[07:52:54] <Thiemo_WMDE>	 Adrian_WMDE: https://integration.wikimedia.org/ci/job/mwext-qunit-composer/452/consoleFull
[07:53:03] <Adrian_WMDE>	 Thiemo_WMDE: https://phabricator.wikimedia.org/T126512
[08:44:13] <hoo>	 aude: Have some birthday cats [[File:{{#property:P18|from=Q146}}]]
[08:44:13] <AsimovBot>	 Caracteres inválidos en el enlace «File:{{#property:P18»; no están permitidos: #<>[]|{}
[10:31:26] <hashar>	 Tobi_WMDE_SW: aude: addshore: jzerebecki:  hello!  if you see any strange tests error such as memory allocation errors in Jenkins.  Please poke me
[10:31:41] <hashar>	 the CI slaves got overloaded yesterday, we pooled six new ones but apparently they dont have enough RAM :-(((
[10:32:57] <Tobi_WMDE_SW>	 hashar: indeed, there was a strange error today here in PS1: https://gerrit.wikimedia.org/r/#/c/269922/
[10:33:16] <Tobi_WMDE_SW>	 but it was gone after a recheck
[10:34:13] <hashar>	 Tobi_WMDE_SW: that one looks like the TimingTest::testMark test has a race condition :-)
[10:34:25] <aude>	 hashar: ok thanks
[11:14:07] <Adrian_WMDE>	 hashar: Wikibase doesn't pass at all since yesterday 6 PM
[11:14:13] <Adrian_WMDE>	 See https://gerrit.wikimedia.org/r/#/c/269919 for a nice history
[11:21:36] <Reedy>	 aude: Yup, it's a big bug in PhpStorm. lol
[11:21:45] <Reedy>	 https://youtrack.jetbrains.com/issue/WI-30685
[11:22:56] <hashar>	 Adrian_WMDE: oh that is the browser tests running against the patch doo
[11:23:24] <Adrian_WMDE>	 Not the WikidataBrowserTests, though, but a small set of tests in Wikibase.git
[11:24:56] <hashar>	 maybe it is a regression in one of the dependencies
[11:25:35] <aude>	 Reedy: :(
[11:26:23] <Adrian_WMDE>	 hashar: doesn't always fail, see https://gerrit.wikimedia.org/r/#/c/268387/
[11:26:26] <hashar>	 https://integration.wikimedia.org/ci/job/mwext-mw-selenium-composer/1395/artifact/log/Non%20existing%20item%3A%20Edit%20tab.png :D
[11:28:50] <Adrian_WMDE>	 The notice shouldn't be an issue imho
[11:29:01] <Adrian_WMDE>	 As for the skin, that's pretty bad
[11:29:23] <hashar>	 yeah looking at the console, the Vector skin is cloned
[11:29:32] <aude>	 Adrian_WMDE: daniel saw issues related to that yesterday, afaik
[11:29:36] <aude>	 DanielK_WMDE: ^
[11:30:00] <Adrian_WMDE>	 I always have issues with regard to how skins are loaded locally
[11:30:01] <aude>	 but then they got fixed.  i maybe core was broken and he restarted apache also
[11:30:04] <hashar>	 there is definitely a skins/Vector dir at least
[11:31:38] <aude>	 Adrian_WMDE: the sites table thing... i think one entry for the wiki could be added in update.php / during install
[11:32:03] <Adrian_WMDE>	 I don't have an entry for my local wiki either. Lots of notices, no actual issue
[11:32:37] <jzerebecki>	 Adrian_WMDE: yes we should do that
[11:33:06] <Adrian_WMDE>	 Did you mean aude?
[11:33:37] <aude>	 https://phabricator.wikimedia.org/T126596
[11:34:02] <jzerebecki>	 err yes. irrsis autocomplete still trips me up.
