Chat log from Ansible Contributor Summit Day 2, October 15, 2020

ChatLog_Ansible+Contributor+Summit+-+Current+Contributor+Workshops+%2F+Unconference_15Oct2020.txt

EVENT NAME: Ansible Contributor Summit - Current Contributor Workshops / Unconference
EVENT DATE: THURSDAY, 15 OCTOBER 2020
-----------------------------------------------

EVENT CHAT

[2020-10-15 09:13:08] [David Moreau-Simard] o/ I join early just for the upbeat on-hold music
[2020-10-15 09:13:46] [Carol Chen] lol you're a champ
[2020-10-15 09:14:00] [Carol Chen] this is REALLY early
[2020-10-15 09:17:36] [David Moreau-Simard] but the song is so good
[2020-10-15 09:20:16] [Gregory Sutcliffe] /me gets a book
[2020-10-15 09:30:49] [Himadri Ganguly] Hi everyone from India.
[2020-10-15 09:37:22] [Tadej Borovšak] Hi all! o/
[2020-10-15 09:38:38] [Carol Chen] Hello! \o
[2020-10-15 09:45:19] [Carol Chen] Thanks for joining! We will start in 15 minutes.
[2020-10-15 09:51:09] [James Cammarata] o/
[2020-10-15 09:52:17] [Manuel Millan] hi from Spain
[2020-10-15 09:55:07] [Dmytro Rabotiahov] Hey. Does it _really_ require flash?
[2020-10-15 09:55:48] [Ton Kersten] Nice tune. Should have connected earlier :-)
[2020-10-15 09:55:58] [Erik Widholm] reminds me of ROL
[2020-10-15 09:56:01] [Ton Kersten] Hello all \o
[2020-10-15 09:56:08] [Dmytro Rabotiahov] It's end of 2020 outside lol
[2020-10-15 09:56:21] [Tadej Borovšak] @Dmytro Rabotiahov No? I do not have flash for a few years now and BlueJeans workds just fine.
[2020-10-15 09:56:37] [Dmytro Rabotiahov] I have `To watch the video, you need Adobe Flash Player 11.1 or higher`.... weird then
[2020-10-15 09:57:57] [Dmytro Rabotiahov] Ok, seems it decided that my browser do not support html5
[2020-10-15 09:58:43] [Carol Chen] Dmytro: which browser are you using? Chrome or Firefox should both work
[2020-10-15 10:01:20] [Brian Coca] flash should only be fallback to html5, check if you have any javascript restrictions
[2020-10-15 10:01:28] [vinod pandey] Hi from India
[2020-10-15 10:01:34] [Carol Chen] Hello everyone!
[2020-10-15 10:01:48] [Carol Chen] You can follow along with the etherpad: https://etherpad.opendev.org/p/virtual-ansible-contributor-summit-october-2020
[2020-10-15 10:01:49] [Dmytro Rabotiahov] was using yandex browser, which is runnin chromium 84.0.4147
[2020-10-15 10:02:05] [Gundalow ] https://etherpad.opendev.org/p/ansible-contributor-summit-october-2020-morning
[2020-10-15 10:02:10] [Mitchell Whalen] morning from NC
[2020-10-15 10:02:29] [Carol Chen] IRC:https://webchat.freenode.net/#ansible-community
[2020-10-15 10:02:51] [Gabriel Benselum] Hi!!!!
[2020-10-15 10:03:03] [baptiste mille-mathias] Hi there
[2020-10-15 10:04:04] [Srinivasa Nagaraja] can you please send  the link for the session held on oct 12
[2020-10-15 10:04:32] [Gregory Sutcliffe] monday's recordings will be up on youtube next week
[2020-10-15 10:04:59] [Carol Chen] We will share them next week as we were really busy this week with Contributor Summit and AnsibleFest. Thanks for your patience!
[2020-10-15 10:05:05] [Srinivasa Nagaraja] okay any specfic channel we need to search for to find the recording ?
[2020-10-15 10:05:22] [Lonnie Holcomb] morning from NC as well
[2020-10-15 10:05:37] [Adefemi Micheal] Hi, Adefemi here from Nigeria
[2020-10-15 10:06:25] [Brian Coca] qalthos
[2020-10-15 10:06:34] [Brian Scholer] hello from Queens, NYC
[2020-10-15 10:06:47] [baptiste mille-mathias] bonjour James
[2020-10-15 10:06:49] [Brian Coca] jimi-c
[2020-10-15 10:06:54] [Brian Coca] jimi|c
[2020-10-15 10:07:06] [James Cammarata] jimi-c on github, jimi|ansible on irc
[2020-10-15 10:07:07] [Carol Chen] the youtube channel is https://www.youtube.com/c/AnsibleCommunity but we will also email all the attendees
[2020-10-15 10:07:12] [James Glynn] hello, hailing from South Florida, US
[2020-10-15 10:07:12] [Jeff Geerling] @Matt Davis - Very dramatic lighting
[2020-10-15 10:07:21] [James Cammarata] it's still the middle of the night there
[2020-10-15 10:07:26] [Patrick Toal] David: Shadowy figure :)
[2020-10-15 10:07:29] [Joel Farthing] Hello again from Saskatoon
[2020-10-15 10:07:36] [Gregory Sutcliffe] very halloween, needs a torch under then chin :)
[2020-10-15 10:07:39] [kefayat ur rehman] Hi, everyone  evening from Pakistan.
[2020-10-15 10:07:40] [David Newswanger] I go by newswangerd on github and IRC. Forgot to mention that
[2020-10-15 10:07:41] [Gregory Sutcliffe] *the
[2020-10-15 10:07:50] [Jeff Geerling] Middle of the night == I need giant spotlights to try to stay awake :D
[2020-10-15 10:08:05] [Amin Vakil] hello felix!
[2020-10-15 10:08:21] [Felix Fontein] hi amin! :)
[2020-10-15 10:09:01] [Adam Morris] Another Adam  :-)
[2020-10-15 10:09:54] [Jeff Geerling] ansible-playbook diet-dr-pepper-me.yml
[2020-10-15 10:10:13] [baptiste mille-mathias] @jeff did you think to add become: true ?
[2020-10-15 10:10:17] [Brian Coca] ansible -m soda -a'staate=cold'
[2020-10-15 10:10:32] [Gregory Sutcliffe] nice green screen sorin :)
[2020-10-15 10:10:34] [Jeff Geerling] @baptiste heh, that gets you the 64oz fountain soda
[2020-10-15 10:11:11] [baptiste mille-mathias] you confused me with pounds and ounces
[2020-10-15 10:11:12] [baptiste mille-mathias] :)
[2020-10-15 10:11:27] [scott boss] sorry I can speak right now
[2020-10-15 10:11:40] [scott boss] have too much background. sorry.
[2020-10-15 10:12:02] [baptiste mille-mathias] Jeffi s grabing some soda ?
[2020-10-15 10:12:08] [Jeff Geerling] Weird
[2020-10-15 10:12:09] [Adam Miller] oh, I forgot to mention that I'm maxamillion on irc and github ... because I'm bad at this game
[2020-10-15 10:12:12] [Jeff Geerling] I'm Jeff (geerlingguy)
[2020-10-15 10:12:13] [scott boss] I can not speak I have a class in th erom with me
[2020-10-15 10:12:20] [Jeff Geerling] No clue why it's not working, worked fine Monday!
[2020-10-15 10:12:39] [Felix Fontein] @maxamillion you're not the only one ;)
[2020-10-15 10:12:49] [Jeff Geerling] I tried three mics and two cameras, and none of them worked this morning :(
[2020-10-15 10:13:01] [Manuel Martins] What's you role? I watch your videos, they are great. (Jeff)
[2020-10-15 10:13:03] [Lonnie Holcomb] Hello Jeff Geerling, enjoy your ansible-galaxy work :)
[2020-10-15 10:13:04] [scott boss] problem with working at a university is there is always lots of background noise from the students/clases
[2020-10-15 10:13:11] [Gregory Sutcliffe] Jeff - technology++
[2020-10-15 10:13:23] [Toshio Kuratomi] I just saw you, Jeff.
[2020-10-15 10:13:24] [Brian Coca] nitzmahone
[2020-10-15 10:13:35] [Jeff Geerling] I had to switch cameras two more times apparently... now it works
[2020-10-15 10:13:38] [Adam Morris] "You want the 32oz or the large?"
[2020-10-15 10:13:47] [Brian Coca] abadger/abadger99
[2020-10-15 10:13:49] [Adam Miller] A LITER IF CIKA
[2020-10-15 10:13:51] [Adam Miller] COLA*
[2020-10-15 10:13:56] [Jeff Geerling] @Manuel - I've been working a lot on the Kubernetes and some other Cloud collections this year
[2020-10-15 10:13:57] [Adam Miller] .... I really messed that up :P
[2020-10-15 10:14:35] [Gregory Sutcliffe] ifty has a great "bond villian" vibe going on there ...
[2020-10-15 10:14:51] [Tim A] Oh and I'm @tima in github and IRC.
[2020-10-15 10:14:58] [Manuel Martins] Your raspberry pi videos are great :D
[2020-10-15 10:15:13] [Tim A] I guess I pulled an Adam. ;)
[2020-10-15 10:15:25] [Adam Miller] woot!
[2020-10-15 10:15:31] [Jeff Geerling] My mic might work now, who knows :P I can get my video to work at least
[2020-10-15 10:16:56] [Gregory Sutcliffe] working with words is super-underappreciated
[2020-10-15 10:17:03] [David Moreau-Simard] +1
[2020-10-15 10:17:08] [Jeff Geerling] +1
[2020-10-15 10:18:34] [Jeff Geerling] Oh well, tried another mic too, that didn't work
[2020-10-15 10:18:39] [Brian Coca] seen but not heard
[2020-10-15 10:18:40] [Lonnie Holcomb] Jeff, pull and plug inyour mic
[2020-10-15 10:18:43] [Patrick Toal] It's not meant to be, Jeff. :)
[2020-10-15 10:18:51] [Brian Coca] stop and start again your raspi
[2020-10-15 10:18:53] [Tim A] read the playbook over it again.
[2020-10-15 10:18:56] [Jeff Geerling] I tried plugging in a net new mic and that didn't work either :P
[2020-10-15 10:18:56] [Brian Coca] ;-p
[2020-10-15 10:18:57] [Patrick Toal] Hello IT, Have you tried turning it off and on again?
[2020-10-15 10:18:59] [Ompragash Viswanathan] Hi Everyone :)
[2020-10-15 10:18:59] [baptiste mille-mathias] these words fronm Jeff had to be said
[2020-10-15 10:19:04] [Eustaquio Guimarães] Good Morning
I'm from Brasília-DF - Brazil. Let's learn.
Thank you
[2020-10-15 10:19:25] [Carol Chen] The BullHorn: http://bit.ly/thebullhorn
[2020-10-15 10:19:42] [Invitado] Hi Grimaldo Santiago - Chile
[2020-10-15 10:19:52] [David Blackwell] still waiting for you all to need a new TMM
[2020-10-15 10:20:05] [Alicia Cozine] Wow, it's so great to see such a wide geographic distribution! Welcome everyone and thank you for putting up with the time differences.
[2020-10-15 10:20:50] [Ton Kersten] But we don;t see it
[2020-10-15 10:21:02] [Ton Kersten] There it is
[2020-10-15 10:21:11] [Jeff Geerling] Elbows!
[2020-10-15 10:21:12] [Carol Chen] Sorry, there's a slight lag for attendees
[2020-10-15 10:21:13] [Toshio Kuratomi] Aye
[2020-10-15 10:21:25] [baptiste mille-mathias] yep
[2020-10-15 10:26:04] [Konrad Rzentarzewski] can you share today's agenda link with us?
[2020-10-15 10:26:12] [Jeff Geerling] https://etherpad.opendev.org/p/virtual-ansible-contributor-summit-october-2020
[2020-10-15 10:26:17] [Konrad Rzentarzewski] tnx!
[2020-10-15 10:26:25] [Gundalow ] @Konrad: https://etherpad.opendev.org/p/ansible-contributor-summit-october-2020-morning
[2020-10-15 10:26:43] [Gundalow ] ^ for now till 2000UTC
[2020-10-15 10:26:56] [baptiste mille-mathias] I cannnot answer to Questin in Q&a serction is it limited to moderator >
[2020-10-15 10:27:36] [Adam Morris] And some "treatments" are palliative
[2020-10-15 10:28:05] [Carol Chen] @baptiste yeah sorry it's limited to moderators/presenters. if you have answer to add I'd be happy to copy it there for you
[2020-10-15 10:28:05] [Brian Coca] but you are!!
[2020-10-15 10:29:55] [Kevin Breit] I may have missed this but are these stats going to be publically available?
