17:00:30 #startmeeting fedora_atomic_wg 17:00:31 Meeting started Wed Jun 14 17:00:30 2017 UTC. The chair is yzhang. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:00:31 Useful Commands: #action #agreed #halp #info #idea #link #topic. 17:00:31 The meeting name has been set to 'fedora_atomic_wg' 17:00:42 #topic Roll Call 17:00:54 .hello yzhang 17:00:55 yzhang: yzhang 'Yu Qi Zhang' 17:01:41 .hello jberkus 17:01:42 rather quiet today 17:01:42 jberkus: jberkus 'Josh Berkus' 17:02:03 .hello dustymaber 17:02:04 dustymabe: Sorry, but you don't exist 17:02:05 sigh 17:02:08 .hello mnguyen 17:02:10 .hello dustymabe 17:02:11 mnguyen: mnguyen 'Michael Nguyen' 17:02:13 dustymabe: dustymabe 'Dusty Mabe' 17:02:15 there was an attempt 17:02:20 .hello sinnykumari 17:02:21 ksinny: sinnykumari 'Sinny Kumari' 17:02:23 trishnag: hi 17:02:26 * trishnag lurks 17:02:26 ksinny: hi :) 17:02:32 dustymabe: Hi 17:02:37 hi dustymabe :) 17:03:00 .hello walters 17:03:01 walters: walters 'Colin Walters' 17:03:14 .hello jasonbrooks 17:03:15 jbrooks: jasonbrooks 'Jason Brooks' 17:03:22 #chair yzhang dustymabe mnguyen ksinny jberkus walters jbrooks 17:03:22 Current chairs: dustymabe jberkus jbrooks ksinny mnguyen walters yzhang 17:03:28 * jbrooks in another mtg 17:03:49 .hello rubao 17:03:50 rubao: rubao 'rubao' 17:03:59 rubao: \o/ - welcome 17:04:02 oh wow you have a fas account o.o 17:04:06 #chair rubao 17:04:06 Current chairs: dustymabe jberkus jbrooks ksinny mnguyen rubao walters yzhang 17:04:10 apologies, double-timing until 10:30, so will be laggy 17:04:26 #topic Action items from last meeting 17:04:35 dustymabe to propose a VFAD for wiki and docs discussion 17:04:35 * maxamillion to update container guidelines to enforce ENV and label 17:04:35 maintainer usage https://pagure.io/atomic-wg/issue/277 17:05:17 dustymabe: hows the VFAD proposal going 17:05:17 #action dustymabe to propose a VFAD for wiki and docs discussion 17:05:22 yzhang: not done 17:05:31 gotcha 17:06:01 maxamillion probably in another meeting 17:06:07 I guess we can come back to him once he joins? 17:06:12 yeah re-action for now 17:06:23 #action maxamillion to update container guidelines to enforce ENV and label maintainer usage https://pagure.io/atomic-wg/issue/277 17:06:35 moving on 17:06:45 dustymabe: what's up? 17:06:59 adamw: there is a meeting item we have that i'd like you to be here for 17:06:59 to tickets 17:07:05 if you don't mind 17:07:09 i'll ping you when we get to it 17:07:10 quick info: This is on progress https://pagure.io/atomic-wg/issue/273 . sorry to keep this hanging for long. 17:07:28 trishnag: thanks for the update 17:07:29 dustymabe: I think Adam updated the guidelines 17:07:40 maxamillion: welcome 17:07:45 rgr 17:07:45 #chair trishnag mailga 17:07:45 Current chairs: dustymabe jberkus jbrooks ksinny mailga mnguyen rubao trishnag walters yzhang 17:07:46 thanks 17:07:47 whoops 17:07:50 #chair maxamillion 17:07:50 Current chairs: dustymabe jberkus jbrooks ksinny mailga maxamillion mnguyen rubao trishnag walters yzhang 17:07:50 .hello maxamillion 17:07:51 maxamillion: maxamillion 'Adam Miller' 17:08:00 mis-ping my bad 17:08:14 :) 17:08:15 maxamillion: how goes: maxamillion to update container guidelines to enforce ENV and label maintainer usage https://pagure.io/atomic-wg/issue/277 17:08:32 bah 17:08:33 I forgot 17:08:35 I'll do that today 17:08:46 ok, we re-actioned it for now 17:08:49 yzhang: thanks 17:09:06 ok, moving on to tickets 17:09:16 #topic clarify policy on atomic host support for older Fedora "number" releases 17:09:16 #chair jlebon 17:09:16 Current chairs: dustymabe jberkus jbrooks jlebon ksinny mailga maxamillion mnguyen rubao trishnag walters yzhang 17:09:21 #link https://pagure.io/atomic-wg/issue/228 17:09:41 so, that's done, no? 17:09:45 Is there anything more to do after that blog post 17:09:50 Or can we close it 17:09:51 except for approval by the Council 17:09:54 jberkus: i think so mostly - jbrooks was also going to do a fedmag post 17:10:00 we want to make sure we reach both audiences 17:10:07 dustymabe, I've got a draft in there 17:10:33 jbrooks: have you received any feedback? 