19:06:09 <gregdek> #startmeeting community-working-group 19:06:09 <zodbot> Meeting started Wed May 4 19:06:09 2016 UTC. The chair is gregdek. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:06:09 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic. 19:06:09 <zodbot> The meeting name has been set to 'community-working-group' 19:06:17 <gregdek> Hi everybody! 19:06:21 <gregdek> #topic Roll Call 19:06:25 <gregdek> Who's about? Say hello! 19:06:32 <gundalow> hello! 19:07:11 <rbergeron> hello! 19:07:43 <gregdek> #info https://www.youtube.com/watch?v=tdTRq3lxnXA 19:07:44 <gundalow> hey rbergeron :) 19:07:51 <rbergeron> howdy howdy 19:08:09 <gundalow> gregdek: haha, ace 19:08:26 <gregdek> As always, our agenda: https://waffle.io/ansible/community 19:08:59 <gregdek> #topic community/80 19:09:04 <gregdek> SLA for Shipit! 19:09:10 <gregdek> Lots of discussion of this one yesterday. 19:09:29 <bcoca> sorry 19:09:38 <gregdek> No, it's good, actually. 19:09:45 <rbergeron> indeed 19:09:49 <gregdek> Helps to clarify, for me, that auto-close and auto-merge are both problematic. 19:10:10 <gregdek> Agreeing that 2 weeks == review during core meeting seems like a sensible compromise. 19:10:16 <bcoca> i've had 'auto-close' before, it ended up just 'deffering problems until they blew up spectacularly' 19:10:23 <jtanner> hello (late again) 19:11:24 <gregdek> So we should maybe wait for more consensus, but I think 2 weeks and then review is certainly good enough, and a better assurance than what we have now. 19:11:40 <gregdek> And no auto-anything there. 19:12:49 <gregdek> Anyway, that's all occurring in the issue, so we can let it play out there and move on. 19:14:12 <gregdek> Next: 19:14:42 <gregdek> community/63 19:15:16 <gregdek> https://github.com/ansible/community/issues/63 19:15:18 <gundalow> what was the context/requirement for this? 19:15:29 <gundalow> (out of interest) 19:15:46 <rbergeron> context is: we want to grow things and make sure stuff is useful to folks, and need a way to measure it :) 19:15:50 <gregdek> We want to know how many "orgs" there are in Galaxy so we know whether we're successfully getting Whole Organizations to use Galaxy as their Official Thing. 19:15:53 <rbergeron> in some way or another 19:15:58 <rbergeron> even as a starting point 19:16:24 <gregdek> But we've now determined that this isn't easily done without a change to the code, so moving this over to the Galaxy stuff. 19:16:35 <gundalow> ok, thanks for context :) 19:17:06 <gregdek> Next: 19:17:22 <gregdek> #topic community/40: code of conduct 19:17:28 <gregdek> https://github.com/ansible/community/issues/40 19:17:39 <gregdek> Still waiting on that meeting with Red Hat legal. No update. Will ping again. 19:18:15 <rbergeron> (side note, created issue in galaxy for tracking orgs -- https://github.com/ansible/galaxy-issues/issues/154) 19:18:37 <gregdek> Yay! Then you can note that in community/40 and close it or move it to backlog as you prefer. 19:18:45 <gregdek> Next: 19:19:08 <gregdek> community/51: weekly new module report 19:19:18 <gregdek> #topic community/51: weekly new module report 19:19:29 <gregdek> https://github.com/ansible/community/issues/51 19:19:43 <gregdek> Done! Now "automated enough" that it runs every Tuesday. 19:19:57 <rbergeron> yay 19:20:24 <gregdek> I FINISHED A THING 19:20:29 <gundalow> \o/ 19:20:50 * bcoca mails lolypop 19:20:51 <gregdek> Next: 19:21:08 <gregdek> #topic community/36: issue triage process 19:21:32 <gregdek> So rbergeron has this complete, and it's not being implemented. I started on it, but fortunately resmo has stepped in :) 19:21:40 <gregdek> (Thank you resmo) 19:21:45 <rbergeron> s/not/now ? 