[00:18:27] 6MediaWiki-Core-Team, 10Wikidata-Query-Service: Wikidata Query: Experiment with RDF exports - https://phabricator.wikimedia.org/T91691#1115420 (10Smalyshev) [03:27:14] legoktm: What color should I paint this bikeshed? I want a new phab project to track debug logging things. People keep adding them to MediaWiki-Logging. So a tag for "debug-logs" or a "MediaWiki-Debug-Logs" project? Or something else? [03:42:03] 6MediaWiki-Core-Team, 10Librarization, 10MediaWiki-Installer, 5MW-1.25-release: Missing log library when trying to install MediaWiki via WebInstaller (mw-config/index.php) - https://phabricator.wikimedia.org/T88951#1115675 (10bd808) [03:46:48] bd808: A MediaWiki-Debug-Logs project sounds good [03:47:23] 6MediaWiki-Core-Team, 6Release-Engineering, 10Wikimedia-Logstash, 7HHVM: Log php fatals with full backtraces again (fatal.log on fluorine) - https://phabricator.wikimedia.org/T89169#1115698 (10bd808) [03:47:24] 6MediaWiki-Core-Team, 5Patch-For-Review: Convert JobRunner.php to PSR-3 logging and add levels - https://phabricator.wikimedia.org/T87521#1115697 (10bd808) [03:47:26] 6MediaWiki-Core-Team, 10Incident-20150205-SiteOutage, 10Wikimedia-Logstash, 6operations, 5Patch-For-Review: Decouple logging infrastructure failures from MediaWiki logging - https://phabricator.wikimedia.org/T88732#1019000 (10bd808) [03:47:55] legoktm: k. I'll file a request for one [03:53:43] {{done}} https://phabricator.wikimedia.org/T92599 [03:54:15] +1'd [06:08:44] 6MediaWiki-Core-Team, 10Continuous-Integration, 7Composer: Come up with non sucky solution for running "composer test" on repos that have vendor/ checked in - https://phabricator.wikimedia.org/T92605#1115821 (10Legoktm) 3NEW [06:09:06] bd808: ^ [14:07:51] * anomie is sorry for saying something dumb yesterday [14:12:34] <^d> anomie: We all do dumb things :) [14:16:49] 6MediaWiki-Core-Team, 10MediaWiki-API, 5Patch-For-Review: Clean up ApiResult and ApiFormatXml, create new formatversion - https://phabricator.wikimedia.org/T76728#1116604 (10Anomie) [14:52:17] anomie: I say at least 3 dumb things before breakfast each day. [14:55:44] 6MediaWiki-Core-Team, 10Continuous-Integration, 7Composer: Come up with non sucky solution for running "composer test" on repos that have vendor/ checked in - https://phabricator.wikimedia.org/T92605#1116654 (10bd808) One approach would be to split the deployable vendor repo from the project repo and only co... [15:25:54] bd808: do you know if/how during scap, things in bits get expired in varnish (e.g. files that were touched)? [15:26:12] stuff like http://bits.beta.wmflabs.org/static-master/extensions/Wikidata/extensions/Wikibase/lib/resources/jquery.wikibase/themes/default/jquery.wikibase.entityview.css [15:26:57] aude: I'm honestly not sure. If they do get purged it is from something that RL does and not a function of scap [15:27:13] https://github.com/wikimedia/operations-puppet/blob/2ebe5a3bbd6e3cb2adac8a31a67aa491bd4aedf2/modules/scap/manifests/scripts.pp#L62 [15:27:14] They may just have low cache TTLs [15:27:22] i see a script but don't see where it used at all [15:28:01] even with debug=true, i get old stuff from bits [15:28:06] that may be the script used by l10nupdate each night... [15:28:14] hm [15:28:35] nope [15:28:49] l10nupdate does this -- https://github.com/wikimedia/operations-puppet/blob/2ebe5a3bbd6e3cb2adac8a31a67aa491bd4aedf2/modules/scap/files/l10nupdate-1#L125-L130 [15:28:50] if i explicitly purge a url on beta in varnish [15:28:58] then the right things happen [15:29:03] ok [15:29:23] purge-varnish looks to be a helper script for manual purges [15:29:28] i see [15:29:41] Ask Roan/Timo about RL cache expiration [15:29:44] ok [15:30:02] it