11:55:36 #startmeeting Ansible Contributor Summit 2021.03 11:55:36 Meeting started Tue Mar 9 11:55:36 2021 UTC. 11:55:36 This meeting is logged and archived in a public location. 11:55:36 The chair is gundalow. Information about MeetBot at http://wiki.debian.org/MeetBot. 11:55:36 Useful Commands: #action #agreed #halp #info #idea #link #topic. 11:55:36 The meeting name has been set to 'ansible_contributor_summit_2021.03' 11:56:24 o/ 11:58:01 is there supposed to be sound? 11:58:12 (I don't hear anything :) ) 11:59:04 #chair abadger1999 andersson007_ alinabuzachis cybette dericcrago dmsimard jborean93 jillr jimi|ansible lmodemal matburt mattclay maxamillion mkrizek nitzmahone ompragash Qalthos relrod samccann sdoran shertel Shrews sivel webknjaz felixfontein Tas-sos tadeboro cyberpear gwmngilfen pabelanger zbr samccann briantist resmo odyssey4me ganeshrn aminvakil geerlingguy sshnaidm ysandeep|afk ysandeep cidrblock 11:59:04 Current chairs: Qalthos Shrews Tas-sos abadger1999 alinabuzachis aminvakil andersson007_ briantist cidrblock cyberpear cybette dericcrago dmsimard felixfontein ganeshrn geerlingguy gundalow gwmngilfen jborean93 jillr jimi|ansible lmodemal matburt mattclay maxamillion mkrizek nitzmahone odyssey4me ompragash pabelanger relrod resmo samccann sdoran shertel sivel sshnaidm tadeboro webknjaz ysandeep ysandeep|afk zbr 11:59:34 felixfontein: there are people speaking now 11:59:34 cybette and gwmngilfen seem to be talking, but I cannot hear them 11:59:53 hmm 12:00:03 we cannot hear you either 12:00:06 felixfontein: three dots on the bottom right. then `settings` 12:00:36 gundalow: sound works, the test tone is playing, but I don't hear anything :) 12:00:42 I'll try to leave and join again 12:00:59 ah, NOW there is voice :) 12:04:34 usual problem I get on google meet 12:06:09 #topic Introductions 12:06:22 #info People are saying hello 12:07:39 * samccann waves 12:08:45 morning samccann :) 12:08:56 *wave 12:09:02 must be pretty early for some of you :) 12:09:43 bcoca: who are you and what did you do with brian? :D 12:09:45 pure luck I got up this early ;-) 12:10:18 hehe :) 12:12:09 i forgot to install the human speak module 12:12:19 haha 12:12:28 in the computer or in the coffee? 12:12:42 brain 12:12:48 coffee is just fuel 12:12:53 * gwmngilfen concedes 12:17:43 I know there are a lot of other people here that we will not get a chance to introduce themselves. 12:17:43 If you to say your 1) GitHub/IRC name. 2) Your interest in Ansible 3) Where in the world you are 4) What you are hoping to get from today 12:19:21 #topic Ansible Community Team Update 12:20:44 bcoca/skynet development/secret lair/plans to T-600 12:23:21 #topic Steering committee 12:23:46 #info ompragash Is introducing the Ansible Community Steering Committee 12:24:07 #info This was launched yesterday. 12:24:13 #chair justjais 12:24:13 Current chairs: Qalthos Shrews Tas-sos abadger1999 alinabuzachis aminvakil andersson007_ briantist cidrblock cyberpear cybette dericcrago dmsimard felixfontein ganeshrn geerlingguy gundalow gwmngilfen jborean93 jillr jimi|ansible justjais lmodemal matburt mattclay maxamillion mkrizek nitzmahone odyssey4me ompragash pabelanger relrod resmo samccann sdoran shertel sivel sshnaidm tadeboro webknjaz ysandeep ysandeep|afk zbr 12:25:49 #info See the announcement email https://groups.google.com/g/ansible-devel/c/FRbn4qT3Jf0 12:27:33 #info Process for requesting a new collection to be included in the `ansible` package is via https://github.com/ansible-collections/ansible-inclusion/discussions 12:29:15 Contrib summit agenda: https://hackmd.io/@ansible-community/contrib-summit-202103 12:30:17 #info Public Agenda for the Wednesday Community Meeting: https://github.com/ansible/community/issues/539 (there is an `.ics` file for your calendar) 12:34:32 hi 12:34:41 hi resmo! 12:35:04 hi everyone! 12:35:16 hi andersson007_! 12:37:00 coffeeeee! 