19:21:47 <rbergeron> i assume? 19:21:49 <gregdek> So I'm expecting something by middle of next week. 19:21:54 <gregdek> LOL NOW BEING IMPLEMENTED 19:22:02 <gregdek> NOT NOT BEING IMPLEMENTED 19:22:09 <gregdek> STOOPID ENGLISH 19:22:13 <gregdek> Sigh 19:22:13 <rbergeron> bad fingers 19:22:14 <rbergeron> bad 19:22:22 <gregdek> Anyway. It's moving. :) 19:22:27 <rbergeron> well: i am happy to test the things as they come along 19:22:27 <jtanner|t420> nice 19:22:30 <rbergeron> or whatever 19:22:45 <rbergeron> otehrwise, SO looking forward to mucking out the stalls of corner cases after this is implemented 19:22:58 <rbergeron> lmk if you need my address for shipping me alcohol to go with that 19:23:01 <gregdek> jtanner|t420: we need to meet. You're going to drink from this firehose soon. :) 19:23:12 <rbergeron> ahem. anyway :) 19:23:27 <jtanner|t420> gregdek, i was going to come to the office tomorrow, if that works for you 19:23:34 <gregdek> Nope! Next week. 19:23:38 <jtanner|t420> very well 19:23:52 <gregdek> I'm in Maryland, out of the office until next Tuesday. I'll set up a calendar invite if yours is up-to-date. 19:24:01 <jtanner|t420> yep 19:24:40 <bcoca> let there be beer! 19:24:49 <gundalow> o/ 19:24:56 <gregdek> Next: 19:25:16 <gregdek> #topic community/47: travis performance 19:25:28 <gregdek> https://github.com/ansible/community/issues/47 19:25:39 <gregdek> Ok gundalow, lots of chatter! What's actually going on? 19:25:51 <gundalow> Progress has been made! I've redone the ticket so you can actually tell what's been done 19:25:59 <gregdek> w00t 19:26:17 <gundalow> following the items in the ticket 19:26:34 <gundalow> a) fired a detailed set of questions to Travis, waiting to here back 19:26:52 <gundalow> b) We have stats now http://buildtimetrend.herokuapp.com/dashboard/ansible/ansible/index.html 19:27:20 <gregdek> And do those stats lead us to an action? 19:27:20 <jtanner|t420> i might be to blame for the git test expansion, but it was a necessary addition 19:27:50 <gundalow> jtanner|t420: I don't disagree that the tests add value. I may be able to tweak the repos they use to be slightly smaller 19:27:53 <gundalow> gregdek: YES! 19:28:07 <gregdek> Even better! 19:28:07 <gundalow> TASK: setup_mysql_db : install mysqldb_test debian dependencies ------- 134.92s 19:28:09 <jtanner|t420> gundalow, the tests i added create and use local repos 19:28:10 <gundalow> TASK: setup_postgresql_db : install dpkg dependencies for postgresql test -- 39.58s 19:28:13 <gundalow> TASK: test_unarchive : Ensure zip is present to create test archive (apt) -- 29.42s 19:28:16 <gundalow> TASK: test_apache2_module : install apache via apt --------------------- 17.47s 19:28:26 <gundalow> jtanner|t420: ah, may be other tests then that pull from github 19:28:50 <jtanner|t420> i loose track of what i do after a day or so 19:29:22 <gundalow> the above durations are from my laptop with SSD on 14.04. I'm in the process of rebuilding docker images and testing with those packages installed, I hope they will all drop to < 0.5seconds 19:29:29 <gundalow> jtanner|t420: that's what git whatchanged is for 19:30:59 <gundalow> so I hope to shave ~ 3 minutes off each of the test runs 19:31:08 <jtanner|t420> maybe it wasn't git 19:31:33 <gregdek> That's good stuff there. 19:32:01 <gundalow> Thanks to goes to robinro for the initial idea 19:32:29 <gregdek> Thx for the update. Next actions are represented in the issue, right? 