[11:34:14] <aude>	 Adrian_WMDE: probably just some things might not work fully if the entry is missing
[11:34:23] <DanielK_WMDE>	 hashar: load.php returned an empty document for me. i think it's related to a new version of Less being used after the switch to 5.5
[11:34:27] <aude>	 nothing should be horribly broken
[11:34:51] <DanielK_WMDE>	 hashar: i'm not sure what fixed it, but running composer update, deleting the cached Less files in /tmp, and restarting apache did it for me
[11:35:00] <DanielK_WMDE>	 it did *not* work until i restarted apache
[11:35:02] <hashar>	 that is crazy
[11:35:14] <DanielK_WMDE>	 seems like something bad got stuck in the byte code cache
[11:35:24] <Adrian_WMDE>	 Or memcache
[11:35:26] <hashar>	 the less files are supposedly generated in a temp dir which is garbage collected between builds
[11:35:31] <DanielK_WMDE>	 Adrian_WMDE: i didn't restart memcached
[11:35:38] <hashar>	 there is no memcache on ci
[11:35:44] <hashar>	 but yeah maybe some byte code cache hits us
[11:35:50] <hashar>	 I can try mass restarting apache
[11:37:30] <Adrian_WMDE>	 DanielK_WMDE: I meant accel
[11:40:58] <hashar>	 at least it started falling fairly recently https://integration.wikimedia.org/ci/job/mwext-mw-selenium-composer/buildTimeTrend
[11:41:08] <hashar>	 and on various slaves
[11:42:04] <DanielK_WMDE>	 hashar: check how it correlates with composer pulling in the newer version of Less and OOJS
[11:43:29] <DanielK_WMDE>	 hashar: also, try to rm /tmp/lessphp_*
[11:43:49] <DanielK_WMDE>	 may not be needed. but it was part of the magic ritual that made it work for me
[11:44:59] <hashar>	 DanielK_WMDE: I have verified, it is properly garbage collected
[11:46:26] <hashar>	 salt -v '*' cmd.run '/etc/init.d/apache2 restart'
[11:51:52] <addshore>	 aude: do you have flow checked out by any chance?
[11:51:57] <addshore>	 could I get you to try something?
[12:04:57] <aude>	 addshore: i might
[12:05:00] <aude>	 what is it?
[12:05:13] <aude>	 not sure i have it setup correctly
[12:05:14] <addshore>	 ./extensions/Flow/tests/phpunit/Notifications/NotifiedUsersTest.php
[12:05:40] <addshore>	 well, as far as I can see on the extension page it just needs Flow, Echo and thats it...
[12:05:49] <addshore>	 but I get Flow\Exception\InvalidInputException: Unknown input to UUID class
[12:05:56] <addshore>	 also, happy aude day! ;)
[12:06:10] <aude>	 hmmm
[12:06:49] <jzerebecki>	 Thiemo_WMDE: why did this https://gerrit.wikimedia.org/r/#/c/269919/ get merged while https://gerrit.wikimedia.org/r/#/c/269948/1 failed in selenium?
[12:06:55] <addshore>	 I just tried it on a totally clean sqlite db too and same error :/
[12:07:03] <Thiemo_WMDE>	 good question. :(
[12:07:20] <addshore>	 it seems to be passing a 14 char UUID to something that wont accept a 14 char input...
[12:07:28] <jzerebecki>	 Thiemo_WMDE: also how did the selenium selector get broken?
[12:07:35] <aude>	 addshore: is this on master?
[12:07:43] <addshore>	 aude: yeh, master of flow MW and echo
[12:08:20] <Thiemo_WMDE>	 jzerebecki: The two CSS selectors I'm fixing are not used anywhere. I do not think this is the actual reason for the failures. Good to have them fixed anyway.
[12:09:07] <Thiemo_WMDE>	 jzerebecki: The message with the two links these selectors pick have changed, but this was a long time ago.
