE: well, it's an internal class to classify exceptions by deadliness :) Contained one is not very deadly as we can recover from it
[22:32:12] <SMalyshev>	 but that really doesn't matter too much if you see one something's wrong :)
[22:32:59] <SMalyshev>	 I added comment to github one, hopefully he'll see it.
[22:34:12] <SMalyshev>	 "We hope you enjoy using Taiga.io, your Open Source, Agile Project Management Tool."
[22:34:28] <SMalyshev>	 so apparently it's an agile project management tool :)
[22:36:42] <DanielK_WMDE__>	 heh
[22:37:45] <DanielK_WMDE__>	 SMalyshev: btw, thank you for the image meta-data search. it's nice to see something that people have been wanting for a decade just happen :)
[22:38:39] <DanielK_WMDE__>	 it's magic when all the pieces finally drop into place
[22:38:47] <SMalyshev>	 DanielK_WMDE: heh, np :) I think once we've refactored to make search filters sane and making indexing sane making this work became not hard at all
[22:39:10] <DanielK_WMDE__>	 yea, good tools make good work!
[22:39:40] <SMalyshev>	 yeah I think without structure in indexing and searching it'd be much harder... I wouldn't dare to touch query parser before :)
[22:40:09] <DanielK_WMDE__>	 heh, i have never ventured there...