[01:17:14] PROBLEM - wikidata.org dispatch lag is higher than 300s on wikidata is CRITICAL: HTTP CRITICAL: HTTP/1.1 200 OK - pattern not found - 1547 bytes in 0.154 second response time [02:19:25] who knows why this query returns duplicate Bangkok? https://query.wikidata.org/#PREFIX%20wd%3A%20%3Chttp%3A%2F%2Fwww.wikidata.org%2Fentity%2F%3E%0APREFIX%20wdt%3A%20%3Chttp%3A%2F%2Fwww.wikidata.org%2Fprop%2Fdirect%2F%3E%0APREFIX%20wikibase%3A%20%3Chttp%3A%2F%2Fwikiba.se%2Fontology%23%3E%0A%0ASELECT%20DISTINCT%20%3Fcity%20%3FcityLabel%20%3Fpopulation%20%3Fgps%0AWHERE%20%0A%7B%0A%20%20%3Fcity%20wdt%3AP31%2Fwdt%3AP279*%20wd%3AQ515%20.%0A%20%20%3Fcity%20wdt% [02:19:25] 3AP1082%20%3Fpopulation%20.%0A%20%20%3Fcity%20wdt%3AP625%20%3Fgps%20.%0A%20%20SERVICE%20wikibase%3Alabel%20%7B%0A%20%20%20%20bd%3AserviceParam%20wikibase%3Alanguage%20%22en%22%20.%0A%20%20%7D%0A%7D%0AORDER%20BY%20DESC(%3Fpopulation)%20LIMIT%2015 [02:19:57] https://goo.gl/i9y0so [02:20:02] much nicer url :) [02:38:19] HakanIST, benestar|cloud - for Bangkok, it returns two different results because one is census and the other is registration. Is there a "preferred" value somehow? [02:49:14] RECOVERY - wikidata.org dispatch lag is higher than 300s on wikidata is OK: HTTP OK: HTTP/1.1 200 OK - 1532 bytes in 0.164 second response time [03:42:52] https://goo.gl/EwqH7U gives me only 4 disasters :( [12:58:17] search broken again? [12:58:54] things are not instant... [12:59:00] shouldn't be [12:59:05] will check out [12:59:17] was just a hiccup I think [13:22:58] such an awesome tool by Hay! https://tools.wmflabs.org/hay/propbrowse/ [16:14:02] Weird stuff. https://www.wikidata.org/wiki/Q22038858 [16:14:21] nikki: any idea about this? ^ [16:16:06] that doesn't look right [16:17:28] The contraints agree with you. https://www.wikidata.org/wiki/Special:WhatLinksHere/Q22038858 [16:18:39] if it's a list article, there should be a sitelink, notable work statements can be added to the person and if you want a list, we can just do a query for it [16:22:21] I don't know anything about the babel of the user besides German... [16:22:54] So I'm not sure if me writting a message has any sense. [16:37:18] kopiersperre: it is lagging, it seems this edit https://www.wikidata.org/w/index.php?title=Q156420&type=revision&diff=293280646&oldid=292884252 is not yet represented [17:07:54] * nikki wonders if there's any sensible way of adding coordinates with a different globe [17:08:10] the ui doesn't support it, quick statements doesn't support it... :/ [17:10:48] and harvest templates can't do coordinates at all [17:51:07] https://www.wikidata.org/w/index.php?title=Q79037&type=revision&diff=303724171&oldid=301205065 [17:51:14] how about this one nikki ? [17:53:30] I would use P106 only for real persons [18:05:42] looks like handful of them has p106 [19:07:29] HakanIST: the link is for an album, so that's definitely wrong [19:36:24] benestar: oh dear, I merged the thing [19:36:35] time to place bets on the amount of rage? :D [19:48:44] JeroenDeDauw: over 9000 [19:50:23] benestar: over 9 million? [20:17:28] any sparql experts? goo.gl/EwqH7U returns just disasters in wikidata [20:17:36] just 4 [20:21:27] PROBLEM - wikidata.org dispatch lag is higher than 300s on wikidata is CRITICAL: HTTP CRITICAL: HTTP/1.1 200 OK - pattern not found - 1545 bytes in 0.139 second response time [20:36:59] yurik: I have no idea why, but putting { } around "?event wdt:P31/wdt:P279* wd:Q3839081 ." seems to make it return all the results [20:38:59] nikki, thanks! do you know how to make it return "anything that has at least one of the property "death, missing, injured, casulties" ? [20:41:30] adding "filter(bound(?casualties) || bound(?deaths) || bound(?missing) || bound(?injured))." seems to work [20:43:04] awesome! [20:44:44] JeroenDeDauw: I see your point though I cannot think of a better solution [20:44:46] nikki, sorry to bug you again but you seem to know all the answers - is it possible to add them together? [20:44:56] so that the result has just the sum? [20:44:59] of all 4 [20:45:14] this way i could order by it [20:45:20] and limit to the top N [20:45:22] only some special case handling in Wikibase.git for Item/Property stuff [20:45:43] benestar: which PR are you talking about?