[12:09:40] <jzerebecki>	 oh ok
[12:11:22] * aude installs echo
[12:11:28] <addshore>	 :D
[12:11:31] <aude>	 tests fail horribly w/o it
[12:11:36] <addshore>	 haha!
[12:12:11] <addshore>	 also found a bug with sqlite and flow, cant run update.php a second time as it complains about stuff already being in the db it is trying to add again
[12:12:19] <addshore>	 https://phabricator.wikimedia.org/T126587
[12:12:55] <addshore>	 trying to debug my core change to see what the Flow tests fail from it. but running the tests even on master fails for me, but with a different error ;)
[12:13:05] * aude hates not knowing now if i can/should do wfLoadExtension or the old way
[12:13:15] * addshore is still doing it the old way ;)
[12:13:17] <aude>	 has to check which it is
[12:13:19] <aude>	 ok
[12:13:35] <jzerebecki>	 mirgrate all the extensions :)
[12:21:52] <aude>	 addshore: tests for flow still fail horribly for me
[12:22:09] <addshore>	 how about just the NotifiedUsersTest ?
[12:22:23] <aude>	 some stuff optionally needs liquid threads, spam blacklist and checkuser
[12:22:33] <aude>	 but those tests are skipped
[12:23:05] <aude>	 Request to parsoid for 'wikitext' to 'html' conversion of content connected to title "Topic:Sy6pv3qq8a3evwot" failed:
[12:23:10] * aude needs to start parsoid
[12:23:21] <addshore>	 wait, the tests need parsoid....?
[12:23:45] <aude>	 something does, but i have it configured ot use my local parsoid
[12:23:53] <addshore>	 ahhhh
[12:23:53] <aude>	 probably the default config is wmf parsoid
[12:24:35] <aude>	 i see some notices and warnings
[12:24:47] <aude>	 but already looking better
[12:26:26] <aude>	 addshore: NotifiedUsersTest passes for me
[12:26:56] <aude>	 DomCrawler component is not available. is also apparently optionally used
[12:27:16] <addshore>	 wahhh
[12:27:22] <addshore>	 how about if you use the wmf parsoid?
[12:27:58] <jzerebecki>	 i think domcrawler is pullet in by composer
[12:29:55] * addshore will have to solve the failure of his before he can work out why the core patch makes it fail....
[12:31:17] <aude>	 i might have to add flow to my composer.local.json
[12:31:26] * aude has not read the install instructions
[12:35:36] <addshore>	 looks like abuse filter is required by flow for tests too otherwise some fail hard
[12:35:46] <addshore>	 Error: 1 no such table: unittest_abuse_filter
[12:35:55] <addshore>	 or i guess at least tables need to be there
[12:36:07] <aude>	 i have abusefilter
[12:36:19] * aude has disabled spamblacklist for some reason
[12:36:40] <addshore>	 I just restarted with a totally clean db ;)
[12:37:26] <addshore>	 *keeps digging*
[12:37:30] <aude>	 ok
[12:37:49] <Adrian_WMDE>	 hashar: 'Source file /mnt/jenkins-workspace/workspace/composer-php55/client/tests/phpunit/includes/Hooks/DataUpdateHookHandlersTest.php does not exist' in https://gerrit.wikimedia.org/r/#/c/269951/
[12:38:09] <aude>	 addshore: hashar i think the directory got renamed
[12:38:34] <addshore>	 aude: for the full run of Flow tests I get Tests: 516, Assertions: 1476, Failures: 7, Errors: 2, Skipped: 58
[12:38:44] <aude>	 oh, maybe not. ( i just saw https://gerrit.wikimedia.org/r/#/c/269643/ but doesn't rename)
[12:38:52] <addshore>	 SpamBlacklist would indeed mean less skips!
[12:39:05] <addshore>	 Parsoid would also mean less skips!
[12:39:33] <addshore>	 and yeh LQT
[12:40:42] <addshore>	 https://www.irccloud.com/pastebin/LXRdKapc/
[12:41:10] <aude>	 i think you have to add config for parsoid
[12:41:15] <aude>	 but can use wmf's parsoid
[12:41:31] <addshore>	 random guess is these other failures are due to windows?
