19:01:26 <gregdek> #startmeeting community-working-group
19:01:26 <zodbot> Meeting started Wed Sep  7 19:01:26 2016 UTC.  The chair is gregdek. Information about MeetBot at http://wiki.debian.org/MeetBot.
19:01:26 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
19:01:26 <zodbot> The meeting name has been set to 'community-working-group'
19:01:36 <gregdek> #topic Roll Call
19:01:53 <gregdek> Agenda, as always: https://waffle.io/ansible/community
19:01:58 <rbergeron> oh
19:02:01 * rbergeron wakes up
19:02:13 * rbergeron scowls at her lag
19:02:19 <gregdek> LOL
19:02:21 <gregdek> Hiya!
19:02:28 <rbergeron> hey!
19:02:30 <linuxdynasty> hello
19:03:01 * rbergeron counts herself as present for roll call
19:03:14 <_shaps_> hi
19:04:04 <gregdek> Hi linuxdynasty _shaps_ :)
19:04:07 <gregdek> And of course rbergeron
19:04:25 <gregdek> #topic Planning for Contributor Conference Brooklyn
19:04:34 <gregdek> https://github.com/ansible/community/issues/115
19:05:03 <gregdek> Looks like we got out our blog post and our emails to let people know what's happening.
19:05:12 <gregdek> Now we need to nail down the agenda, I imagine.
19:05:23 <gregdek> rbergeron: you've been doing some thinking about that?
19:05:41 <gregdek> We should prolly take some time to check out / update the etherpad...
19:06:04 <rbergeron> gregdek: yeah, I think i have some things to add to the etherpad
19:06:16 <rbergeron> along with folks i have reached out to that i will poke again
19:06:36 <gregdek> Want to grab some time to iterate over it in the next day or two?
19:06:52 <rbergeron> yeah, let's do that... /me looks at schedule
19:07:15 <linuxdynasty> I was going to ask when will people know if they have been chosen to speak ?
19:08:03 <rbergeron> linuxdynasty: i am not sure on that front. we're mostly looking at the contributor summit portion (the day before ansiblefest) --
19:08:16 * rbergeron doesn't seen notting in here to poke at
19:08:52 <rbergeron> linuxdynasty: i know it's being reviewed so I would expect Real Soon Now, esp. since the event is fairly soon
19:09:12 <gregdek> Will ping Bill for that.
19:09:43 <rbergeron> gregdek: anytime between 12:00 - 1:30, after 2pm your time is fine tomorrow if you want to get together
19:09:50 <rbergeron> or we can ad-hoc and acdtually just remember to get together at some point
19:10:00 <gregdek> Let's do noon tomorrow for me.
19:10:43 <rbergeron> okay! i will set a thing.
19:10:50 <gregdek> OK!
19:10:56 * gundalow waves
19:11:41 <gregdek> Hi gundalow :)
19:11:55 <gregdek> #topic contributor summit sf wrapup
19:12:06 <gregdek> https://github.com/ansible/community/issues/93
19:12:21 <gregdek> gundalow rbergeron anything left here? The final wrap-up blog is out, I think.
19:12:46 <gundalow> gregdek: all done
19:12:49 <rbergeron> i think all done
19:13:14 <gregdek> ok, closing that one.
19:13:26 <gundalow> \o/
19:13:29 <gregdek> :)
19:13:33 <gregdek> #topic Module Maintainer Guidelines
19:13:35 <gregdek> https://github.com/ansible/community/issues/81
19:13:58 <rbergeron> i bet i was supposed to do something i didn't do, heh
19:13:59 * rbergeron sighs
19:14:41 <gregdek> I think it's just updating.
19:14:48 <rbergeron> okay, i'll do that. i think i did some of it but not all :)
19:15:21 <gregdek> We can always finish that up in tomorrow's meeting and get a twofer!
19:15:33 <rbergeron> i also note that we have some stuff on the contributor summit agenda / plan list to enhance some of this in the docs sprint
19:15:50 <rbergeron> or at least "improve new module development docs" iirc
19:15:58 <rbergeron> okay, i'll add that to our little meeting tomorrow
19:16:16 <rbergeron> done
19:16:55 <gundalow> FYI Scot has some stuff in flight to update developing_modules, and as part of Testing Working Group we are updating notes on what we want to see from it
19:17:14 <gundalow> (docs scot)
19:17:19 * rbergeron nods
19:17:24 <gregdek> Oh, nice!