> [20:45:51] the one in internal serialization [20:46:12] benestar: I made another comment [20:46:34] benestar: unfortunately my rage level is currently under 9000, I'm so sorry [20:47:03] yes, I read that one [20:47:19] and I'm not happy about making the encapsulation worse [20:48:43] benestar: I'm thinking it's odd that Wikibase has so many places where it deals with "entities" (been thinking that for a long time, which is why the new interface was added) [20:48:52] Those serializers are just one case of that [20:49:03] You do not have this in other apps... [20:49:27] If you have different types of (DDD) entities, you don't make general code that can deal with all of them [20:49:54] Such as a RepositoryThatCanStoreAllEntities::save(), which then needs an SerializerForAllEntities [20:50:31] benestar: if you try to make Wikibase work for new types of entities, I think it's important to keep that in mind [20:50:49] Often code does not need to be able to do that (and is better structured to handle specific parts of the domain) [20:51:09] This seems to be made even worse due to the lack of context boundaries in Wikibase [20:51:11] ... [20:51:42] * benestar notices some Jeroen rant about Wikibase architecture ^^ [20:51:57] hm, it sounds like something that would be possible but I can't figure out how [20:52:01] * nikki continues looking [20:53:00] nikki, i tried xsd:float(?casualties) + xsd:float(?deaths) + xsd:float(?missing) + xsd:float(?injured) (and also xsd:integer) but that didn't work [20:55:08] yurik: try COALESCE [20:55:49] ie COALESCE(?casualties, 0) etc [20:57:41] benestar: [20:58:14] ah, yeah, coalesce seems to do it [20:59:07] nikki, benestar -- https://goo.gl/zdrT0w [20:59:13] it gives me 0 for some reason [20:59:31] I just put "(coalesce(?casualties,0) + coalesce(?deaths,0) + coalesce(?missing,0) + coalesce(?injured,0) as ?total)" in the select [20:59:56] ah, let me see [21:00:58] yurik: doing it in the header works for me [21:01:06] but i wonder why it doesn't in the body [21:01:14] maybe because it's defined before the variables are? [21:01:36] yeah, moving the bind lines to the bottom makes it work [21:01:49] haha, position is relevant [21:01:55] that makes me wtf [21:02:27] yurik: btw as far as I can see you can get rid of the type casting with xsd:float [21:04:00] benestar, nikki thanks, it seems to work now! https://goo.gl/z900sL - but why do we only have 200 disasters??? [21:04:36] yurik: probably because nobody went and classified them... [21:04:52] we have 4000 and something, but you've limited it to those which have coodinates and at least one property saying how many people were affected [21:05:16] it's a DanielK_WMDE :) [21:05:21] yurik: autolist can do it semi-automatically based on categories https://tools.wmflabs.org/autolist/autolist1.html [21:05:24] DanielK_WMDE, http://en.wikipedia.beta.wmflabs.org/wiki/Sparql :) [21:05:25] hey benestar! [21:05:28] (but yeah, the answer is also because people haven't added all the information yet) [21:05:30] this is what you can do now :))) [21:05:45] yurik: nice! [21:05:50] graphs for wikidata! [21:05:53] \o/ [21:06:07] DanielK_WMDE: I worked on the serialization