[12:41:41] <addshore>	 as you dont get thouse in that paste?!
[12:42:21] <addshore>	 bah, but aude, if I run the whole set of tests the tests that did fail no longer do....
[12:42:39] <addshore>	 only if I run them by themselves....
[12:42:49] <aude>	 addshore: on Special:EnableFlow I get The error message received was: Could not open '/tmp/mw-UIDGenerator-UID-88'.
[12:42:55] <aude>	 when i click submit
[12:43:24] <addshore>	 I get a Permission error ;)
[12:43:25] <aude>	 could be something wrong with my settings or permissions
[12:43:45] <addshore>	 I think I saw something about an unreadable file in the docs
[12:43:50] <hashar>	 oh my
[12:43:56] <aude>	 $wgGroupPermissions['sysop']['editcontentmodel'] = true;
[12:44:15] * aude notes setting up parsoid and visual editor were both super easy in comparisoin
[12:44:22] <aude>	 probably way easier than wikibase also
[12:45:29] <aude>	 addshore: i don't get any of $wgGroupPermissions['sysop']['editcontentmodel'] = true;
[12:45:33] <aude>	 errr https://www.irccloud.com/pastebin/LXRdKapc/
[12:45:52] <addshore>	 same 7 fails as me?
[12:46:17] <addshore>	 aude: what about if you just do this?
[12:46:18] <addshore>	 php ./tests/phpunit/phpunit.php ./extensions/Flow/tests/phpunit/Notifications/NotifiedUsersTest.php
[12:46:26] <hashar>	 Adrian_WMDE: I have no idea why client/tests/phpunit/includes/Hooks/DataUpdateHookHandlersTest.php  would suddenly disappear. That does not make any sense to me
[12:48:54] <Adrian_WMDE>	 I'm quite against sudden disappearances, too
[12:53:37] <aude>	 addshore: i documented all my failures in phabricator
[12:53:44] <aude>	 they all seem different than yours
[12:53:54] <addshore>	 aude: cool! I just added the main one of mine to https://phabricator.wikimedia.org/T126603?workflow=create
[12:54:20] <addshore>	 thanks for your help aude!
[12:55:16] <aude>	 ok
[12:55:19] <aude>	 good luck
[12:55:45] <aude>	 i do generally have the php things installed that are recommended in mediawiki's composer.json
[12:56:05] <aude>	 also i am using hhvm (if somehow that is different)
[13:32:43] <hoo>	 ugh https://www.wikidata.org/wiki/Special:Contributions/87.64.190.16
[13:32:49] <hoo>	 looked out bot doing mass creations
[13:32:55] <hoo>	 also not sure baout notability :S
[13:35:41] <aude>	 hoo: https://josm.openstreetmap.de/ticket/9775?cversion=0&cnum_hist=18
[13:36:16] <aude>	 would be better if they logged in
[13:37:12] <aude>	 (and think he/she is mostly logged in doing this)
[13:37:56] <hoo>	 I see
[13:38:00] <hoo>	 thanks for the hint
[13:38:08] <hoo>	 do you have any way to contact them?
[13:38:10] <aude>	 if streets in the netherlands are notable, then i think schools are
[13:38:32] <hoo>	 Sure, but might need some initial discussion, depending on how many schools there are
[13:39:00] <aude>	 yeah
[13:39:09] <aude>	 see pm
[13:39:21] <HakanIST>	 in istanbul there're 1281 high scools 
[13:45:15] <intgr>	 Is Wikipedia watchlist integration working for anyone? I have "Show Wikidata edits in your watchlist" checked in my enwiki preferences, but I've never seen any Wikidata edits show up.