19:17:32 <rbergeron> gundalow: awesome, thanks for the fyi
19:18:08 <gundalow> Testing notes https://public.etherpad-mozilla.org/p/ansible-testing-working-group
19:19:02 * gregdek looks
19:19:27 <gregdek> That's awesome :)
19:20:40 <gregdek> OK, moving on:
19:21:04 <gregdek> #topic Public Metrics
19:21:15 <gregdek> https://github.com/ansible/community/issues/38
19:21:21 <gregdek> jtanner: you around?
19:22:01 <jtanner> gregdek: yep
19:22:23 <jtanner> trying out weechat, so i'm not seeing notifications quickly
19:22:37 <gregdek> Ah, ok
19:22:53 <gregdek> Anything new to share on ^^^? I know you've been working on a bunch of stuff
19:23:25 <jtanner> not really ... i did make a chart for total comments per month
19:23:44 <jtanner> https://jctanner.github.io/ansible/stats/total_comments.png
19:24:13 <jtanner> i suppose contributors/stars/etc is something i should do next
19:24:28 <jtanner> been working on ansibullbot issue triage for ansible/ansible this week though
19:24:58 <gregdek> So.
19:25:06 <gregdek> What are we to make of these numbers? :)
19:25:36 <rbergeron> charts and graphs, of course
19:25:37 <rbergeron> :)
19:25:41 <rbergeron> oh, that's *with*
19:25:44 <jtanner> well, for the comments, it's safe to say we have quadrupled the github traffic since 2013
19:26:55 <gregdek> True.
19:27:20 <gregdek> My real question is: what data do we have that tells us what we could do better?
19:28:10 <jtanner> my personal inclination from this data is that employees can't scale enough to handle the volume and we have to find more ways to source community
19:28:11 <rbergeron> (are you looking for suggestions or actual concrete things that we are already keeping track of as data points?)
19:29:56 <jtanner> rbergeron: me or greg?
19:29:59 <gregdek> I just want to make sure we don't lose the purpose of keeping track of these numbers. Everyone agrees that "there's more users".
19:30:41 <gregdek> Just musing about what we want out of this. How do we present it? Do we do a regular blog post?
19:30:52 <jtanner> did you have something in mind? i don't really know the original purpose?
19:31:07 <gregdek> Yeah. That's what I'm realizing. :)
19:31:16 <rbergeron> jtanner: asking greg, mostly, but you too
19:31:18 <gregdek> We've sort of played telephone with this one.
19:31:27 <rbergeron> i think the purpose is ... multiple
19:31:27 <jtanner> i collect data to make myself feel better when i work on ansibot instead of triaging the issue queue by hand
19:31:49 <rbergeron> #1: understand where our bottlenecks are
19:32:38 <rbergeron> #2: set some goals to improve the bottlenecks, wheever they are (time from PR to merge, to first comment, whatever)
19:33:22 <rbergeron> #3: general transparency / acknowledgement of throughput (if we want to improve any of these statistics as a community, we have to be able to point to something that shows we're actually committed to makingan effort, and shows people where the improvements can be made / how they can help, whatever)
19:33:26 <jtanner> 2 is typical devops sort of concept but we have to keep in mind that when you improve a singular metric, you typically worsen others
19:33:51 <rbergeron> I think in general, super general, having any sort of dashboard showing metrics for community is soft of a standard thing
19:34:17 <jtanner> i wish it were more standard, but github is fairly lacking in that dept
19:35:02 <gregdek> So.
19:35:15 <jtanner> our backlog is also increasing due to lower "quality" of issues being submitted
19:35:40 <gregdek> Then why aren't we closing those low-quality issues and demanding higher-quality issues?