19:32:33 <gundalow> yup 19:32:40 <gundalow> nothing else to report on Travis 19:32:54 <gregdek> OK, that's all of our In Progress ones. 19:33:19 <gregdek> We also just went through and dragged a few of the "ready" into in progress just now. 19:33:30 <gregdek> So we'll start work on those this week. 19:33:34 <gregdek> #topic Open Floor 19:33:43 <gregdek> Anything anyone want to chat about in particular? 19:33:43 <gundalow> hum, maybe off topic but is travis broken on devel still? 19:33:53 <gregdek> oh, dunno. 19:34:32 <jtanner|t420> get_uri test falls over for all my PRs 19:35:01 <gregdek> bcoca abadger1999 sivel ^^^ ? 19:35:12 <jtanner|t420> it's a timeout in the remote connection 19:35:18 <gregdek> Ohhhh. 19:35:40 <gregdek> Is that a travis problem? 19:35:43 <bcoca> isn't someone setting up 'local http' to test SSL? 19:35:47 <gregdek> Like, with travis itself? 19:35:48 <gundalow> external resource I think 19:35:54 <bcoca> ^ use that to test get_url and avoid external resource issues 19:36:04 <sivel> alikins: I think is working on setting up the "local" test endpoints 19:36:09 <abadger1999> alikins: Were you working on moving those tests into a web server we startup as part of the integration tests? 19:36:34 <gundalow> issues with https://sni.velox.ch I think 19:36:47 <abadger1999> Ugh. 19:37:15 * gundalow mutters about test systems having external access 19:37:20 <abadger1999> I seem to recall that what alikins was working on would not cover SNI. 19:37:28 * gundalow fetches his QA hammer 19:37:34 <abadger1999> So that would be even further down the pipe. 19:37:49 <gregdek> So what's calling sni.velox.ch? 19:37:53 <alikins> https://github.com/ansible/ansible/pull/15688 has (non ssl) httpbin 19:38:08 <jtanner> gregdek: tasks in test_uri role 19:38:18 <gundalow> 228 - name: Verify SNI verification fails on old python without urllib3 contrib 1 uri: 2 url: 'https://sni.velox.ch' 3 ignore_errors: true 19:38:23 <gundalow> 4 when: not ansble_python.has_sslcontext 5 register: result 19:38:27 <gundalow> hum, that didn't past well, sorry 19:39:19 <gregdek> Anything we can do here? 19:39:26 <gundalow> kill the test for a bit? 19:39:44 <gregdek> Would be nice not to break travis for everyone while we wait... 19:41:01 <gregdek> Do all travis tests fail right now, or only certain ones, or...? 19:41:10 <abadger1999> Hmm... https://sni.velox.ch is responding for me. 19:41:15 <gundalow> travis stops when it hits a failing test 19:41:28 <abadger1999> Maybe it's not an external problem... Maybe code is broken. 19:41:30 <gregdek> And this test is present for every run for every PR? 19:41:34 <alikins> how often does it fail? in batches or some subset? sni.velox.ch aparently rotates it's ssl config and certs 19:41:41 <gundalow> gregdek: correct (for ansible/ansible 19:41:54 <gregdek> But not for modules? 19:42:30 <gundalow> we don't run the integration tests on ansible-modules-core/extra. Only the basic linter type validators 19:43:30 <alikins> do we whitelist the external hosts? or could a pr add any random http request? 19:44:01 <abadger1999> This passes on ubuntu12 and fails on everything else? That implies to me that hte code is broken. 19:45:03 <gregdek> Perhaps this is better for #ansible-devel at this point? 19:45:18 <gregdek> Sounds like not much us poor community folk can do about that :) 19:45:36 <abadger1999> yeah. 19:45:47 <rbergeron> :) 19:45:52 <gregdek> But I wish you good luck and much success! 19:45:56 * rbergeron can get out her stick and bang it a lot 19:46:25 <gregdek> Anything else for our open floor? 19:46:40 <gregdek> Because if not... 19:47:07 <gregdek> I will adjourn in 5... 19:47:12 <gregdek> 4... 19:47:18 <gregdek> 3... 19:47:22 <gregdek> 2... 19:47:26 <gregdek> 1... 19:47:29 <rbergeron> so good at counting backwards 19:47:30 <gregdek> 1/2... 19:47:34 <gregdek> #endmeeting