[00:25:10] <halfak|Mobile>	 Zppix: sorry to drop off.  Currently change prop only works in production so we have work arounds for labs in ORES.
[00:25:50] <halfak|Mobile>	 Was hoping to get a mirror installation in labs.  Folks a similar niche as icinga as general support for services
[00:25:58] <halfak|Mobile>	 *fills
[00:26:36] <halfak|Mobile>	 Also holy crap I'm finally done coping things up.  It's amazing how much less *stuff* seems to be there after processing the tree
[00:57:36] <Zppix>	 ok i will talk to paladox in a bit if hes around still
[01:00:57] <Zppix>	 paladox are you about?
[01:01:02] <Zppix>	 halfak|Mobile:  ^
[01:04:50] <Zppix>	 so you want a mirror of prod? i mean just setup scap have a use the deployment-prep server and scap it to both prod and labs?
[01:05:04] <Zppix>	 wmnet can  scap to wmflabs but wmflabs cannot go to wmnet
[01:05:36] <Zppix>	 halfak^
[01:11:07] <Zppix>	 Amir1:  is https://github.com/wiki-ai/editquality/pull/24 ready for merge? 
[03:14:14] <Amir1>	 Zppix: nope
[03:15:58] <Zppix>	 ack
[03:16:25] <Zppix>	 Amir1:  your on the coc committee no?
[03:16:38] <Amir1>	 Yup
[03:17:07] <Zppix>	 Amir1:  We need a better explaintation on how it needs to be formatted, attribution, etc
[03:17:53] <Zppix>	 for example https://github.com/wiki-ai/ores/pull/206 was denied due to the unclearness of repo inclusion of COC
[03:18:50] <Amir1>	 Zppix: Well, as a CoC member my duty is to handle reports of violation. These kind of stuff should be tackled by talking in community 
[03:19:08] <Amir1>	 the workflow I had in mind was using code review in mediawiki core
[03:19:13] <Amir1>	 which happened 
[03:19:24] <Zppix>	 Amir1:  i know the workflow, im talking about attributon and such
[03:19:46] <Zppix>	 Amir1:  i also know commenting on that task is like talking to teens in a high school
[03:20:11] <Amir1>	 :))) That's a fair point 
[03:20:29] <Amir1>	 The text is negligible to have attribution, etc. and it only links to the  real CoC which has attribution (obviously) 
[03:20:36] <Amir1>	 so I don't think this would be needed
[03:20:56] <Zppix>	 Amir1:  That should be made *clear*
[03:21:39] <Amir1>	 Yeah, but again, in those cases I'm not a CoC committee member as it means nothing here, I'm just a community member suggesting things
[03:22:02] <Zppix>	 considering, mediawiki license its required to attribute
[03:22:05] <Amir1>	 Zppix: Where do you think we need to make it clear 
[03:22:22] <Zppix>	 hell anywhere, a billboard, a simple notice, anything
[03:25:22] <Amir1>	 hmm, okay. Let me see 
[05:21:09] <travis-ci>	 wiki-ai/revscoring#1022 (pytz - 4dee4f7 : Amir Sarabadani): The build passed. https://travis-ci.org/wiki-ai/revscoring/builds/241661004
[11:51:48] <paladox>	 Zppix hi i am now.
[15:36:02] <halfak>	 o/
[15:36:03] <halfak>	 glorian_wd, 
[15:36:08] <halfak>	 Saw you pinging me on hangouts
[15:36:16] <glorian_wd>	 halfak: o/
[15:36:47] <glorian_wd>	 yeah. Basically, I am curious if it's possible to extract/print the value of the features that we create on wikidatawiki
[15:37:11] <glorian_wd>	 That's what I sought last night. It seems that the extractor function in wikiclass works differently than the one in revscoring
[15:37:48] <halfak>	 That doesn't make any sense. 
[15:37:55] <halfak>	 You shouldn't be using an extractor. 
[15:38:04] <halfak>	 Just use solve() like in the tests I linked you to. 
