[00:11:28] <halfak>	 saltmaster and minion master!?  How many systems are we interacting with!?
[08:35:47] <Amir1>	 This sounds fun: https://github.com/mewwts/addict
[13:33:01] <Amir1>	 OK. I'm stuck now
[13:33:17] <Amir1>	 getting some fresh air to see why it gives me error
[13:43:51] <halfak>	 o/
[13:44:00] <halfak>	 Hey Amir1 
[13:52:14] <Amir1>	 halfak: hey
[13:52:19] <Amir1>	 how are you?
[13:52:46] <Amir1>	 I'm working on running ores in beta using scap and puppet master
[13:52:58] <Amir1>	 https://gerrit.wikimedia.org/r/#/c/280403/
[13:53:13] <Amir1>	 it still gives me error
[13:53:38] <halfak>	 Amir1, seems like this could be handed off to releng at some point soon.  What do you think?
[13:54:20] <Amir1>	 releng don't do this part in their own
[13:54:24] <Amir1>	 they only help 
[13:54:31] <Amir1>	 I tried that before :D
[13:54:47] <Amir1>	 halfak: also I found a bug in scap itself!
[13:55:00] <Amir1>	 fixed, it's not merged
[14:05:23] <halfak>	 Amir1, hmm..  That's a bit weird.  It seems like there should be a point at which a releng engineering fills in the last bits. 
[14:33:38] <halfak>	 Amir1, is this a *last bits* situation?
[14:34:38] <Amir1>	 I'm asking in #wikimedia-releng
[14:35:02] <Amir1>	 trying to understand why the puppet fails but the command doesn't
[14:35:20] <Amir1>	 It gave me enough headaches already
[14:35:31] <Amir1>	 but I think it's the last parts
[14:36:03] <Amir1>	 halfak: I want to push some stuff related to scap to the master
[14:36:07] <Amir1>	 nothing will break
[14:36:09] <Amir1>	 is it okay?
[14:36:23] <halfak>	 Amir1, if it is minor and just scap, yeah. 
[14:36:28] <Amir1>	 yes
[14:36:32] <Amir1>	 awesome
[14:38:35] <Amir1>	 why it pushed it to the scap_again
[14:38:41] <Amir1>	 in local I'm on master
[14:38:43] <Amir1>	 am i?
[14:39:40] <halfak>	 git status?
[14:40:09] <Amir1>	 yup
[14:40:13] <Amir1>	 I was at scap
[14:40:24] <halfak>	 heh
[14:40:47] * halfak updates editquality models for revscoring 1.1.8
[14:41:43] <Amir1>	 made a pr and then merged it
[14:42:11] <halfak>	 That works. 
[14:44:04] <halfak>	 Amir1, I think I might draft announcement of new models for ORES at the hackathon
[14:44:07] <halfak>	 or maybe on the plane
[14:44:19] <halfak>	 So we can switch to all the gradient_boosting & random forest models
[14:44:43] <Amir1>	 it would be great
[14:44:56] <Amir1>	 I hope we can announce moving to prod in the hackathon
[14:44:59] <halfak>	 Will be nice to ditch the SVC models.  
[14:45:17] <halfak>	 I don't want to update them for revscoring 1.1.8 since they took so long to build. 
[14:47:53] <halfak>	 OK time to do 1000 things in fab
[14:47:58] <halfak>	 *phab
[14:48:10] <halfak>	 I'm setting up a sorted backlog board
[14:48:17] <halfak>	 So that we don't have 1 huge column
[14:48:33] <Amir1>	 nice
[14:52:35] <halfak>	 Now to experiment with moving cards!
[14:54:43] <halfak>	 "You are about to apply a batch edit to 128 tasks" 
[14:54:44] <halfak>	 :D
[15:20:25] <halfak>	 Woah!  Old ORES paths are broken!
[15:20:32] <halfak>	 v1 has some broken stuff. 
[15:20:35] * halfak fixes ASAP
[15:38:42] <halfak>	 Yeah... this is surprisingly broken :/
[15:53:14] <halfak>	 OK going to staging
[15:53:26] <halfak>	 Yikes.  I wonder why this didn't get reported
[15:53:28] <halfak>	 :/
[15:54:32] <halfak>	 https://phabricator.wikimedia.org/T131251
[15:54:37] <halfak>	 "ORES pre-deploy test -- validate responses against swagger"
[16:14:22] <halfak>	 OK.  We're fixed
[16:14:27] <halfak>	 Yikes.  That was gross. 
[16:15:01] <halfak>	 Maybe we can ask schana to help us shore up our CI and deploy process. 
[16:15:14] <halfak>	 I'd really like to transition back to doing research primarily at some point :) 
[16:15:34] <halfak>	 Lots of research *of* ORES and its context that is currently getting ignored. 
