16:03:03 <sanja> #startmeeting atomic-community
16:03:03 <zodbot> Meeting started Mon Jan  8 16:03:03 2018 UTC.  The chair is sanja. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:03:03 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
16:03:03 <zodbot> The meeting name has been set to 'atomic-community'
16:03:03 <centbot> Meeting started Mon Jan  8 16:03:03 2018 UTC.  The chair is sanja. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:03:03 <centbot> Useful Commands: #action #agreed #help #info #idea #link #topic.
16:03:05 <sanja> ah yeah to start it you mean
16:03:11 <sanja> inviting zodbot?
16:03:29 <dustymabe> sanja: what's the question?
16:03:34 <dustymabe> zodbot is already in the channel
16:03:42 <sanja> woah why twice
16:03:48 <jberkus> centos bot too
16:03:54 <dustymabe> for some reason centos bot is in here too
16:04:07 * dustymabe wonders if that is necessary
16:04:12 <dustymabe> jbrooks: ^^ ?
16:04:14 <sanja> why it's showing that it started twice
16:04:21 <sanja> #topic roll call
16:04:23 <ashcrow> .hello smilner
16:04:24 <zodbot> ashcrow: smilner 'None' <smilner@redhat.com>
16:04:24 <sanja> .hello sanja
16:04:26 <dustymabe> .hello2
16:04:26 <zodbot> sanja: sanja 'Sanja Bonic' <sanja@redhat.com>
16:04:32 <ashcrow> sanja: both bots respond, it's alright :-)
16:04:33 <zodbot> dustymabe: dustymabe 'Dusty Mabe' <dustymabe@redhat.com>
16:04:45 <walters> starting a meeting with both the centos and fedora bots simutaneously is very appropriate 😆
16:04:46 <giuseppe> .hello gscrivano
16:04:47 <zodbot> giuseppe: gscrivano 'Giuseppe Scrivano' <gscrivan@redhat.com>
16:04:56 <walters> .hello walters
16:05:00 <jberkus> .hello jberkus
16:05:00 <zodbot> walters: walters 'Colin Walters' <walters@redhat.com>
16:05:01 <miabbott> .hello miabbott
16:05:01 <jligon> .hello jligon
16:05:03 <zodbot> jberkus: jberkus 'Josh Berkus' <josh@agliodbs.com>
16:05:06 <zodbot> miabbott: miabbott 'Micah Abbott' <miabbott@redhat.com>
16:05:09 <zodbot> jligon: jligon 'Jeff Ligon' <jligon@redhat.com>
16:05:18 <jbrooks> .hello jasonbrooks
16:05:20 <zodbot> jbrooks: jasonbrooks 'Jason Brooks' <jbrooks@redhat.com>
16:05:57 <sanja> Ok, shall we start with the topics then?
16:06:13 <rubao> .hello rubao
16:06:13 <zodbot> rubao: rubao 'rubao' <rubao.net@hotmail.com>
16:06:46 <ashcrow> sanja: go for it
16:07:19 <sanja> #chair jberkus miabbott jligon
16:07:19 <zodbot> Current chairs: jberkus jligon miabbott sanja
16:07:19 <centbot> Current chairs: jberkus jligon miabbott sanja
16:07:37 <sanja> #chair dustymaybe jbrooks walters
16:07:37 <zodbot> Current chairs: dustymaybe jberkus jbrooks jligon miabbott sanja walters
16:07:37 <centbot> Current chairs: dustymaybe jberkus jbrooks jligon miabbott sanja walters
16:07:56 <sanja> #chair gscrivano ashcrow
16:07:56 <zodbot> Current chairs: ashcrow dustymaybe gscrivano jberkus jbrooks jligon miabbott sanja walters
16:07:56 <centbot> Current chairs: ashcrow dustymaybe gscrivano jberkus jbrooks jligon miabbott sanja walters
16:08:37 <sanja> Ok, let's start then.
16:08:47 <sanja> #topic community Leads change
16:08:56 <sanja> That capitalization hurts. Sorry.