[15:40:18] <glorian_wd>	 hmm ok let me try it again
[15:40:29] <glorian_wd>	 :)
[15:46:29] <paladox>	 halfak i saw you talking about something todo with icinga and icinga monotoring services. Not sure what you meant though?
[15:46:32] <paladox>	 also hi
[15:47:19] <halfak>	 Hi paladox !
[15:47:37] <paladox>	 :)
[15:47:52] <halfak>	 So I was thinking about your maintenance of labs-based icinga (which I think is great because it mirrors a critical service generally only available in production)
[15:48:08] <halfak>	 And was asking myself about other types of services I wanted from production.  
[15:48:16] <halfak>	 And I thought ChangeProp would be a good idea. 
[15:48:23] <paladox>	 Ah yep.
[15:48:48] <halfak>	 What do you think about maintaining another generalized service> 
[15:48:50] <halfak>	 *?
[15:49:09] <paladox>	 Yep
[15:49:37] <paladox>	 You can have it monotore what ever services you want :)
[15:49:38] <paladox>	 No limits :)
[15:50:14] <paladox>	 I wonder if in puppet do they add the check for ChangeProp since i could copy
[15:50:20] <paladox>	 (migrate it to icinga2 format)
[15:50:41] <halfak>	 paladox, ahh.  I wasn't clear.  I was hoping to convince you to maintain ChangeProp in labs itself. :)  
[15:50:52] <halfak>	 SInce you're maintaining icinga in labs. 
[15:50:56] <paladox>	 Ah
[15:51:00] <paladox>	 what's ChangeProp?
[15:51:20] <paladox>	 Is it done by icinga?
[15:51:32] <halfak>	 It allows applications (like ORES) to respond to changes as they happen.  E.g. a revision is saved, we score it and cache the score for future requests. 
[15:51:45] <halfak>	 ChangeProp is entirely separate from icinga
[15:51:56] <halfak>	 But I figure that many apps in labs will want to use changeprop
[15:52:02] <halfak>	 Just like they will want to use icinga. 
[15:52:08] <paladox>	 oh
[15:52:19] <paladox>	 how does that get setup?
[15:52:30] <halfak>	 That's a good question.  I haven't looked into it. 
[15:52:34] <paladox>	 Ok
[15:52:46] <halfak>	 I'm sure it has a nice scap/puppet setup
[15:52:55] <halfak>	 I had one other thought that you might be more interested in.  git-lfs 
[15:53:04] <paladox>	 is it this https://www.mediawiki.org/wiki/Change_propagation ?
[15:53:11] <halfak>	 Yup.  that's it. 
[15:53:37] <halfak>	 https://github.com/wikimedia/change-propagation
[15:54:56] <paladox>	 Will need help to setup that :)
[15:57:14] <halfak>	 I recomment asking about it in #wikimedia-services.  Generally a nice bunch :) 
[15:57:56] <halfak>	 I have a list of things that labs is lacking and should really have for basic prod-like support. 
[15:58:21] <halfak>	 (ChangeProp, Scap, some git versioned file storage, etc.)
[15:59:44] <paladox>	 ok
[16:00:00] <paladox>	 halfak there's scap in labs
[16:00:10] <paladox>	 git-lfs i thought is also in labs
[16:00:34] <halfak>	 The scap doesn't seem to work for any-ol deployment (e.g. ores deploys outside of beta-labs)
[16:00:41] <halfak>	 Last I checked anyway. 
[16:00:48] <halfak>	 I didn't know about git-lfs being available. 
[16:01:39] <paladox>	 Oh
[16:01:52] <paladox>	 halfak anything in the wikimedia apt repo is installable on labs. Also the deployment puppet role seems to work for me.
[16:10:42] <Zppix>	 halfak:  im here fyi
[16:14:17] <Zppix>	 halfak:  you wanted memeory checks from icinga didnt you?
[16:19:16] <halfak>	 o/ Zppix 
[16:19:29] <halfak>	 yeah.  I suppose so.  