[2020-10-15 10:30:09] [Gundalow ] @Kevin: Yes
[2020-10-15 10:30:18] [baptiste mille-mathias] kevin: https://stats.eng.ansible.com/
[2020-10-15 10:30:26] [Gundalow ] DING DING DING, Please take the time to do the Poll
[2020-10-15 10:31:17] [Adam Miller] something came up, I need to drop ... be back asap
[2020-10-15 10:31:19] [Adam Miller] <3
[2020-10-15 10:32:02] [Adam Morris] @Gregory If nobody else has questions then please explain why you can't use a survival curve
[2020-10-15 10:32:23] [Adam Morris] Because I am curious.
[2020-10-15 10:32:26] [Paul Belanger] I thought github api for user had type fileld to say if user or bot
[2020-10-15 10:32:37] [David Blackwell] there was a question about "do we have to use containers".  No, but it makes testing different releases trivial, as well as moving your whole ansible config from system to system.
[2020-10-15 10:34:47] [Carol Chen] Thanks David, added.
[2020-10-15 10:35:57] [Adam Morris] That makes sense, Thank you.
[2020-10-15 10:38:55] [Todd Lewis] re: monitoring -- we use splunk to examine our Tower/AWX logs, and have it email daily for some things, hourly for others, looking for specific things we've found useful.
[2020-10-15 10:39:33] [Adam Morris] But what if someone missed the first question but answered the second one?
[2020-10-15 10:41:59] [Adam Morris] Fair enough.
[2020-10-15 10:42:17] [Gundalow ] I hope this helps explains why we do surveys, we really need this feedback.
[2020-10-15 10:42:24] [Adam Morris] Thank you for answering my odd questions though.
[2020-10-15 10:44:06] [Carol Chen] And we will randomly select survey responders and send you swag ;)
[2020-10-15 10:44:40] [baptiste mille-mathias] perhaps finding new contributors on github having done a PR could a good target for survey
[2020-10-15 10:45:47] [baptiste mille-mathias] overall do you know if collections brought more contributions or contributors ?
[2020-10-15 10:46:09] [Ton Kersten] Alaska it is :-)
[2020-10-15 10:46:26] [Carol Chen] Ton: haha yes!
[2020-10-15 10:46:46] [] None in Africa
[2020-10-15 10:46:55] [Lonnie Holcomb] well I have a work meeting
[2020-10-15 10:46:57] [Adam Morris] @Ton, Hawaii if they get enough people there it could cover the entire pacific with a blue circle!
[2020-10-15 10:47:46] [Ton Kersten] @Adam: Weather wise that would be a better idea :-)
[2020-10-15 10:49:15] [Gundalow ] Current bubble plot: https://github.com/ansible/community/blob/master/decks/images/community_contributions_bubbleplot.png
[2020-10-15 10:49:35] [Adam Morris] (My nearest meetup is 100 miles away but I doubt that there are enough people within about 30 miles that would be interested to make a meetup happen closer...)
[2020-10-15 10:50:06] [baptiste mille-mathias] thanks !
[2020-10-15 10:50:13] [Adam Morris] Do you have a survival curve for your to do list items?
[2020-10-15 10:50:20] [Gregory Sutcliffe] dont even go there
[2020-10-15 10:50:30] [Jefferson  Santos] Amazing stats! Very fun and interesting.
[2020-10-15 10:50:52] [baptiste mille-mathias] I love stats
[2020-10-15 10:51:09] [Mark Chappell] For AWS collections I'd be interested in "member" vs non-member contributions
[2020-10-15 10:51:10] [Gregory Sutcliffe] me too :)
[2020-10-15 10:51:21] [Adam Morris] only slightly?
[2020-10-15 10:51:24] [scott boss] thank you for the stats.
[2020-10-15 10:51:35] [James Glynn] great stats presentation, thanks greg
[2020-10-15 10:51:38] [Srikanth Chandrasekaran] Are there any users of Palo Alto Ansible collections here?
[2020-10-15 10:51:40] [baptiste mille-mathias] I don't understand graphes stats but I like the indea of crossing data
[2020-10-15 10:51:43] [Gregory Sutcliffe] mark, so ther challenge there is, can you define the difference? for that bubble plot I literally had to get gundalow to give me a list of names
[2020-10-15 10:51:53] [Jefferson  Santos] I really like to know what PostgreSQL community  is focused ritght know.
[2020-10-15 10:52:26] [Felix Fontein] Andersson007 can tell :)
[2020-10-15 10:52:49] [Matt Davis] Who moved my cheese?
[2020-10-15 10:53:15] [Mark Chappell] @Gregory - Contributions from folks who don't have commit access.  I suspect I'm skewing some of the staff/non-staff stats
[2020-10-15 10:53:42] [Dylan Silva] bbl folks!
[2020-10-15 10:54:31] [baptiste mille-mathias] kudo to felix
[2020-10-15 10:54:41] [Mark Chappell] Felix rocks.
[2020-10-15 10:54:43] [Matt Davis] Hear hear- thanks so much Felix!
[2020-10-15 10:54:49] [Brian Scholer] ++ for Felix!
[2020-10-15 10:54:57] [Felix Fontein] thanks!
[2020-10-15 10:54:58] [Carol Chen] \o/ for Felix
[2020-10-15 10:55:47] [Iftikhar Khan] Felix rocks!
[2020-10-15 10:57:21] [David Blackwell] what, if any version of 'Ansible' will not auto include collections?
[2020-10-15 10:57:22] [Felix Fontein] Ansible 2.11.0 will depend on ansible-base 2.10.x
[2020-10-15 10:57:44] [Ton Kersten] Will there be RPM's for RHEL[78]
[2020-10-15 10:58:20] [scott boss] checkout a tool called "fpm" for making packages.. it can make "clean" packages for many platforms.
[2020-10-15 10:58:40] [scott boss] and it is easy to automate/CI
[2020-10-15 10:59:00] [Tim A] Aren't these criss-crossing version numbers going to confuse the world?
[2020-10-15 10:59:04] [Ton Kersten] I know fpm, but in the 2.9 code you could do "make rpm" and that's gone in 2.10
[2020-10-15 10:59:39] [Adam Morris] @Tim A, People are already confused by AWX/Tower vs Ansible Core versions.
[2020-10-15 11:00:31] [Tim A] @Adam I get it, but those are two distinct projects.
[2020-10-15 11:00:51] [Matt Davis] (heh, wasn't sure if @Adam Morris' comment was in favor or not)
[2020-10-15 11:00:52] [David Blackwell] @Tim A.  I teach an Ansible workshop weekly.. most people don't see them as different projects
[2020-10-15 11:01:04] [Richard Elrod] @Ton we're not planning on doing RPMs for 2.10. 2.11 plans are still under discussion.
[2020-10-15 11:01:15] [Adam Morris] I would imagine that the collections will basically become their own "projects"
[2020-10-15 11:01:15] [David Blackwell] Thats why I make sure to emphasize that the engine is the doer.. Tower doesn't 'do' anything, it added enterprise features.
[2020-10-15 11:01:38] [Ton Kersten] @Adam: Just need to cope with pip then
[2020-10-15 11:03:13] [Adam Morris] @Matt, I don't see it as a problem moving forward.  It would become more like Python libraries vs python itself... but maybe that's just me.
[2020-10-15 11:04:00] [Ton Kersten] /me needs to go now. Maybe will be back later.
[2020-10-15 11:04:17] [Carol Chen] Ton: thanks for joining! see ya
[2020-10-15 11:04:49] [Joshua Robinett] Wait - are you suggesting a 3.0 release may break backwards compatibility?
[2020-10-15 11:04:51] [Ton Kersten] See ya. Enjoy
@Cybette: Laphroigh beer has been ordered :-)
[2020-10-15 11:05:04] [Carol Chen] Ton: o/
[2020-10-15 11:05:13] [Ton Kersten] Bye
[2020-10-15 11:05:30] [Adam Morris] Laphroaig beer? I need to know more.
[2020-10-15 11:06:34] [Toshio Kuratomi] Tima, yes.  I think it is confusing.  I think as the version numbers diverge further and further, it will actually become less confusing but that's not where we started, unfortunately.
[2020-10-15 11:06:36] [Ton Kersten] @Adam. Local brewer lets his beer mature in old Laphroaigh barrels. And I have ordered some
[2020-10-15 11:07:01] [Joel Farthing] Oh my^^
[2020-10-15 11:07:07] [Carol Chen] Adam: to look forward to the day when we can have F2F meetups again.
[2020-10-15 11:07:15] [Adam Morris] @Ton, Ah, interesting.
[2020-10-15 11:07:38] [Carol Chen] Bullhorn: 2 weeks, but 3 weeks this time due to AnsibleFest/Contributor Summit
[2020-10-15 11:07:55] [baptiste mille-mathias] F2F meetup and hackathon !
[2020-10-15 11:09:40] [Adam Morris] (I have part of a bottle of Laphroaig quarter cask here.  :-)   )
[2020-10-15 11:10:47] [Carol Chen] @Adam: ooh that's nice
[2020-10-15 11:11:39] [Adam Morris] (Next poll is going to be favourite tipple...)
[2020-10-15 11:12:38] [Sorin Sbarnea] don't add any more garbage, only stuff that is actively maintained should be included. others should use galaxy install.
[2020-10-15 11:12:43] [Sascha Schmidt] Basically I would want to be asked but I do not believe it's a good idea to ask to many ppl
[2020-10-15 11:13:10] [Adam Morris] I want a "it depends" option...
[2020-10-15 11:14:14] [Carol Chen] Adam: what will be your considerations before getting involved? feel free to respond in chat!
[2020-10-15 11:14:51] [Felix Fontein] True :)
[2020-10-15 11:15:00] [Brian Coca] or three
[2020-10-15 11:17:01] [Adam Morris] Carol: A bunch of different factors.  What the policy is about.  How much time it will take.  How much work I have currently on my plate.
[2020-10-15 11:18:30] [Patrick Toal] Maybe spin up a quick Google Form?
[2020-10-15 11:18:52] [Carol Chen] Adam: good points
[2020-10-15 11:23:23] [Gundalow ] QUESTION: What do people think about more content from community.general into it's dedicated collection?
[2020-10-15 11:25:11] [Gundalow ] If you want to speak there is a "wave hand" option and we can enable your microphone
[2020-10-15 11:25:47] [Adam Morris] Carol: If it is a policy that I feel is important to me then I will make time.  If It won't take much time I will put forward my opinion if I feel I have something to say.  The problem with that is all discussions could just become bogged down  with people who drastically disagree.
[2020-10-15 11:26:42] [scott boss] I do think smaller collections is a good idea.  over a few very large ones.
[2020-10-15 11:26:43] [Alicia Cozine] I thought it was maintainer-dependent:  if a maintainer or group of maintainers steps forward for modules in *.general, she/he/they can move that content into a separate collection.
[2020-10-15 11:28:20] [James Glynn] I am a newbie to community, just now wanting to start contributing, so take my answer with a grain of salt lol... but I kind of lean towards the idea that if specifically maintained, should be in its own collection
[2020-10-15 11:28:26] [Adam Morris] I would have thought the long term goal would be to shrink community.general but maintaining backwards compatibility will be tricky.
[2020-10-15 11:28:48] [David Blackwell] agreed.. the point was to not have a huge collection of modules especially when many of them might not be used
[2020-10-15 11:28:53] [Sandra McCann] I'm not following how c.g 2.0 in Ansible package, along with c.postgres (or whatever the example was) - how is that breaking backward compatibility for Ansible users?
[2020-10-15 11:29:28] [Sandra McCann] (in 2.10.x)
[2020-10-15 11:29:32] [David Moreau-Simard] I share Adam's concern in that moving things in and out of collections (and in and out of ansible/ansible-base) could end up breaking things and making it complicated for end users
[2020-10-15 11:30:03] [baptiste mille-mathias] I'd say I prefer to have a smaller (as a user) ansible where I know all module are properly supported and where additional functionalities are in collections but people may disagree
[2020-10-15 11:30:28] [Sorin Sbarnea] true: there is a lot of extra cost of maintain each new repository. split makes sense when there are many maintainers (>3) and work is very active.
[2020-10-15 11:31:30] [baptiste mille-mathias] can we see what project like puppet, which has the same kind of split features, did ?
[2020-10-15 11:31:31] [Tadej tadeboro Borovšak] For people who prefer minimal setups, ansible-base + selected collections is the way to go. Ansible is a "kitchen sink included"-kind of package that people comming form 2.9 expect.
[2020-10-15 11:31:32] [Adam Morris] Sorin: But does that mean that when a collection becomes Stable because it actually works well that it should be rolled back into community.general?
[2020-10-15 11:31:36] [Mitchell Whalen] part of the issue is the limitation of the 2 point namspace "community.general". Expanding the namespace to "community.general.databases" would give a much more manageable stucture IMO
[2020-10-15 11:32:21] [James Glynn] yes that helped, thank you
[2020-10-15 11:32:26] [Sorin Sbarnea] based on my experience with splitting stuff... some will thrive, must most will become abandonware.
