[00:00:11] <Reedy>	 it has /$ZUUL_PROJECT -m checkstyle
[00:00:13] <Reedy>	 which becomes /mediawiki/extensions/PdfHandler -m checkstyle
[00:00:28] <Reedy>	 oh
[00:00:29] <legoktm>	 that's inside the docker container
[00:00:31] <Reedy>	 mounting
[00:00:38] <legoktm>	 mhm
[00:00:54] <legoktm>	 src/extensions/$EXT_NAME/... should work
[00:01:56] <Reedy>	 https://gerrit.wikimedia.org/r/#/c/429367/3/jjb/mediawiki-extensions.yaml
[00:07:48] <legoktm>	 :D
[00:07:51] <legoktm>	 are you going to deploy it?
[00:08:41] <Reedy>	 I... can
[00:09:00] * Reedy looks for the instructions again
[00:11:37] * Reedy waits for ../.local/bin/jenkins-jobs --conf etc/jenkins_jobs.ini test config/ -o output/
[00:12:14] <Reedy>	 if [ ! -f &quot;src/extensions/$EXT_NAME/tests/phan/config.php&quot; ]; then
[00:12:14] <Reedy>	     echo &quot;Phan config not found\n&quot;
[00:12:14] <Reedy>	     exit 0
[00:12:14] <Reedy>	 fi
[00:12:53] <legoktm>	 seems right
[00:13:16] <Reedy>	 mmm
[00:14:04] * Reedy rechecks some jobs
[00:15:08] <Reedy>	 00:14:50 [mwext-php70-phan-docker] $ /bin/bash -eu /tmp/jenkins4697845656797878585.sh
[00:15:08] <Reedy>	 00:14:50 + '[' '!' -f src/extensions/PdfHandler/tests/phan/config.php ']'
[00:15:08] <Reedy>	 00:14:50 + echo 'Phan config not found\n'
[00:15:08] <Reedy>	 00:14:50 Phan config not found\n
[00:15:09] <Reedy>	 00:14:50 + exit 0
[00:15:11] <Reedy>	 00:14:50 Finished: SUCCESS
[00:15:14] <Reedy>	 :D
[00:15:22] <Reedy>	 Not really a success
[00:15:25] <Reedy>	 But we'll take it ;)
[00:15:28] <legoktm>	 test a job that should ran phan too?
[00:15:44] <Reedy>	 I've rechecked brion's master patch too
[00:15:55] <Reedy>	 I did think of htat :P
[00:16:19] <Reedy>	 https://integration.wikimedia.org/ci/job/mwext-php70-phan-docker/5475/console
[00:16:38] <Reedy>	 LGTM
[00:16:45] <legoktm>	 :D
[00:16:53] <brion>	 \o/
[00:17:06] <Reedy>	 Could almost switch the phan job on everywhere now ;)
[00:17:18] <brion>	 phan-tastic!
[00:25:37] <Reedy>	 brion: I did the other backports, but as they're slightly modified, I'm not self merging them
[00:25:45] <Reedy>	 Did 1.31 as it's the same as master
[00:25:51] <Reedy>	 https://gerrit.wikimedia.org/r/#/q/Ib4ee9cf12ac04304c576087727eff5dc521ae751
[00:25:59] <brion>	 sounds good
[00:26:01] <brion>	 thanks!
[00:28:25] <brion>	 i'll test and merge down the branch ones tomorow
[14:43:04] <dmaza>	 hello everyone.. If anyone has a few minutes, this PR needs some love. It modify blocks/autoblocks. Any comment/suggestion is more than welcome
[14:43:06] <dmaza>	 https://gerrit.wikimedia.org/r/#/c/425313/
[18:51:04] <MaxSem>	 dmaza: in https://gerrit.wikimedia.org/r/#/c/425313/10/includes/user/User.php does loadFromSession() still need trackBlockWithCookie()?
[18:59:24] <no_justification>	 I don't like the list of open bugs for 1.31 still :\
[18:59:32] <no_justification>	 A little much for me to cut a non-broken rc
[19:01:31] <dmaza>	 MaxSem: well that was one of my questions in gerrit. We were setting the cookie before on every session load for logged in users. But I think it is an overkill
[19:46:02] <MaxSem>	 dmaza: so right now the workflow is like: you get cookies when you try to edit, create an account AND whenever MW decides it needs a session
[19:48:08] <dmaza>	 yes
[19:49:07] <dmaza>	 MaxSem: I don't think we need to set the cookie when we load from session, 'cause we don't use the block anywhere unless you either try to edit or create an account
[19:50:34] <dmaza>	 Unless I'm missing something something, I believe we can remove it from loadFromSession.
[20:11:33] <anomie>	 dmaza: I think removing it would also resolve T180050
[20:11:33] <stashbot>	 T180050: [Russian] [Chinese] PHP Warning: Recursion detected in RequestContext::getLanguage MW v1.31.0-wmf.6 - https://phabricator.wikimedia.org/T180050
[20:36:55] <dmaza>	 anomie, MaxSem: I'm gonna go ahead and remove it. Is there anywhere else where I would need to set the cookie other than EditPage->edit() and SpecialCreateAccount->checkPermissions() ?
[20:48:02] * MaxSem scratches head
[21:13:58] <MaxSem>	 dmaza: honestly, I'm not sure myself: ideally, we would want this on every blocked check - however, lookig at the code, adding something like that e.g. to Title::checkUserBlock() would be too messy due to a 
[21:14:03] <MaxSem>	 mix of concerns
[21:19:07] <dmaza>	 MaxSem: and even then Title::checkUserBlock() is probably not called when creating an account
[21:19:15] <dmaza>	 ¯\_(ツ)_/¯
[21:19:31] <Platonides>	 what is called there, then?
[21:19:45] <Platonides>	 it should be checking blocks there…
[21:19:51] <dmaza>	 where?
[21:20:15] <Platonides>	 when (attempting to) create an account
[21:21:26] <dmaza>	 AuthManager->checkAccountCreatePermissions()
[21:21:34] <dmaza>	 among other things
[21:23:02] <dmaza>	 ideally, something like this that sets a cookie should be in a middlewhere before the response is delivered. I guess that MW that would mean on a hook somewhere right before rendering the page