[16:19:52] <Zppix>	 halfak:  also i reviewed the pytz pr amir created i think its good to merge? you agree?
[16:20:13] <halfak>	 link?
[16:20:18] <paladox>	 halfak that will requiring you to add the check_mem file to all the instances that you want the check on.
[16:20:40] <paladox>	 https://github.com/wikimedia/puppet/blob/e959321aa620b77403cc9379db2e86080323c6e8/modules/icinga/files/check_ram.sh
[16:20:42] <Zppix>	 halfak:  https://github.com/wiki-ai/revscoring/pull/323
[16:20:51] <halfak>	 paladox, will need to find a way to do it with puppet. 
[16:20:57] <paladox>	 Oh
[16:21:14] <halfak>	 Zppix, merged
[16:21:15] <paladox>	 add this https://github.com/wikimedia/puppet/blob/d0f5eaba58416f9d15e8fb05203cf08a6deebc3e/modules/icinga/manifests/plugins.pp#L100 puppet class
[16:21:30] <Zppix>	 halfak:  lol i could done that :P
[16:21:39] <paladox>	 halfak  ^^
[16:21:58] <halfak>	 Zppix, that readme still renders badly. 
[16:22:03] <halfak>	 paladox, looking
[16:22:09] <paladox>	 ok thanks :)
[16:22:25] <Zppix>	 halfak: what exactly are you not  liking?
[16:22:42] <halfak>	 Compare the renderings side-by-side
[16:22:52] <halfak>	 There's a lot of obvious differences. 
[16:23:15] <Zppix>	 will do ill try to make it look similar if not make it look pretty
[16:24:47] <halfak>	 Sure!  Improvements are welcome, but given the squash, you should put it in a separate PR
[16:25:07] <halfak>	 paladox, trying to figure out what exactly I need to do here.  I don't do that much puppet. 
[16:25:24] <halfak>	 Do I just require icinga::plugins in our base role?
[16:26:36] <paladox>	 yep
[16:27:16] <halfak>	 WTF puppet repo
[16:27:18] <halfak>	 Seriously
[16:27:22] <halfak>	 This thing is such a mess. 
[16:27:56] <paladox>	 halfak got an error?
[16:28:47] <halfak>	 Na.  Just trying figure out the right place to add it. 
[16:29:06] <paladox>	 ok
[16:29:21] <paladox>	 halfak is your repo public?
[16:29:38] <paladox>	 puppet repo
[16:30:04] <halfak>	 https://github.com/wikimedia/puppet/tree/d0f5eaba58416f9d15e8fb05203cf08a6deebc3e/modules
[16:30:14] <paladox>	 oh
[16:31:24] <halfak>	 So I think I'm adding "require icinga::plugins" to https://github.com/wikimedia/puppet/blob/d0f5eaba58416f9d15e8fb05203cf08a6deebc3e/modules/ores/manifests/base.pp
[16:31:27] <halfak>	 Does that seem right?
[16:31:38] <paladox>	 yeh
[16:31:45] <paladox>	 hope this works :) and dosent fails :)
[16:32:13] <halfak>	 Looking more, it seems that I need the prepend "::", so more like "require ::icinga::plugins"
[16:32:21] <paladox>	 ah
[16:32:41] <paladox>	 try that :)
[16:36:29] <glorian_wd>	 halfak: in order to call the test class, do I need to pass MediaWiki user ID? I keep getting this error 'AttributeError: module 'mwtypes' has no attribute 'User'' :/
[16:36:37] <glorian_wd>	 which points to my mwxml
[16:38:03] <halfak>	 MediaWiki user ID
[16:38:07] <halfak>	 What
[16:38:10] <halfak>	 Why
[16:38:16] <halfak>	 You're testing wikibase processing. 
[16:43:26] <halfak>	 "test class" doesn't make sense either
[16:43:28] <halfak>	 glorian_wd, ^
[16:43:41] <halfak>	 https://gerrit.wikimedia.org/r/358240
[16:43:57] <halfak>	 Zppix, would you make a task on the Scoring-platform-team workboard for setting this up? 