[2020-10-15 11:32:32] [Gary Dhadwal] I am traditional unix  and new ansible #
[2020-10-15 11:33:23] [Alicia Cozine] we could have a "community.databases" collection, but we already have "community.mysql", so that set the pattern for specific database-related collections
[2020-10-15 11:33:37] [Adam Morris] Mitchell: But I need community.general.databases.mariasql.V5.perl_api.
[2020-10-15 11:33:37] [Gary Dhadwal] Need to understand whole process around ansible
[2020-10-15 11:33:52] [Alicia Cozine] Sorin ++ for "abadonware"
[2020-10-15 11:33:57] [James Gill] Cool Then I'm a "no" on the last poll
[2020-10-15 11:34:22] [Kim Covil] Not sure if this is a stupid concept... but is there a possibility of collection aliases that get deprecated when collections get renamed/moved?
[2020-10-15 11:34:33] [Sorin Sbarnea] CPAN ~= pypi for perl.
[2020-10-15 11:34:45] [Mitchell Whalen] community.databases-general could encompass those not large enough to be on their own.
[2020-10-15 11:35:09] [Brian Coca] Kim that already works, you can leave the old collection just having a meta/runtime.yml that redirects all content
[2020-10-15 11:35:32] [Kim Covil] ok thanks Brian
[2020-10-15 11:35:41] [Brian Coca] https://github.com/bcoca/db.mysql example of 'alias collection'
[2020-10-15 11:36:18] [Jeff Geerling] smaller collections++
[2020-10-15 11:36:18] [Todd Lewis] Is there a facility for the equivalent of an http 302? I would allow to get the moved resources but throw warnings...
[2020-10-15 11:36:28] [Gundalow ] Please add thoughts into https://etherpad.opendev.org/p/ansible-contributor-summit-october-2020-morning
[2020-10-15 11:37:36] [Anthony Johnson] An example of confusion I would like this to address is gluster_heal_info being present in both community.general and gluster.gluster. By pulling collections out, will these duplicates be consolidated or overwritten?
[2020-10-15 11:37:46] [Gundalow ] Strawman Proposal 2: move things out of c.g after the last 1.x.0 minor release (see release schedule: end of November 2021)
[2020-10-15 11:37:50] [Adam Morris] Mitchell: But at what point does a thing become "large enough"?
[2020-10-15 11:37:51] [Sandra McCann] strawman 2 - would the older modules be marked as deprecated in 1.x c.g ?
[2020-10-15 11:38:36] [scott boss] have to step away for a meeting.
[2020-10-15 11:38:43] [Carol Chen] Thanks for joining Scott!
[2020-10-15 11:39:17] [Gundalow ] @Scott: Thanks for your time
[2020-10-15 11:39:49] [Adam Morris] Mitchell: If the intention when started is that community.widget is going to be pretty big btut it turns out it only needs a couple of pieces then should it have been split off?
[2020-10-15 11:40:13] [Gundalow ] afk: Getting a drink
[2020-10-15 11:40:18] [Mitchell Whalen] Adam: That would require research. What gave MySQL their own? There would have to be some justification such as a new release that adds a huge number of features. or possiblly increased growth in use
[2020-10-15 11:40:29] [Sorin Sbarnea] i would personally allow users to define their implicit lookups, so they can switch from one to another with changing a single place.
[2020-10-15 11:40:34] [Brian Coca] well, that only happens if 'badly' moving the collection
[2020-10-15 11:40:41] [Brian Coca] ^ what toshio said
[2020-10-15 11:42:12] [Adam Morris] Mitchell: I don't have any real answers, but I think there does need to be a balance between size of individual collections and total number of collections.
[2020-10-15 11:43:02] [Brian Coca] well, 'total number' is not something we'll control in the end, as for size .. that will be up to authors
[2020-10-15 11:43:09] [Todd Lewis] Maybe "community.general" should be thought of as a meta collection, a short-cut way of saying "this list of collections that keeps changing".
[2020-10-15 11:43:27] [Brian Coca] we can control collections included and make rules for inclusion in ansible, but in general they are '3rd party' at thsi point
[2020-10-15 11:43:47] [Mitchell Whalen] Adam:  When new someone adds a new Tech, normally not all the features would be available. There for it should start in a more general community.  then again if size and use grows, splitting off would be a more viable option. But like stated, backwards compatibility would be an issue.
[2020-10-15 11:45:02] [Todd Lewis] As long as there are clear warnings that say how things should be expressed, uses can update their playbooks accordingly.
[2020-10-15 11:45:23] [Adam Morris] Mitchell: Splitting them out later is going to cause further issues down the road (In my opinion)
[2020-10-15 11:45:24] [Anthony Johnson] @Todd +1
[2020-10-15 11:45:30] [Alicia Cozine] If folks use Fully Qualified Collection Names (FQCN) and upgrade to a version where modules have moved, would running their playbooks in check mode give them the right warnings to update them?
[2020-10-15 11:45:41] [Todd Lewis] There should be no "magic" w/o explicit messaging how to remove the magic.
[2020-10-15 11:46:17] [Brian Coca] or moved mulitple times i.e msql.mysql, db.mysql, db.maria ....
[2020-10-15 11:46:30] [Mitchell Whalen] Adam: it could, unless like mention in the chat changing community.general as a metadata place holder to download the other
[2020-10-15 11:46:44] [Mitchell Whalen] the largest issue is oinformin the users of th change tbh.
[2020-10-15 11:47:02] [Jeff Geerling] I can't speak, but yeah the idea would be (IMO) once something moves out of c.g, it never goes back in
[2020-10-15 11:47:04] [Todd Lewis] "never" is a nice idea, until it isn't.
[2020-10-15 11:47:30] [David Moreau-Simard] never say never -- see: ansible-modules-core/ansible-modules-extras split/merge :)
[2020-10-15 11:47:43] [Toshio Kuratomi] dmsimard, very true :-/
[2020-10-15 11:47:44] [Adam Morris] Would net new be allowed to go into community.general?
[2020-10-15 11:47:56] [Jeff Geerling] @Adam Morris I sure hope not
[2020-10-15 11:48:55] [Adam Morris] Jeff: Me too, but that changes things with regards to the discussion mitchell and I have been having.
[2020-10-15 11:49:29] [Toshio Kuratomi] adam morris, I believe so.  Because not everyone is going to want to manage the whole work of creating releases, etc for a single module, plugin ,etc.
[2020-10-15 11:49:33] [Jeff Geerling] Basically the more we treat community.general as a product in and of itself, the more we (IMO) fail to meet the main goal of moving to collections
[2020-10-15 11:50:03] [Todd Lewis] +1 Jeff
[2020-10-15 11:50:04] [Toshio Kuratomi] If community.general doesn't allow that, there will probably be a different "community.general"-like collection that aggregates things down the road.
[2020-10-15 11:50:07] [Jeff Geerling] which was, "we don't have this massive horrible unmaintainable thing that everyone relies on and also hates because its hard to get changes into it" (which was ansible/ansible, now is community.general)
[2020-10-15 11:50:28] [Adam Morris] Is community.network any less arbitrary than community.databases?
[2020-10-15 11:50:34] [Jeff Geerling] sorry I can't talk, my mic still doesn't work :(
[2020-10-15 11:51:12] [Felix Fontein] @Adam: `network` was just the largest group of modules :)
[2020-10-15 11:51:35] [Felix Fontein] (and they are a bit special)
[2020-10-15 11:51:55] [Sorin Sbarnea] multiple moves can and will happen but i doubt they will be common
[2020-10-15 11:52:13] [baptiste mille-mathias] Can't Galaxy help us manage this module lifecycle and moving  problem ?
[2020-10-15 11:52:24] [Adam Morris] Felix: Everyone thinks that they are special :)
[2020-10-15 11:52:33] [Brian Scholer] BlueJeans question: if I have to drop off temporarily, am I going to be able to re-join?
[2020-10-15 11:52:36] [Todd Lewis] Movement shouldn't be encouraged, but it should be well-supported when appropriate.
[2020-10-15 11:52:49] [Carol Chen] @Brian: yes, just use the same link you joined with earlier
[2020-10-15 11:52:57] [Brian Scholer] ty!
[2020-10-15 11:53:23] [Felix Fontein] Adam: true ;) there's all that infrastructure specially for network modules though, like terminal plugins, netconf plugins, cliconf plugins, ...
[2020-10-15 11:53:54] [baptiste mille-mathias] toshio +1
[2020-10-15 11:54:27] [David Moreau-Simard] redirects are nice but what if it redirects to a collection that isn't installed by default as community.general might be ?
[2020-10-15 11:55:04] [Adam Morris] I think that there is currently a lot of confusion about how collections are going to be managed.  Given that there are at least two different opinions on net new going into community.general?
[2020-10-15 11:55:04] [Mitchell Whalen] Adam: Breaking it down into smaller sections will be more manage for commits and for those places that don't want/ or feel comfortable to manage everything with ansible
[2020-10-15 11:55:59] [Sandra McCann] +1 for  a policy that moves out of c.g etc to what should be a permanent home. (aka no c.databases)
[2020-10-15 11:56:52] [Brian Coca] keep 1 dump, not multiple smal dumps
[2020-10-15 12:01:16] [Adam Morris] What about community adding support for an open source product?  If I wanted to add support for "widget" would that go in a separate widget collection or community.widget or community.general?
[2020-10-15 12:01:31] [Adam Morris] (i.e. there is no company involved)
[2020-10-15 12:01:44] [Jeff Geerling] @Adam Morris I would say community.widget (or adam_morris.widget)
[2020-10-15 12:01:54] [Jeff Geerling] (or open_source_project.widget)
[2020-10-15 12:02:03] [Adrian Teri] The challenge of 1 dump versus many ...but what if not all things are junk in the dump? In my mind the dead weight will have to be dropped at some point
[2020-10-15 12:02:38] [Adam Morris] Adrian: A fuel dump contains fuel, just because it's in a dump doesn't make it junk.
[2020-10-15 12:03:36] [Amin Vakil] we can't hear you carol
[2020-10-15 12:03:45] [Carol Chen] yeah it seems like it
[2020-10-15 12:04:11] [Adam Morris] Jeff: I agree with you on this but I'm curious what others think.  how much should go into community moving forward?
[2020-10-15 12:04:39] [Adam Morris] But shouldn't it be postgres.postgres... rather than community.postgres?
[2020-10-15 12:05:32] [Alicia Cozine] +1 for presenting the choices in writing
[2020-10-15 12:05:45] [Alicia Cozine] maybe in the next issue of the Bullhorn?
[2020-10-15 12:06:03] [Adam Morris] I think we need some clear proposals...
[2020-10-15 12:06:43] [Alicia Cozine] Adam Morris: the namespace for the collection reflects the maintainers - and the community maintains the postgres modules
[2020-10-15 12:06:44] [Adam Morris] Including guidelines for whether things should move into community.widget versus widget.whatever
[2020-10-15 12:07:51] [Adam Morris] Alicia: How many maintainers are needed before it becomes community though?
[2020-10-15 12:08:12] [Sorin Sbarnea] openstack makes decisions using the review process using specific repos, could be an option.
[2020-10-15 12:09:26] [Robbie Fontenot] As a end user, I would not expect going from 2.9->2.10, 2.11, etc. to require rewriting playbooks.  Going from 2.x -> 3.x I would expect to have to make changes.
[2020-10-15 12:09:37] [James Glynn] can you provide the links for those choices - newbie, i don't know all the links yet :-)
[2020-10-15 12:09:54] [Adrian Teri] has break started? How long?
[2020-10-15 12:10:01] [Adam Morris] James: You weren't born with the innate knowledge?
[2020-10-15 12:10:13] [Adrian Teri] Oh I see it 15 mins
[2020-10-15 12:10:25] [Alicia Cozine] Adam: using "community" as the maintainer is the default - non-community-maintained modules are either maintained by staff or by a company
[2020-10-15 12:10:37] [Alicia Cozine] Carol: you're still muted
[2020-10-15 12:10:52] [David Blackwell] @Alicia the problem with that is anyone can setup a GitHub and add their own collections.
[2020-10-15 12:11:11] [James Glynn] @Adam LOL nope. I have been using Ansible for maybe a year, but the Monday session was my first introduction to "contributors"
[2020-10-15 12:11:34] [David Blackwell] makes it look like community. are the "open" modules and <uniquename>.collection are special.  Even going so far as making the line between supported and not more blurry
[2020-10-15 12:12:11] [Adam Morris] James: No big deal, I contributed once many years ago now and have just been trying to get active again this week.
[2020-10-15 12:12:13] [Carol Chen] yeah my audio is messed up. just now mic works but not speakers. now speakers work but not the mic.
[2020-10-15 12:12:28] [baptiste mille-mathias] I hear someone mic though
[2020-10-15 12:12:35] [Alicia Cozine] David: yes, anyone can create a collection, but namespaces on Galaxy must follow certain rules and adding collections to existing namespaces is also restricted
[2020-10-15 12:13:14] [James Glynn] @Adam my biggest problem right now is trying to remember what I knew about IRC from the early 90s hahahaha
[2020-10-15 12:19:37] [Toshio Kuratomi] Aye, you are loud and clear now :-)
[2020-10-15 12:19:39] [baptiste mille-mathias] yes
[2020-10-15 12:19:49] [Mark Chappell] yes
[2020-10-15 12:20:36] [Carol Chen] thank you!
