13:01:09 #startmeeting Team Silverblue 13:01:09 Meeting started Mon Aug 20 13:01:09 2018 UTC. 13:01:09 This meeting is logged and archived in a public location. 13:01:09 The chair is mclasen. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:01:09 Useful Commands: #action #agreed #halp #info #idea #link #topic. 13:01:09 The meeting name has been set to 'team_silverblue' 13:01:19 #topic roll call 13:01:34 who do we have today ? 13:02:47 o/ 13:02:53 .hello2 13:02:55 dustymabe: dustymabe 'Dusty Mabe' 13:03:20 .hello2 13:03:21 misc: misc 'None' 13:03:30 * juhp is lurking 13:04:12 .hello2 13:04:13 alciregi: alciregi 'Alessio Ciregia' 13:04:36 * alciregi is lurking too 13:04:49 :me is just here for the food 13:05:06 alright, some people are here 13:05:09 do we have topics ? 13:05:12 #topic agenda 13:05:27 misc: lol :) 13:05:39 I have a few informational items 13:05:41 mclasen: /me mostly lurking 13:05:53 but can help answer releng questions if anyone has any 13:06:06 * aday mostly cheerleader 13:07:23 any other topics suggestions ? if not , I'll just go with what I have 13:07:45 I also asked rishi to come over to tell us about the toolbox, but it might have been too short notice for a monday morning... 13:09:08 well, I would like to discuss the case of inkskape ( https://discussion.fedoraproject.org/t/inkscape-on-silverblue-a-few-caveat-to-keep-in-mind/ ), not as a specific case, but as the general workflow when people have suboptimal flatpak and what we can do for it 13:09:24 (and by discuss, I just mean "bring awareness of the issue", cause I suspect we can't fix it) 13:09:25 first big info is that today is the 1.0 release of flatpak. congrats! 13:09:40 \o/ 13:09:41 nice 13:09:47 misc: ok, go first with that 13:09:53 flatpak 1.0 ++ 13:09:56 #info flatpak 1.0 released today 13:09:56 (see, that's why we have cake) 13:10:08 #topic issues with inkscape flatpak 13:10:15 * mclasen clicks on that link 13:10:51 so, that's not so much about inkskape more that some software upstream might have bad default or badly interact when in a flatpak 13:11:09 and while people shouldn't use random packages, we likely all know they will 13:11:14 * dustymabe wonders if sanja is around 13:11:15 yeah, so we should perhaps help the inkscape team be a better player in a flatpak world 13:11:29 and I suspect that those random package would reflect badly on silverblue 13:11:43 dustymabe: I pinged her 13:12:21 mclasen: yup, but my main fear is that this one was found serendipitously, and that's not scalable :/ 13:12:35 well, software has bugs 13:12:45 using xdg-dirs was best practice before flatpak came along... 13:13:01 * dustymabe brb 13:13:26 I'll look over that thread after the meeting and file a few inkscape issues 13:13:31 we could certainly add something to the flatpak docs about how to handle it 13:13:32 yeah, and anything asking to modify /usr is user unfriendly by design 13:13:40 aday: don't we have that already? 13:13:48 I thought we have a whole section on integration ? 13:13:53 but as a user, I can't really see that before it is too late 13:14:03 mclasen: from a developer point of view. not so much from a user documentation point of view 13:14:12 I see 13:14:22 it would basically be a note to say "you need to update your user docs" 13:14:29 I also suspect that people who did the flatpak package didn't used inkskape as I did 13:14:40 cause even the designers I spoke with didn't knew that option 13:15:08 you could argue that apps shouldn't require users to dig into their config data, but that's maybe besides the point 13:15:25 yeah 13:15:50 * sanja is around, just unfortunately held up and lurking. Scrolling back to read. 13:16:36 ok, moving on from inkscape 13:16:45 #topic renames 13:17:12 we renamed the github repositories 13:17:37 Yes, we'll reflect that in the links on the website still. 13:17:37 it is now https://github.com/fedora-silverblue 13:17:41 misc: it's certainly true that there will inevitably friction for users in the transition; docs and guides for users will be important 13:18:18 #info github repositories are now at https://github.com/fedora-silverblue , links will be updated 13:18:37 weird -- what is https://pagure.io/teamsilverblue ? 