17:10:44 oh, speaking of which, there's been no feedback from teh council, let me move that forwards 17:11:03 jberkus: +1 17:11:05 Want me to action that jberkus 17:11:09 as a reminder 17:11:20 dustymabe, I haven't, done know if I've looked in the right place / done it right 17:11:26 lemme check the issue, will get update before the end of the meeting 17:11:40 alright 17:11:41 jbrooks: ping me after meeting - i can help 17:11:48 ok 17:11:55 .hello miabbott 17:11:56 miabbott: miabbott 'Micah Abbott' 17:12:00 #chair miabbott 17:12:00 Current chairs: dustymabe jberkus jbrooks jlebon ksinny mailga maxamillion miabbott mnguyen rubao trishnag walters yzhang 17:12:14 #info waiting on fedora magazine blog post in order to close this ticket 17:12:16 is pagure down? 17:12:36 seems like it 17:12:40 good think I already have everything open 17:12:45 works for me 17:12:54 works for me as well 17:12:59 oh we're back 17:13:22 #chair clint eastwood 17:13:22 Current chairs: clint dustymabe eastwood jberkus jbrooks jlebon ksinny mailga maxamillion miabbott mnguyen rubao trishnag walters yzhang 17:13:22 anyways we'll leave this ticket open then, moving on unless there are objections 17:13:30 dang it walters 17:14:00 don't inflate our numbers 17:14:08 #topic fedimg: don't use 'builder' instance for uploading AMIs 17:14:15 hehe 17:14:16 #link https://pagure.io/atomic-wg/issue/269 17:14:21 I don't think sayan is here 17:14:36 I suppose we just wait for him to update? 17:14:54 pretty much 17:14:58 so, I think our PRD and versions policy are approved by the Council 17:15:02 awaiting verification 17:15:37 keep us updated 17:15:45 moving on 17:15:48 put an action in for next week 17:15:55 ok 17:15:58 #action jberkus to update wg on final approval of PRD 17:16:10 #topic talk/workshop proposals for Flock 17:16:31 #link https://pagure.io/atomic-wg/issue/279 17:16:46 so we've made some progress on this - i'll update my sessions with abstracts today and submit them 17:16:58 please add notes to the ticket when you've submitted your sessions 17:17:10 jbrooks and I have a draft for our proposal 17:17:19 feel free to review 17:17:30 as a reminder I think the submission deadline is tomorrow? 17:17:42 jberkus maxamillion how goes your sesions 17:17:46 sessions* 17:17:55 the original submission deadline is tomorrow - i don't know if they moved it back or not 17:17:56 mine's submitted and the one we're joint on jberkus was submitting 17:18:08 ooops, thanks. we have a description, I was submitting and ran into a timeout last night 17:18:11 will resubmit today 17:18:14 jberkus: +1 17:18:37 do we want to do a doc workshop for Fedora Atomic docs? 17:18:43 or is the VFAD plenty? 17:19:11 jberkus: i don't know - i think the VFAD could be more 'organizational' 17:19:32 while a workshop for docs could be 'actually create docs for stuff we've identified' 17:20:11 speaking of which, do you need to register before you submit a session 17:20:19 i really think we still need someone to curate docs - i.e. someone tells me, hey we need some docs on kickstart. i give them a text file with commands or a screencast showing how to do it. they create some pretty docs 17:20:58 i can spit content out all day long, but formatting and placing it where it needs to go is a different story 17:21:16 well its not on the issue, does anyone want to submit that talk? 17:21:20 dustymabe: that would assume that we can have the VFAD before Flock 17:21:21 When is the VFAD getting planned for wiki/Docs? 17:21:28 I can submit it if we're ready to do it 17:21:35 jberkus: i think we can have the VFAD before flock 17:21:38 trishnag: its still in planning 17:21:46 if we don't then we'll just have the VFAD at flock 17:21:47 :) 17:22:02 its just a FAD then 17:22:05 ACK. 17:22:07 IRLFAD 17:22:09 trishnag, dustymabe: the prerequisite is having the CI and publication pipeline working 17:22:10 :) 17:22:17 which is *possible*, but it's not done 17:22:56 jberkus: is there a ticket tracking that work? 17:22:57 jberkus: +1 to publication :) 17:23:26 dustymabe: kinda-sorta. could use a reboot. this will be tracked on github 17:23:44 #action jberkus to create new issues, specs for docs.projectatomic.io publication 17:23:59 jberkus: would be nice to have a ticket in our atomic-wg tracker just for visibility maybe? 17:24:12 +1 17:24:19 OK, but I'm not going to match ticket-for-ticket 17:24:30 that's just insane ... 17:24:32 jberkus: sure, links to other tickets are fine 17:24:43 i've done that quite a bit 17:24:56 it's a trail of crumbs basically 17:25:03 but we always have one starting point 17:25:12 or we can just create ticket here: https://github.com/projectatomic/atomic-host-docs/issues 17:25:28 maintaining this on both pagure and github is hard :) 17:25:32 trishnag: something else to track, and that doesn't go to a mailing list 17:26:05 So back to the topic, do we feel ready to do a docs workshop? 