[2020-10-15 12:20:56] [Cruz Pitre] \o/
[2020-10-15 12:20:59] [Toshio Kuratomi] jillr, do you want to test your mic as well?
[2020-10-15 12:21:18] [Carol Chen] good idea to do mic test
[2020-10-15 12:21:20] [Carol Chen] hi Cruz!
[2020-10-15 12:21:35] [Cruz Pitre] Hello!
[2020-10-15 12:22:01] [Carol Chen] Cruz: too bad we won't see your lovely map this time :)
[2020-10-15 12:25:05] [Cruz Pitre] I have not been able to log into the BlueJean app.   Keeps asking for an account password.
[2020-10-15 12:25:54] [Cruz Pitre] I guess the provided link is not meant for the blue jean app.
[2020-10-15 12:27:37] [Jill Rouleau] https://github.com/ansible/community/wiki/Diversity/FAQs
[2020-10-15 12:30:27] [Jill Rouleau] #ansible-diversity
[2020-10-15 12:30:42] [Gundalow ] REMINDER: We are continually updating https://etherpad.opendev.org/p/ansible-contributor-summit-october-2020-morning with a summary
[2020-10-15 12:31:22] [Toshio Kuratomi] You can use timanddate.com to try narrowing things down ahead of time.
[2020-10-15 12:31:39] [Toshio Kuratomi] https://www.timeanddate.com/worldclock/meeting.html
[2020-10-15 12:32:53] [Carol Chen] thanks Toshio
[2020-10-15 12:33:47] [Brian Scholer] I'm a native English speaker and prefer subtitles for everything as well, on TV and movies, presentations, etc.
[2020-10-15 12:34:09] [James Gill] Agreed doesn't seem to be a down side to having them.
[2020-10-15 12:34:11] [baptiste mille-mathias] closed captions can be helpful for foreigner, generally it helps me because I've problem hearing. Let's try them
[2020-10-15 12:35:04] [baptiste mille-mathias] what's D&I ?
[2020-10-15 12:35:08] [Todd Lewis] D&I? (Sorry, but stepped back in from lunch, and have no context.)
[2020-10-15 12:35:12] [baptiste mille-mathias] ah
[2020-10-15 12:35:13] [James Glynn] yes, in the poll i answered "had a conflict", but i already scheduled time tomorrow to go through the sessions :-)
[2020-10-15 12:35:16] [Sandra McCann] Diversity and Inclusion
[2020-10-15 12:35:26] [Michael ​] yep there is
[2020-10-15 12:35:31] [Adam Morris] I think the Poll question is incorrect.
[2020-10-15 12:35:32] [Sandra McCann] making ansible community a welcoming place to everyone interested
[2020-10-15 12:35:36] [Michael ​] you go to the bottom right, there is a option
[2020-10-15 12:35:40] [Brian Scholer] ah there is, I just turned on CC
[2020-10-15 12:35:42] [James Cassell] I also use Closed Captions all the time.  One of the nice features of Google Meet, as much as I try to avoid Google
[2020-10-15 12:35:43] [James Gill] Lower left hand corner and it looks like it works pretty well
[2020-10-15 12:35:53] [baptiste mille-mathias] no problem, thanks !
[2020-10-15 12:35:58] [Adam Morris] "Did anybody go to the D&I sessions" Yes, but I didn't.
[2020-10-15 12:36:13] [Michael ​] it is sometime incorrect (see my carbonated openshift quote on irc), and a tad laggy sometime
[2020-10-15 12:36:20] [James Gill] slight lag on captions.
[2020-10-15 12:36:45] [Toshio Kuratomi] August 13 and Sep 22
[2020-10-15 12:37:31] [Michael ​] sometime it also give the caption in advance
[2020-10-15 12:37:53] [Carol Chen] there must be minions typing them and some are from the future
[2020-10-15 12:38:19] [Alicia Cozine] For anyone who's confused by the "version-independent" idea, we publish version-specific documentation like https://docs.ansible.com/ansible/2.9/
[2020-10-15 12:38:44] [Alicia Cozine] and we mark the most recent version as "latest":  https://docs.ansible.com/ansible/latest/
[2020-10-15 12:38:50] [Alicia Cozine] currently 2.10 is "latest"
[2020-10-15 12:39:51] [James Cassell] What changes are in store for ansible-base 2.11?
[2020-10-15 12:40:05] [Toshio Kuratomi] Yep, exactly right
[2020-10-15 12:40:06] [sivel ] https://github.com/ansible/ansible/pull/72222
[2020-10-15 12:40:13] [sivel ] That is the PR for the updated roadmap ^
[2020-10-15 12:40:20] [sivel ] It includes the schedule and planned work
[2020-10-15 12:43:26] [James Cassell] I think the old dev guides should remain available, even if not the default google search result
[2020-10-15 12:43:52] [sivel ] Starting at ~17:15 we'll be talking about some of those 2.11 planned features
[2020-10-15 12:45:49] [Todd Lewis] By "persona", do you mean target audience?
[2020-10-15 12:46:40] [Sandra McCann] UX folks also use 'personas' but yes, it's identifying types of users of the documentation and what do they need that might be different from other users
[2020-10-15 12:47:09] [Jeff Geerling] I think my microphone might finally be working, had to restart everything :P
[2020-10-15 12:47:40] [Brian Coca] persona is easier than 'target audience grouped by interests and usage'
[2020-10-15 12:49:34] [Natalie Wong] UX perspective on a persona: "... personas are
composite archetypes based on behavior patterns uncovered during the course of our
research, which we formalize for the purpose of informing the product design" - p. 62, "About Face" by Alan Cooper, Rober Reimann, David Cronin, Chris Noessel
[2020-10-15 12:50:05] [Gundalow ] Documentation Working Group Agends & Meeting invite https://github.com/ansible/community/issues/521
[2020-10-15 12:50:48] [Felix Fontein] @Natalie thanks for the definition!
[2020-10-15 12:50:56] [Brian Scholer] dcos WG is great! I regret I haven't attended recently due to being really busy at work but I remember realy liking the meetings
[2020-10-15 12:52:01] [Brian Coca] i can spew a lot of words too, but you do it coherently
[2020-10-15 12:52:58] [Mark Chappell] Docs wg have always been friendly and responsive when I've interacted with them
[2020-10-15 12:53:41] [Carol Chen] Jeff: I used an external mic today and it gave me problems. everything was fine with just the laptop mic on Monday :P
[2020-10-15 12:55:16] [Gundalow ] Any questions for Content Team?
[2020-10-15 12:58:00] [James Cassell] Why does networknig need something other than regular ssh?
[2020-10-15 12:58:09] [Kim Covil] Are resource modules just for networking?
[2020-10-15 12:59:33] [Brian Coca] James, mostly cause networking uses a persistent session (not just auth) which ssh control perisst does not allow for
[2020-10-15 13:00:04] [Brian Coca] also other issues with openssh being slightly incompatible with older network appliance's proprietary ssh servers
[2020-10-15 13:00:11] [Brian Coca] but that is less of an issue now
[2020-10-15 13:00:59] [Gundalow ] @James Cassell Most of the network modules work via raw SSH (ie no Python on the network device). Some network devices also have a HTTP interface, ie Arista EOS has an eAPI over HTTP interface
[2020-10-15 13:01:43] [Gundalow ] See the table at the end of https://docs.ansible.com/ansible/latest/network/user_guide/platform_index.html for which platforms support
[2020-10-15 13:01:49] [Felix Fontein] what exactly IS a resource module? :)
[2020-10-15 13:02:16] [Brian Coca] basicaly reentrant facts
[2020-10-15 13:02:35] [Sandra McCann] @felix - https://docs.ansible.com/ansible/latest/network/user_guide/network_resource_modules.html  :-)
[2020-10-15 13:02:37] [Felix Fontein] sounds like I already implemented some without knowing the name :)
[2020-10-15 13:03:01] [Jose Jorge Atuncar Chong] excellent information
[2020-10-15 13:03:04] [Joshua Robinett] There a link to the index filter?
[2020-10-15 13:03:22] [Kim Covil] great... I can see that being useful to update the ipa modules for example
[2020-10-15 13:03:34] [Felix Fontein] Brian, Sandra: thanks!
[2020-10-15 13:03:48] [Tadej tadeboro Borovšak] We mostly build such modules, but we often modify the public API of the module because upstream API are not always the nicest thing to use.
[2020-10-15 13:04:41] [Gundalow ] If you are interested in using Ansible with Networking join the Networking Breakout at 2000UTC https://etherpad.opendev.org/p/ansible-contributor-summit-october-2020-network
[2020-10-15 13:04:57] [Bradley Thornton] index_of: https://github.com/ansible-collections/ansible.utils
[2020-10-15 13:05:18] [Joshua Robinett] Thanks!
[2020-10-15 13:06:21] [Anthony Johnson] If it commits, it ships!
[2020-10-15 13:06:40] [Kim Covil] how do you see cumulus linux and vyos fitting in with the network stuff
[2020-10-15 13:06:53] [Kim Covil] where they are both linux boxes and network devices
[2020-10-15 13:07:08] [Kim Covil] but I can ask that all again in the networking breakout later :)
[2020-10-15 13:07:17] [Bradley Thornton] Resource modules are developed for vyos at the same pace as the other platform
[2020-10-15 13:07:41] [Bradley Thornton] Yes, let's talk more in the network breakout
[2020-10-15 13:07:48] [Sorin Sbarnea] my OCD: list should be alphabetically sorted!
[2020-10-15 13:08:17] [Bradley Thornton] Feel free to add any network topics here we should talk about https://etherpad.opendev.org/p/ansible-contributor-summit-october-2020-network
[2020-10-15 13:08:19] [Gundalow ] @Kim: Could you please add a note in https://etherpad.opendev.org/p/ansible-contributor-summit-october-2020-network it's a good question and I wouldn't want to get it lost
[2020-10-15 13:09:08] [Jeff Geerling] We are actually testing openshift stuff on an openshift cluster now!
[2020-10-15 13:09:23] [Jeff Geerling] (we were testing openshift stuff on a super old janky openshift 3 test container before)
[2020-10-15 13:10:01] [Patrick Toal] :thumbsup:
[2020-10-15 13:12:07] [James Cassell] wasn't the old mantra ,"don't be just a wrapper around the API"?
[2020-10-15 13:12:16] [Ken Dreyer] James :)
[2020-10-15 13:14:43] [Gundalow ] Reminder, Cloud breakout is at 2000 UTC: https://etherpad.opendev.org/p/ansible-contributor-summit-october-2020-cloud
[2020-10-15 13:15:02] [Tadej tadeboro Borovšak] I hate auto-generated Ansible content because it is really awkward to use most of the time.
[2020-10-15 13:16:34] [Jill Rouleau] There are trade offs aith generated content to be sure, but in this case we felt confident we could do something that would work and be maintainable. Time always tells of course. :)
[2020-10-15 13:16:47] [Jill Rouleau] If you're using VMware we'd love to get feedback on the new collection
[2020-10-15 13:17:28] [Tadej tadeboro Borovšak] I am using Google and forti stuff, and that content is ... not the best to put it nicely ;)
[2020-10-15 13:17:31] [Ulf Tigerstedt] Heavy vmware/vsphere user here, I'll look into the new rest modules.
[2020-10-15 13:17:48] [Ken Dreyer] I'm watching this experiment with interest because I have a couple Ansible modules that use a complicated REST API with many parameters. I'm more interested in automating the test cases than I am in automating the module code generation.
[2020-10-15 13:18:31] [Jill Rouleau] The generator is here https://github.com/ansible-collections/vmware_rest_code_generator
[2020-10-15 13:18:36] [Brian Coca] like fixing vault issue and breaking callbacks...
[2020-10-15 13:19:01] [Matt Davis] *cough* import/include *cough*
[2020-10-15 13:19:12] [sivel ] I broke the crap out of that!
[2020-10-15 13:19:18] [Matt Davis] lol
[2020-10-15 13:19:24] [Felix Fontein] it's not a feature, it's a bug!
[2020-10-15 13:19:26] [Brian Coca] well, but that covers all categories, intended and unintended
[2020-10-15 13:19:28] [Matt Davis] Well, you *fixed* the crap out of it
[2020-10-15 13:19:33] [Todd Lewis] +1 import / include is a hot mess.
[2020-10-15 13:19:36] [Sorin Sbarnea] i am really curious about minimal version of python.... dropping old python ok but jumping to py38 instead of py36?
[2020-10-15 13:19:53] [sivel ] @Todd it *was* a hot mess, it's now very predictable
[2020-10-15 13:20:37] [Todd Lewis] predictable, yes, but understandable, no. The final answer is always "test and see what the surprises are."
