[08:41:56] <dcausse>	 ejoseph: hey, this message ^ above from Erik might perhaps explain why you were having difficulties to see deprecation warnings
[09:28:49] <gehel>	 errand, back in 20'
[10:20:09] <ejoseph>	 Yh, I was wondering what I was doing wrong
[11:22:12] <dcausse>	 lunch+errand
[14:21:39] <inflatador>	 greetings
[14:22:45] <dcausse>	 o/
[14:26:20] <inflatador>	 <o/
[14:31:37] <dcausse>	 inflatador: Moritz did create a deb package for jvmquake that we manually installed on wdqs1010 for testing
[14:32:31] <dcausse>	 he ran into tests issues as well which I failed to reproduced using buster+docker (https://phabricator.wikimedia.org/P22365)
[14:33:08] <dcausse>	 if the tests on wdsq1010 is proven successfull we agreed to relax the test suite a bit to make the build pass
[14:40:31] <inflatador>	 dcausse cool, thanks for the update. If you are OK w/not all tests passing we can probably push out the changes today or early next wk. I'll probably post an issue on the upstream Github as well
[14:42:35] <dcausse>	 thanks but I think it's fine to wait and adapt the MR with some more test silenced once we're confident it does what we need
[14:44:06] <dcausse>	 as to understand why these tests fail in some places and not others I'm a bit puzzled, only thing I found is the use of "fakeroot" which seems to trigger some failures but I don't understand why it's being used in some cases and not others
[15:22:10] <dcausse>	 errand
[15:45:41] <ebernhardson>	 \o
[15:55:18] <dcausse>	 o/
[15:58:57] <ebernhardson>	 it's not clear that changing the saneitizer profile generated less update jobs, still ~25% of pages flagged old. i guess i gotta read that code better :)
[15:59:38] <ebernhardson>	 meh, my bluetooth borked. restarting before joining meeting
[16:02:02] <gehel>	 ryankemper, ebernhardson, ejoseph: we're talking about hiring in https://docs.google.com/document/d/1ZUN6JtkbDqZQ857YWHavkUfmLs0ijfvygqqoBwywIJs/edit
[16:09:28] <inflatador>	 ebernhardson was looking at the sanetizer phab ticket just now, if you need help LMK
[16:31:45] <ebernhardson>	 inflatador: it turns out that in this specific case the problem is mostly PEBKAC :)  The patch to change profiles was merged but it didn't make it into the march 8th branch cut, i think i was expecting a mid-day US branch cut, but instead the branch was cut at 2am UTC so the patch wasn't shipped this week
[16:32:09] <ebernhardson>	 (if i had realized that i would have manually shipped the patch all the way to prod, oh well)
[17:08:05] <inflatador>	 quick workout, back in ~30
[17:45:50] <inflatador>	 back
[18:00:55] <ebernhardson>	 hmm, not seeing how the Elastica library expects us to generically manage the deprecation headers. most obvious would be to extend \Elastica\Client and replace the _log method. But they explicitly document that it's deprecated to override this method and you should instead provide a custom LoggerInterface. Except they don't pass the warning headers to the logger :P
[18:01:08] <inflatador>	 Lovely
[18:03:34] <ebernhardson>	 maybe will extend the transport instead, might be more direct
[18:49:07] <inflatador>	 Lunch/errands, back in ~1 hr
[19:27:27] * ebernhardson doesn't find it promising when the first page of results from google includes a link to CirrusSearch open tasks
[20:09:27] <inflatador>	 back
[22:59:04] <inflatador>	 leaving in a few, happy weekend all!