[16:28:21] <schana>	 yes halfak, I'd love to work on that -- I've been distracted lately with the whole working overseas thing :(
[16:28:46] <halfak>	 Yeah.  no worries.  Once you get all that settled 
[16:29:03] <halfak>	 I figured you'd have some insights into making ORES resilient :) 
[17:02:16] <halfak>	 Amir1, when you get a chance, could you check out https://phabricator.wikimedia.org/T115785 and update it for your current work/thoughts on the extension?
[17:05:40] <Amir1>	 sure
[17:05:41] <Amir1>	 just give some minutes to figure it out why my ssh doesn't work
[17:05:41] <Amir1>	 :(
[17:05:41] <Amir1>	 this whole situation is changing from tiring to frustrating 
[17:05:41] <halfak>	 No worries. 
[17:05:41] <halfak>	 Keep your mind on ssh stuff 
[17:05:41] <halfak>	 If you end up forgetting about this, no worries. 
[17:39:16] <halfak>	 Amir1, https://phabricator.wikimedia.org/tag/revision-scoring-as-a-service-backlog/
[17:39:22] <halfak>	 A thing of beauty
[17:39:42] * Amir1 looking
[17:40:28] <Amir1>	 halfak: this is so great
[17:40:30] <Amir1>	 wow
[17:43:30] <halfak>	 https://gist.github.com/halfak/b5e45af03a8bfe3d5570eceeb51a1f5d
[17:43:34] <halfak>	 ^ 218 actions
[17:45:12] <Amir1>	 I think my inbox is exploding 
[17:53:21] <halfak>	 Sorry about that 
[17:54:11] <halfak>	 OK Lunch time!
[17:54:24] <Amir1>	 happy that it's exploding :)
[17:54:31] <Amir1>	 it's exploding for a good reason
[18:01:04] <halfak>	 o/ awight 
[18:01:12] <halfak>	 sorry for a massive pile of inbox spam
[18:01:57] <awight>	 I should adjust my Phab settings or something... but don't worry, your spam was buried amidst FR-tech spam :)
[18:02:45] <awight>	 I wish I had more time to help...
[18:03:11] <awight>	 happily, Amir1 did an amazing job fixing the tech debt I left behind!
[18:03:39] <awight>	 How's the productionization going?  Is there still .deb busy work?
[18:04:26] <Amir1>	 hey, awight :)
[18:04:37] <Amir1>	 I have something you need to take a look
[18:04:40] <Amir1>	 it's very simple
[18:04:44] <awight>	 Hi!  sure
[18:05:23] <Amir1>	 we are relicencing the extension from CC0 to GPL
[18:05:32] <awight>	 hehe k
[18:05:42] <Amir1>	 https://gerrit.wikimedia.org/r/271947
[18:05:54] <Amir1>	 we are waiting for you :)
[18:06:22] <grrrit-wm1>	 (03CR) 10Awight: [C: 031] "Sounds good..." [extensions/ORES] - 10https://gerrit.wikimedia.org/r/271947 (owner: 10Legoktm)
[18:07:22] <Amir1>	 awesome awight :)
[18:07:24] <Amir1>	 thanks
[18:07:40] <Amir1>	 we are moving to prod, the only thing left is the scap3 work
[18:07:46] <grrrit-wm1>	 (03PS2) 10Legoktm: Relicense as GPL-3.0+ [extensions/ORES] - 10https://gerrit.wikimedia.org/r/271947 
[18:10:02] <grrrit-wm1>	 (03CR) 10Legoktm: [C: 032] "PS2: Rebased. +2 per Amir and Adam." [extensions/ORES] - 10https://gerrit.wikimedia.org/r/271947 (owner: 10Legoktm)
[18:10:02] <grrrit-wm1>	 (03Merged) 10jenkins-bot: Relicense as GPL-3.0+ [extensions/ORES] - 10https://gerrit.wikimedia.org/r/271947 (owner: 10Legoktm)
[18:10:02] <Amir1>	 \o/
[18:11:36] <schana>	 is there a common opinion regarding copy-left licenses vs permissive?
[18:38:56] <halfak>	 schana, I feel strongly about permissive and all of the ORES/revscoring/wikilabels stuff is MIT
[18:39:20] <halfak>	 But if there's constraints or someone else feels equally strongly, I'll go GPL and the like 
[18:39:50] <schana>	 I agree about the permissive, I was just curious about the discussion regarding GPLv3 for the extension
[18:40:31] <halfak>	 Gotcha.  I think there are norms/policy/legal-bits that require GPL there
[18:41:29] <yuvipanda>	 schana: I used to have feelings about it, but my current only feelings are: HAVE A LICENSE! I personally prefer Apache2 but people disagree.
[18:41:58] <yuvipanda>	 and I don't usually care enough to protest. all my Jupyter project stuff is 3-clause-bsd licensed since that's common in that community, kubernetes stuff is apache2 for similar reasons, etc
[18:46:34] <halfak>	 MIT == Apache2 + Lawyer happiness
[18:46:40] <halfak>	 Woops.  Other way around
[18:46:49] <halfak>	 MIT + lawyer happiness == Apache2
[19:44:07] * halfak --> Israel
[19:44:08] <halfak>	 o/