[2020-10-15 13:20:39] [sivel ] @Sorin we'll get there, but yes. Part of that is by the time that 2.12 releases, Py 3.6 will be EOL, and we desire features in Python 3.7
[2020-10-15 13:21:56] [Tadej tadeboro Borovšak] So running Ansible on RHEL will not be supported? IIRC; RHEL 8 ships 3.6
[2020-10-15 13:22:03] [sivel ] @Todd that would indicate a lack of documentation or some issue in describing it. The 2.10 documentation worked to clarify that much more
[2020-10-15 13:22:09] [Matt Davis] Appstream for 3.8
[2020-10-15 13:22:16] [sivel ] Ultimately there are a few simple rules that explain everything
[2020-10-15 13:22:17] [James Cassell] Isn't 3.6 going to be supported for the life of RHEL 8?
[2020-10-15 13:22:28] [Ken Dreyer] It is
[2020-10-15 13:22:29] [sivel ] See above, about appstream
[2020-10-15 13:22:42] [Brian Coca] Todd i would argue that include: behaving widly different on seemingly unrelated changes to the task is not 'understandable'
[2020-10-15 13:22:42] [Sorin Sbarnea] yes, it is py36 is the new py27, the python that never dies.
[2020-10-15 13:22:48] [Matt Davis] Ansible's not on RHEL's support cycle, so appstreams are totally workable for us
[2020-10-15 13:23:06] [James Cassell] I guess RHEL 9 probably will have 3.10 (or 3.9)
[2020-10-15 13:23:25] [Felix Fontein] for the remote end, py 2.6/7 will still be supported for some time
[2020-10-15 13:23:26] [Ken Dreyer] Matt, I guess that works until Ansible PMs want an LTS :)
[2020-10-15 13:23:40] [sivel ] @Ken that is happening too, but for now 2.9 is that thing
[2020-10-15 13:23:59] [sivel ] but the "S" is important, it's just "support", not features or otherwise
[2020-10-15 13:24:01] [baptiste mille-mathias] @relrod are you confident enough people will test  rc releases ?
[2020-10-15 13:24:15] [Sorin Sbarnea] IMHO that the leap is too big, and I am an early adopter.
[2020-10-15 13:24:19] [Todd Lewis] It's not a doc problem. It's a problem of the quirks of the implementation language percolating into the app layer. Put it this way: if you re-wrote Ansible in C, the only path to the current behavior would be to first implement python.
[2020-10-15 13:24:20] [Matt Davis] If not, we've given them the option
[2020-10-15 13:24:24] [sivel ] @baptise that is to be determined
[2020-10-15 13:24:32] [Ken Dreyer] Agreed Sorin, I think 3.6 makes a lot more sense
[2020-10-15 13:24:55] [Sorin Sbarnea] i am aware of bugs with py38 which are still there which prevented zuul itself from running stable with it.
[2020-10-15 13:25:25] [Sorin Sbarnea] py36 is sweet, you can spent the next year introducing type hitns to ansible codebase.
[2020-10-15 13:25:32] [sivel ] Todd, imagine them from the perspective of macros vs dynamic functions. Imports are pre-processors that happen during parsing, whereas includes are runtime triggers
[2020-10-15 13:26:03] [James Glynn] anyone else lose video?
[2020-10-15 13:26:07] [sivel ] Nope
[2020-10-15 13:26:08] [Alicia Cozine] @Todd feedback would be great on https://docs.ansible.com/ansible/latest/user_guide/playbooks_reuse.html
[2020-10-15 13:26:20] [Brian Coca] the previous include worked both like macro and function w/o warning
[2020-10-15 13:26:40] [Sorin Sbarnea] what i do not like is that ansible is not longer installable as editable, which works against running bleading edge, something I used to do for development.
[2020-10-15 13:27:42] [Brian Coca] it still is, just depends on 'which part' you want to be editable
[2020-10-15 13:27:58] [Brian Coca] core/base works as before, for collections, you need to do a bit more work
[2020-10-15 13:28:03] [sivel ] editable installs have never been something we've indicated was supported or recommended, as our recommendation is the "running from source" or just installing. But it's possible we could look into it
[2020-10-15 13:28:17] [Brian Scholer] also `ansible-chaos` commands could play an ominous tone on each run  👍
[2020-10-15 13:28:35] [Felix Fontein] /me is a happy user of ansible-base editable installs :)
[2020-10-15 13:28:50] [Gundalow ] QUESTION: QUESTION: How can we get more people to test ansible & ansible-base RCs?
[2020-10-15 13:28:54] [Alicia Cozine] Do we need a special cowsay cow for ansible-chaos?
[2020-10-15 13:29:13] [Brian Scholer] 😆🐄
[2020-10-15 13:29:16] [Adam Morris] A cowsay chao?
[2020-10-15 13:29:26] [Sorin Sbarnea] while editable installs from git are clearly not for mainstream, they are very useful for keeping developers in sync with main branch changes, they will find bugs that CI does not.
[2020-10-15 13:29:31] [Alicia Cozine] @Felix: I mis-read that as "edible installs" and got very confused
[2020-10-15 13:29:41] [Sorin Sbarnea] is like first line of human defense
[2020-10-15 13:29:43] [Pierre-Louis Bonicoli] @bcoca: could you detail the bit more work ?
[2020-10-15 13:29:52] [Alicia Cozine] @Adam Morris: +1 for cowsay chao
[2020-10-15 13:29:54] [Ken Dreyer] I think we should document "here's how to run your tests with the RC"
[2020-10-15 13:30:20] [baptiste mille-mathias] well I feel it's a bit harsh to answer to people you should have test the RC
[2020-10-15 13:30:20] [Ton Kersten] Ken: +1
[2020-10-15 13:30:21] [Ken Dreyer] Like give a couple of examples of projects that have integrated RC testing
[2020-10-15 13:30:27] [Alicia Cozine] @Ken: good idea - would you be willing to open an issue so we don't forget?
[2020-10-15 13:30:31] [baptiste mille-mathias] ... as a user
[2020-10-15 13:30:33] [Ken Dreyer] I put a link to https://github.com/ktdreyer/koji-ansible/pull/173 in the Etherpad
[2020-10-15 13:30:52] [Ken Dreyer] @Alicia, sure, where should I open this?
[2020-10-15 13:31:02] [Alicia Cozine] on ansible/ansible, please @Ken
[2020-10-15 13:31:20] [Alicia Cozine] https://github.com/ansible/ansible/issues
[2020-10-15 13:31:31] [Jeff Geerling] I believe we have been discussing this since 2015 (role argspec / validation) :) — I don't even care at this point *how* it happens, I just want it *to* happen
[2020-10-15 13:31:46] [Jeff Geerling] like if the argspec has to be written in Ruby, so be it :P
[2020-10-15 13:31:50] [Toshio Kuratomi] Jeff you say that now.... ;-)
[2020-10-15 13:31:54] [Matt Davis] Yeah, it's been so close to done like 3 times now, but it's definitely happening this time
[2020-10-15 13:31:56] [Alicia Cozine] with "documentation" in the title it should automatically get tagged as docs and we'll see it
[2020-10-15 13:32:05] [xavier  renaut] role validation : awesome idea  to gain time !
[2020-10-15 13:32:26] [Jeff Geerling] And my other requirement though is a role is not required to have an argspec (so all existing roles can just work and don't need to be updated until they have the time to be updated)
[2020-10-15 13:32:46] [sivel ] @Jeff yep, you'll be good there
[2020-10-15 13:32:53] [Jeff Geerling] woop woop
[2020-10-15 13:33:57] [Joshua Robinett] Would it be possible to manually validate ALL roles before running any roles?
[2020-10-15 13:33:57] [Matt Davis] @Jeff the one caveat to that is probably around Tower direct role execution- I think they'll only support running role entrypoints from the Tower UI that are "listed"
[2020-10-15 13:34:21] [sivel ] And by "listed" we are extending ansible-doc to list them
[2020-10-15 13:34:22] [Matt Davis] (because we can't tell the difference between a task file and a role entrypoint)
[2020-10-15 13:34:49] [Jeff Geerling] @Matt I didn't even know there was a way to run a role directly in Tower
[2020-10-15 13:35:05] [sivel ] @Jeff not yet, but tower 4.0 maybe?
[2020-10-15 13:35:10] [Tim A] @Geerliong There isn't yet.
[2020-10-15 13:35:13] [Matt Davis] They'll be adding it once we have it- runner can already do it though
[2020-10-15 13:35:33] [Tim A] That wasn't announced yet. (ahem)
[2020-10-15 13:35:34] [Matt Davis] (it generates a playbook, was before adhoc supported include_role et al)
[2020-10-15 13:36:02] [Matt Davis] https://media.tenor.com/images/9f43bb4f0c40e4730d415a7220991937/tenor.png
[2020-10-15 13:36:06] [Dylan Silva] bluejeans ftw
[2020-10-15 13:36:09] [Patrick Toal] Must be on a Mac.
[2020-10-15 13:36:12] [Brian Coca] screenshot or it ididnt happen
[2020-10-15 13:36:20] [James Gill] :)
[2020-10-15 13:36:35] [James Glynn] i feel like the kid in jurrasic park ... "he left us"
[2020-10-15 13:36:47] [Sloane Hertel] there is a WIP for docs https://github.com/ansible/ansible/pull/72120
[2020-10-15 13:36:49] [Brian Coca] there was proposal + long discussion on ansible/propsals
[2020-10-15 13:36:50] [Tim A] One of the biggest hold up has been the lack of what Sam. is talking about.
[2020-10-15 13:37:09] [Sloane Hertel] I also have a very rough draft branch, that's quite incomplete https://github.com/ansible/ansible/compare/devel...s-hertel:role_validation
[2020-10-15 13:37:12] [Tim A] I'm painful aware bcoca.
[2020-10-15 13:37:39] [Jeff Geerling] yay we can see now!
[2020-10-15 13:37:40] [Patrick Toal] larger text?
[2020-10-15 13:38:22] [Gundalow ] @Patrick: Is it readable now?
[2020-10-15 13:38:40] [Patrick Toal] Yes.  much better, tnx.
[2020-10-15 13:38:46] [xavier  renaut] @gundalow awesome readibility
[2020-10-15 13:38:53] [Jeff Geerling] Some of this seems to be duplication of data inside meta/main.yml, no? (Description, short description)
[2020-10-15 13:38:59] [sivel ] This is kind of like DOCUMENTATION in a module, but used as the arg spec.  A little bit of a mashup
[2020-10-15 13:39:09] [Brian Coca] like plugins do
[2020-10-15 13:39:13] [sivel ] @Jeff there may be multiple argspecs, for multiple entrypoints into a role
[2020-10-15 13:39:40] [sivel ] such as for different `tasks_from` for example
[2020-10-15 13:39:45] [Jeff Geerling] @sivel :-/
[2020-10-15 13:40:04] [Jeff Geerling] Can we get module argspecs in yaml too? :P
[2020-10-15 13:40:14] [David Shrewsbury] It very closely follows module spec, except some changes for the entry points. This is because it will share validation codes.
[2020-10-15 13:40:18] [sivel ] @Jeff that is somewhat planned for the future
[2020-10-15 13:40:28] [sivel ] We'll see where it leads us
[2020-10-15 13:40:40] [Jeff Geerling] I might've missed it, but can you add a 'description' to each parameter?
[2020-10-15 13:40:45] [Sloane Hertel] yes
[2020-10-15 13:40:47] [David Shrewsbury] yep
[2020-10-15 13:40:56] [Jeff Geerling] okay good. I was glancing then Sam jumped back to this file :D
[2020-10-15 13:41:21] [Tim A] Wondering how this spec maps to the OpenAPI spec.
[2020-10-15 13:41:27] [Patrick Toal] I find this format is tough to follow sometimes. Jumping between chat, screen, IRC, etc.
[2020-10-15 13:41:36] [Jeff Geerling] Please make sure it works with the top-level roles: keyword too — I know some people think that's gooky, but that's how I at least build like 98% of my playbooks.
[2020-10-15 13:41:47] [David Blackwell] how much time overhead will this add?
[2020-10-15 13:42:05] [Toshio Kuratomi] Is that toggle going to be via a comment or will it be part of the yaml?
[2020-10-15 13:42:05] [Brian Scholer] yay! +1 for pluggable
[2020-10-15 13:42:09] [Joshua Robinett] Can we have a "check" mode where we only validate the role?
[2020-10-15 13:42:11] [Ton Kersten] Is there a way to connect with an iPad and have sound?
[2020-10-15 13:42:11] [James Cassell] custom validation sounds good
[2020-10-15 13:42:21] [V K] So we have a role standard internally that every tasks/main.yml includes another tasks/assert_vars.yml that has all validation tasks for role vars. Mostly a bunch of asserts. Each role also has a  pykwalify schema that we use to validate  inventory in git. I'm interested in the proposal as an evolution of what we're doing now. (BTW: We also use pyinvoke to provide a pluggable CLI to "run" any role.)
