[01:47:11] <legoktm>	 CindyCicaleseWMF: https://pingback.wmflabs.org/ is awesome! 
[01:48:23] <CindyCicaleseWMF>	 Thanks :-) I'm pretty happy with it! I want to add one more tabs with the total unique wiki count so it doesn't have to be inferred from the other data.
[01:50:37] <CindyCicaleseWMF>	 And finer granularity on MediaWiki versions will be possible once https://phabricator.wikimedia.org/T126279 is done
[01:52:07] <legoktm>	 CindyCicaleseWMF: where is the configuration stored? I found a small typo (32-bir)
[01:52:40] <CindyCicaleseWMF>	 Oh! Thanks! https://meta.wikimedia.org/wiki/Config:Dashiki:Pingback
[01:53:31] <legoktm>	 that was easy :D
[01:54:15] <CindyCicaleseWMF>	 legoktm: I know, right? I'm really happy with reportupdater and dashiki. And, the analytics folks were super helpful!
[01:55:12] <CindyCicaleseWMF>	 legoktm: Currently, the data gets generated weekly. Do you think it would be useful to do so daily? It would be easy to change. It wouldn't result in a ton more data. I guess we could see if more wikis get installed on Wednesdays or Saturdays . . .
[01:55:42] <legoktm>	 I don't see much value
[01:56:05] <legoktm>	 I'm really curious as to what the 1.27/LTS usage is like compared to other versions
[01:56:29] <CindyCicaleseWMF>	 yes, that would be good to know!
[01:56:59] <CindyCicaleseWMF>	 oddly, there is one 1.25 data point . . .
[03:58:53] <greg-g>	 CindyCicaleseWMF: that is awesome :) Thank you!
[14:45:28] <CindyCicaleseWMF>	 greg-g: you're welcome! And, the MediaWiki version graph is now annotated with when MediaWiki releases occurred. There will be finer granularity on the releases in the future.
[15:22:58] <anomie>	 legoktm: Yay, parser tests finally pass for me locally! But I guess CI is still not using Remex for other tests, since TidyTest fails locally after switching my config to Remex?
[17:05:05] <legoktm>	 anomie: hmm, probably that needs a setMwGlobals to for remex?
[17:08:25] <anomie>	 legoktm: Or we could just set remex in TestSetup.php. Either way we'd need to adjust that test too for remex.
[17:08:50] <legoktm>	 that would work too
[17:11:39] <legoktm>	 though I wonder why I didn't see that as a test failure in CI's php70 job...
[17:14:15] <anomie>	 Does the CI config set remex, or does it leave $wgTidyConfig null so MWTidy::singleton() defaults to raggett?
[17:15:07] <legoktm>	 the latter, all config overrides are set in DevelopmentSettings.php now
[17:18:53] <anomie>	 Apparently that one doesn't change between libtidy 0.99 and libtidy 5, but does change with remex.
[17:24:05] <legoktm>	 ah
[19:57:19] <no_justification>	 Do we have stated style/policy preferences for $GLOBALS['foo'] vs global $foo?
[19:57:24] <no_justification>	 And is it sniff-enforced?
[19:57:30] <no_justification>	 I know MW core tends to prefer the latter.
[19:58:46] <Reedy>	 You could infer from https://www.mediawiki.org/wiki/Manual:Coding_conventions/PHP that we do
[19:59:00] <Reedy>	 AS that only mentiones $wgFoo, not $GLOBALS['wgFoo']
[20:00:32] <no_justification>	 True
[20:01:47] <Reedy>	 But AFAIK, no and no to your questions
[20:05:26] <no_justification>	 I guess why I was getting at it was a better sense of how widely a global is used. But I'm looking at it wrong... It's not global keyword vs $GLOBALS superglobal... 
[20:05:32] <no_justification>	 Its local assignment.... 
[20:06:31] <no_justification>	 When you assign $wgTitle to $title to save two keystrokes later on you mask the usages somewhat 
[20:07:01] <no_justification>	 But you can assign $GLOBALS['wgTitle'] just the same
[20:11:21] <legoktm>	 no_justification: there are places where $GLOBALS['wgWhatever'] will break
[20:11:32] <no_justification>	 That's true
[20:11:45] <legoktm>	 installer related stuff mostly
[20:12:36] <legoktm>	 there's no sniff enforcement against it (we do have them for $_POST/$_GET etc)
[20:13:25] <no_justification>	 Do we use $_ENV anywhere?
[20:13:29] <no_justification>	 I know we use $_SERVER
[20:14:50] <legoktm>	 https://codesearch.wmflabs.org/search/?q=%5C%24_ENV&i=nope&files=&repos= OAuth apparently, not core
[20:15:21] <legoktm>	 https://gerrit.wikimedia.org/r/plugins/gitiles/mediawiki/tools/codesniffer/+/master/MediaWiki/Sniffs/Usage/SuperGlobalsUsageSniff.php
[20:15:30] <legoktm>	 		'$_POST' => true,
[20:15:30] <legoktm>	 		'$_GET' => true
[20:24:40] <no_justification>	 I don't think anyone should use $_COOKIE directly
[20:24:49] <no_justification>	 We have methods in WebRequest/WebResponse
[20:25:08] <James_F>	 Or $_SESSION except inside MW, but we're slowly fixing those.
[20:25:15] <James_F>	 Reedy: Do we want to land https://gerrit.wikimedia.org/r/#/c/392217/ (GWToolset extension.json) and ask people to test?
[20:25:16] <no_justification>	 Yeah
[20:56:02] <legoktm>	 no_justification: submit a patch for $_COOKIE? :)
[21:09:18] <no_justification>	 No tiem
[21:24:29] <legoktm>	 MaxSem: do you know about https://phabricator.wikimedia.org/T187147 ? (since you investigated it for hhvm)
[21:27:55] <legoktm>	 subbu: one more please :) https://gerrit.wikimedia.org/r/#/c/417052/ (I forgot about that originally)
[21:41:33] <MaxSem>	 legoktm: haven't investigated the situation for PHP7
[21:54:00] <legoktm>	 MaxSem: want to? :)
[22:01:14] <DanielK_WMDE>	 #startmeeting
[22:01:33] <DanielK_WMDE>	 huh, no meetbots?
[22:01:39] <DanielK_WMDE>	 or did i mistype?
[22:01:51] <kchapman>	 after 2 last week even :(
[22:02:00] <legoktm>	 DanielK_WMDE: wrong channel!
[22:02:08] <DanielK_WMDE>	 oooh :)
[22:02:27] <DanielK_WMDE>	 legoktm: thanks :D
[22:02:39] <legoktm>	 :)
[22:15:37] <greg-g>	 haha
[22:25:45] <bd808>	 I love that every week you are surprised at two bots and yet nobody logs into toolforge to fix that ;)
[22:55:20] <DanielK_WMDE>	 bd808: it's the little joys in life...