[00:13:13] <paladox>	 DanielK_WMDE: the blue button :)
[00:13:28] <paladox>	 The start review button is the unwip button!
[08:55:21] <tgr>	 Krinkle: it's called importWiring now
[11:27:07] <legoktm>	 addshore, DanielK_WMDE: https://phabricator.wikimedia.org/T197450 sounds like it could be MCR related
[11:28:41] <addshore>	 *reads up*
[11:29:00] <addshore>	 legoktm: that just sounds like some sort of very odd crazy bug
[11:29:24] <legoktm>	 somehow the message cache is polluted
[11:30:02] <legoktm>	 nothing in Gadgets has changed recently 
[11:30:23] <addshore>	 hmmmm
[11:30:34] <addshore>	 where does mcr touch message caching? :P
[11:30:50] <legoktm>	 page content retrevial?
[11:31:01] <legoktm>	 I should probably look at the patches themselves before wildly blaming :p
[11:32:20] <addshore>	 https://gerrit.wikimedia.org/r/#/c/mediawiki/core/+/439449/ is in the .999 branch i guess, and touched caching
[11:32:30] <addshore>	 is the same issue visible between mw.org and test wiki?
[11:33:23] <legoktm>	 mw.o gadgets look fine
[11:33:49] <legoktm>	 same with test.wd
[11:34:06] <addshore>	 interesting that it is only between test2 and test then, if it was a general .999 caching issue Iwould have thought it would end up being shared between all of group0
[11:35:05] <addshore>	 The revisionStore caching defintly still includes $db->getDomainID() in the key
[11:36:07] <legoktm>	 legoktm@terbium:~$ mwscript eval.php --wiki=testwiki
[11:36:07] <legoktm>	 > var_dump(wfMessage('gadgets-definition')->text());
[11:36:12] <legoktm>	 returns test2wiki message content
[11:36:58] <legoktm>	 ohh
[11:37:01] <legoktm>	 it's the entire message cache
[11:37:10] <legoktm>	 > var_dump(wfMessage('common.css')->text());
[11:37:10] <legoktm>	 string(34992) "/* Test for proposed enwiki small change/*
[11:37:20] <legoktm>	 that's from https://test2.wikipedia.org/wiki/MediaWiki:Common.css
[11:37:29] <addshore>	 O_o
[11:40:17] * legoktm lucnh
[11:40:19] <legoktm>	 lunch*
[11:46:43] <addshore>	 lunch legoktm ? where are you?
[12:26:35] <anomie>	 addshore: Hmm, what branch was your .999 based on?
[12:26:49] <legoktm>	 addshore: I'm in Italy!
[12:27:13] <legoktm>	 addshore: also _joe_ told me that testwp has mcrouter enabled, so that could potentially be another point of variance
[12:29:25] <Krinkle>	 addshore: (What is .999?)
[12:32:32] <tgr>	 anomie: it was branched from master
[12:33:32] <legoktm>	 addshore: I'm 98% sure it's mcrouter, not MCR
[12:33:38] <legoktm>	 still debugging
[12:33:46] <_joe_>	 mcouter or the config for it?
[12:33:57] <legoktm>	 the config
[12:34:32] <legoktm>	       ["keyspace":protected]=>
[12:34:32] <legoktm>	       string(5) "local"
[12:34:48] <_joe_>	 ok
[12:40:31] <addshore>	 legoktm: yay, no MCR explosion :)
[12:41:48] <legoktm>	 addshore: I apologize for pre-emptively blaming you :p
[13:41:34] <legoktm>	 anomie: do you want to write up a few sentences for https://www.mediawiki.org/wiki/MediaWiki_1.31#Interwiki_usernames ?
[13:46:32] <anomie>	 legoktm: I gave it a try: https://www.mediawiki.org/w/index.php?title=MediaWiki_1.31&diff=2807068&oldid=2807064
[13:47:22] <legoktm>	 thanks :D
[15:45:30] <Krinkle>	 legoktm: Do you think it would make sense to create a sniff for this? https://phabricator.wikimedia.org/T7176
[15:45:46] <Krinkle>	 I suppose it might be implicit through the new/upcoming rejecting of unknown annotations
[15:45:56] <Krinkle>	 but not sure if that would cover this and/or if we can address them in time for that
[15:46:05] <Krinkle>	 Do we have a sniff yet for requiring visibility on members and methods?
[16:22:52] <bd808>	 anomie: you ran away as I was saying "see ya"
[16:48:13] <Krinkle>	 addshore: eh? https://wikitech.wikimedia.org/w/index.php?title=Server_Admin_Log&diff=1794709&oldid=1794708
[16:48:31] <Krinkle>	 noticed it due to NOTOC not being set anymore
[17:10:44] <addshore>	 Krinkle: eh, thas evil...
[17:11:07] <addshore>	 (-104,400
[17:11:28] <addshore>	 *undos*
[17:14:04] <addshore>	 Krinkle: fixed in https://wikitech.wikimedia.org/w/index.php?title=Server_Admin_Log&diff=1794820&oldid=1794819
[17:29:00] <Krinkle>	 addshore: thx :)
[17:29:12] <addshore>	 no idea how that happened
[17:29:15] <Krinkle>	 I suppose it's time for an archive anyway
[17:36:07] <addshore>	 Krinkle: .999 was a branch created on thursday evening including the MCR related patches, current on group0 for the week
[17:41:02] <Krinkle>	 addshore: ah, okay :)
[20:27:02] <legoktm>	 Krinkle: we still use @private/@protected in a few places....I'll comment
[20:27:16] <legoktm>	 Krinkle: yes we do have a sniff that requires visibility
[22:03:54] <davidwbarratt>	 what's the best way to test database updates? Do I need to install all these different databases and make sure the change(s) work?
[22:08:59] <Reedy>	 No
[22:09:11] <bd808>	 davidwbarratt: typically we care the most about mysql/mariadb and sqlite. Really nice people will also try to figure out if they work for mssql and oracle, but that is honestly seldom done
[22:09:23] <Reedy>	 And postgres? :P
[22:09:42] <bd808>	 oh yeah that is a thing too ;)
[22:48:20] <MaxSem>	 {{citation needed}}
[22:50:09] <MaxSem>	 I actually think that a) it's not required to update the non-FLOSS DB schemas when changing the main schema; and b) nobody should have to install a non-free DB to contribute to MW
[22:52:55] <Reedy>	 ( $this->callback )( $this->dbw, $this->fname );
[22:52:59] <Reedy>	 We write some weird syntax
[22:53:05] <Reedy>	 Why are we () around the callback part?
[22:53:24] <Reedy>	 because it's a variable?
[22:53:43] <MaxSem>	 yup
[22:53:50] <Reedy>	 it looks weird af
[22:54:06] <MaxSem>	 $this->callback( $this->dbw, $this->fname ) only works if there's a method called callback
[22:54:33] <MaxSem>	 looking at Flow, it overuses callbacks to a scary extent
[23:47:40] <Krinkle>	 Maybe we should strip down the CodeReview extension to just the code needed for read-only, given we don't have any real intent to test or maintain the other part.
[23:47:57] * Krinkle is looking at a depreation fix to CodeReview's udp2irc code, which is not active
[23:48:44] <Reedy>	 I believe it's still actively used elsewhere