[2020-10-15 13:42:34] [Sloane Hertel] Jeff, the one thing about the `roles:` is that it ignores extra options. We were thinking old-style roles would use the global toggle to disable validation if needed.
[2020-10-15 13:42:39] [Todd Lewis] Can you specify patterns/regex for the var names? We do that a lot in our roles.
[2020-10-15 13:42:44] [Jeff Geerling] /me has still never used tasks_from
[2020-10-15 13:42:55] [Brian Scholer] lol same Jeff
[2020-10-15 13:43:06] [Sloane Hertel] (for example, can't use tasks_from with roles:)
[2020-10-15 13:43:07] [Gundalow ] @Patrick Toal Any suggestions on how we can give a better experience for viewers?
[2020-10-15 13:43:09] [Jeff Geerling] @Sloane - I mean with `roles:` you can have list items like { name: rolename, validate: no }
[2020-10-15 13:43:13] [Brian Scholer] I still use the main.yml import_tasks based on soe variable
[2020-10-15 13:43:41] [sivel ] @Jeff tasks_from is a great way of implementing a single role that may have different goals. install vs uninstall.  But you must use it with import_role/include_role
[2020-10-15 13:43:42] [Tim A] I have Jeff. It can really cut down on the number of includes and conditionals in a complex role.
[2020-10-15 13:43:59] [Patrick Toal] @gundalow I think in this case, there's just too much info too fast for the medium. :)
[2020-10-15 13:44:00] [Kim Covil] will there be a way of referring internally within the arg spec for the different entry points to save duplication of the same arg definitions?
[2020-10-15 13:44:16] [David Shrewsbury] @Joshua The plan is to be able to run 'include_role' via adhoc and that will run the role and do validation as well.
[2020-10-15 13:44:30] [Felix Fontein] will there be version_added for roles (and role parameters)?
[2020-10-15 13:44:37] [Joshua Robinett] It would be useful to be able to validate many roles before executing any
[2020-10-15 13:44:50] [Sloane Hertel] since it will be an action plugin, you can call it directly
[2020-10-15 13:45:04] [Brian Coca] its hard cause vars can dynamically 'appear'
[2020-10-15 13:45:14] [Jeff Geerling] version_added would definitely be helpful for roles-in-collections
[2020-10-15 13:45:17] [Joshua Robinett] yah .. that makes sense
[2020-10-15 13:45:31] [Jeff Geerling] (no use outside of that but would be helpful inside collections)
[2020-10-15 13:46:05] [Felix Fontein] @Jeff definitely! Making roles 'documentable' would also allow to properly integrate them into antsibull-docs and antsibull-changelog
[2020-10-15 13:46:38] [Jeff Geerling] Ideally someday, far far far in the future, Galaxy could show docs for plugins and collection roles too :D
[2020-10-15 13:46:47] [Tim A] There's another long standing proposal: standard documentation for roles.
[2020-10-15 13:47:20] [Jeff Geerling] @Sam - It would be really nice to have the argspec provide 'default:' and then if no defaults/main.yml is provided, it comes from spec
[2020-10-15 13:47:20] [Tim A] I may have missed this. Does this arg spec and the ansible-doc changes work for standalone roles?
[2020-10-15 13:47:34] [James Cassell] maybe have ansible-doc read it from defaults/main.yml?
[2020-10-15 13:47:35] [Sloane Hertel] Tim, yes
[2020-10-15 13:47:37] [Sorin Sbarnea] ansible-lint needs information exposed by the roles about what they can accept or not, the same way we have for modules.
[2020-10-15 13:47:38] [Brian Coca] tima yes
[2020-10-15 13:48:13] [David Shrewsbury] standalone and collection roles will be supported
[2020-10-15 13:48:16] [James Cassell] new lowest precedence level - "role spec"
[2020-10-15 13:48:42] [Sorin Sbarnea] openstack already has an informal RST way to document roles arguments, but having a standard way to define what is required/optional and format would be a huge improvement
[2020-10-15 13:48:48] [Jeff Geerling] My ideal would be "new system defaults to meta/argspec if present, then falls back to defaults/main.yml"
[2020-10-15 13:49:03] [Jeff Geerling] And that way I could just drop defaults/main.yml once I add argspecs
[2020-10-15 13:49:15] [Jeff Geerling] 2.11.1 ;)
[2020-10-15 13:49:19] [Gundalow ] /me -> afk
[2020-10-15 13:49:22] [Brian Coca] you CAN use both, but sanity would have you choose one and stick with it
[2020-10-15 13:49:31] [Jeff Geerling] vars precedence is so easy there are like 8 blog posts on Medium that all say different things about them
[2020-10-15 13:49:50] [James Cassell] often I have jinja in my defaults/main.yml -- would the role spec support jinja?
[2020-10-15 13:49:55] [Tim A] This is good. Now we need standardized docs for roles so everyone knows how to (re)use these roles.
[2020-10-15 13:50:17] [sivel ] people try to think too in depth about precedence. It was meant to be easy, and people should think less about it :)
[2020-10-15 13:50:25] [Sorin Sbarnea] ansible-lint can do this! just expose the argspec first
[2020-10-15 13:50:27] [Jeff Geerling] @James - good point. I avoid it as much as possible, but I do have a couple jinja vars in defaults
[2020-10-15 13:50:41] [James Cassell] if not, we'd need to keep support for defaults/main
[2020-10-15 13:50:50] [Todd Lewis] ...because it's still a hot mess :)
[2020-10-15 13:51:05] [David Shrewsbury] ansible-doc will output the role documentation in text format (or json for 3rd party parsing)
[2020-10-15 13:51:10] [Jeff Geerling] Either that, or force role authors to use a set_fact for that use case. But could just leave support for defaults/main.yml and anything in there overrides the base default in argspec
[2020-10-15 13:51:10] [Toshio Kuratomi] Will you want any of this docs stuff to appear in the web docs?
[2020-10-15 13:51:18] [Brian Coca] or even on include_tasks
[2020-10-15 13:51:21] [Toshio Kuratomi] Or no because it is only about roles?
[2020-10-15 13:51:21] [Jeff Geerling] @Toshio YES but on galaxy
[2020-10-15 13:51:27] [Toshio Kuratomi] heh :-)
[2020-10-15 13:51:31] [Jeff Geerling] For years and years and years we've wanted docs on galaxy
[2020-10-15 13:51:36] [Toshio Kuratomi] Col but... I have no control over that website ;-)
[2020-10-15 13:51:37] [James Cassell] otherwise, I tend to often abuse "{{ omit }}" to indicate an unset value in my roles
[2020-10-15 13:51:44] [V K] I second the idea of a "check" mode for validation only. We have inventory in git and it would be useful to run just the validation of vars used by roles ahead of actual deployments, so users can fix in advance.
[2020-10-15 13:51:54] [Brian Coca] for galaxy/AH it will make sense to show/build
[2020-10-15 13:51:57] [Jeff Geerling] we need a galaxy person in here to complain to :)
[2020-10-15 13:52:05] [Jeff Geerling] All the cool features go to Automation Hub for now
[2020-10-15 13:52:20] [Brian Coca] since you already have --json and galaxy/ah alraedyu uses that, it should not be hard to build, just requries resources
[2020-10-15 13:52:24] [Sorin Sbarnea] lets make this right and come up with a spec format that works the same way for modules and roles.
[2020-10-15 13:53:46] [Mitchell Whalen] what about a simple scan variables and show default that get used since defaults can be defined in multiple places?
[2020-10-15 13:55:53] [Brian Coca] disable auto, use your own and control the failed_when
[2020-10-15 13:56:32] [Brian Coca] why having it as an 'action' makes it very flexible
[2020-10-15 13:57:01] [Brian Coca] role meta?
[2020-10-15 13:57:32] [Brian Coca] yes, but for case that role author does the explicit action
[2020-10-15 13:57:40] [Sloane Hertel] we also discussed role meta. If the role defines an argspec, it should validate
[2020-10-15 13:57:49] [Sloane Hertel] yeah, what sivel said
[2020-10-15 13:58:11] [James Cassell] "{{ (var) if (var is string) else var_is_not_string() }}" I've done access-time validation by doing something like this, causing it to bail at the jinja level
[2020-10-15 14:00:18] [Ken Dreyer] Jeff, I'm at the point where I'm planning to just push my docs for my collections to readthedocs.org
[2020-10-15 14:00:39] [Ton Kersten] Has anyone connected with the iPad app? Keeps asking to register.
[2020-10-15 14:00:45] [Brian Coca] also having it as an action allows for easy testing of diff implementations
[2020-10-15 14:00:46] [Jeff Geerling] @Ken :( yeah
[2020-10-15 14:01:01] [Ken Dreyer] I haven't figured out how to make it pretty yet
[2020-10-15 14:01:29] [Ken Dreyer] I've been meaning to look into using https://github.com/ansible-community/sphinx_ansible_theme
[2020-10-15 14:01:43] [Brian Coca] its been discussed for loong time, see proposal
[2020-10-15 14:02:03] [Sloane Hertel] https://github.com/ansible/proposals/issues/39
[2020-10-15 14:02:04] [Felix Fontein] @Ken here's a minimal CSS file which makes the antsibull-docs generated docs look good with a normal sphinx theme: https://github.com/felixfontein/felix-ansible-docsite/blob/main/_static/ansible.css
[2020-10-15 14:02:33] [Felix Fontein] @Ken I extracted it from there, since the theme there was too heavy for my use-case
[2020-10-15 14:02:36] [Ken Dreyer] Felix, any tips or examples of generating docs for a collection that's no in Ansible?
[2020-10-15 14:04:01] [Felix Fontein] @Ken I'm using https://github.com/felixfontein/felix-ansible-docsite for https://ansible.fontein.de/ which hosts docs for some collections not in Ansible (and one that is).
[2020-10-15 14:04:38] [Felix Fontein] @Ken it assumes some changes to antsibull-docs though that haven't been merged yet
[2020-10-15 14:04:40] [Ken Dreyer] Do you run antsibull-docs? Or what code is generating and pushing to your website?
[2020-10-15 14:04:52] [Gundalow ] Back in 55 minutes for Minimum version of Python controller increase
[2020-10-15 14:04:56] [Gundalow ] Thanks everybody
[2020-10-15 14:05:06] [Ken Dreyer] Felix: Doh, I see the Makefile now :)
[2020-10-15 14:05:09] [Jose Jorge Atuncar Chong] than´k
[2020-10-15 14:05:15] [Ken Dreyer] Thank you
[2020-10-15 14:05:16] [Jose Jorge Atuncar Chong] thank´s
[2020-10-15 14:05:30] [Felix Fontein] @Ken the repo has a makefile, which runs antsibull-docs. The branch of antsibull-docs I'm using is https://github.com/ansible-community/antsibull/compare/main...felixfontein:faster-ansible-doc-2-additions which combines several of my PRs
[2020-10-15 14:05:39] [Ken Dreyer] Excellent
[2020-10-15 14:06:00] [Ken Dreyer] This is exactly what I was looking for
[2020-10-15 14:06:32] [Jeff Geerling] For documentation on galaxy: https://github.com/ansible/galaxy/issues/2225 and https://github.com/ansible/galaxy/issues/1937
[2020-10-15 14:09:45] [Todd Lewis] Let me toss an idea up in the air here: hierarchical tags. So "--tags=abc.xyz" would apply the "xyz" tag only in the context of where "abc" was active, such as an included task file or role. (Or is that imported? Hot mess!)
[2020-10-15 14:16:33] [Sorin Sbarnea] https://www.reddit.com/r/ansible/comments/jbsdbm/molecule_31_was_released/
[2020-10-15 14:17:34] [Kim Covil] Slight off-topic question regarding roles... Is there a way of doing a lookup to determine whether a role has a particular task file (for use in import_role:, tasks_from:) ?
[2020-10-15 14:31:36] [Brian Coca] lookup('pipe', 'ls ' ~ role_path ~ '/tasks')
[2020-10-15 14:32:46] [Kim Covil] but would we know what the role_path is from outside a role (i.e. where we would import it)?
[2020-10-15 14:36:05] [sivel ] @Kim, I'm not sure what you are attemting, but it sounds too dyanmic.  These should be explicit choices by playbook authors, about what entry points they want to use.  It's the role authors responsibility to document those entry points
[2020-10-15 14:37:59] [Kim Covil] I guess it might be too dynamic... I have a bunch of host roles that a setup playbook applies to hosts based on host_roles variable... what I want to do is run through the same list of host_roles in a decommission playbook but call a decommission task in that role if it exists
[2020-10-15 14:39:01] [Kim Covil] I don't want to just ignore errors as I want to differentiate between the decommission task not existing and the decommission task failing
[2020-10-15 14:52:52] [Todd Lewis] Good, because ignore_errors doesn't work for include_tasks, and your playbook won't compile with a bogus import_tasks.