[16:44:06] <halfak>	 And assign it to yourself :D 
[16:44:29] <halfak>	 That workboard is the way we track and report progress, so I want to make sure you make it into our next progress update for working on this. 
[16:48:57] <Zppix>	 halfak:  task for what exactly sorry i've been in github
[16:49:45] <Amir1>	 halfak: 1- I made a document for UX fixes, hope you saw it
[16:50:26] <Zppix>	 Amir1:  o/ sorry if i seem liked i was angry towards you last night this coc stuff is quite annoying to me :/
[16:50:33] <Amir1>	 2- the blocker for one of the oldest bug is resolved now, I can fix it 
[16:51:06] <Amir1>	 Zppix: none at all, CoC discussions is heated and I don't think you were angry 
[16:51:51] <Zppix>	 Amir1:  okay i just felt as i read the convo earlier again today  i felt like i was being a bit out of line but whatever as long as we both agree no harm no foul :)
[16:52:27] <Amir1>	 Exactly :)
[16:53:34] <Zppix>	 Amir1:  fyi i didnt know wb was defunct til after i made that pr  im going to close it (if its not already) i already disabled wb-vandalisms ci cause its just a waste of resources :)
[16:54:10] <Amir1>	 Zppix: that's great. Thanks. We should close it at some point at time, I don't know if github allows that
[16:54:53] <Zppix>	 Ill look into that
[16:55:08] <glorian_wd>	 halfak: sorry, I meant test function.
[16:56:11] <Zppix>	 Amir1:  i could disable all types of merging for prs on that repo, effecively disabling the repo unless a change is just pushed to master w/o pr and i coould add a notice in the master branch to use revscoring?
[16:56:41] <Amir1>	 That would be great 
[16:57:02] <Zppix>	 ok doing it
[16:57:26] <Zppix>	 ok i can disable them all but i disabled 2/3 of them ill add the notice
[16:58:42] <Zppix>	 Amir1:  halfak  am i okay to directly push to wb-vandalism's master branch to put a notice to use revscoring repo instead since wb-vandalism is defunct?
[16:59:55] <Amir1>	 okay on my side but IDK about halfak, specially this might be used somewhere I'm not aware of (and he would definitely know)
[17:00:16] <Zppix>	 hence is why i asked both of you :)
[17:00:54] <glorian_wd>	 halfak: So, I am trying to write test functions for wikidatawiki, similar to the one which you showed me last night. And I keep getting the error that I've mentioned above on the line where "import wikidatawiki" is placed. 
[17:10:24] <halfak>	 Zppix, the task I was referring to was for memory checks
[17:10:32] <halfak>	 (sorry stepped away for a minute)
[17:11:17] <halfak>	 Amir1, I think wke should move that repo 
[17:11:27] <halfak>	 So that it is Ladsgroup/wb-vandalism again
[17:11:46] <halfak>	 We don't use it at all. 
[17:12:02] <halfak>	 glorian_wd, you need to be more specific about the error. 
[17:12:14] <halfak>	 No one can help you with a paraphrased error message and no code. 
[17:12:48] <halfak>	 Zppix, when you create the phab task, let me know and I'll post https://gerrit.wikimedia.org/r/358240 to it. 
[17:16:11] <Zppix>	 halfak:  -backlog or the main board?
[17:16:25] <halfak>	 Main board.  Active column. 
[17:17:35] <Zppix>	 halfak:  https://phabricator.wikimedia.org/T167602
[17:17:48] <glorian_wd>	 halfak: The code is https://gist.github.com/GlorianY/f29e80f9a1c2843d22fd46b05aa09ce4. The line import wikidatawiki yields to an error "  File "/usr/local/lib/python3.5/dist-packages/mwxml/iteration/user.py", line 6, in <module>class User(mwtypes.User):AttributeError: module 'mwtypes' has no attribute 'User'".  
[17:19:13] <halfak>	 What is in wikidatawiki.py?
