[00:43:01] <legoktm>	 bpirkle: no, I got stuck at a city council meeting, it's still on my list for tonight
[00:57:24] <legoktm>	 bpirkle: did you see my suggestion about creating an unpolyfill for the getallheaders library?
[00:58:20] <legoktm>	 hmm
[00:58:28] <legoktm>	 does that only exist in the apache2 SAPI?
[00:59:00] <legoktm>	 https://secure.php.net/manual/en/function.getallheaders.php
[00:59:06] <legoktm>	 > 7.3.0 	This function became available in the FPM SAPI. 
[00:59:11] <legoktm>	 so we do need the polyfill
[01:00:58] <legoktm>	 it doesn't look like it exists in the CLI either anyways
[01:03:49] <TimStarling>	 legoktm: do you know why wmf-config still has a PHP 5.5 lint test?
[01:05:17] <paladox>	 i think for HHVM
[01:07:00] <TimStarling>	 we don't have that for core, we use HHVM to lint for HHVM
[01:08:28] <legoktm>	 TimStarling: probably because we forgot to remove it. as long as it has a recent version of PHPCS plus hhvmlint, we should replace with with a 7.2 lint test
[01:08:50] <TimStarling>	 I tried to use ?? in a config file
[04:11:00] <bpirkle>	 @legoktm regarding getallheaders, is there an example of where already have something similar that I could look at?
[05:28:19] <legoktm>	 bpirkle: I was thinking of something like https://github.com/BackEndTea/un-poly-all but that won't work since we do actually need the polyfill in this case
[06:30:19] <legoktm>	 > usleep($options['delay'] * 1000);
[06:30:24] <legoktm>	 why not use sleep()? :|
[06:32:15] <_joe_>	 legoktm: usleep is more performant!!
[06:32:27] <legoktm>	 it is?
[06:32:31] <legoktm>	 wait...
[06:32:35] <_joe_>	 lol
[06:32:47] <legoktm>	 heh, you got me
[06:34:21] <_joe_>	 legoktm: since you're here, any reason why calling openlog(..) in the auto-prepend file would be an issue? ref https://phabricator.wikimedia.org/T211184
[06:39:22] <legoktm>	 _joe_: I don't think so, though I'm not super familiar with syslog. Reading the code, the PHP openlog() looks like a straightforward wrapper around the C openlog()
[06:39:51] <legoktm>	 it does look like the simplest solution compared to the other ideas
[06:40:13] <_joe_>	 it's also the only one (besides just patching fpm)
[06:40:23] <_joe_>	 that yields correct results
[06:40:40] <legoktm>	 https://bugs.php.net/bug.php?id=76229 claims this is fixed in 7.3?
[06:40:45] <_joe_>	 yes
[06:40:49] <_joe_>	 I was about to add
[06:40:53] <_joe_>	 it's temporary
[06:44:46] <_joe_>	 (and every time I say that, I remember https://tools.wmflabs.org/bash/quip/AU7VTzhg6snAnmqnK_pc)
[06:46:52] <legoktm>	 hehe
[08:20:16] <legoktm>	 Reedy: https://phabricator.wikimedia.org/T202110#4802571
[08:25:54] <legoktm>	 bpirkle: I finished reviewing the entire mw/vendor patch, left comments in Gerrit and Phab, I'll save your mw/core guzzle patch for tomorrow (today I guess now)
[20:18:22] <dmaza>	 legoktm, anomie: what else can we do or who else can we get in touch with to get this moving ?https://phabricator.wikimedia.org/T211048