[13:46:12] <hoo>	 intgr: That only works in the normal watchlist, not the enhanced one
[13:46:43] <hoo>	 Not sure how that's called in the settings
[13:47:21] <intgr>	 Ah, thanks! Can I turn that off using for example GET parameters?
[13:47:36] <hoo>	 Yes, but that's not nice
[13:47:51] <hoo>	 aude: ^
[13:54:26] <intgr>	 I tried toggling a few settings, but none of them had the effect.
[13:56:18] <hoo>	 I need to go
[13:56:29] <hoo>	 aude will surely be able to help
[13:56:39] <intgr>	 Ok, thanks
[14:03:37] <aude>	 intgr: https://en.wikipedia.org/wiki/Special:Watchlist?enhanced=0 disables enhanced mode (or can be set / unchecked in preferences)
[14:04:40] <aude>	 https://phabricator.wikimedia.org/T46874 is the tracking bug 
[14:05:20] <aude>	 https://en.wikipedia.org/wiki/Special:Preferences#mw-prefsection-rc (uncheck group changes)
[14:06:19] <intgr>	 Ahh, it's changed under "Recent Changes" tab in prefs, but applies to watchlist as well, that's why I didn't find it
[14:06:22] <intgr>	 Thanks
[14:16:36] <aude>	 yeah, it's confusing :(
[14:42:54] <intgr>	 mdupont: Your bot is creating entries for enwiki talk pages, please shut it off
[14:46:31] <intgr>	 Any admins around who could block it? See https://www.wikidata.org/wiki/Special:Contributions/MdupontDataBot , appears to violate WD notability policy.
[14:51:08] <intgr>	 HakanIST: Ping?
[14:53:03] <intgr>	 Thanks
[14:53:28] <sjoerddebruin>	 mdupont: ping me when you need a nuke
[15:22:16] <aude>	 i think we need a new wikidata build....
[15:31:20] <DanielK_WMDE>	 Lydia_WMDE: https://en.wikipedia.org/wiki/Wikipedia:Wikipedia_Signpost/2016-02-03/Op-ed
[16:14:42] <aude>	 Lydia_WMDE: i am proposing https://phabricator.wikimedia.org/T126596 for the next sprint
[16:33:53] <JulianH>	 Hi there! Me and a fellor student helped to develop a new datatype 'Mathematical expression', which went live on 09.02. Naturally there is no property at the moment with this datatype, but we started a discussion here: https://www.wikidata.org/wiki/Wikidata:Property_proposal/Natural_science#defining_formular 
[16:35:03] <JulianH>	 Is there a possibility to speed this process up, so we become able to get some feedback about the new datatype within next week?
[17:01:40] <JulianH>	 The correct link is https://www.wikidata.org/wiki/Wikidata:Property_proposal/Natural_science#defining_formula
[17:21:09] <DanielK_WMDE>	 hashar, jzerebecki: full disk? https://integration.wikimedia.org/ci/job/mwext-Wikibase-client-tests-sqlite-hhvm/8217/console
[17:21:17] <DanielK_WMDE>	 lots of huge logs from all teh failures, perhaps?
[17:22:10] <jzerebecki>	 DanielK_WMDE: -> https://phabricator.wikimedia.org/T126615 ; checking
[17:22:58] <jzerebecki>	 DanielK_WMDE: nope 187M available on /var/lib/mysql
[17:25:10] <DanielK_WMDE>	 jzerebecki: then why does the sqlite check die with a disk IO error?
[17:25:28] <jzerebecki>	 DanielK_WMDE: link?
[17:25:34] <DanielK_WMDE>	 jzerebecki: still the same patch, btw
[17:25:38] <DanielK_WMDE>	 https://gerrit.wikimedia.org/r/#/c/269718/
[17:25:53] <aude>	 i think paladox posted it in the releng channel
[17:26:13] <jzerebecki>	 oooh thought that was mysql
[17:26:23] <aude>	 oh
[17:26:32] <aude>	 no, sqlite
[17:26:41] <DanielK_WMDE>	 first it was qunit failing, then sqlight, then mysql, now it's sqlite again
[17:27:00] <DanielK_WMDE>	 oh, and selenium was failing too, due to some skin fuckup
[17:36:14] <jzerebecki>	 DanielK_WMDE: should work now. new instances with 2GB for 1 job went OOM :( .