12:37:50 \o/ 12:37:52 #info We will share a link to ompragash's slides 12:38:20 #info We will do a blog post on the Steering Committee shortly which will include more information 12:39:06 'info The objective is to increase the number of non-Ansible employees in the group 12:39:38 #info The objective is to increase the number of non-Ansible employees in the group 12:41:02 #info We will help mentor and train anyone that has interest in being more involved in any part of the Ansible Community 12:45:13 #topic Stats update - 1 year on, is Collections achieving it's goals? 12:45:32 #info gwmngilfen presents a review of the past year 12:50:23 #info The stats repo is https://github.com/ansible-community/stats-collections/ 12:51:25 #info The list of potential statistics we are interested in, and other questions we have are tracked here https://github.com/ansible-community/stats-collections/issues 12:54:04 seeing gwmngilfen with "Gregory" as name is confusing 12:54:20 at least both start with 'g' :) 12:54:24 #info If you are interested in the graphs and details that gwmngilfen is currently presenting, take a look at the more detailed blogs on https://emeraldreverie.org/ 12:54:44 Zhenech: haha, yes. I've changed my work email so I'm `gundalow Barker`, no one seems to know who `John` is 12:55:14 gundalow, that's why I own zhenech@redhat, so that people don't have to think ;) 12:55:33 Zhenech: yup, I have gundalow@redhat 12:58:13 QUESTION: What thoughts and ideas do people have 12:58:34 what collections were looked at ? community, vendor, content team, all? 12:59:07 bcoca: Good question, it's roughly the collections that are included in the `ansible` package 12:59:52 bcoca: full list is here https://github.com/ansible-community/stats-crawler/blob/main/config/collections.yml PRs welcome to add other repos in 13:00:46 We do track repos outside of gh/ansible-collection such as https://github.com/openstack/ansible-collections-openstack 13:01:19 looks like a year wrapping issue 13:02:09 felixfontein: what does? 13:02:25 #info You can see the stats at https://stats.eng.ansible.com/ 13:02:55 #topic Ansible 3.0.0: How we got here and what's next 13:03:24 #info dmsimard Is giving an update on what's been happening with the `ansible` package versioning 13:05:26 #info Blog posts on ansible recent version changes https://www.ansible.com/blog/announcing-the-community-ansible-3.0.0-package & https://www.ansible.com/blog/ansible-3.0.0-qa 13:07:35 #info Slides that dmsimard is currently presenting http://ansible.github.io/community/decks/3.0.0-community-update.html#/ (had lots of links) 13:09:33 #info ansible and ansible-core roadmaps https://docs.ansible.com/ansible/devel/roadmap/ansible_roadmap_index.html 13:16:36 Q: what will ansible-core be renamed to for 2.12? :P 13:18:37 we have not run the random name generator yet 13:19:25 Sorry I'm a bit late and groggy. Had a late night and will be online in a bit! 13:19:33 good morning geerlingguy! 13:19:38 greetings 13:19:44 o/ 13:20:36 Hello all o/ 13:20:55 geerlingguy: Morning, know it's early for you. Great to see you here. Everything has been recorded and will be on YouTube. Kept a summary at the end of https://hackmd.io/uZDSLOOdS1Kx0xfZVIATmQ 13:21:17 hello everyone, i got into something at last minute, sorry i'm late 13:21:50 aminvakil: Welcome. Lots of people come and go through the day. Feel free to ask questions here (or in Google Chat) at any point 13:22:01 :thumbs-up: 13:23:19 #info FYI we are keeping a summary of the discussions at the end of https://hackmd.io/uZDSLOOdS1Kx0xfZVIATmQ 13:23:20 The session is being recorded and will be on YouTube. There will be blog posts containing a summary of the discussions and call to actions. Subscribe to The Bullhorn to know when they are published https://us19.campaign-archive.com/home/?u=56d874e027110e35dea0e03c1&id=d6635f5420 13:23:47 then we can have ARA: ansible releases ansible ;) 13:25:24 lol 13:25:27 :D 13:25:45 too bad the name is already taken ;) 13:26:27 it's funny over the years the permutations that people have come up with for the ara acronym 13:27:57 another reporting anagram 13:28:32 it's actually another recursive acronym :p 13:28:45 do we have to choose just one? 13:29:01 aleatory random acronym 13:29:31 i just thought you were manga fan ... ara ara ... 13:30:35 or maybe it's about ara parrots 13:31:05 FYI gwmngilfen presentation `The year in review Yay? Did Collections work?