13:19:14 * otaylor also thought that was where issues were tracked 13:19:35 we decided to stay on github for now, since pagure is lacking proper namespacing and renaming 13:19:43 and is just generally frustrating 13:19:48 yeah, we are still tracking on there - but we need to rename it - which isn't possible in Pagure. So there is no option other than either creating a new one on Pagure or the one on GitHub 13:20:07 to remove the team? 13:20:10 for now we created issue-tracker on GitHub and can still create one on Pagure but called fedora-silverblue 13:20:12 yes 13:20:24 it was because of how things were pre-Summit 13:20:46 * dustymabe would prefer for things to be in one place wherever that place may be 13:21:06 I think if silverblue is integrated into Fedora, having the issue tracker (at least) outside of fedora is considerably more annoying than whatever quirks pagure has 13:21:34 yeah. the reason we did that for coreos was because of a large community being on GH already for container linux 13:21:36 ok, well - since the docs for silverblue will be on fedora-docs, we'll create fedora-silverblue then 13:21:43 and we seriously gotta do something about pagure 13:22:05 sanja: i.e. the rename ? 13:22:08 it's furstrating to the point of me not wanting to work with it at all - it takes 4 seconds to load any given page, the search bar doesn't work and several other things. 13:22:11 yes, doing it now @dustymabe 13:22:29 it has no grouping, can't rename things, its slow... 13:22:35 but its ours, so yay ? 13:22:39 mclasen: it does have gropuing ?? 13:23:10 just flat namespace from what I can see ? 13:23:12 so the namespace thing i figured out and put the docs stuff under fedora-docs namespace 13:23:17 oh, ok 13:23:26 * mclasen takes it back 13:23:29 pagure is cool, it does have grouping 13:23:52 * aday chuckles 13:23:57 slow is likely something that will be fixed once they start to deploy repospanner to share the load, IIRC 13:24:29 haha - i'm ok with github, don't get me wrong. I just want to have reasons for using one vs the other 13:25:09 and definitely think having two places is the worst scenario 13:25:17 the reasons are it's unusable and either we take a chunk of time to fix it or it's always one side saying we gotta use it and the other side of being against it 13:25:35 no, the worst is having 5 place, with 1 being jira and a 2nd being bugzilla :p 13:25:38 lol 13:26:18 all the levels of hell 13:26:32 yeah 13:26:36 well so what do we do 13:26:45 i've deposited clear usability problems before already with several people 13:26:47 ehh. unstable isn't a word I would use to describe pagure.. slow at times, yes. lacking a few features, yes. 13:26:55 unusuable not unstable 13:26:57 it's very stable 13:26:58 but you don 13:27:02 sorry enter press 13:27:09 ahh. sorry - i misread that word 13:27:12 no problem 13:27:22 still, either we fix things or ...well, you know it's just not fun to use at all 13:27:29 I don't think we need to rush into moving things from github to pagure. the rename is good, in any case 13:27:35 I suspect it might also be slower from Europe than US ? IIRC, that's hosted in the Oregon univrsity lab and connectivity wasn't great :/ 13:27:38 there's not reward or incentive for outside of fedora users and even some fedora people use it grudgingly. counting myself as fedora people. 13:27:44 misc: good point 13:28:03 and given that sanja is doing most of the docs work, I would think her opinion on the usability of it matters... 13:29:48 any more thoughts on this ? 13:30:19 just pick one 13:30:34 well can we improve pagure? 13:30:38 what are the incentives for using it? 13:30:53 other than someone at some point decided it's what we use in fedora and then some people decided not to anyway? 13:31:27 i'm happy to use it if we can have an actual roadmap of we're fixing the search bar, we're fixing the connectivity, we're allowing renaming and it doesn't take 2 years to fix all of those 13:31:38 its not like we're the odd ones out ... anaconda is on github too, afaik 13:31:45 if you were really concerned about the choice, you could write up your requires, review the options, speak to whoever it is maintains pagure 13:31:54 I guess someone should bring that to Fedora infra 13:31:56 but yea, i'd just pick one :) 13:32:11 sanja: i think generally the two benefits people see are that 13:32:13 Centos is going to use pagure too, so