stuff and I think it should work now as I implemented it [21:06:19] to make registering of additional serializers possible [21:06:23] now we just have to make them cacheable, visible without js, and sane to edit ;) [21:06:42] the good thing - it will be merged to production next week, the bad thing - sparql has no caching :)) [21:06:52] yep [21:07:13] yurik: btw - we are thinking about allowing users to store "query templates", so all you need to put into your url is the template name, and some parameters. [21:07:20] the templates would be on wiki pages, like lua mondules [21:07:35] or on Wikidata? [21:07:37] DanielK_WMDE, sure, but not really needed imo [21:07:53] Jonas and Stas are working on sparql caching [21:07:54] its another level of indirection, which is nice of course :) [21:08:01] yurik: oh hell yes needed. [21:08:15] for readability alone. but also to make the cache more efficient. [21:08:20] DanielK_WMDE, look at the graph definitions - it looks perfectly fine without it :-P [21:08:32] there'S hundreds of ways to formulate the same query. having a standard way means only one cache entry [21:08:48] "wikidatasparql:///?query=PREFIX%20wikibase%3A%20%3Chttp%3A%2F%2Fwikiba.se%2Fontology%23%3E%0APREFIX%20wd%3A%20%3Chttp%3A%2F%2Fwww.wikidata.org%2Fentity%2F%3E%20%0APREFIX%20wdt%3A%20%3Chttp%3A%2F%2Fwww.wikidata.org%2Fprop%2Fdirect%2F%3E%0A%0ASELECT%20%3Fid%20%3Fvalue%20WHERE%20%7B%0A%20%20%20%7B%20SELECT%20DISTINCT%20%3Fid%20%20WHERE%20%7B%0A%20%20%20%20%20%3Fplace%20wdt%3AP17%20%3Fid%20.%0A%20%20%20%7D%20%7D%0A%20%20%20%3Fid%20wdt%3AP299%20% [21:08:49] (graph definitions are complex enough so that sparql gets lost in there) [21:08:50] 3Fcode%20.%0A%20%20%20%3Fid%20wdt%3AP1082%20%3Fvalue%20.%0A%7D%0A" [21:08:55] that'S not something i want people to edit [21:08:57] or see [21:09:08] exactly - that's just one line out of a hundred! :) [21:09:29] benestar: query templates would live on wikidata, yes. [21:09:32] DanielK_WMDE, i agree with you :) [21:09:51] DanielK_WMDE: so one could embed them like images from commons, right? [21:10:11] that will get really interesting :) [21:10:18] yurik: the worst lines. but yes, the graph definition should also live on a separate page. and be combined with the query, to make a graph. [21:10:20] anyway [21:10:32] that i'm not sure about [21:10:56] DanielK_WMDE: what do you think about http://pastebin.com/U6TifCpb ? [21:11:00] DanielK_WMDE, israel? [21:11:10] benestar: i saw your patches. seems to go into the right direction. will have to look at the details again later. [21:11:16] yurik: oh yea! [21:11:20] awesome! [21:11:24] ok :) [21:11:46] benestar, use 5.5 notation :) [21:12:02] yurik: we didn't decide on that yet in Wikibase [21:12:14] core has it already ) [21:12:14] but perhaps we will do the same as core does? [21:12:46] as far as i know, core has already started adding it [21:13:10] is there a compact syntax for lambdas? [21:15:28] * DanielK_WMDE -> food [21:16:05] yurik: there is not, not even in PHP 7 [21:16:13] bummer [21:16:22] DanielK_WMDE, that's =>, not -> [21:16:33] and $food [21:16:41] yurik: related rant https://www.entropywins.wtf/blog/2016/01/30/missing-in-php7-function-references/ [21:17:08] short lambas would mitigate that to some extend as well [21:17:41] heeeeeyys [21:17:43] *a [21:22:55] oh my god. double lazowik and benestar at the same time on IRC! What does it mean! [21:23:16] it means… [21:23:21] it's valentines! [21:23:33] haha [21:27:04] ok, off to bed, thanks everyone! [21:27:08] will finish the disaster graph later [21:27:22] :D [23:36:56] RECOVERY - wikidata.org dispatch lag is higher than 300s on wikidata is OK: HTTP OK: HTTP/1.1 200 OK - 1538 bytes in 0.173 second response time