[06:11:37] <Guest36026>	 <hello> [checking in with everyone]
[09:33:09] <ciss>	 hi, i've noticed that the label service fails on a few occasions. e.g., the label for https://www.wikidata.org/wiki/Q16069637 is not retrieved. what is the (likely) reason for this, and what are common strategies to work around the problem?
[09:37:04] <ciss>	 ... oh. it's because no label has been provided in the language i requested.
[09:40:27] <nikki>	 you can list more than one language (comma separated) if you want it to try multiple languages
[09:44:52] <Alphos>	 (as fallbacks)
[09:45:18] <ciss>	 i love you guys.
[09:45:29] <Alphos>	 didn't do much, but yw ^^
[09:45:56] <ciss>	 picked up a rather cumbersome coalesce solution at first - this is so much better
[09:46:10] <Alphos>	 you'd be surprised
[09:46:30] <Alphos>	 if you had rdfs:label + COALESCE, it's possibly faster than SERVICE
[09:47:14] <Alphos>	 (which means you could end up below the timeout instead of above)
[09:47:21] <ciss>	 yeah, i've noticed anything involving wikibase: statements seems to increase query runtime by quite a bit
[09:48:17] <ciss>	 i take that back, i only noticed it for service. the other time with the statement i simply had a false implementation
[09:53:07] <nikki>	 I think you also have to use rdfs:label if you want to use the content of the label somewhere else in the query
[09:53:22] <ciss>	 sounds plausible
[09:54:14] <ciss>	 tbh i dislike that implicit addition of ?*Label variables a bit
[09:54:39] * nikki nods
[09:54:54] <nikki>	 I personally just use rdfs:label all the time
[09:55:26] <ciss>	 i'll probably end up there as well, but for now i try to keep my queries as clean as possible
[14:39:50] <Stryn>	 is there any gadgets to copy statements from item to other?
[14:44:10] <sjoerddebruin>	 nope
[14:53:43] <Stryn>	 sad
[15:24:38] <berroxl>	 anyone knows where to find data sets around curated problems of e.g. a corporation or a county?
[15:24:42] <berroxl>	 country
[16:05:18] <sjoerddebruin>	 hm, interesting development this week https://grafana.wikimedia.org/dashboard/db/wikidata-datamodel?refresh=30m&orgId=1&from=now%2Fy&to=now%2Fy
[17:02:32] <Lucas_WMDE>	 Hi SMalyshev, do you have a few minutes for me?
[17:02:39] <Rodejong>	 Question: I am wondering about "{{property:P27}}" as this effects only the most recent centuries. In the 1400's there was not something like citizinship like we understand it today. It's kind of useless to use P27 for a person who lived in 1401 to place him in Antwerp in Belgium as it is today. It would be better to have something like Nationality, in a more free perspective. We see loads of errors made on persons from
[17:02:39] <Rodejong>	 history. People are named citizens of stats that even didn't exist then. How do we correct this? 
[17:02:40] <AsimovBot>	 10[1] 04https://www.wikidata.org/wiki/Template:property:P27
[17:05:32] <SMalyshev>	 Lucas_WMDE: on meetings now but will be done in an hour probably
[17:05:58] <Lucas_WMDE>	 okay, then I think I’ll go home and get back to you as WikidataFacts in an hour :) thanks
[17:46:02] <SMalyshev>	 hmm I see there's no rdf dump for the week
[17:46:17] <SMalyshev>	 addshore: do you know anything about it/can check?
[17:47:08] <addshore>	 *reads up*
[17:47:26] <addshore>	 hmm, no, is that in production or something that is done on labs?
[17:47:54] <addshore>	 ooh, sjoerddebruin interesting massive jump in item redirects
[17:48:00] <SMalyshev>	 it's somewhere in production but I have little idea about the setup
[17:48:20] <addshore>	 alos a massive jump in property size a couple of weeks ago sjoerddebruin
[17:48:58] <SMalyshev>	 json dump is there but rdf is not... I wonder if my latest patch to it broke anything :(
[17:51:11] <addshore>	 where does the rdf dump surface / get exposed to the world?
