[08:22:37] Jonas_WMDE, i addressed your comment. Let me know if you see any other issues [08:28:40] yurik thanks! I will have a look later ... [09:13:02] hi [09:13:36] Is there a way with the sparql endpoint to get all the people from wikidata with their country of birth born in the last n years ? [09:14:10] yes [09:14:36] Nazral: I'm debugging something and I try to write you this query in a little time [09:14:55] can't say when because *debugging* [09:15:05] Thank you a lot! [09:15:56] Nazral: last 5 years: http://tinyurl.com/y8f8tc25 [09:16:05] but that already took 27 s to run, might be too inefficient for longer timespans [09:17:32] thanks Lucas_WMDE :) [09:17:49] Thank you Lucas_WMDE, I will work with that! [09:17:59] np :) [09:25:11] Lucas_WMDE: is there a way to avoid the timeout if I change the person and country by personLabel and countryLabel ? [09:25:56] it's heavier to use the service:wikibase… [09:27:01] it’s possible, but only with some ugly optimization by hand http://tinyurl.com/y8lplxfl [09:27:27] (there’s an open task to do that optimization automatically, but we haven’t implemented it yet) [09:27:49] oh nice thanks [09:37:56] Nazral: actually, with some more optimization by hand, the query gets a lot faster :) http://tinyurl.com/y9npno8v [09:39:51] I love how well all thrive to make more efficient queries [09:39:58] we* [09:49:24] Dafuq. https://www.wikidata.org/w/index.php?title=Q1007134&type=revision&diff=574241233&oldid=534129739 [09:56:16] sjoerddebruin: I don't understand half the merges due to the game [09:56:30] I've pinged Magnus again on the admin noticeboard. [09:56:35] This shouldn't be possible. [13:06:49] Lucas_WMDE, i looked - CONFIG.location.index is used as a root everywhere [13:07:00] I don't think it should be changed [13:07:21] renaming it into .root or .site would only confuse things [13:07:47] yurik: are there any places that append something to it? [13:08:31] Lucas_WMDE, yes, in two places - but only hashes [13:08:52] ah, okay [13:09:01] that works in either case [13:09:34] but appending a different filename will break if it contains 'index.html', and there is already one location that initializes it with that value [13:10:16] Lucas_WMDE, yes, but its not being used in that way - it has always been just the root. Do you want me to rename it to "site" ? [13:10:49] index is not the best name, but everyone understands its meaning in this context, so changing might be a worse thing [13:11:35] yurik: no, it has not always been just the root, that’s what I keep saying! there’s already one location that initializes it with '/index.html', not just '/'! :D [13:12:03] ah, sorry, I didn't catch that. looking... [13:12:09] btw, i reverted the blanks [13:12:16] configLocal in wikibase/config.js [13:12:19] ok, thanks [13:13:23] (btw, if you don’t mind, I could fix trivial requests like that myself, just to speed things up. just wasn’t sure if you’d be okay with it in this case) [13:13:47] ah yes, i see. The localhost use case. And yes, I'm always ok with others fixing minor things :) [13:13:59] ok good to know :) [13:14:45] and I think the localhost use case might actually need to be /index.html, because if you use file://, the browser doesn’t automatically pick index.html, it shows you a directory listing [13:27:19] Lucas_WMDE, done [13:27:40] thanks, I’ll have a look [13:28:06] Lucas_WMDE, how well do you know rollup/webpack and babel? [13:28:14] not at all [13:28:35] bummer. WDQS app badly needs a build env upgrade :( [13:28:46] so that we can finally use proper ES6 things [13:28:49] I just know that they would let us write nicer code if anyone went through the trouble of setting it up :) [16:20:21] Ehm, what's going on here? https://grafana.wikimedia.org/dashboard/db/wikidata-dispatch?refresh=1m&orgId=1 [16:21:01] Are we dispatching in advance now? :P [16:23:25] huh [16:25:32] Amir1: do you know anything about this? no new dispatch statistics in Grafana since 11:21 today (Berlin time – 7 hours ago) [16:26:01] at least https://www.wikidata.org/wiki/Special:DispatchStats still works and says everything’s okay [16:27:44] https://www.wikidata.org/w/api.php?action=query&meta=siteinfo&format=json&siprop=statistics looks okay as well [16:28:19] Lucas_WMDE: sjoerddebruin: Due to the dataset NFS failure on Saturday cron jobs have been disabled on a bunch of machines [16:28:32] including stat1005 where this runs on [16:29:01] :/ [16:30:24] I was working on beers yesterday and if we keep the current suggester setup, we might get suggestions for % of alcohol on paintings. :P [16:31:25] hoo: okay, so it’s not related to the recentchanges issues? [16:31:57] Lucas_WMDE: No, not at all [16:32:05] ok, thanks [16:37:27] Lucas_WMDE: sjoerddebruin Data should be back [16:37:39] :D [16:37:51] Looks very positive now. :P [17:54:56] https://www.wikidata.org/wiki/Q41780208#P2846 :D [20:35:17] Just spent half an our creating these... https://www.wikidata.org/wiki/Q41785110#P2868 [20:45:01] I spent hours on beers last night, Josve05a. [20:45:01] Virtual beers, to be clear. [20:50:58] 'suuure' [22:06:46] Seems like most example queries use wdt for population, which may return any historical ones. What is the best way to get the latest population? And we should probably update the examples with that approach. http://tinyurl.com/y9pggx94 [22:06:49] cc: SMalyshev [22:42:59] yurik: Well, wdt should only include the latest data [22:43:10] if it doesn't the rank on the data is not correctly set [22:43:14] which may very well be [22:44:19] hoo http://tinyurl.com/ybhl2tvw [22:44:35] i don't think we rank different years [22:45:57] yurik: Fixed it https://www.wikidata.org/w/index.php?title=Q223177&type=revision&diff=574717317&oldid=574644704 [22:46:35] hoo, are you sure this is a good approach? I think most of the population data in Wikidata doesn't have that [22:46:45] it might be a bot job [22:47:00] it's the one the community chose to go with [22:47:35] Yeah, could be… but maybe not an easy one… hm [22:47:47] in many cases maybe, but not in all [22:48:00] sadly, bot can only do it once - once a new data value is added, it won't be easy to switch an old prefered to the new one [22:48:23] yeah, the one adding the new value needs to do taht [23:35:11] apparently there are almost 10,000 multi-valued populations http://tinyurl.com/yauw77m3