[17:36:23] <jzerebecki>	 but those were disabled
[17:36:53] <DanielK_WMDE>	 let's hope for the best :)
[17:37:27] <DanielK_WMDE>	 more fun
[17:37:40] <DanielK_WMDE>	 jzerebecki: DB connection error: Access denied for user 'jenkins_u0'@'localhost' (using password: YES) (127.0.0.1:3306).
[17:37:58] <DanielK_WMDE>	 oh and ERROR 1007 (HY000) at line 2: Can't create database 'jenkins_u0_mw'; database exists
[17:38:02] <jzerebecki>	 link?
[17:38:24] <DanielK_WMDE>	 https://gerrit.wikimedia.org/r/#/c/270007/
[17:39:02] <DanielK_WMDE>	 also "DB connection error: Cannot return last error, no db connection", but i guess that's the same as the "access denied" one
[17:39:43] <jzerebecki>	 DanielK_WMDE: from before those low mem slaves where disabled
[17:40:47] <DanielK_WMDE>	 jzerebecki: you mean these errors have the same root cause as the sqlite one?
[17:41:06] <jzerebecki>	 DanielK_WMDE: I think so
[17:41:18] <DanielK_WMDE>	 i don't know enough about the ci infrastructure to understand what "low mem slaves where disabled" means. are these mysql slaves?
[17:41:33] <DanielK_WMDE>	 but if they went offline, i'd expect different error messages
[17:44:42] <jzerebecki>	 DanielK_WMDE: what I meant is that some new slaves to run jobs where added. those had only 2GB but also only one job slot. seems that was not enough memory for many mediawiki jobs. so to not cause more failures a few minutes ago they h_ashar disabled them. now those failures should happen anymore.
[17:46:47] <DanielK_WMDE>	 looks like there is quite a backlog. takes forever for jobs to even start
[17:46:56] <DanielK_WMDE>	 or do we have to trigger them again?
[17:50:48] <jzerebecki>	 DanielK_WMDE: no backlog, one 270007,1 seems to somehow be stuck which I don't understand
[17:53:57] <DanielK_WMDE>	 jzerebecki: but tests don't run. nothing is happening on https://gerrit.wikimedia.org/r/#/c/269718/ or https://gerrit.wikimedia.org/r/#/c/270007/ or https://gerrit.wikimedia.org/r/#/c/270008/
[18:01:10] <jzerebecki>	 DanielK_WMDE: one of them was the problem I didn't understand. so retried. the others have results on gerrit. now things are queued...
[18:03:24] <DanielK_WMDE>	 jzerebecki: what do you think of https://gerrit.wikimedia.org/r/#/c/270008/ - should i just retry?
[18:04:46] <jzerebecki>	 DanielK_WMDE: it is queued and waiting for job slots, see https://integration.wikimedia.org/zuul/
[18:06:07] <DanielK_WMDE>	 oh, didn't see you triggering a recheck.
[18:06:11] <DanielK_WMDE>	 i wonder why it failed in the first place
[18:06:31] <d3r1ck>	 Hi DanielK_WMDE 
[18:06:45] <jzerebecki>	 yea that that one recheck was somehow dropped was very weird
[18:08:27] <DanielK_WMDE>	 jzerebecki: i found it more strange that the php55 job died without an error message, while the hhvm run passed
[18:08:30] <DanielK_WMDE>	 d3r1ck: hi
[18:17:39] <Lorena_>	 Hello
[18:24:35] <andre__>	 hi Lorena
[18:25:50] <d3r1ck>	 andre__: Hi
[18:27:19] <andre__>	 hi d3r1ck
[18:38:31] <jzerebecki>	 DanielK_WMDE: "Build timed out (after 30 minutes)" was the error. jenkins killed it after 30min.