makes it frustrating and difficult to use beta when bits don't update [15:30:21] (https://phabricator.wikimedia.org/T65034) [15:30:35] and thought maybe something was missing in the scap process there [17:37:30] 6MediaWiki-Core-Team, 10MediaWiki-Debug-Logging, 6Release-Engineering, 10Wikimedia-Logstash, 7HHVM: Log php fatals with full backtraces again (fatal.log on fluorine) - https://phabricator.wikimedia.org/T89169#1117046 (10Legoktm) [17:37:35] 6MediaWiki-Core-Team, 10Incident-20150205-SiteOutage, 10MediaWiki-Debug-Logging, 10Wikimedia-Logstash, and 2 others: Decouple logging infrastructure failures from MediaWiki logging - https://phabricator.wikimedia.org/T88732#1117047 (10Legoktm) [17:37:37] 6MediaWiki-Core-Team, 10Librarization, 10MediaWiki-Debug-Logging, 10MediaWiki-Installer, 5MW-1.25-release: Missing log library when trying to install MediaWiki via WebInstaller (mw-config/index.php) - https://phabricator.wikimedia.org/T88951#1117050 (10Legoktm) [17:37:39] 6MediaWiki-Core-Team, 10MediaWiki-Debug-Logging, 5Patch-For-Review: Convert JobRunner.php to PSR-3 logging and add levels - https://phabricator.wikimedia.org/T87521#1117048 (10Legoktm) [17:39:52] 6MediaWiki-Core-Team, 7HHVM, 7Upstream: HHVM with FastCGI does not support streaming output - https://phabricator.wikimedia.org/T91468#1117065 (10bd808) [17:52:00] legoktm: https://gerrit.wikimedia.org/r/#/c/195842/ [17:55:58] autismcat: I don't really know what to do about https://phabricator.wikimedia.org/T92455 btw...should we just have User / CentralAuth not use master when caching? [17:58:48] 6MediaWiki-Core-Team, 7Epic: General authentication improvements for MediaWiki - https://phabricator.wikimedia.org/T90925#1117182 (10bd808) [17:58:49] 6MediaWiki-Core-Team, 10MediaWiki-Authentication-and-authorization, 7Epic: Modernize MediaWiki authentication system - https://phabricator.wikimedia.org/T89459#1117183 (10bd808) [18:00:13] legoktm: ideally functions that get User objects for parsing, special pages lists, and most stuff could just use a slave (the cache could be checked thought not backfilled) [18:00:35] for now, the main User for the request should keep the cache/master logic to avoid weirdness though [18:13:23] ok, I'll note that on the bug [18:45:15] AaronSchulz: https://gerrit.wikimedia.org/r/195448 [18:47:55] also, what would be a good sleep value for that script to avoid overloading stuff? it's a little under 3 million notifications [19:20:11] legoktm: 1 second per 100 jobs? in theory jobqueue has waitForBackups() but that does nothing with redis [19:20:21] you could also check the job queue count [19:36:37] AaronSchulz: ok, at how many jobs would I want to stop? [19:42:12] maybe 100k [19:49:01] AaronSchulz: implemented in https://gerrit.wikimedia.org/r/196629 [20:03:21] 6MediaWiki-Core-Team, 10MassMessage, 5Patch-For-Review: Strict standards: Declaration of MassMessageServerSideJob::makeText() should be compatible with MassMessageJob::makeText($stripTildes = false) in MassMessage/includes/job/MassMessageServerSideJob.php on l... - https://phabricator.wikimedia.org/T92650#1117618 [20:04:40] legoktm: if the script fails at 110k jobs, can you resume it properly? [20:05:20] AaronSchulz: yeah, it uses the database as state so it is resumable [20:17:44] legoktm, Keegan: \o/ SUL is getting really real! [20:17:59] Verily [20:18:18] Hopefully it will pass by and occur before I even accept the reality [20:19:25] There should be champaign and dancing in the streets when the last unattached account it gone :) [20:30:49] bd808: I might cry. [20:30:51] It might happen. [20:31:47] tears of joy are acceptable [20:39:35] 6MediaWiki-Core-Team, 10Wikimedia-General-or-Unknown, 