16:09:37 * ashcrow chuckles
16:11:13 <sanja> So, hi everyone - happy 2018 and let's announce the switch. So Josh is from now on the Kubernetes Community Lead
16:11:25 <sanja> jberkus, wanna say anything?
16:11:28 <jligon> congrats Josh!
16:11:59 <walters> welcome again!  On this topic I'd like to figure out how to keep up our PA outreach into the Kube community
16:12:16 <ashcrow> jberkus: Congrats!
16:12:43 <ashcrow> walters: +1
16:12:46 <dustymabe> walters: ideally, to make it better than it has been
16:13:05 <sanja> Well, we're not dividing things completely since I'll definitely be taking care to get PA out in the open in the DevOps and container communities anyway
16:13:09 <dustymabe> we've done a decent job with openshift, but broader kube. we aren't very well integrated
16:13:20 <sanja> So anything where Kubernetes appears, PA should be close anyway
16:13:35 * dustymabe is talking mainly about atomic host* I should stop doing that
16:14:02 <ashcrow> dustymabe: :-)
16:14:02 <sanja> you got your priorities, dustymaybe
16:14:13 <jberkus> sorry, double-timing
16:14:15 <dustymabe> :) - it's just a slip up I make a lot
16:14:25 <dustymabe> PA ~= PAH
16:14:31 <dustymabe> Project Atomic Host
16:14:33 <jberkus> this is super-exciting, becuase we're going to get so much more done
16:14:33 <dustymabe> :)
16:14:50 <miabbott> dustymaybe is my new favorite nickname
16:14:54 <sanja> walters - if you have suggestions for me in general about PA outreach, welcome anytime - that's for everyone, if you happen to have an idea or suggestion, drop me an email or write to the mailing list, regardless of whether it's docs or other resources or any forms of outreach
16:14:56 <dustymabe> jberkus: hopefully you can funnel a lot more people our way
16:15:01 <sanja> lol
16:15:11 <dustymabe> thanks mirabbott
16:15:25 <walters> yeah, it feels like for example the containers we build in fedora are also part of the PA umbrella right?
16:15:40 <walters> and a major point of those is to work nicely in kube
16:15:41 <dustymabe> walters: yes and no
16:15:41 <sanja> yeah, definitely
16:15:46 <ashcrow> I'd say yes
16:16:01 <dustymabe> i think some of the containers we build in fedora definitely come from the PA umbrella (i.e. the upstream tech)
16:16:03 <sanja> what's the no part?
16:16:09 <dustymabe> but if someone makes a cowsay container, i'd say not
16:16:20 <ashcrow> dustymabe: that's fair
16:16:52 <sanja> oh yeah
16:17:06 <walters> yeah, it's fuzzy
16:17:45 <walters> we're not going to figure out the clear dividing line in all these things on the first topic in this meeting, i was just trying to note that overlap exists
16:17:47 <sanja> ok, let's move on with this then and if we have more we roll it to open floor and add action items if necessary? thanks for the welcome by the way :)
16:18:09 <sanja> #topic documentation and resources
16:19:01 <sanja> So, I'm working on a rehaul of the docs, including parts of how we explain Atomic - I've discussed this in calls with ashcrow, jberkus, gscrivano, dustymabe already but wanted to announce it a little broader
16:19:23 <sanja> And ask whether anyone has any objections or clear ideas of how they *don't* want things
16:19:57 <sanja> that is for example, I want to move fast now, especially in the wake of DevConf. we have 3 talks just about FAW, and one about new distros in general ostree/flatpak
16:20:14 <ashcrow> No objections from me!
16:20:24 <jberkus> sanja: I'd suggest inclusion of dustymabe's tutorial into the AH docs
16:20:44 <sanja> I want to have a rudimentary documentation for FAW online by then and it should be somewhat in the same style as the docs for PA in general. For that I'm working on a rehaul and I wanted to ask whether you have any objections
16:21:07 <jligon> no objections that I can think of
16:21:31 <dustymabe> sanja: no objections. and I can say if you get our docs in order we all owe you a beer
16:21:31 <sanja> Plan is to have it ready by the next community meeting which would be on the 22nd - and then finish to add contnet for FAW until devconf, then work on additional content after fosdem when more time frees up
16:21:40 <jberkus> different doc repo, or just another dir in the same repo?