[17:19:14] <halfak>	 Seriously?
[17:20:33] <halfak>	 what's the full error?
[17:20:47] <halfak>	 So I can figure out what line in wikidatawiki.py is erroring. 
[17:22:31] <icinga2-wm>	 CUSTOM - check disk_space on Ores-worker-05 is OK: DISK OK 6gb left on /
[17:25:35] <glorian_wd>	 halfak: the content of wikidatawiki is similar to https://github.com/wiki-ai/wikiclass/blob/master/wikiclass/feature_lists/wikidatawiki.py. 
[17:25:44] <glorian_wd>	 the full error is  https://usercontent.irccloud-cdn.com/file/iEVaBaEu/the%20full%20error
[17:25:49] <glorian_wd>	 seems pointing to mwxml
[17:27:50] <halfak>	 glorian_wd, are your requirements up to date? 
[17:28:20] <Zppix>	 halfak:  (bug: Txxxx in the commit message is your friend) cough cough https://gerrit.wikimedia.org/r/#/c/358240/2//COMMIT_MSG
[17:28:35] <halfak>	 Zppix,  you created the task after I uploaded the patchset
[17:28:50] <Zppix>	 halfak:  it doesnt matter the bot will still pick it up
[17:29:24] <Zppix>	 i fixed it for you :P
[17:29:48] <halfak>	 Nice 
[17:30:54] <Zppix>	 glorian_wd:  fyi revscoring has a requirement that was upgrade to a newer version please update accordingly that may be the issue
[17:31:08] <Zppix>	 that was set today
[17:31:20] <glorian_wd>	 halfak, Zppix : yea. I am trying on it now
[17:31:49] <Zppix>	 halfak:  should i send a email to wiki-ai mailing list about the pytz upgrade?
[17:32:18] <halfak>	 Zppix, na.  But I think filing a task in "Done" would be good.  It can go out with our next periodic update. 
[17:32:24] <Zppix>	 ok
[17:32:45] <halfak>	 E.g. https://phabricator.wikimedia.org/phame/post/view/52/status_update_june_3rd_2017/
[17:33:54] <Zppix>	 halfak:  https://phabricator.wikimedia.org/T167604
[17:34:41] <halfak>	 Perfect. 
[17:35:10] <glorian_wd>	 halfak: I have installed the latest packages on requirements.txt, and the same error still occurs. Strange
[17:36:22] <Zppix>	 what file is this happening in ?
[17:36:29] <halfak>	 glorian_wd, https://gist.github.com/halfak/5c49a36d80f6e59fa1665dc898f2f2a8
[17:36:43] <halfak>	 mwtypes==0.2.0
[17:39:21] <glorian_wd>	 My mwtypes==0.2.0
[17:41:57] <halfak>	 mwxml==0.2.2
[17:43:35] <glorian_wd>	 halfak: https://gist.github.com/GlorianY/4f524642ff29ed7e164f1539bed54b0f
[17:44:03] <glorian_wd>	 Mine is 0.3.1. I think it should be fine
[17:44:11] <glorian_wd>	 hmm
[17:48:11] <halfak>	 Confirmed that all works with mwxml 0.3.1
[17:49:26] <glorian_wd>	 Let me try to degrade the mwxml
[18:05:20] <paladox>	 halfak is scb2005 your instance?
[18:05:38] <paladox>	 it is showing as down in #wikimedia-operations
[18:06:24] <glorian_wd>	 halfak: I think the error is already solved once I degraded my mwxml. I will have a closer check after dinner
[18:06:32] <glorian_wd>	 thanks!