there will be incentive to fix it 13:32:14 - it's open source 13:32:22 - we can add features that we need to it 13:32:34 i'm not arguing for or against those reasons, just stating what I think most people believe 13:32:56 there are a lot of projects still on github, but I think most of them were around before pagure existed 13:32:56 adding features is a theory unless resources are supplied 13:33:26 but we do not know until we ask 13:33:37 pagure did got feature requested 13:33:37 again. i'm not saying we should use pagure. just I think the decision should be justified and documented somewhere. because people are going to ask the question 13:34:01 I guess slowness could be documented and measured 13:34:11 ok I think this should be a priority in general for Fedora - to not tell people to use Pagure without fixing the issues. such things are important as it's an everyday tool and shouldn't just be imposed without giving clear benefits 13:34:24 we did document and justify for coreos 13:34:40 let's do an overall pagure thing ...i already startd that when other people asked me 13:34:40 having signle account management is a benefit, IMHO 13:34:45 indeed. we had a community vote, which I think makes sense 13:34:46 misc yes I do like that 13:34:56 and if we want to be Fedora, that also mean using Fedora tool, or all kind of stuff break 13:34:58 ok, action item for me 13:35:00 ok, so lets take the action to justify and document 13:35:17 I'll write that up, send it to everyone in workstation group, silverblue people here, and maybe some others 13:35:19 including Matthew and Bex 13:35:24 #action matthias and sanja to justify and document staying with github for now 13:35:35 nnnonono not justifying to stay with github 13:35:41 rather initiate discussion about pagure 13:35:44 #undo 13:35:44 Removing item from minutes: ACTION by mclasen at 13:35:24 : matthias and sanja to justify and document staying with github for now 13:35:51 and make a clear decision where there is no more questions about oh but why not pagure 13:36:00 #action sanja to initiate discussion around pagure usability issues 13:36:11 some actual research and measurement including for APAC people, UTC, not just US 13:36:15 yeah 13:36:27 good point about asia 13:36:30 ok, moving on from the pagure discussion, we'll not have a decision yet then 13:36:42 I have created docs and will populate them with the docs template 13:37:00 this week we should have a kubernetes on fedora guide published which is already written, just needs to be in the right format in the containers repo and it's good to go 13:37:02 chris negus wrote it 13:37:06 and several other docs we'll use 13:37:18 there's fedora-docs/coreos /silverblue and /containers 13:37:21 which will be in the main fedora docs 13:37:26 it's asciidoc format 13:37:35 and i'll write a separate email about it to the mailing lists and forums once it's all done 13:37:39 this week or next at the latest 13:37:40 do we have 'how to contribute?' hints on the docs website ? 13:37:48 see above - end of this week, maybe next 13:37:51 those will need updating, I guess (if they exit) 13:37:52 proper readme and how to 13:37:55 exist 13:39:00 sanja: any more on docs from you ? 13:39:57 moving on then 13:40:02 #topic composes 13:40:06 still broken :( 13:40:10 :? 13:40:10 with a different reason every day 13:40:11 :/ 13:40:41 i'd like to thank mclasen for starting to look at them 13:40:59 unfortunately this is the most turbulent time (right after branching) 13:41:02 I guess there's nothing we can do other than fixing breakages every day until we get lucky 13:41:19 mclasen: yeah it's a good idea to get all our ducks in a row before branching 13:41:32 and that way we know what breakage is because of branching and what isn't 13:41:38 but it's ok, we're starting to look at it now 13:41:58 we pretty much just have to follow along and fix issues we find.. the releng team will probably be churning out two a day for a while 13:42:01 todays culprit is grub, so I'll find pjones later 13:42:30 mclasen: if you find any issues please add them to the issue tracker even if it's just comments 13:42:35 we need to also rename everything everywhre 13:42:36 ok, sure 13:42:45 sanja: we've started the rename in the releng side 13:42:50 mohan helped us a bit there 13:42:53 including