` is now available online http://ansible.github.io/community/decks/2021-03-contrib-summit-stats-collections-1yr-review.html 13:31:08 the acronym not having a single definition reminds me of RDO which did a shirt about it once https://www.rdoproject.org/images/blog/rdo-tshirt.jpg?1460046750 13:35:53 #topic Execution Environments via Ansible Builder 13:39:37 While we are still in break, I encourage people to fill in a doc about contributor and maintainer experience for a discussion later: https://hackmd.io/z0LmOzdBSQ2MkecSIf5gxw 13:40:26 Sign in with GitHub then you can edit ^ 13:41:17 ========================== 13:42:02 dmsimard: are you doing surveys behind my back? :) 13:42:49 gwmngilfen: not technically a survey but I can see how it could be interpreted that way :p 13:43:58 gundalow, You could try using the "Polls" functionality in Google Meet.... 13:44:38 * tremble laughs, ok it's done now 13:44:45 tremble: looks like cybette has just done that 13:44:54 cybette++ 13:45:04 I read your mind ;) 13:45:17 * gwmngilfen dons a tinfoil hat 13:45:18 One of the downsides of Google Meet is there is only a single host 14:02:17 #topic How do we improve the experience of contributors and maintainers 14:03:35 #info dmsimard is now leading an interactive discussion about how can we improve the experience of contributors and maintainers (ie YOU) 14:03:53 morning all 14:04:08 maxamillion: Morning :) 14:04:45 #info The notes that dmsimard is presenting are at https://hackmd.io/z0LmOzdBSQ2MkecSIf5gxw?both (you are welcome to edit this) 14:06:01 QUESTION: What had caused you pain, what's been difficult, what has been difficult to understand 14:09:08 If we could wave a magic wand and fix one thing to make it easier to contribute to Ansible, what would that be? 14:09:49 our survey data suggests that the process of contributing is one of the main issues - so I'd love to hear some experiences of that, perhaps understand how we could improve that 14:12:36 #info Interactive Training Scenario. If you are new to contributing to Ansible it's really worth the 20 minutes to run through the session. It's all in the website so nothing needed on your computer. https://www.katacoda.com/ansible-community/scenarios/fixing-a-bug 14:13:40 tadeboro: there is https://docs.ansible.com/ansible/devel/dev_guide/developing_collections.html#contributing-to-collections - though probably not visible enough? or not detailled enough? 14:20:42 felixfontein: There is some information there, but what is missing is collection-specific information. 14:21:55 tadeboro: yes, definitely 14:32:09 o/ 14:33:20 Finally at my desk, and it's break time! 14:33:34 No problem though, I'm debugging a network cable that might be bad. 14:33:52 There's a tester for that. ;) 14:34:04 #topic Munch break, back at 15:20 UTC 14:34:24 Network cable tester has saved me so many hours of pain. 14:34:38 Also, the TDR built into Cisco switches is super handy. 14:39:29 sdoran: yeah my dad has one of those expensive fluke testers, and I often have a couple cables with me when I visit. For some reason I've been hitting a lot of weird cable problems lately. I guess it's because I'm starting to use 2.5/10G and some of my weaker Cat6 cables just can't cope (bad crimp, kink in the line...) 14:39:45 welcome jgoldschrafe to IRC who I've been trying to get to log in for months 😏 14:40:20 geerlingguy: fiber 14:40:22 welcome o/ 14:41:12 The transceivers are super cheap and so is the fiber 14:41:32 @geerlingguy I found a pretty affordable one from LAN Shack that has worked very well for years. The Fluke ones are just too expensive. 14:42:05 The 2.5/10G over copper does seem to be more sensitive to layer 1 issues. 14:48:09 jgoldschrafe: welcome to IRC \o/ 14:49:48 cyberpear: Would you like slides uploaded somewhere? 14:49:53 Sorry cyberpear 14:50:04 cybette: Would you like slides uploaded somewhere? 