17:26:42 * dustymabe yields to jberkus 17:27:15 it would give us a deadline. and I finally have design help for the HTML/CSS 17:27:24 so that's + 17:27:49 dustymabe: ok. 17:27:59 the main risk is that I'm up for jury duty soon 17:28:12 so if I get picked, I'll need others to put in more work 17:28:20 but seems like a good idea 17:28:27 however ... when will we hold the VFAD? 17:29:12 soon™ 17:29:26 haha - i'm thinking after Fedora 26 release 17:29:47 when's that? 17:29:56 2nd week of july i think 17:30:09 well, I'll know by then about jury duty 17:30:15 can we plan on the last week of July? 17:30:16 * dustymabe has a lot of stuff in flight before then 17:30:22 sounds reasonable to me 17:30:46 Should we put that in an existing ticket somewhere 17:30:56 yzhang: there is no ticket yet :) 17:30:57 is there a ticket? 17:31:04 Also, please update the ticket for the existing talks once they've been submitted 17:31:09 #action create ticket to plan docs VFAD 17:31:10 haha - haven't proposed it yet, but have done a good job of alluding to it 17:31:31 I think we're mostly in consensus about this topic 17:31:48 moving on 17:31:57 #topic Fedora 26 Atomic/Cloud Test Day 17:32:04 #link https://pagure.io/atomic-wg/issue/280 17:32:34 looks like the planning is mostly done 17:33:12 do we have anything else to talk about? 17:33:14 * ksinny will be participating in Test Day 17:33:22 paging roshi 17:33:31 roshi is AFK today 17:33:33 i have an update 17:33:47 #info test day is on the 20th - dusty will be publishing blog post on fedmag soon 17:33:59 i have the post written up, just have to send it over for review 17:34:19 quick question though - do we have any wiki page like https://fedoraproject.org/wiki/Cloud_SIG for the atomic wg? 17:34:58 don't think so? 17:35:01 dustymabe: not to the best of my knowledge, no 17:35:15 i guess i'll create one - wanted to link to it from the post 17:35:22 dustymabe: +1 17:35:24 ok EOM for this ticket 17:35:27 #action dustymabe to create SIG page for wiki 17:35:34 dustymabe: https://fedoraproject.org/wiki/Atomic 17:36:01 ok I think we're moving on 17:36:04 #topic Atomic Host images omit many common locales that all other flavors include 17:36:05 trishnag: there's also https://fedoraproject.org/wiki/Cloud, which is different than CloudSIG 17:36:12 #link https://pagure.io/atomic-wg/issue/282 17:36:19 ah yes 17:36:21 adamw: wanted to include you for this ticket 17:36:22 dustymabe: you just added the link 17:36:30 aha 17:36:40 walters and jlebon are here too i think 17:36:53 what's a good solution to this problem? 17:37:18 fwiw, the obvious impact of this is that if you run the atomic host installer in Canada, you get a bunch of locale errors every time you boot up. though apparently there's a change in hand to have the installer not use locales that aren't available in the installed system, somehow. 17:37:56 adamw: is there a link to that change somewhere? 17:38:03 let me see if i can find it again 17:38:58 adamw: huh what? why would the locales not be available? 17:39:14 i believe https://github.com/rhinstaller/anaconda/pull/871 17:39:20 jberkus: because we limit what locales are put into atomic host 17:39:38 mostly for space I think 17:39:45 but anaconda doesn't know about that and offers / automatically uses locales that won't actually be in the installed system 17:39:57 (it guesses your locale via geoip) 17:40:10 right, this works fine for the cloud base image because we don't really offer an ISO image to do installs 17:40:16 for atomic, we do offer an ISO 17:40:40 IIRC the size difference was on the order of 70MB but i can check quickly 17:40:45 doing so now 17:41:02 walters: cool, adding that info to the ticket will be nice 17:41:21 jlebon: also, the link adamw just posted shows a change to rhel7 anaconda, did that code make it into Fedora? 