[18:06:35] * glorian_wd starving
[18:06:39] <paladox>	 halfak #wikimedia-services say there's an instance for changeprops
[18:06:43] <paladox>	 in beta cluster
[18:06:57] <paladox>	   <Pchelolo>	paladox: you mean beta cluster? there's an instance on deployment-changeprop.deployment-prep.eqiad.wmflabs
[18:24:53] <halfak>	 paladox, we share that instance with a few others
[18:25:01] <paladox>	 Oh ok
[18:25:25] <halfak>	 paladox, the beta cluster changeprop propagates changes to the beta wikis.  We want to propagate changes from the prod wikis to labs systems
[18:25:41] <paladox>	 oh
[18:33:37] <halfak>	 glorian_wd, that's weird.  Mine worked with the current version of mwxml
[18:47:28] <Zppix>	 do you need anything else  halfak?
[18:48:32] <halfak>	 Been talking to paladox about setting up some other services in labs.  
[18:48:46] <halfak>	 We'd use changeprop if it existed. 
[18:48:59] <halfak>	 I'm also interested in talking about git-lfs or some other type of versioned large file storage. 
[18:49:16] <Zppix>	 halfak:  https://lists.wikimedia.org/pipermail/wikitech-l/2017-June/088297.html
[18:50:02] <halfak>	 We have a lot of large files in the project (models and datasets for training/testing models)
[18:50:18] <halfak>	 It's a pain to check out the repo when you need to download all the history of all the models :( 
[18:50:40] <halfak>	 Zppix, yeah.  that's the person who sort of checked out and stopped responding on phab.  Maybe he'll pick it back up again soon :) 
[18:50:57] <Zppix>	 gerrit i think doesnt do that paladox can you confirm that gerrit doesn force you to dl large files to preview them?
[18:51:16] <paladox>	 Zppix what's dl?
[18:51:21] <Zppix>	 download
[18:51:38] <paladox>	 Nope gerrit dosent
[18:52:07] <halfak>	 um... this is git
[18:52:23] <paladox>	 gerrit 2.13.8+ supports git lfs with the git lfs plugin.
[18:52:34] <halfak>	 Sure.  But we need a lfs server
[18:52:37] <paladox>	 https://github.com/GerritCodeReview/plugins_lfs
[18:54:27] <Zppix>	 I'm sure releng would be willing to add that plugin?
[18:54:49] <Zppix>	 surely we're not the only ones who would use lfs
[18:56:40] <halfak>	 Zppix, I've talked to them in the past and they were less enthusiastic. :\
[18:57:17] <halfak>	 I'd really like to use git-lfs and it is critical that it is available to the public so you don't need prod access to contribute to our projects. 
[18:57:34] <Zppix>	 ill look into it if i find a lfs src control that would be combatible with our existing setup, and with wikimedia development standards ill let you know
[18:57:52] <halfak>	 We get a ton of contribs through github, so I want to be able to use it from there too. 
[18:57:54] <halfak>	 Cool :) 
[18:58:34] <Zppix>	 i could ask github if they would be willing to work with  us on that as well if you would like
[18:58:38] <halfak>	 Crap.  Forgot to squash. 
[19:00:36] <halfak>	 And now it's mixed together with other commits :\
[19:01:59] <Zppix>	 halfak:  sorry i wish github allowed my cmd line to work instead of making me do 30 steps extra to just do commit
[19:02:14] <Zppix>	 why revert?
[19:02:18] <halfak>	 Damn.  There's no easy way to fix this. 
[19:02:30] <halfak>	 Because the history is now full of bullshit
[19:02:34] <Zppix>	 halfak:  you cannot rewrite git history
[19:02:40] <halfak>	 Oh sure you can
[19:02:53] <Zppix>	 rephrase:you cannot EASILY rewrite history
[19:03:15] <Zppix>	 you would have to edit git objects etc etcf
[19:03:43] <Zppix>	 Ill make sure from now on my prs always have a squash this comment :P
[19:04:25] <halfak>	 Damn damn damn.  Why does it have to mix together all of the commits? 
[19:05:01] <travis-ci>	 wiki-ai/revscoring#1032 (revert-321-md-from-rst - afb4781 : Aaron Halfaker): The build passed. https://travis-ci.org/wiki-ai/revscoring/builds/241794458
[19:06:45] <Zppix>	 halfak:  it  just takes like a picture of the commits in the origin branch of the pr and photoshops it to master (i know thats the worst explaination ever)
[19:08:17] <Zppix>	 halfak:  dont forget to schedule your puppet patch for  deployment :)
[19:09:22] <Zppix>	 how the hell did you do that?