16:21:43 <dustymabe> or wine or coffee
16:22:04 <sanja> you say the beer thing because I'll give the beer back and just drink my tasty hot chocolate :D so yay, beers for you all
16:22:38 <walters> I personally had felt that FAW was a bit of a side project but its existence definitely forces some issues in the project structure
16:23:28 <sanja> yes it does, but also we really have 3 talks on it that were separately submitted and I think it's gonna become more than a side project but that's just a feeling right now, so let's see where it goes, I guess
16:23:32 <dustymabe> sanja: hot chocolate then!
16:23:44 <walters> like for example the whole "bundled container runtimes" issue...i personally would find it weird if we didn't have at least one pre-installed in FAW...in fact being agnostic to container runtimes makes it harder to have pre-installed tools (e.g. GUIs) that use them; like virt-manager + libvirt
16:23:46 <ashcrow> walters: I agree. I think it's worth getting some decent 'how to' docs out there folks can easily follow to try it out.
16:23:53 <dustymabe> walters: this might be an indication that there will be some growing pains to come with FAW
16:24:27 <dustymabe> since we are doing most of our innovative work in fedora first and fedorra users tend to be more desktop/workstation users
16:24:33 <walters> yeah a lot; anyways I think my feedback is i agree we should doc it obviously but let's just caveat prominently that it's "somewhat experimental" or something?
16:24:36 <dustymabe> they seem to have quite a bit of interest in FAW
16:24:37 <sanja> so maybe we'll end up with FAW+rkt and FAW+docker like Fedora+Gnome or Fedora+Openbox?
16:25:09 <dustymabe> sanja: ehh. not sure about that
16:25:13 <dustymabe> at least not initially
16:25:23 <ashcrow> yeah, that would be hard to maintain
16:25:36 <dustymabe> i think *investment* will probably be pretty low in FAW for now
16:25:39 <sanja> walters: yes definitely going to say it in my talk, and have a call with jlebon and klember about theirs so will say it there to mention that it's experimental still
16:25:44 <ashcrow> but that's specifics which we can work out later
16:25:44 <walters> yeah the whole "fedora is for desktops, centos/rhel for servers" thing is a deep issue...
16:25:45 <sanja> yeah definitely not initially but that's how I'd imagine it in the future :D
16:26:15 <dustymabe> walters: i think with atomic host we overcome some of those reasons people think that
16:26:23 <dustymabe> the *testing* is mainly what helps that
16:26:31 <sanja> Yeah, we just gotta make potential users aware of that, I'll document it and say it
16:26:36 <dustymabe> +1
16:27:09 <sanja> this actually brings me to the next topic planned, I'll roll over
16:27:12 <sanja> #topic Atomic Workstation
16:27:19 <walters> dustymabe: yes although e.g. the python3 vs ansible thing is a classic example of where admins have perfectly working config management scripts and don't want to necessarily port all of them every 6 months, and we still have that issue with servers
16:27:24 <sanja> sine we're already on it
16:27:36 <sanja> s/since/since
16:29:47 * dustymabe assumes sanja is typing
16:29:47 <walters> so just briefly on this topic i think we need to outline more of a vision; one thing that's been on the back of my mind is to make the goal "like https://www.qubes-os.org/ but with containers and virt"; specifically sharing data with containers is way easier
16:30:00 <sanja> Do we have any hidden docs/resources on Workstation other than what's found in the wiki and its links to the mailing list? Is someone using Atomic Workstation as their main OS now, including all the pain points?
16:30:37 <dustymabe> sanja: I know of at least 5 people using it as thier main OS
16:30:42 * miabbott is using Atomic Workstation full time
16:30:47 <walters> a few people here are; i'm not aware of any docs other than those linked from https://pagure.io/workstation-ostree-config etc.