17:41:45 dustymabe: well, the anaconda thing is only one aspect, really 17:41:54 dustymabe: not yet, no 17:41:59 people might still actually want to *use* the typical Fedora locales on an atomic host system, regardless of how it's deployed, no? 17:42:33 jlebon: ok, we should track that work somehow and not drop it 17:42:45 would it make sense to split this by server vs workstation. for the latter i'd vote for just including them all 17:42:58 you keep using that word "Fedora" like there's one thing ;) 17:43:11 among other things for example, the container image also doesn't have locales 17:43:15 how much space do locales take? 17:43:34 jberkus: 13:40:41 walters | IIRC the size difference was on the order of 70MB but i can check quickly 17:43:37 walters: yeah, I'm more concerned about that; it's rather a PITA to add in the locales you need 17:43:51 yeah, definitely true 17:44:01 adamw: +1 17:44:16 among other things we have the same problem with ostree deltas that deltarpm has with this situation 17:45:09 walters: for the container image, we want to load the locales as part of the user building an application image, but we really need a tool for that 17:45:26 since Fedora expects to ship with all of the locales, the tooling for adding missing locales is *terrible* 17:45:54 walters: what is that problem that both deltas share? 17:46:02 (note that this is not unique to us, Debian has the same issue) 17:46:32 maxamillion, deltarpm needs to reconstruct the original RPM bit for bit to do GPG verification, which it can't do if the locale files have been stripped 17:46:44 walters: ohhhh ok, gotchya 17:46:46 it's why if you install deltarpm in the fedora base container image it will complain and mostly not be useful 17:47:18 note i filed an RFE for rpm to not try deltarpm if files were missing https://bugzilla.redhat.com/show_bug.cgi?id=1458035 17:47:20 ostree deltas have the converse issue in this case; since we're stripping locales, defining some sort of "extra layer" for them is...harder 17:47:28 but it was closed :( 17:48:51 So do we have consensus on any actionable items 17:48:52 ok - anyone want to summarize the challenges this ticket gives us for the ticket? 17:49:38 i'll evaluate the size difference again and i vote we discuss more once we have that data back in hand 17:49:59 can you update the ticket at https://pagure.io/atomic-wg/issue/282 to summarize walters ? 17:50:04 ok - i will note that our fedora 26 images are larger than the 25 ones - haven't really investigated it 17:50:07 yes 17:50:10 but we are trending large :( 17:50:14 and then we can discuss how much we care about image sizes :) 17:50:17 walters: +1 17:50:23 #action walters to update https://pagure.io/atomic-wg/issue/282 with the discussed points 17:50:29 moving on then 17:50:36 #topic Open floor 17:50:50 anyone have any open floor topics? 17:50:56 test day next week :) 17:51:06 please give things a spin if you can 17:51:15 we will be around on the 20th for any questions 17:51:25 ksinny: anything going on with you and alt arch? 17:52:03 dustymabe: nothing new in terms of Atomic 17:53:09 Rawhide nightlies are mostly getting DOOMED from past 2 weeks I believe and hence don't see much in composes for Atomic 17:53:25 ksinny: cool - i think last nights worked 17:53:40 * ksinny needs to keep looking at it 17:53:42 anyone else with anything? 17:53:50 I guess since we have no other pressing concerns, does anyone want to give thoughts on the proposal I wrote up for flock: https://pagure.io/atomic-wg/issue/279#comment-444783 17:54:43 yzhang: almost seems like two sessions 17:54:56 yzhang: I like it but I worry you'll run out of time 17:55:11 mm, well there is a 3 hour option 17:55:12 1. how to install kube/origin 2. the nitty gritty of system containers and docker vs crio 17:55:16 for a hands on 17:55:22 make 3 proposals 17:55:27 1 3-hour workshop 17:55:33 then two 1-hour sessions 17:56:07 as far as I can tell the "Do-sessions" are 2 or 3 hours 17:56:17 whereas the talks are 30/60 min 17:56:30 I think this would be better for a do session no? 17:56:32 jbrooks: are you here 17:56:37 yes 17:56:52 BUT if it's not a do-session, then it could be one or two preso sessions 17:57:32 Hm, we can discuss further in #fedora-cloud or #atomic when jbrooks gets back from his other meeting 17:57:36 Since we're running out of time 17:57:45 I'd like to get his input too 17:57:55 Alright we're running out of time 17:58:08 Anybody else with questions/concerns 17:58:26 Speaketh now or hold your peace until whenever I guess 17:58:32 we're always around 17:58:40 especially dustymabe that man doesn't sleep 17:58:56 yzhang, sorry, split attn, I'll ping you in fedora-cloud 17:59:04 :) - /me needs sleep 17:59:07 sounds good, will end meeting in 5 17:59:08 4 17:59:10 3 17:59:11 2 17:59:12 1 17:59:14 #endmeeting