[17:51:33] <addshore>	 I see https://github.com/wikimedia/puppet/blob/ccd01ee53bc84d4cec3e22fcb71ac353d9416ce9/modules/snapshot/manifests/cron/wikidatadumps/rdf.pp
[17:51:38] <SMalyshev>	 addshore: https://dumps.wikimedia.org/wikidatawiki/entities/
[17:52:08] <SMalyshev>	 addshore: yeah those ones. So rdf should be done by now but no sign of it
[17:52:39] <addshore>	 https://github.com/wikimedia/puppet/blob/7734e5fdf63c72419879282e36d6d5c47ae2b3e8/hieradata/hosts/snapshot1006.yaml
[17:52:48] <addshore>	 im guessing it would be on snapshot1006
[17:53:36] <SMalyshev>	 aha... I don't think I have access there
[17:54:01] <addshore>	 apparently I do
[17:54:02] <addshore>	 heh
[17:54:16] <addshore>	 no idea via which group
[17:55:16] <SMalyshev>	 in /var/log/wikidatadump may have some clues...
[17:55:38] <addshore>	 no such directory
[17:55:41] <addshore>	 let me look in logstash
[17:56:50] <SMalyshev>	 huh that's interesting - this says there should be https://github.com/wikimedia/puppet/blob/ccd01ee53bc84d4cec3e22fcb71ac353d9416ce9/modules/snapshot/manifests/cron/wikidatadumps/common.pp
[17:57:02] <SMalyshev>	 is there some puppet fail?
[17:58:40] <addshore>	 unless im looking at the wrong host
[17:59:04] <addshore>	 maybe the wikidatawiki there means the regular dumps for wikidatawiki run there and the rdf are still somewhere else
[18:00:00] <SMalyshev>	 hieradata/hosts/snapshot1007.yaml:snapshot::cron::misc: true
[18:00:04] <SMalyshev>	 maybe this one?
[18:00:27] <SMalyshev>	 not sure I understand puppet magic correctly....
[18:00:31] <addshore>	 oohl, i see a directory
[18:01:17] <addshore>	 hmm, although it looks like there are only logs for the json dump on production there
[18:01:25] <addshore>	 but it looks like truthy ones should be there
[18:01:44] <addshore>	 https://usercontent.irccloud-cdn.com/file/ZOPDEwbo/image.png
[18:02:41] <SMalyshev>	 so what the 0703 logs say?
[18:03:48] <addshore>	 lots :p
[18:04:01] <addshore>	 I grepped for rdf and saw nothing
[18:05:11] <addshore>	 addshore@snapshot1007:/var/log/wikidatadump$ cat dumpwikidatajson-wikidata-20170703-all-* |grep -v Processed
[18:05:12] <addshore>	 Dumping shard 0/5
[18:05:12] <addshore>	 Dumping shard 1/5
[18:05:12] <addshore>	 Dumping shard 2/5
[18:05:12] <addshore>	 Dumping shard 3/5
[18:05:12] <addshore>	 Dumping shard 4/5
[18:05:25] <addshore>	 most of the output is just spamming "Processed 900578 entities."
[18:05:34] <addshore>	 but these areonly json, there is nothing for rdf :/
[18:05:43] <addshore>	 the only truthy ones are for BETA
[18:06:07] <SMalyshev>	 hmm weird so what's going on with rdf?
[18:06:09] <addshore>	 so something must be going wrong before the log file even gets started
[18:08:12] <addshore>	 I would have thought someone would have got an email on failed crons?
[18:08:44] <SMalyshev>	 probably... I'll ask ops. 