16:30:48 * jbrooks raises hand
16:31:37 <jbrooks> The "why" is something we need to define
16:31:40 <sanja> Yeah, I'll switch this week as well. So need to know who to poke. ;)
16:31:45 <jbrooks> My why is: dogfooding this stuff
16:31:53 <jbrooks> which isn't broadly applicable
16:32:02 <sanja> thanks for the link to Qubes OS walters
16:32:11 <sanja> just quickly checked their own docs also
16:32:21 <sanja> very simplified and very user-oriented
16:32:23 <walters> we have this issue a bit too with Atomic Host that it's really not very differentiated from the "set of RPMs" that go into it - for example one thing I've thought about is changing the default bash prompts in the host and container to show which you're in
16:32:29 <jbrooks> qubes is very clear
16:32:51 <jberkus> sanja: a few people, including some non-fedora folks, actually
16:33:02 <jberkus> one of the heptio guys has switched his laptop to FAQ
16:33:05 <jberkus> er, FAW
16:33:40 <miabbott> walters: i've been doing that bash prompt change on my FAW
16:33:44 <miabbott> very helpful
16:33:45 <sanja> #link https://pagure.io/workstation-ostree-config
16:34:39 <jbrooks> miabbott, is that documented somewhere?
16:34:50 <sanja> do you have a config file to share for that bash prompt?
16:34:57 * jbrooks uses flatpak and layering almost exclusively
16:35:11 <miabbott> jbrooks: just a hack i found...basically checks for an env var
16:36:22 <miabbott> jbrooks: https://github.com/miabbott/files/blob/master/dotfiles/bashrc#L189
16:36:33 <jbrooks> thx
16:36:56 <miabbott> requires 'ENV container docker` in all your containers, tho
16:37:22 <jlebon> miabbott: might be more generic to check /.dockerenv
16:37:26 <sanja> thank you, noted for own usage!
16:37:39 <jlebon> i use [ vs {: https://github.com/jlebon/files/blob/master/dotfiles/.bashrc#L222
16:37:50 <miabbott> jlebon: that is the super hack i was hoping to find
16:39:03 <sanja> neat. ok, anything else to note about Atomic Workstation for now?
16:39:32 <sanja> ok, then next topic
16:39:42 <sanja> #topic minimal image for IoT & other purposes
16:40:22 <sanja> jberkus told me about this, don't know how much of this has already been discussed
16:40:43 <dustymabe> there is an open issue
16:40:49 <walters> #link https://pagure.io/atomic-wg/issue/403
16:40:53 <sanja> but in essence, minimal image with only necessary packages - I could work on this and already talked to ashcrow about it. any pointers to people or resources to kick this off?
16:40:53 <ashcrow> thanks walters
16:41:13 <dustymabe> so..
16:41:28 <dustymabe> i think this is similar to FAW, where we probably need to define our scope before we start working on it
16:41:33 <ashcrow> I think it's pretty neat, but I worry about spreading oursevles too thin
16:41:42 <dustymabe> what we don't want is to create a bunch of stuff and then not maintain it
16:41:56 <sanja> yeah, just thought I'd ask here as a start to identifying people
16:42:08 <giuseppe> sanja, I am interested in discussions about minimization of the image
16:42:17 <sanja> ok then I'd say we probably best have a starting conversation about it at DevConf since most of us are there and then summarize that in written form to share?
16:42:20 <dustymabe> so I know pbrobinson has had interest in this in the past (i.e. Atomic for IOT)
16:42:27 <jberkus> giuseppe: for FAW or AH?
16:42:30 <ashcrow> dustymabe: the issue itself reminds me of an epic which needs to be broken into spikes and units of work
16:42:43 <giuseppe> jberkus, AH
16:43:05 <dustymabe> ashcrow: sure, but who is interested in doing that? where does it play in priorities, etc
16:43:29 <giuseppe> jberkus, would it be very different effort for FAW?