[19:09:33] * halfak flexes muscles. 
[19:10:08] <Zppix>	 halfak:  ive seen your picture xD just kidding 
[19:11:00] * halfak is not sure if the joke is that I don't have muscles or that I do :S  
[19:11:39] <Zppix>	 both
[19:13:00] <halfak>	 lol
[19:13:04] <paladox>	 lol
[19:13:30] <Zppix>	 you cant be mad if i say both now can you :P
[19:15:50] <halfak>	 Oh you can always be mad.  ;) 
[19:15:56] <halfak>	 But no of course not :D
[19:16:55] <paladox>	 lol
[19:17:14] <Zppix>	 we need an ores bot i cant tell if you said that in good faith halfak  :P
[19:18:12] <halfak>	 AGF
[19:28:01] <glorian_wd>	 halfak: yeah that's a strange error.
[19:39:07] <glorian_wd>	 halfak: I am seeing wikibase.py under /wikiclass/feature_lists now. I notice that there's "+1" on each attribute (e.g. wikibase.revision.claims + 1, wikibase.revision.properties + 1). I'm wondering if the underlying reason behind this. Hmm
[19:41:30] <halfak>	 log(0) is undefined. 
[19:42:19] <glorian_wd>	 Ok. And, why "log"? why not retrieve the original value (i.e. length of list)?
[19:43:24] <halfak>	 glorian_wd, there's a big difference between 2 and 3 statements. 
[19:43:36] <halfak>	 There's not that much of a difference between 20 and 21 statements. 
[19:43:46] <halfak>	 In a linear scale, the distance is the same. 
[19:43:50] <halfak>	 In a log scale it isn't. 
[19:44:08] <halfak>	 log(21) - log(20) = 0.02
[19:44:22] <halfak>	 log(3) - log(2) = 0.18
[19:44:30] <halfak>	 21 - 20 = 1
[19:44:33] <halfak>	 3 - 2 = 1
[19:49:12] <glorian_wd>	 halfak: okay makes sense. Thanks :D
[19:53:13] <Zppix>	 halfak:  is there date for the wikilabels/ores deploy?
[19:54:25] <halfak>	 Zppix, not yet.  But I'll likely do it this week.
[19:55:17] <Zppix>	 k
[19:56:21] <Zppix>	 halfak:  https://git-lfs.github.com/
[19:56:33] <halfak>	 Hey Zppix yeah.  
[19:56:37] <halfak>	 I like it. 
[19:56:43] <halfak>	 Not OK for deploys though :( 
[19:57:13] <halfak>	 Would be great if we could somehow mirror the lfs into labs and/or our production environment
[19:57:14] <Zppix>	 how are you currently doing deployments?
[19:57:24] <halfak>	 scap in prod, fabric on labs. 
[19:57:30] <halfak>	 With models in the repo itself. 
[19:57:33] <halfak>	 *model files
[19:58:44] <Zppix>	 halfak:  do you know if your prod server can talk to wmflabs servers?
[20:00:44] <Zppix>	 if so you can scap to wmflabs from prod but for wmflabs to prod, maybe use a bot that keeps them sync'd similar to translatewiki bot
[20:01:32] <Zppix>	 change prop from wmflabs to prod and prod to wmflabs isnt going to be easy
[20:01:43] <Zppix>	 you can think firewalls and networking engineers for that
[20:01:49] <halfak>	 Zppix, no changeprop from labs to prod. 
[20:01:56] <Zppix>	 irregarldess its not easy
[20:01:59] <halfak>	 Instead from prod to labs to tools
[20:02:14] <halfak>	 prod to labs *should* be easy. 
[20:02:28] <Zppix>	 that is yes
[20:02:44] <Zppix>	 prod iirc can talk to labs but labs cannot directly talk to prod
[20:02:53] <halfak>	 Right. 