5Patch-For-Review: Existed pages without ability to reach and obviously wrong namespace - https://phabricator.wikimedia.org/T87645#1117764 (10Mdann52) Just possibly happened again - see [[ https://en.wikipedia.org/wiki/Special:Log?action=view&page=Wikiped... [20:43:01] 6MediaWiki-Core-Team, 10Wikimedia-General-or-Unknown, 5Patch-For-Review: Existed pages without ability to reach and obviously wrong namespace - https://phabricator.wikimedia.org/T87645#1117788 (10Betacommand) I really doubt this is a mediawiki issue. Odds are its a issue with title parsing and namespace assu... [20:49:52] legoktm: I don't see any --start param in the script [20:50:05] if it fails and you re-run it, won't there be duplicated mass-messages? [20:50:57] AaronSchulz: no, once a user's job is queued, it marks them as notified, which updates the utr_status column. And it only pulls users out of the db which have a utr_status of 0 [20:52:03] 6MediaWiki-Core-Team, 6operations, 7Wikimedia-log-errors: rbf1001 and rbf1002 are timing out / dropping clients for Redis - https://phabricator.wikimedia.org/T92591#1117815 (10aaron) [20:52:11] ah, OK [20:53:15] 6MediaWiki-Core-Team, 6operations, 7Wikimedia-log-errors: rbf1001 and rbf1002 are timing out / dropping clients for Redis - https://phabricator.wikimedia.org/T92591#1117816 (10chasemp) side note, I think the replication config was moved to hiera and maybe this just isn't working? https://phabricator.wikimed... [21:03:03] 6MediaWiki-Core-Team, 6CA-team, 10SUL-Finalization, 6operations: db1068 (s4/commonswiki slave) is missing data about at least 6 users - https://phabricator.wikimedia.org/T91920#1117862 (10Keegan) [21:03:34] 6MediaWiki-Core-Team, 10Wikimedia-General-or-Unknown, 5Patch-For-Review: Existed pages without ability to reach and obviously wrong namespace - https://phabricator.wikimedia.org/T87645#1117865 (10Cyberpower678) >>! In T87645#1117788, @Betacommand wrote: > I really doubt this is a mediawiki issue. Odds are it... [21:42:02] ori: https://gerrit.wikimedia.org/r/#/c/196507/ [22:00:11] The_Real_NSA: can you spot check https://phabricator.wikimedia.org/P398 and make sure it looks the same as yours, except that the headers are gone? [22:09:19] legoktm: doing [22:11:36] legoktm: Looks good to me [22:11:45] * The_Real_NSA is still annoyed that Turkish didn't get updated [22:20:54] Keegan: ok, I'm going to start out with testwikidatawiki because it only has 5 users, one of which is me [22:22:34] legoktm: :) [22:23:03] I c/p the text into my talk page on various language wikis and previewed to check the subst: and rtl/ltrs [22:23:24] I found a mistake in sq the other day doing that. Anyway, it's all good as far as rendering goes [22:24:26] https://test.wikidata.org/wiki/User_talk:Lego-test [22:24:56] oh, I added a hidden to the end of the message [22:26:32] Keegan: ^ if that looks good I'll do the rest of the test wikis [22:27:38] It does look good [22:27:51] * Keegan goes to request a global rename to User:Lego-test~testwikidatawiki [22:32:29] umm [22:32:34] https://test2.wikipedia.org/wiki/Special:GlobalRenameRequest [22:32:38] wut [22:33:55] bd808: ^ do you know why that doesn't work on test2? [22:34:47] just checked a bunch of random wikis and it works everywhere else... [22:37:36] it's showing up on https://test2.wikipedia.org/wiki/Special:Version... [22:38:13] The plot thickens [22:38:26] > var_dump(SpecialPageFactory::exists('SpecialGlobalRenameRequest')); [22:38:26] bool(false) [22:44:29] https://ta.wikipedia.org/wiki/%E0%AE%9A%E0%AE%BF%E0%AE%B1%E0%AE%AA%E0%AF%8D%E0%AE%AA%E0%AF%81:GlobalRenameRequest [22:44:53] https://or.wikipedia.org/wiki/%E0%AC%AC%E0%AC%BF%E0%AC%B6%E0%AD%87%E0%AC%B7:GlobalRenameRequest [22:44:58] ok wtf is going on [22:45:11] if this has to do with $wmgUseShortUrl, that's crazy [22:46:49] sigh [22:46:50] yes [22:46:52] it is ShortUrl [22:48:58] Keegan: ok, so, GlobalRenameRequest is broken on wikis with ShortUrl enabled. I'll have a patch in a few minutes. (cc: greg-g) [22:50:02] thanks :) [22:50:08] * greg-g sighs [22:50:09] ok [22:50:24] greg-g: it's Friday! [22:50:50] must mean it's "day we make exceptions to the no deploy friday rule"! [22:51:11] No no, it's the "Naturally this happens. It's Friday." [22:51:30] potato, potato [22:51:36] :P [22:51:57] legoktm: I'm leaving soon, you've got my blessing, godspeed [22:52:07] ty greg [23:09:41] bd808, csteipp, hoo: could one of you review https://gerrit.wikimedia.org/r/196708 ? [23:10:25] fun, hidden test failures [23:12:30] legoktm: can you fix this in the German text? DerHexer caught it https://meta.wikimedia.org/w/index.php?title=Single_User_Login_finalisation_announcement%2FPersonal_announcement%2Fde&diff=11539408&oldid=11383261 [23:12:43] * bd808 is trying to get cable modem online [23:12:45] Oh wait [23:12:49] Hang on [23:13:09] I think I ran a find and replace for that [23:13:34] and I did [23:13:35] So nm [23:14:08] ok [23:14:38] legoktm: Looks sane to me, can't vote atm. [23:14:56] Looks ok to me as well [23:15:05] But didn't test [23:15:41] 1 + 1 = +2? [23:15:53] ;) [23:17:20] You tested it, right? [23:17:30] * bd808 has a modem! [23:18:09] legoktm: Mh... Can we have both that and wgSpecialPages? That would make it more confortable to me [23:18:16] * comfortable [23:19:17] hoo: yes tested. As in don't remove the $wgSpecialPages stuff? [23:19:19] legoktm: This is magic I know nothing about [23:19:37] legoktm: Yeah [23:20:00] Also you should add a comment on why you use the hook, otherwise someone might change that at some point to simplify [23:20:01] bd808: https://github.com/wikimedia/mediawiki/blob/master/includes/specialpage/SpecialPageFactory.php#L249 [23:20:19] ok [23:22:10] // This hook can be used to remove undesired built-in special pages [23:22:36] 'SpecialPage_initList': Called when setting up SpecialPageFactory::$list, use this [23:22:37] hook to remove a core special page. [23:22:51] it works both ways [23:22:54] While this wont break anything (nothing else in core uses wgSpecialPage for anything [23:22:59] MobileFrontend uses it to conditionally add special pages [23:23:02] we're doing something undocumented [23:23:11] So I feel like we should update the docs? [23:23:19] at least [23:24:44] updated the CA patch, doing a core patch now... [23:25:56] Ok, +2ed that one :) [23:26:43] hoo: core patch is https://gerrit.wikimedia.org/r/196711 [23:27:49] {{done|+2}} [23:28:53] thanks [23:34:30] * legoktm hugs YuviPanda for deploying ShortUrl on a testwiki [23:35:15] ON FRIDAY?! [23:35:46] friday is the new monday [23:48:46] Keegan: deployed, and all wikis now have a GlobalRenameRequest [23:48:54] (I hope) [23:50:40] legoktm: heh, about ShortUrl on a testwiki https://meta.wikimedia.org/wiki/User_talk:Keegan_(WMF)#About_the_title_of_the_page [23:51:06] lol [23:51:54] ok, I'm going to do testwiki now [23:52:24] So can't people query the API and get their own copy of users to be renamed? [23:53:55] um, not currently [23:54:05] Okay [23:55:24] hmm [23:55:27] something went wrong [23:55:28] https://test.wikipedia.org/w/index.php?title=Special:RecentChanges&hidebots=0 [23:55:53] https://test.wikipedia.org/wiki/Special:AbuseLog [23:55:54] wow [23:56:12] Oh god [23:56:22] I'm going to get dinner. STRESS. [23:56:51] Of the last 811 actions, this filter has matched 801 (98.77%). On average, its run time is 7.87 ms, and it consumes 348 conditions of the condition limit.