16:43:30 <jbrooks> We can look to the workstation as an example
16:43:30 <dustymabe> if we don't have anybody that says they can work on something then I don't even want to do the work to break it up
16:43:40 <jbrooks> It started unoffical, and then got taken up
16:44:00 <jbrooks> So, someone needs to just get rolling
16:44:24 <jberkus> ok, thats two other topics, can we move on from FAW?
16:44:30 <ashcrow> dustymabe: I hear you
16:44:39 <dustymabe> i'm not so sure - workstation started as a shadow project (mainly by walters) and then patrick from releng/infra took interest
16:44:46 <jberkus> giuseppe: I think so, you don't really want to "minimize" a desktop env
16:44:51 <dustymabe> and that helped immensely
16:44:56 <ashcrow> In this room there sounds like there are 3 people with *active* want to do the minimal image ... jberkus, sanja, giuseppe
16:45:02 <dustymabe> i would say next steps would be to talk to pbrobinson
16:45:11 <jberkus> sanja: hey, giuseppe has a topic, and I have a topic
16:46:04 <ashcrow> Are we punting minimal iso till next meeting or still on that topic?
16:46:57 <jbrooks> I say punt until someone does some tests/prototype
16:47:07 <jbrooks> There's nothing stopping that
16:47:28 <ashcrow> I'm fine with that
16:48:00 <jberkus> giuseppe: ?
16:48:12 <giuseppe> I am fine as well
16:49:01 <jberkus> sanja: I'd like to move on to post-meltdown releases and CAH?
16:50:30 * dustymabe wonders if we lost her
16:50:34 <jberkus> I think so
16:50:35 <ashcrow> I think so
16:50:41 <dustymabe> she mentioned she was having some issues earlier
16:50:47 <dustymabe> jberkus: feel free to change topic
16:50:51 <jberkus> #topic CentOS Atomic Host and post-Meltdown
16:50:58 <jberkus> jbrooks: do we have a new CAH release yet?
16:51:03 <dustymabe> hmm that didn't work
16:51:08 <dustymabe> #topic CentOS Atomic Host and post-Meltdown
16:51:18 <jbrooks> I did a test build for the release on fri, KB is doing the official build right now, we'll release today
16:51:24 <jberkus> did we lose the bots?
16:51:35 <jberkus> keen
16:51:41 <dustymabe> they seem to be here
16:51:42 <dustymabe> oh well
16:51:45 <dustymabe> just keep moving
16:51:48 <ashcrow> they respond to msg pings
16:51:49 <jberkus> so, question: do we actually need to update the container images?
16:51:52 <jbrooks> bot lag?
16:52:00 <jbrooks> for what?
16:52:01 <jberkus> seems like we shouldn't need to, given that this is a kernel patch ...
16:52:17 <dustymabe> jberkus: only if the images contain an kernel rpms
16:52:32 <dustymabe> i know the origin guys had to rebuild some containers because they included the kernel-headers packages
16:52:35 <jberkus> do any images?