[18:17:41] <Rodejong>	 Question: I am wondering about "{{property:P27}}" as this effects only the most recent centuries. In the 1400's there was not something like citizinship like we understand it today. It's kind of useless to use P27 for a person who lived in 1401 to place him in Antwerp in Belgium as it is today. It would be better to have something like Nationality, in a more free perspective. We see loads of errors made on persons from
[18:17:41] <Rodejong>	 history. People are named citizens of stats that even didn't exist then. How do we correct this? 
[18:17:41] <AsimovBot>	 10[2] 04https://www.wikidata.org/wiki/Template:property:P27
[18:22:12] <SMalyshev>	 addshore: looks like dumps are broken due to https://phabricator.wikimedia.org/T169680...
[18:22:25] <addshore>	 aaaah
[18:41:36] <Lucas_WMDE>	 SMalyshev: do you have some time now? (it’s related to https://phabricator.wikimedia.org/T102752 btw, where you just commented)
[18:41:47] <SMalyshev>	 Lucas_WMDE: sure
[19:59:57] <sjoerddebruin>	 hey hoo :D
[20:00:17] <hoo>	 hi sjoerddebruin :)
[20:00:25] <sjoerddebruin>	 Hope you are doing alright
[20:00:47] <hoo>	 Yeah, just very busy irl, but that will improve after this week
[20:00:54] <sjoerddebruin>	 Good to hear.
[20:25:25] <d3r1ck>	 hoo: Hello :)
[20:26:03] <hoo>	 hi d3r1ck
[20:26:15] <d3r1ck>	 hoo: Hope you are doing great?
[20:26:41] <d3r1ck>	 Its been a while
[20:26:55] <hoo>	 I am indeed… how about you?
[20:47:24] <multichill>	 JD|cloud: Interesting discussion about the self promotion items. I wonder how many users created items like https://www.wikidata.org/wiki/Q15136093 or https://www.wikidata.org/wiki/Q20895241 to promote themselves
[20:47:45] <multichill>	 On the Dutch Wikipedia you get puked out if you write an article about yourself
[20:49:50] <JD|cloud>	 look for the item on someone named "Pavan Kumar N R"
[20:50:17] <JD|cloud>	 I'd have deleted it if wikiquote wasn't dragging their feet closing the local deletion discussion >:(
[20:50:17] <sjoerddebruin>	 https://www.wikidata.org/wiki/Q3809586 shouldn't be a occupation either
[20:50:42] <Alphos>	 why not ?
[20:50:56] <sjoerddebruin>	 It's a function.
[20:51:47] <multichill>	 Andy is very aggresive as always
[20:52:07] <Alphos>	 hmmm
[20:52:52] <multichill>	 Very understandable if you look at https://www.wikidata.org/wiki/Q15136093 . He's basically been unemployed for many years and Wiki stuff is his only income
[20:53:05] <sjoerddebruin>	 ORCID is of course much more important than actual trusted authority control.
[20:54:26] <sjoerddebruin>	 Removed some invalid VIAF entries from his item.
[20:57:02] <multichill>	 JD|cloud: Looked at his userpage. Isn't he in violation of the paid editing policy? 
[20:57:50] <JD|cloud>	 who
[20:57:54] <JD|cloud>	 Andy?
[20:58:03] <JD|cloud>	 is he not disclosing it?
[20:58:19] <multichill>	 He got paid by TED, that's not on his user page isn't it?
[20:58:23] <JD|cloud>	 If you're in doubt you can always contact WMF legal but I doubt there's an issue her
[21:00:26] <multichill>	 JD|cloud: Isn't https://en.wikipedia.org/wiki/Wikipedia:Paid-contribution_disclosure enforced by enwp admins?
[21:01:44] <multichill>	 And he already hates me, so no, I'm not bringing this up with legal. Being hounded by Andy is something I don't want to ever experience again
[21:02:27] <JD|cloud>	 We don't have a policy like enwiki does, we're technically not empowered in the same fashion
[21:02:40] <JD|cloud>	 Keep in mind that this channel is publicly logged too
[21:10:18] <multichill>	 Sure, let it be logged the first person I thought about when I read http://wondermark.com/1k62/ was Andy