[2020-10-15 14:54:56] [Gundalow ] DING DING DING Starting back in 5 minutes
[2020-10-15 14:59:08] [Jeff Geerling] _the final countdown_
[2020-10-15 15:00:30] [Ompragash Viswanathan] hii
[2020-10-15 15:00:59] [Carol Chen] Hi Ompragash!!
[2020-10-15 15:01:32] [Ompragash Viswanathan] Hey Carol \o/
[2020-10-15 15:01:45] [James Glynn] when is the cloud breakout?
[2020-10-15 15:02:02] [Jill Rouleau] @James the breakouts are at 20:00 UTC
[2020-10-15 15:02:26] [James Glynn] @jill thanks!
[2020-10-15 15:02:50] [James Cassell] ubuntu 20.04 python?
[2020-10-15 15:03:30] [sivel ] We are not actively packaging for Ubuntu 20.04, but Python 3.8 is default on 20.04
[2020-10-15 15:03:42] [Carol Chen] Please check the etherpad links for the various breakouts for a *different* bluejeans link: https://etherpad.opendev.org/p/virtual-ansible-contributor-summit-october-2020
[2020-10-15 15:03:45] [Carol Chen] line 123
[2020-10-15 15:03:50] [Toshio Kuratomi] james yeah, that's the only place it will natively work.
[2020-10-15 15:04:16] [sivel ] Ubuntu 18.04 also has Py3.8 packaged
[2020-10-15 15:04:30] [David Moreau-Simard] centos7 is 3.6 iirc
[2020-10-15 15:04:49] [Toshio Kuratomi] But the tooling is for an earlier version, no?
[2020-10-15 15:04:55] [sivel ] We will not be packaging for EL7 any longer
[2020-10-15 15:05:12] [Mitchell Whalen] Will people have support for ansible-base on say 3.6 if something does break? Many large companies, won't  move to a version that isn't supported by the OS vendor support
[2020-10-15 15:05:23] [Toshio Kuratomi] yeah, it's 18.04 is 3.6
[2020-10-15 15:05:27] [baptiste mille-mathias] wow my current company still use rhel6
[2020-10-15 15:05:44] [sivel ] Also note that this is controller, and not targets
[2020-10-15 15:06:01] [Toshio Kuratomi] It is also for configuing the controller.
[2020-10-15 15:06:07] [Adam Miller] @baptiste - you're still using RHEL6 as your Ansible controller?
[2020-10-15 15:06:26] [Adam Miller] this change only affects the control node, not the remote hosts in the inventory
[2020-10-15 15:06:45] [Jeff Geerling] @Adam - I have a few different controllers running on 16.04 and Debian 9 still :P
[2020-10-15 15:06:46] [sivel ] The controller can be a target, but there are different requirements
[2020-10-15 15:06:58] [Mitchell Whalen] controller or no, vendor OS support will still be required for  some companies
[2020-10-15 15:07:05] [Jeff Geerling] and I have a ton of playbooks that are meant to run on the controller and manage the controller
[2020-10-15 15:07:08] [baptiste mille-mathias] Adam Miller it may happens, not my decision
[2020-10-15 15:07:46] [baptiste mille-mathias] we try to move most what we can to awx but for ad-hac action I can do something from the current node I'm on
[2020-10-15 15:09:00] [Toshio Kuratomi] Adam Miller, The important thing to me is that it affects configuring the controller too.
[2020-10-15 15:09:24] [sivel ] It doesn't though, we still support back to py2.6 for targets
[2020-10-15 15:09:33] [Adam Morris] Why do we have to change anything?  Life was so much easier last Millennium.  You young people and your "improvements".  Now get off my lawn.  (Was that anti-change enough to cover all possible objections?)  :-D
[2020-10-15 15:09:36] [Toshio Kuratomi] sivel: It does though.
[2020-10-15 15:09:47] [Adam Miller] Toshio, I don't follow
[2020-10-15 15:10:11] [Toshio Kuratomi] Adam So this came up when we were discussing supporting RHEL8.
[2020-10-15 15:10:52] [Jeff Geerling] The 3.6/3.8 decision seems to be a pretty tough fork in the road in terms of anyone running Ansible on things outside of Tower
[2020-10-15 15:11:13] [Jeff Geerling] (And the answer shouldn't, IMO, be 'everyone gets to learn how to use pyenv')
[2020-10-15 15:11:35] [Toshio Kuratomi] Because modules that are run to  the controller host use sys.executable of the ansile-playbook script, you end up using the python that you run with.  That would mean that on a RHEL8 controller, you would no longer be using the system python (3.6) but some newer version.
[2020-10-15 15:12:07] [sivel ] You would need to stop assuming that controller and target are the same, and actively think differently about the scenario
[2020-10-15 15:12:23] [Toshio Kuratomi] The newer version doesn't have the python modules that are running in the system python.  So you either end up not being able to use those tools at all or else you install a different instance of those tools for the newer python version.
[2020-10-15 15:12:29] [Toshio Kuratomi] If you do the latter, you still have ssues
[2020-10-15 15:12:56] [Brian Coca] requires overriding implicit localhost
[2020-10-15 15:12:58] [Toshio Kuratomi] For instance, any dnf plgins that were installed by the sysadmin for the system dnf won't be present n the dnf that you installed for your newer python.
[2020-10-15 15:13:14] [Adam Miller] yes
[2020-10-15 15:13:33] [Sorin Sbarnea] there is decent type information in py36
[2020-10-15 15:13:38] [Sorin Sbarnea] quite usable in fact
[2020-10-15 15:14:03] [baptiste mille-mathias] so fun time ahead
[2020-10-15 15:14:51] [David Moreau-Simard] python-yum/python-dnf/selinux have historically been a pain due to venvs and stuff :)
[2020-10-15 15:14:52] [Adam Morris] Isn't this just an argument for putting Ansible Controller into a container in all cases?
[2020-10-15 15:15:10] [Sorin Sbarnea] tested on packed only for == will work only on, we all know that.
[2020-10-15 15:15:24] [Brian Coca] not really, if you are going to use against localhost, you basically want to 'define' localhost vs using implicit behaviour
[2020-10-15 15:15:35] [Jeff Geerling] @Adam - Maybe but that's a major architecture shift for ansible, the project, if we want everyone everywhere to do that, and not 'install ansible and use it'
[2020-10-15 15:16:13] [Sam Doran] @Adam No, not really.
[2020-10-15 15:16:35] [Sorin Sbarnea] selinux cannot be repackaged because it depends on specific kernel, it would not be portable at all
[2020-10-15 15:17:16] [David Moreau-Simard] There is https://pypi.org/project/selinux/ but it's kind of a hack :)
[2020-10-15 15:17:18] [Matt Davis] repackaging selinux bindings in downstreams for a local Python is my last choice- I have a couple other ideas to try first
[2020-10-15 15:17:19] [Sorin Sbarnea] that is why https://pypi.org/project/selinux/ is only a shim
[2020-10-15 15:17:30] [sivel ] Also, ansible 2.9 EOL is 2023 iirc
[2020-10-15 15:17:30] [Matt Davis] dmsimard: no "kind of" about it ;)
[2020-10-15 15:17:55] [Jeff Geerling] @sivel so the answer is we should all use 2.9 forever ;)
[2020-10-15 15:18:07] [Richard Elrod] if chrome it let me install it without restarting
[2020-10-15 15:18:10] [Sorin Sbarnea] selinux is highly dependent not only on python.h but also kernel,.... if someone is fixing it, i build him a mini shrine.
[2020-10-15 15:18:16] [Sam Doran] Screen sharing is hard. :)
[2020-10-15 15:18:29] [David Moreau-Simard] does screen sharing work in bluejeans with wayland yet ?
[2020-10-15 15:18:41] [Michael ​] @sorin: I am sure we can do some magic stuff with UML and unikernel
[2020-10-15 15:18:43] [Matt Davis] IIRC fulscreen yes, windowed, no
[2020-10-15 15:18:48] [Brian Coca] *grumble* bluejeans *mumble* *fist shaking*
[2020-10-15 15:19:14] [Adam Morris] @Jeff, I get that, it just seems like a simpler solution than an infinite number of "well, we can deal with this objection by doing...."
[2020-10-15 15:19:41] [Richard Elrod] @Brian after using the other platform... I'll take bluejeans any day.
[2020-10-15 15:20:45] [Carol Chen] we'll be evaluating more event platforms. while bluejeans isn't perfect, it checks a lot more boxes than many others.
[2020-10-15 15:21:09] [Carol Chen] suggestions are also welcome!
[2020-10-15 15:21:15] [Brian Coca] @Richard ascii cinema over irc is not that bad!
[2020-10-15 15:21:41] [Jeff Geerling] /me still reads that as "ascii enema"
[2020-10-15 15:21:51] [Sorin Sbarnea] I think that zoom by far offers the most rich options but I am not sure about scalability aspect.
[2020-10-15 15:21:52] [Michael ​] ah ah
[2020-10-15 15:22:10] [Brian Coca] @jeff its pretty much a synonim
[2020-10-15 15:22:11] [Adam Morris] @Brian, it beats MS Teams any day.
[2020-10-15 15:22:16] [scott boss] zoom really scales...
[2020-10-15 15:22:26] [Andrew Gaffney] @Jeff now I refuse to read it any other way
[2020-10-15 15:22:46] [Brian Coca] @adam i find google meet as 'slightly' better, but really want a good alternative (still searching)
[2020-10-15 15:22:56] [Sam Doran] Glad I'm not the only one that _always_ reads it that way.
[2020-10-15 15:23:34] [Sorin Sbarnea] i hate meet because even with app it forces me to open any link in Chrome instead of default browser.
[2020-10-15 15:24:05] [Adam Morris] @Brian, Unfortunately the powers that be have decreed that we must use MS Teams.  This is what we get when we hire one former MS exec in a high position.
[2020-10-15 15:24:37] [Brian Coca] @adam  https://www.youtube.com/watch?v=mmo3HFa2vjg
[2020-10-15 15:24:54] [sivel ] I really liked Vidyo, which is apparently the technology under the hood for Google Meet, or at least used to be for Google Hangouts several years back (~2015)
[2020-10-15 15:25:25] [James Glynn] we use slack, webex, zoom, teams ... bit of a multi-personality company lol
[2020-10-15 15:27:03] [Alicia Cozine] looks good
[2020-10-15 15:27:10] [baptiste mille-mathias] looks good now
[2020-10-15 15:27:12] [Gabriel Benselum] +1
[2020-10-15 15:28:33] [baptiste mille-mathias] mecanical keyboard isn't it ?
[2020-10-15 15:28:52] [Michael ​] the perk of working remotely, you can use them :p
[2020-10-15 15:28:56] [Andrew Gaffney] click click clank click click
[2020-10-15 15:29:20] [Andrew Gaffney] I didn't even notice the noise, because I've got a mechanical, too
[2020-10-15 15:29:22] [Brian Coca] /me uses brown switches
[2020-10-15 15:29:49] [Carol Chen] /me uses blue switches (but not right now)
[2020-10-15 15:30:02] [Michael ​] /me use butterflies
[2020-10-15 15:30:18] [Gabriel Benselum] /me uses a VERY legacy lenovo desktop keyboard
[2020-10-15 15:30:54] [Joshua Robinett] /me cherry mx
[2020-10-15 15:31:47] [Andrew Gaffney] Me and my wrists love my Kinesis Advantage2
[2020-10-15 15:32:44] [Sorin Sbarnea] holy pandas?
[2020-10-15 15:32:45] [V K] Kinesis fan here as well.
[2020-10-15 15:32:57] [Jeff Geerling] It would be nice if EEs could be on Galaxy :P
[2020-10-15 15:33:01] [Toshio Kuratomi] It looks nice.
[2020-10-15 15:33:09] [Alicia Cozine] Shane: your cat is totally stealing the show.
[2020-10-15 15:33:17] [scott boss] /me cherry mx browns
[2020-10-15 15:33:18] [Michael ​] +1 for the cat
[2020-10-15 15:33:18] [James Gill] blue clicky
[2020-10-15 15:33:20] [baptiste mille-mathias] sorry for opening the discussion about that ?
[2020-10-15 15:33:21] [Jeff Geerling] Galaxy adds Quay.io on top of Pulp and whatever custom stuff it does for roles :D
[2020-10-15 15:33:28] [baptiste mille-mathias] :)
[2020-10-15 15:33:31] [Brian Coca] KEYBOARD BRO!
[2020-10-15 15:33:31] [Jeff Geerling] hehe
[2020-10-15 15:33:35] [Toshio Kuratomi] Does it handle non-python deps?
[2020-10-15 15:33:44] [Matt Davis] it does
[2020-10-15 15:33:47] [Jeff Geerling] Quay.io is soooo hard to browse/discover
[2020-10-15 15:33:51] [Paul Belanger] Could you show / demo, talk about how ansible-test would work?
[2020-10-15 15:33:55] [Brian Coca] @toshion system packages afaik
[2020-10-15 15:34:04] [Toshio Kuratomi] What does that look like in the execution env file?