16:52:39 <jberkus> ah, ok
16:52:51 <dustymabe> no real worry, but might fail some security scanning
16:54:09 <jberkus> it seems likely that we'll have multiple security-forced releases related to spectre over the next few months
16:54:21 <sanja> If someone still wants to give one more insight into something that's not covered but Atomic-related, please just put your name in there - I'll send out a link to everyone on the list with a link to a presentation where you can either input your slides or send me yours and i'll put them in there - if someone doesn't want to work with slides, also good, just talk freely since it's only 2-3 mins each. We'll have one cockpit demo
16:54:22 <sanja> and other than that if someone wants to demo theirs, please state it
16:54:54 <dustymabe> jberkus: yeah, we'll spin when they come out
16:54:58 <jberkus> sanja: I think several of your statements to this channel got eaten
16:55:04 <sanja> anything to add for FOSDEM or DevConf? we also have the same FAW talk from me at FOSDEM and one more about distributions in general (featuring ostree and flatpak obviously)
16:55:30 <dustymabe> i really wish we could have someone git the pure Atomic Host talk at FOSDEM
16:55:39 <dustymabe> s/git/give
16:55:49 <jberkus> dustymabe: a bit late for that
16:55:53 <dustymabe> yeah
16:56:12 <jbrooks> sanja, I had a gap in messages from you from 08:42 to 08:54
16:56:30 <jberkus> dustymabe: one problem with submitting AH talks to track conferences is that the OS track people say it's containers and boot it, and the containers people say it's OS and boot it
16:57:05 <dustymabe> jberkus: so we make the container OS track! solved
16:57:09 <jbrooks> Heh
16:57:24 <jberkus> dustymabe: I tried, but FOSDEM picked another devroom organizer
16:57:34 <jbrooks> It's definitely OS
16:58:06 <dustymabe> FYI riot/matrix servers are having issues
16:58:19 <dustymabe> so we lost sanja walters miabbott
16:58:22 <jbrooks> Ah, do we use those
16:58:27 <dustymabe> some people do
16:58:27 <jbrooks> I guess they do
16:59:18 <jberkus> yeah, I think that europe and US are having completly separate discussions right now
16:59:48 <ashcrow> zodbot: ping
16:59:50 <zodbot> pong
17:00:45 <ashcrow> hi sanja_!
17:00:48 <sanja_> Sorry, I thought it was Riot but just got an email from jberkus saying it's Freenode - anyone still here?
17:00:56 <dustymabe> here*
17:01:02 <dustymabe> sanja_: i think it's the bridge
17:01:04 <ashcrow> We are
17:01:09 <sanja_> So, in case I seemed oblivious to all of your messages - I didn't get them.
17:01:16 <ashcrow> :-)
17:01:35 <dustymabe> sanja_: I don't think we responded to your questions
17:01:36 <sanja_> That means I'll read back in the logs once they're there, I was done with my topics though so has anyone ended the meeting already?
17:01:42 <dustymabe> not yet
17:01:49 <sanja_> Did we move on to open floor yet?
17:01:54 <sanja_> what was the last you got from me?
17:02:04 <ashcrow> 11:55:04           sanja | anything to add for FOSDEM or DevConf? we also have the same FAW talk from me at FOSDEM and one more about distributions in general (featuring ostree and flatpak obviously)
17:02:18 <ashcrow> but it came sort of randomly in the middle of other conversations
17:02:29 <sanja_> ah yeah but I noticed before Riot crashed entirely and isn't opening back up that suddenly a lot of unread messages appeared
17:02:34 <sanja_> which I couldn't read anymore
17:02:58 <sanja_> yeah noone was saying anything so I just had a meeting with myself
17:03:01 <sanja_> ;)
17:03:07 <ashcrow> ha :-D
17:03:07 <jbrooks> :)
17:03:10 * dustymabe does that sometimes
17:03:12 <sanja_> hahaha
17:03:29 * jbrooks gives himself a cookie
17:03:31 <sanja_> Well, I hope you guys had some sense from it anyway, I'll read back in the logs then.
17:03:40 <sanja_> open floor now or were you discussing still
17:03:55 <sanja_> I don't think i can even do that since not on with my normal username right now, just a guest atm
17:04:30 <dustymabe> let me see if I can end it
17:04:32 <dustymabe> #endmeeting
17:04:37 <ashcrow> oddly enough the bots were not responding to our commands either ... though they responded to pings
17:04:44 <sanja_> fascinating
17:04:45 <ashcrow> zodbot: ping
17:04:45 <zodbot> pong
17:04:47 <ashcrow> centbot: ping
17:04:47 <centbot> pong
17:04:47 <zodbot> ashcrow: Ping with data, please: https://fedoraproject.org/wiki/No_naked_pings
17:04:48 * dustymabe reads scrollback to see who is chair
17:05:00 <sanja_> *of course* the first meeting I start ends in asynchronous disaster.
17:05:08 <ashcrow> #endmeeting