[06:02:47] <RoanKattouw>	 addshore: What are your thoughts on including opt-out preferences for former beta features in https://grafana.wikimedia.org/dashboard/db/betafeatures?refresh=5m&orgId=1 ? Should that be in the same dashboard or a different one, do you think?
[06:41:17] <TimStarling>	 why does the job queue have two different ways to wait for replication, called on adjacent lines?
[06:42:19] <TimStarling>	 as in, it waits, then on the next line it possibly waits again depending on different criteria
[06:45:32] <TimStarling>	 because "Relying on waitForBackups() alone will not work with non DB queues"
[06:46:29] <TimStarling>	 seriously he introduced both things just weeks apart
[06:49:02] <TimStarling>	 ok that was my bad git blame skills, actually he introduced the first in August 2012 and the second in March 2013
[07:22:11] <TimStarling>	 for the record this rant was brought to you by https://gerrit.wikimedia.org/r/c/mediawiki/core/+/453091
[10:17:14] <addshore>	 RoanKattouw: opt out preferences? I don't quite follow
[11:18:13] <Reedy>	 I guess people that have chosen to disable a beta feature
[17:28:17] <legoktm>	 anomie: welcome back :)
[17:28:33] <anomie>	 legoktm: Thanks. Now I just have to catch up on all the emails from when I was out.
[17:33:26] <legoktm>	 so see you in a week? :P
[17:43:59] <DanielK_WMDE>	 wb anomie! 
[17:44:06] <DanielK_WMDE>	 hope you had a good  vacation
[18:02:12] <RoanKattouw>	 addshore: I meant the type of preference that lets someone opt out of something that used to be a beta feature
[19:38:07] <RoanKattouw>	 Maybe that's weird enough that I should make a separate dashboard
[20:53:50] <Krinkle>	 addshore: Could you make sure https://phabricator.wikimedia.org/T202097 is triaged? Not high prio though.
[21:53:39] <addshore>	 RoanKattouw: hmm, perhaps
[21:53:59] <addshore>	 I mean, the BF data tracked currently is pretty useless once the feature is a default
[22:19:43] <Reedy>	 anomie: tgr|away https://en.wikipedia.org/wiki/Wikipedia:Bots/Noticeboard#Out_of_curiosity..._%28to_all_bot_operators%29
[22:19:52] <Reedy>	 OAuth for "desktop apps" doesn't make sense, does it?
[22:19:57] <Reedy>	 I'm pretty sure we've had this discussion before
[22:21:00] <Reedy>	 and/or we need OAuth2?
[22:23:49] <legoktm>	 yes, we need OAuth2 for that
[22:24:01] <legoktm>	 because you can't have secret keys in an application you distribute
[22:25:33] <legoktm>	 I guess you could have people register their own "owner-only" OAuth apps, and maintain their own secret keys?
[22:25:57] <Reedy>	 Sounds like hassle
[22:26:08] <Reedy>	 "Passwords are insecure, quite simply. There's a reason the world is moving to OAuth."
[22:26:12] <Reedy>	 Still requires passwords
[23:34:10] <Reedy>	 Krinkle: travis is unhappy
[23:55:34] <Krinkle>	 no it isn't :P
[23:55:39] <Krinkle>	 it's just confused.
[23:56:21] <Krinkle>	 But yeah, quite often it seems to time out at the composer install step
[23:56:26] <Krinkle>	 10 min and then it gets killed
[23:56:30] <Reedy>	 I bet it's a composer regression
[23:56:34] <Reedy>	 I note it's using 1.7.2
[23:56:36] <Reedy>	 Which was released today
[23:56:38] <Krinkle>	 I've requeued some of the jobs and they seem to pass eventually
[23:57:37] <Krinkle>	 yeah, or something up with packagist (the site)
[23:58:03] <Reedy>	 Yeah, being linked in one way or antoher isn't a surprise
[23:59:18] <Krinkle>	 Seems like they might need a big corporate sponsor or at least a better CDN and a way of using their server that is more cache-able.