either sunsetting the fedora-atomic-wg on wiki (no rename possible on the page itself) 13:43:00 awesome, thanks dustymabe 13:43:14 * dustymabe just referring to FAW to silverblue renaming 13:43:15 i'm looking to edit the reference on pagure to exchange the names where possible 13:43:24 yeah dustymabe +1 13:43:34 for pagure, I'll look into it, for the wiki..i think otaylor owns it? 13:43:59 so we could delete the page now that we have an actual website and forum? or create a new wiki page with the right name 13:44:05 no team, just fedora silverblue 13:44:54 o/ 13:45:01 .hello x3mboy 13:45:02 * otaylor can take an action item to add redirects and/or notes to all the atomic-workstaiton stuff on the wiki 13:45:02 x3mlinux: x3mboy 'Eduard Lucena' 13:45:12 replacing the wiki content with a stub that just points to the website seems best ? reducing redundancy 13:45:34 action: owen to add redirects and/or notes to all the faw pages on the wiki 13:45:43 #action: owen to add redirects and/or notes to all the faw pages on the wiki 13:46:40 going back to general info items: I gave 2 1/2 silverblue presentations since the last meeting 13:46:47 hooray mclasen 13:46:55 * misc did only 1 13:47:01 at flock, gnome.asia and devconf.us 13:47:36 hooray misc 13:47:40 we should get all those added to the website, as recordings become available 13:47:54 #action matthias to add recent silverblue talks to the website 13:48:05 oh and we've grown quite a bit on twitter both @flatpakapps and @teamsilverblue 13:48:24 in my devconf.us presentation I talked a bit about the toolbox pet container that Debarshi is working on 13:48:28 which means we also have to put in some effort to get good docs up ...and fix bugs 13:48:40 oh any news on that mclasen? 13:49:06 rishi put a tarball out for people to try it. I'll push him to get it into git this week 13:49:21 any write up on this ? 13:49:23 of course that'll bring up the darn pagure-or-github question again :) 13:50:06 misc: no writeup yet, but a blog post about it would be the next priority after getting an initial release out 13:50:47 i know the FCOS group is interested in a toolbox pet container too 13:50:57 fcos ? 13:51:05 fedora coreos 13:51:14 * mclasen hadn't heard that abbreviation yet 13:51:27 and we should probably overlap with the containers SIG as there will be people interested there as well 13:51:32 * misc did heard it 13:51:32 woudl be good to cooperate, yes 13:51:38 * dustymabe has heard walters talk about the toolbox container before 13:51:52 * dustymabe tries to find a link to an issue that is open about it 13:53:11 https://github.com/openshift/os/issues/78 -- we should probably migrate that to https://github.com/coreos/fedora-coreos-tracker/issues 13:53:35 last info item from me: I have a stack of silverblue stickers on my desk here, so if stickers are needed somewhere, let me know 13:54:15 unfortunately, i forgot to bring them to my devconf talk, that would have been smart 13:54:33 #info matthias has silverblue stickers 13:54:42 I also do have them :p 13:55:01 if people can catch me, I can give them (but I am the IRL version of Carmen Sandiago...) 13:55:37 more swag will be ordered this month so if you know you need it somewhere 13:55:41 one more thing from me 13:55:49 container camp uk is coming up, september in london 13:56:02 if anyone wants to go, help be at the booth and demo silverblue or talk about coreos -> please let me know 13:56:19 misc or aday or others who are already in europe preferred 13:56:29 I will remember to bring stickers to my next silverblue talk ... in 3 weeks in Denver 13:56:50 #info misc has stickers too 13:57:08 * ksinny waves 13:57:37 https://2018.container.camp/ 13:57:45 6-7 september 13:57:48 * sanja has to go pick up her son from school 13:57:59 fill me in if there's anything else needed, gotta rush 13:58:02 * sanja waves. 13:58:14 coming up on the hour anyway 13:58:23 (write an email if you wanna do the container camp thing) 13:58:23 last call 13:58:25 * mclasen rings the bell 13:58:45 hi, ksinny! 13:59:15 mclasen: Hey! 13:59:17 ksinny: I gave all the mugs away already 13:59:24 Good to meet some of you at Flock :) 13:59:34 awesome mclasen++ 13:59:35 ksinny: Karma for mclasen changed to 4 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 13:59:47 #endmeeting