14:51:23 abadger1999: I'll ping you with a link 14:52:10 Thanks, folks! Most of my Ansible involvement has been around internal stuff, but I'm looking forward to getting involved in the community 15:00:20 jgoldschrafe: welcome :) What areas interest you? 15:17:54 we'll be resuming in about 2 minutes 15:21:38 #topic Ansible network collections update 15:22:16 #info cidrblock and ganeshrn Are giving an update on using Ansible for managing your Network infrastructure 15:22:45 #info 2021 Spring Roadmap https://github.com/ansible/community/wiki/Network%3A2021-Spring-Roadmap 15:39:23 DING DING DING: Please take the time to do the polls in Google Meet (Top right of the chat window "triangle square circle" icon) 15:42:15 Any questions for Networking, feel free to ping cidrblock or ganeshrn or join #ansible-network 15:42:26 #topic Help maintaining the Ubuntu PPA 15:42:43 #info dericcrago is giving an update on the Ansible Ubuntu PPA 15:50:33 QUESTION: Anyone interested in helping with packaging? 15:51:37 i am happy with ppa 15:55:11 ubuntu admins will look for/expect a ppa or for us to host a deb repo, I don't know how many of them we have represented here vs how much of our overall user base is on ubuntu- 15:55:29 jillr: Morning :) 15:55:38 howdy :) 15:56:08 #info Current Test PPA building what I've done for testing so far (not official): https://github.com/dericcrago/ansible.ppa/tree/main/.github/workflows -> https://launchpad.net/~deric.crago 15:56:34 #topic Docs split (ansible vs ansible-core) 15:57:08 Docs team has been trying to balance 20 spinning dishes while an earthquake and tsunami have hit at the same time. Kudos to docs team 15:57:13 #info samccann gives an updated on the various changes regarding having different release train for `ansible` vs `ansible-core` packages and how we've got separate areas for the documentation 15:57:19 geerlingguy: Well said 15:57:22 geerlingguy: good summary! 16:00:01 ++ 16:05:01 #info Agenda for the Docs Working Group: https://github.com/ansible/community/issues/579 16:18:57 #topic Documentation Survey feedback 16:20:05 The latest release of the Asible 2.9 will be this : 16:20:07 ansible 2.9.20rc1 release (final 2.9 bugfix RC!) ? 16:21:21 #info acozine (Ansible Documentation Lead) gives an update on the recent Documentation Survey 16:22:01 Tas-sos: https://pypi.org/project/ansible/#history 16:22:04 https://pypi.org/project/ansible-base/#history 16:23:26 version 2.9 is the latest version that has the Ansible as it was before the last big changes came... 16:30:16 #info gwmngilfen analysis of the Ansible Documentation Survey https://emeraldreverie.org/2021/02/23/docs-survey-2020-results/ 16:38:09 sivel: You're on-deck for contributor summit presentation 16:38:34 abadger1999: right now? or in 7 minutes? 16:38:55 sivel: on desk is now (but the previous one has also wrapped up) 16:39:20 on deck is now, at bat is officially in 7 minutes. (but the previous one has wrapped up) 16:40:37 sivel: sorry, I assumed baseball idioms but maybe those aren't American-universal. 16:41:04 ;) I got it 16:41:21 more nautical for outside USA 16:41:44 i worked for co named 'ondeck' and didnt realize till year later why they had gifted us baseball bats 16:42:41 true, deck sounds more like the deck of a ship to me 16:42:53 or a card deck 16:43:21 #topic Ansible-core 2.11 potential topics 16:43:32 #info Sivel talks about the Ansible 2.11 Roadmap 16:43:41 #info Ansible-core 2.11 roadmap https://github.com/ansible/ansible/blob/devel/docs/docsite/rst/roadmap/ROADMAP_2_11.rst 16:52:54 I guess/hope ansible-core 2.12 will also have more role docs things, and maybe even filter/test plugin docs 16:54:29 sivel: nitzmahone Could you talk about ^ 16:55:14 it's less a question then more a statement/hope :) 17:03:41 currently for tests/filters we have __doc__ (docstgrings), full doc (plugin == file) and ast parsing chains into original fucntion, the hard part is getting all 3 to work as one 17:07:15 afaik, __doc__ doesn't exist. And if it doesn't, we shouldn't add it. 17:07:51 all 3 exist in the wild 17:07:58 none of em are official 17:08:44 how about having DOCUMENTATION with a top-level dict that maps filter/test name to the actual documentation? 17:09:08 or a multi-doc YAML string in DOCUMENTATION, one doc per plugin? 