[20:02:59] <halfak>	 I've got to head out folks. 
[20:03:09] <halfak>	 Time to do my late lunch again :S
[20:03:12] <Zppix>	 ill talk to ops and bawolff
[20:03:24] <halfak>	 And then I should probably take some time off this afternoon (CDT)
[20:03:26] <halfak>	 :) 
[20:03:36] <Zppix>	 hey cdt is my timezone you thief
[20:03:56] <halfak>	 Zppix, tomorrow, ottomata will come online and will be very interested in helping out with making sure that EventBus works in labs. :) 
[20:04:07] <halfak>	 I'll make sure you get introduced. :) 
[20:04:12] <Zppix>	 ok, ill talk to security about labs ->prod
[20:04:14] <halfak>	 ottomata is EDT 
[20:04:26] <halfak>	 Have a good one!
[20:04:27] <halfak>	 o/
[20:04:31] <Zppix>	 you too
[20:05:25] <glorian_wd>	 halfak: thanks for your help today. Have a good Sunday :)
[20:05:42] <Zppix>	 glorian_wd:  im here if you need me
[20:06:00] <glorian_wd>	 Zppix: okay cool
[20:06:49] <Zppix>	 glorian_wd:  is the wikidatawiki thing a public statement if not thats the import error
[20:07:22] <glorian_wd>	 Zppix: It's already solved after I degrade my mwxml
[20:07:30] <glorian_wd>	 The error was really weird though.
[20:07:42] <Zppix>	 mwxml what ver are you using now?
[20:07:45] <glorian_wd>	 degraded*
[20:07:55] <glorian_wd>	 the same version with halfak 
[20:08:05] <glorian_wd>	 0.2.2
[20:08:21] <Zppix>	 this is wikiclass or revscoring repo?
[20:18:10] <glorian_wd>	 Zppix: wikiclass
[20:18:20] <Zppix>	 ok
[20:19:47] <Zppix>	 glorian_wd:  im setting requirements.txt to always install 0.2.2
[20:24:03] <glorian_wd>	 Zppix: but according to halfak , the latest one also works
[20:24:45] <glorian_wd>	 Nonetheless, I guess it's safer to set mwxml 0.2.2 in the requirements
[20:25:42] <Zppix>	 glorian_wd:  end users should have to experiment thats our job in my opinion
[20:25:53] <Zppix>	 shouldnt*
[20:28:38] <paladox>	 halfak lol a 3:30pm lunch :)
[20:33:02] <Zppix>	 paladox:  got any ideas about change prop?
[20:33:09] <Zppix>	 all mine require talking to ops and security
[20:33:38] <paladox>	 Zppix nope, was talking to halfak today about that.
[20:36:03] <Zppix>	 whoses all in security besides bawolf?
[20:36:58] <paladox>	 Reedy
[20:37:43] <Zppix>	 ill try him
[20:39:25] <glorian_wd>	 Zppix: gotcha
[20:55:24] <Zppix>	 I suggest putting yourself on +g or +r botnets are starting to filter into freenode
[22:16:05] <glorian_wd>	 Zppix: still there?
[22:24:14] <Zppix>	 not for long ill be back around 8 cdt
[22:24:21] <Zppix>	 pm*
[22:25:03] <glorian_wd>	 Zppix: lol. I am already in bed at 8 PM CDT
[22:25:10] <glorian_wd>	 okay then. no worries.
[22:25:34] <Zppix>	 feel free to ask here and i can email you or something
[22:25:37] <paladox>	 thats really early
[22:25:38] <paladox>	 it's 23:25pm here :)
[22:35:15] <glorian_wd>	 Zppix: No. I gotta go to sleep anyway. It's already 00:34 here :). Have a nice evening folks
[22:36:15] <glorian_wd>	 Zppix: I initially wanted to ask you something. But let me try to get the answer myself first. If I get stucked again, I will ask here tomorrow :D