[2020-10-15 15:34:04] [Brian Coca] but should be easy to add 'gems' and others
[2020-10-15 15:34:11] [Brian Coca] packages:
[2020-10-15 15:34:19] [Jeff Geerling] quay could still be backend, but having a UI is nice
[2020-10-15 15:34:22] [Chris Houseknecht] It's an idea we've discussed for on-premise Automation Hub in the future.
[2020-10-15 15:34:50] [Jeff Geerling] And can someone address the question of baking a playbook/project inside the EE?
[2020-10-15 15:35:06] [Jeff Geerling] E.g. that way you can create an EE per project
[2020-10-15 15:35:06] [Sandra McCann] mesmerized by the cat napping in the background
[2020-10-15 15:35:14] [Andrew Gaffney] kitty!
[2020-10-15 15:35:15] [Toshio Kuratomi] jgeerling ++
[2020-10-15 15:35:16] [Adam Morris] It's stuffed
[2020-10-15 15:35:32] [Paul Belanger] https://docs.openstack.org/infra/bindep/
[2020-10-15 15:35:36] [Sandra McCann] heh
[2020-10-15 15:35:50] [sivel ] @Jeff we are working on real support for playbooks in collections, so baking a collection that has a playbook in it will become a supported thing
[2020-10-15 15:36:04] [Jeff Geerling] Ooh good question in Q&A: "Is ansible-test included in the image? How module/plugin developers can use it?" — looks like it uses ansible-runner to run the playbook, could ansible-test be run from inside the EE?
[2020-10-15 15:36:22] [Paul Belanger] https://github.com/ansible-collections/ansible.netcommon/blob/main/bindep.txt is another example
[2020-10-15 15:36:23] [sivel ] @Jeff but in the end, an EE is just a simple container with ansible/ansible-base and ansible-runner in it, and can be built by other means
[2020-10-15 15:36:57] [Matt Davis] Dunno if @mattclay is in here right now, but IIRC the plan was for ansible-test to live inside the EE as it usually would as an ansible component
[2020-10-15 15:37:04] [sivel ] I don't believe ansible-runner has any capability to run ansible-test, so you would have to runt he container via your container execution (docker, podman) and run ansible-test manually
[2020-10-15 15:37:16] [Toshio Kuratomi] Build, install, runtime.
[2020-10-15 15:37:25] [Sorin Sbarnea] bindep is created and maintained by openstack, we already have a two-step image builder that makes use of it.
[2020-10-15 15:37:26] [Jeff Geerling] time to file an issue for ansible-test to support running through an EE :P
[2020-10-15 15:37:32] [Toshio Kuratomi] shanemcd ^ Would be good groups
[2020-10-15 15:37:34] [Carol Chen] https://etherpad.opendev.org/p/ansible-contributor-summit-october-2020-morning
[2020-10-15 15:37:46] [Shane McDonald] Thanks Toshio
[2020-10-15 15:38:02] [sivel ] @Jeff since ansible-runner is responsible for running the EE, it would be a RFE for runner fwiw
[2020-10-15 15:38:08] [Alicia Cozine] we see it
[2020-10-15 15:38:17] [James Glynn] yes
[2020-10-15 15:38:47] [Jeff Geerling] My major question regarding galaxy_ng, Galaxy, and the future: What's the long term goal for galaxy.ansible.com in terms of integration all the fancy new features/design that are in Automation Hub (sorry but I can't post it as a Q&A in the Q&As since I'm logged in here as presenter)
[2020-10-15 15:39:41] [David Blackwell] @Jeff from what I understand.. automation hub is only available for paid for Tower licenses
[2020-10-15 15:40:18] [Gundalow ] Any Execution Environment questions please add to https://etherpad.opendev.org/p/ansible-contributor-summit-october-2020-morning Line 185
[2020-10-15 15:40:24] [Jeff Geerling] I understand that, but it seems like for the past year or so, there's been almost zero development/improvements to the community galaxy.ansible.com
[2020-10-15 15:40:34] [sivel ] @jeff asking the leading questions :)
[2020-10-15 15:41:28] [sivel ] As I understand it, the future is to use galaxy_ng to power community galaxy, some glue, and duct tape
[2020-10-15 15:42:20] [Jeff Geerling] @sivel I'm just nervous about all my poor roles
[2020-10-15 15:42:30] [sivel ] I have no comment at this time
[2020-10-15 15:42:43] [Brian Coca] jgroles.com ...
[2020-10-15 15:43:45] [Ken Dreyer] I am wondering the same thing about galaxy.ansible.com
[2020-10-15 15:43:53] [Jeff Geerling] https://ansible.jeffgeerling.com
[2020-10-15 15:44:38] [Ken Dreyer] Reminds me of the struggles around pypi.python.org over the years
[2020-10-15 15:44:40] [Toshio Kuratomi] geerlinguy: Does it feel like you're living in the opening paragraph of Cinderella?
[2020-10-15 15:45:37] [Jeff Geerling] :D
[2020-10-15 15:46:02] [Toshio Kuratomi] ;-)
[2020-10-15 15:46:22] [Jeff Geerling] qq - with this selection — how do you see what the namespace is?
[2020-10-15 15:46:54] [Jeff Geerling] e.g. would Published be "published.[collection_name]"
[2020-10-15 15:47:29] [Patrick Toal] Is there any plan for signing of collections?
[2020-10-15 15:47:42] [David Blackwell] so no mater the collection if synced from galaxy.ansible.com it shows community even if it is a certified collection?
[2020-10-15 15:47:43] [sivel ] @Patrick: yes
[2020-10-15 15:49:49] [Sorin Sbarnea] building on top of pulp was quite a smart move
[2020-10-15 15:50:02] [Jeff Geerling] don't rebuild, reuse
[2020-10-15 15:50:21] [Patrick Toal] You can see farther, standing on the shoulders of giants. :)
[2020-10-15 15:50:45] [Brian Coca] no need for giants, just stack enough people ...
[2020-10-15 15:51:06] [Tim A] agree Sorin.
[2020-10-15 15:51:55] [Jeff Geerling] "don't re-invent the wheel"
[2020-10-15 15:52:24] [Sorin Sbarnea] i am still frustrated that you need to upload a collection to galaxy in order to discover it fails
[2020-10-15 15:52:34] [Sorin Sbarnea] no way to do a dry upload
[2020-10-15 15:52:47] [Chris Houseknecht] Sorin you can run the importer locally
[2020-10-15 15:52:51] [Jeff Geerling] would be nice to have --dry-run or something, but a lot of that validation happens server-side :/
[2020-10-15 15:53:23] [Sorin Sbarnea] i do not mind so much about the server side, others may do, but to validate before tagging a release is really needed.
[2020-10-15 15:53:31] [Jeff Geerling] (In the UI, it was hard to see what collection namespace a collection was in)
[2020-10-15 15:53:36] [Chris Houseknecht] sorin: https://github.com/ansible/galaxy-importer#run-importer
[2020-10-15 15:54:36] [Chris Houseknecht] Those are your 'buckets' or repositories of available content
[2020-10-15 15:54:53] [David Blackwell] what is the presumed point of the locally run "galaxy"
[2020-10-15 15:55:04] [David Blackwell] why download collections from there except for dark sites
[2020-10-15 15:55:14] [Jeff Geerling] "What's the long term goal for galaxy.ansible.com in terms of integrating new features/design that are in Automation Hub?" (from way earlier)
[2020-10-15 15:55:23] [David Blackwell] are you expecting single locations to run multiple discrete engine environments?
[2020-10-15 15:55:36] [Chris Houseknecht] David Blackwell: to make it easy to share and discover content available to you from RH, Community and from within your organization
[2020-10-15 15:55:46] [Felix Fontein] @Ken Dreyer https://github.com/felixfontein/ansible-basic-sphinx-ext
[2020-10-15 15:55:59] [Jeff Geerling] > What's the long term goal for galaxy.ansible.com in terms of integrating new features/design that are in Automation Hub?
[2020-10-15 15:56:05] [Jeff Geerling] hehe I'll just keep pasting it
[2020-10-15 15:56:12] [baptiste mille-mathias] what is the target to replace galaxy with galaxy-ng ?
[2020-10-15 15:56:17] [Chris Houseknecht] Jeff Geerling: to be determined
[2020-10-15 15:56:21] [David Blackwell] @chris but an organization should theoretically only have one primary Ansible enviornment
[2020-10-15 15:56:27] [Felix Fontein] I'm also curious about that one :) there are a lot of roles on community galaxy, what happens to them?
[2020-10-15 15:56:30] [baptiste mille-mathias] s/what/when/
[2020-10-15 15:56:31] [Chris Houseknecht] It's something we'll be talking about in planning meetings for next year
[2020-10-15 15:56:42] [Jeff Geerling] Need a migration path, and time to migrate roles
[2020-10-15 15:56:52] [baptiste mille-mathias] good question from Felix
[2020-10-15 15:58:18] [Brian Coca] artifactory for collections
[2020-10-15 15:58:49] [Sorin Sbarnea] artifactory is quite cool, but also expensive
[2020-10-15 15:59:11] [Chris Houseknecht] Great job @newswangerd !!
[2020-10-15 15:59:24] [James Glynn] great job david, thanks
[2020-10-15 15:59:48] [Iftikhar Khan] yay! Recharge Day
[2020-10-15 16:00:06] [Carol Chen] https://etherpad.opendev.org/p/virtual-ansible-contributor-summit-october-2020
[2020-10-15 16:01:20] [James Glynn] thank you for all this, it was excellent
[2020-10-15 16:01:21] [Matt Davis] on the flip side, there's no "in room" convo that virtual folks miss out on ;)
[2020-10-15 16:01:22] [Jill Rouleau] Definitely better tea this way!  :)
[2020-10-15 16:01:37] [Jeff Geerling] @Matt - Very, very true
[2020-10-15 16:01:47] [Adam Morris] I'm in America and have decent Tea.
[2020-10-15 16:02:00] [Adam Morris] (But then I'm Scottish, just living in America)
[2020-10-15 16:02:09] [Iftikhar Khan] @jill +1 for better tea
[2020-10-15 16:02:23] [Gary Dhadwal] Thanks for the presentation
[2020-10-15 16:02:29] [Chris Thompson] hmm...not seeing the breakouts.
[2020-10-15 16:02:39] [Felix Fontein] Thanks everyone!
[2020-10-15 16:02:39] [Chris Thompson] weird. good thing I can see screenshare :P
[2020-10-15 16:02:44] [Sandra McCann] link for docs breakout - https://bluejeans.com/324026273
[2020-10-15 16:02:50] [Sorin Sbarnea] Thanks everyone! Time to go to sleep now...
[2020-10-15 16:03:00] [Jeff Geerling] cloud breakout - https://bluejeans.com/439557209
[2020-10-15 16:03:06] [Carol Chen] 20:00 Breakouts
Network
https://etherpad.opendev.org/p/ansible-contributor-summit-october-2020-network
Cloud
https://etherpad.opendev.org/p/ansible-contributor-summit-october-2020-cloud
Documentation
https://etherpad.opendev.org/p/ansible-contributor-summit-october-2020-Docs
Security
https://etherpad.opendev.org/p/ansible-contributor-summit-october-2020-security
[2020-10-15 16:03:09] [Alicia Cozine] Thank you gundalow and cybette for herding all us cats!
[2020-10-15 16:03:35] [Jill Rouleau] /me heads to the cloud breakout, thanks folks  o/
[2020-10-15 16:03:46] [Gary Dhadwal] Please email link to bluejeans video of this session
[2020-10-15 16:03:55] [Chris Thompson] (never mind...I clearly didn't catch on that the pad in use changed since the morning :P)
[2020-10-15 16:03:59] [Gary Dhadwal] and also one from 12/10/20
[2020-10-15 16:04:10] [Gundalow ] Thanks all
[2020-10-15 16:04:14] [Tadej tadeboro Borovšak] o/
[2020-10-15 16:04:16] [Chris Thompson] Thanks! Good shtuff.
[2020-10-15 16:04:21] [Brian Scholer] thanks!
[2020-10-15 16:04:31] [Amin Vakil] thank you all
[2020-10-15 16:04:33] [baptiste mille-mathias] bye
[2020-10-15 16:04:35] [Adam Morris] Thanks all, bye
[2020-10-15 16:04:37] [Marco Braga] Thank you all!
[2020-10-15 16:04:43] [baptiste mille-mathias] thanks
[2020-10-15 16:04:49] [Srinivasa Nagaraja] thank you
[2020-10-15 16:05:06] [Carol Chen] Find the Breakout sessions here - https://etherpad.opendev.org/p/virtual-ansible-contributor-summit-october-2020
[2020-10-15 16:05:22] [Carol Chen] line 136 onwards
[2020-10-15 16:05:25] [Carol Chen] Thank you everyone!
[2020-10-15 16:10:01] [Manuel Millan] thanks for the workshop
[2020-10-15 16:33:04] [Jose Jorge Atuncar Chong] thank´s great event