17:09:10 #topic BREAK 17:09:14 21 minute break 17:09:17 felixfontein: yep. That or separate file are the ways to go. 17:10:21 separate file also improves performance, since the actual Python file gets smaller (same would be true for modules and other plugins) 17:10:57 would still require current code for backwards compat 17:11:19 and that is not even the big issue with filters/tests, its listing them across collections 17:11:27 true... 17:12:35 why the loader needs to be modified as well as consumers, then it becomes 'transparent' to rest of issues downstream, depending on how many/any of existing doc methods we want to support for that casse 17:12:38 * dmsimard grabs lunch 17:13:09 removes the need of using `ast` to get `DOCUMENTATION` too, which is a perf improvement 17:13:32 sivel: we already bypass that for most 17:13:41 ast shoudl be fallback 17:15:41 Sorry, the most important reason why we should go to python 3.8 ? 17:19:03 As opposed to? Keeping Py2 or why we chose py3.8? 17:21:14 We didn't pick 3.6 due to it's lifecycle and when support for ansible-core 2.12 will end. And there fewer distros using py3.7 as their default, than 3.6 or 3.8. 17:21:27 Yep, there are many, but a big one is reducing the test matrix and churn on EOL Pythons. 17:21:45 (esp with Python coming every ~18mo now) 17:22:18 Today it's 2.6, 2.7, 3.5, 3.6, 3.7, 3.8, and 3.9 17:22:52 We spend a ridiculous amount of time dealing with "$upstream just dropped support for $eol_python" 17:23:09 (ie, "what choice to we make to keep CI working?") 17:23:46 until we get a quantum computer that can tests all pythons simultaneouslly ... 17:23:53 If we can get ahead of that a bit with the controller, we can spend more time on y'know, actual features and code instead of fixing busted deps 17:24:44 The target side of things will always have that problem, because we want to support the builtin Pythons, but that's a fairly onerous requirement for the controller 17:24:48 and since surface of modules in core has been brought down to am min, we can just test the general infra (ansiballz, module_utils) against those pythons, each collection is then responsible for theirs 17:24:56 --^ yep 17:26:03 So the hope is that we can keep each version of the controller supported only on non-EOL Pythons, which should minimize the risk from our upstream deps dropping support for those Pythons 17:26:40 and reduce the ammount of code we care about those deps to the 'strictly required' .. why we are also very careful on adding deps in module_utils 17:27:19 to the point we had to build our own version of 'requests' in url.py 17:30:06 ================== 17:30:12 #topic Community Galaxy update 17:30:41 #info nitzmahone and awcrosby give an update on what happening with galaxy.ansible.com 17:32:55 geerlingguy: This session is for you :) 17:33:08 Oh I'm enraptured 17:34:48 Thanks for you answers, 17:41:03 Mmm waffles. 17:48:02 hehe 17:48:36 waffles.... now I get hungry :D 17:50:07 same :( 17:51:04 DING DING DING: Voting time. Top of the Google chat window, click the "Triangle, Square, Circle" then click polls. 17:52:03 heh, I didn't realize that, oops 18:05:09 Please do the polls 18:05:12 #topic Documenting Ansible Collections 18:05:41 #info samccann talks about the future for how we can better document collections 18:30:54 #topic Cloud update 18:31:16 #info jillr provides an update on the what the Cloud Team has been doing 18:45:21 #topic Top n-things we don’t have answers to 18:45:49 #info abadger1999 leads a discussion around some more difficult discussions 18:46:02 AUDIENCE PARTICIPATION REQUIRED 19:49:29 Can we have all slides in HackCMD ? 19:50:00 I will add them 19:50:20 Thanks 19:51:09 What will we do tomorrow ? Do we need to prepare something for HackAthon ? 19:51:31 just show up in this channel between 12 - 20 UTC :) 19:51:43 OK! :-) 19:59:14 @dericcrago: ping about Ubuntu PPA 19:59:47 hi apple4ever / Tas-sos (anyone else interested in Ubuntu PPA) 20:00:08 Tas-sos: It's mostly going to be driven by why you want to do 20:00:29 thanks everyone! 20:02:27 #Info Thank you everybody 20:02:50 #info Join us tomorrow in #ansible-community during the same time for Open Office Hours, Hackathon. 20:02:52 #endmeeting