19:35:41 <jtanner> we continually gain more users who are new to linux/automation/python etc and we have to hand hold
19:35:52 <rbergeron> gregdek: maybe it's just a topic for us to chat about at the contributor summit -- find out what would actually help people
19:36:01 <jtanner> we don't have the data to properly classify "low" quality versus "high"
19:36:15 <rbergeron> ...or providing easier tools / instructions / methods to obtain the quality information we actually need?
19:36:18 <gregdek> Is the templating helping in that rehard?
19:36:21 <gregdek> Regard, even?
19:36:36 <jtanner> i have thought about trying to build a corpus of ngrams that typically imply an issue is going to require code change and giving those issues a higher weight
19:36:42 <gregdek> and yes, this could be a good summit topic
19:36:48 <jtanner> but that's a whole project of research and discovery unto it's self
19:37:13 <jtanner> the templating is actually the best thing since sliced bread
19:37:28 <jtanner> imo
19:37:59 <gregdek> That's good. :)
19:38:02 <jtanner> would prefer real fields+components but template is getting us 90% there
19:38:30 <gregdek> Are we able to apply state to issues that tell us "this issue doesn't have enough info"?
19:39:01 <_shaps_> there is a label on github 'requires_info'
19:39:07 <jtanner> issue workflow includes a path to needs_info when an important subset of template requirements were not given
19:39:25 <jtanner> component name / ansible version / issue type
19:40:04 <gregdek> Maybe we could break that out in this report. How are you building these? Anything we could play with?
19:40:20 <jtanner> building what?
19:40:32 <jtanner> the graphs?
19:41:06 <rbergeron> yeah
19:41:19 <jtanner> heavily modified fork of sivel's pr-trage: https://github.com/jctanner/pr-triage
19:41:31 <jtanner> has a few phases
19:41:44 <gregdek> Ah, yes. That's awesome. Lots of examples
19:42:08 <gregdek> Does it require running your own db?
19:42:30 <jtanner> 1) fetch github issue objects and pickle them 2) serialize objects to json and enumerate some synthetic attributes 3) create stats from json with python pandas
19:42:51 <jtanner> the data is stored on disk rather than in a db
19:43:08 <gregdek> Excellent.
19:43:27 <jtanner> unless you intend to also run the flask dashboard, which does require mysql + elasticsearch
19:44:11 <jtanner> it may be easier to share the processed data via a public ipython notebook so people can play datascientist without having to duplicate the infra
19:44:32 <jtanner> something i had intended to do when time allowed
19:44:52 <gregdek> Now *that* sounds awesome.
19:46:35 <gregdek> Sorry for asking so many questions.
19:46:53 <rbergeron> lol
19:47:55 <gregdek> And yes, i think in the end rbergeron you're right: this is a topic for the contributor summit.
19:48:07 <gregdek> thanks jtanner -- great stuff.
19:48:11 <jtanner> np
19:48:52 <rbergeron> gregdek: i'll plop it on the etherpad and we can expound more tomorrow
19:48:58 <gregdek> kk
19:49:10 <gregdek> Next:
19:49:14 <gregdek> #topic Ansible Durham Meetup
19:49:23 <gregdek> My fault. Will have a date by next week.
19:49:46 <gregdek> https://github.com/ansible/community/issues/104
19:49:57 <gregdek> Have cleared the decks post-vacation. :)
19:50:04 <jtanner> containers containers containers
19:51:02 <gregdek> yes.
19:51:46 <gregdek> OK.
19:51:49 <gregdek> #topic Open Floor
19:51:50 <rbergeron> containerize all the excitement
19:53:28 <jtanner> i heard through the grapevine that someone is putting vault into docker-container
19:53:39 <gregdek> ?
19:54:07 <gregdek> You mean ansible-container?
19:54:12 <jtanner> oops, yeah
19:54:17 <rbergeron> which vault?
19:54:20 <rbergeron> :)
19:54:23 <gregdek> I dunno. :)
19:54:33 <rbergeron> lol
19:55:15 <jtanner> i only speak of the vault i originally wrote =)
19:56:00 <gregdek> *The* vault!
19:56:06 <rbergeron> :)
19:57:42 <gregdek> ok.
19:57:45 <gregdek> And with that:
19:57:47 <gregdek> #endmeeting