[18:38:46] <DanielK_WMDE>	 huh :)
[18:39:13] <jzerebecki>	 probably cpu starvation by other jobs running on the same vm
[19:49:32] <jianli>	 the chinese sitelink on https://en.wikipedia.org/wiki/Hygrophila_%28plant%29 points to the wrong item. I tried to update it to "水蓑衣属" but got "The link zhwiki:水蓑衣属 is already used by item Q11130538. You may remove it from Q11130538 if it does not belong there or merge the items if they are about the exact same topic."
[19:50:14] <jianli>	 I don't know how to fix it. Can you fix it for me?
[19:53:30] <Vogone>	 jianli, hi
[19:53:37] <Vogone>	 you need to remove it from https://www.wikidata.org/wiki/Q11130538 first
[19:53:45] <Vogone>	 before you can add it to the other item
[19:55:53] <jianli>	 i got it. thanks. bye.
[21:04:37] <abian>	 Question - why are translated main pages like https://www.wikidata.org/wiki/Wikidata:Portada still running?
[21:05:25] <abian>	 We should be using https://www.wikidata.org/wiki/Wikidata:Main_Page for all languages, right?
[21:06:59] <sjoerddebruin>	 Hm, yep
[21:09:04] <abian>	 However, https://www.wikidata.org/ still redirects to https://www.wikidata.org/wiki/Wikidata:Portada using a registered account in Spanish
[21:09:58] <sjoerddebruin>	 Ah, yes.
[21:10:01] <sjoerddebruin>	 Let me fix
[21:11:03] <Vogone>	 sjoerddebruin, hm, but somehow this looks worse to me
[21:11:09] <Vogone>	 since the page title is in English then
[21:11:14] <Vogone>	 for all languages
[21:11:15] <sjoerddebruin>	 Page title is hidden
[21:11:25] <Vogone>	 not in the URL :p
[21:11:38] <sjoerddebruin>	 Should that matter?
[21:12:04] <Vogone>	 well, it doesn't look nice and confused at least me before I looked into the source text
[21:12:25] <sjoerddebruin>	 Hm
[21:12:46] <Vogone>	 more a stylistic matter than a practical one
[21:13:49] <abian>	 Commons have all their categories and meta-pages in English, including titles
[21:13:54] <abian>	 That's a bad excuse :)
[21:14:31] <GerardM>	 abian ... that is best practice for now
[21:14:39] <abian>	 But I think that writing URLs in English isn't no worrying
[21:14:47] <abian>	 Okay
[21:14:53] <abian>	 21:04:00 < abian> Question - why are translated main pages like https://www.wikidata.org/wiki/Wikidata:Portada still running?
[21:14:56] <abian>	 21:04:48 < abian> We should be using https://www.wikidata.org/wiki/Wikidata:Main_Page for all languages, right?
[21:15:15] <abian>	 Argh, *isn't worrying
[21:15:19] <Vogone>	 well, in an ideal world it should be our goal to make Wikidata and Commons comprehensible ven to people who label themselves as en-0
[21:15:25] <Vogone>	 *even
[21:28:04] <abian>	 Urgent help is needed in translation
[21:28:10] <abian>	 Even fundraising banners have, in Spanish, terrible typos
[21:28:36] <abian>	 I have received complaints and jokes for typos in these banners
[21:37:00] <abian>	 sjoerddebruin: Do you want I open a task on Phabricator? Or is configuring the main page something fast/easy to do?
[21:37:21] <sjoerddebruin>	 It's easy to do, but need more opinions.
[21:37:52] <abian>	 I thought there was consensus on this :S
[21:40:56] <abian>	 Is it easy to configure localized URLs with "Main Page" content?
[22:43:41] <mdupont>	 shit