/srv/irclogs.ubuntu.com/2018/05/17/#ubuntu-desktop.txt

=== maclin1 is now known as maclin
didrocksgood morning06:34
dufluGood Thursday06:39
jibelGood morning06:41
didrockssalut jibel, hey duflu!06:42
dufluHi jibel06:42
dufluSo today is the day.... :)  http://cdimages.ubuntu.com/daily-live/current/06:42
=== pstolowski|afk is now known as pstolowski
oSoMoNgood morning desktoppers07:15
=== amano_ is now known as amano
didrockshey oSoMoN07:21
oSoMoNsalut didrocks07:22
oSoMoNça va?07:22
didrocksça va, le soleil est revenu ce matin07:23
didrockset toi ?07:23
oSoMoNtrès bien, pas de rush pour aller à l’école ce matin, ma fille est partie en classe verte pour 3 jours, c’est étrangement calme (et il fait beau)!07:26
didrocksah, en effet :)07:27
seb128good morning desktopers07:30
didrockssalut seb12807:30
oSoMoNsalut seb12807:37
seb128lut didrocks oSoMoN, en forme ?07:37
oSoMoNoui, et toi?07:37
seb128nickel07:37
willcookeword07:58
didrockshello08:01
didrocksbad FIFO implementation :)08:01
Laneyey up08:01
didrocksand broken keyboard for Laney :p08:02
didrocksor too much hairs under the "h"08:02
Laneynope08:02
Laneyjust teaching you the nottingham dialect08:02
didrocksthis is like "how mouch?"08:02
Laneyhttps://www.youtube.com/watch?v=NQTGkO5GjD808:03
didrocksok, remind me to never get there or I'll get lost and cry :)08:04
Laney:D08:05
Laneydon't actually hear people talking like that very often sadly :(08:05
didrocksis it consider as slang or a local dialect?08:05
didrocksconsidered*08:05
Laneywell I guess I mostly know people that aren't from here originally08:05
Laneynot many of the local old people :P08:06
seb128good morning U.K08:06
seb128willcooke, Laney, how are you?08:06
Laneyhmm, probably a dialect?08:06
Laneyhey seb12808:06
Laneyfeeling GOOOOOOOOOOOOOOOOOOD08:06
Laneywent to bed at 223008:06
Laneywhat about you?08:06
willcookealright me ducks?08:06
=== maclin1 is now known as maclin
Laneywotcha geeza08:08
seb128I'm doing great08:08
seb128yesterday was tennis night!08:08
didrocksLaney: that's my daily routine to go to bed at that time now :)08:08
seb128it was quite windy and I'm out of shape, didn't play much during the winter between weather/us being sick in turn/weather08:08
seb128but it was still good fun :)08:08
Laney:>08:09
seb128same here, 2230 is our normal bed time now :/08:10
willcookeheh. Rock and roll in the channel today08:10
* willcooke complains that his knees hurt08:11
didrocksfor dancing Rock and roll? :)08:11
dufluMorning willcooke, Laney (I just saw words that didn't immediately translate, but they do aurally)08:13
dufluAnd morning seb12808:19
seb128hey duflu, how are you today?08:20
dufluand oSoMoN08:20
Laneygreetings duflu08:20
dufluseb128, doing OK. I finished a rework of a mutter branch for the Gnome guys just now that I expected to take at least all day.08:20
dufluYou?08:20
Laneycouple of months ago one of the lights in the kitchen started making the fuse trip08:21
seb128nice08:21
seb128I'm good thanks08:21
Laneyunscrewed it to have a look at the wiring08:21
Laneyface and body covered in water /o\08:21
seb128urg08:21
willcookeoof08:21
dufluPresumably with electricity off08:21
Laneythe push pipe fitting from the bath had come loose08:21
Laneyyes08:21
oSoMoNhi duflu08:29
Nafalloo/08:41
dufluHi Nafallo08:50
Nafalloduflu: hey! how's thing going with that annoying purple screen and mouse bug? :-)08:55
dufluNafallo, I think I saw Laney land an upstream fix last night08:56
Nafallooooh. nice :-)08:56
Nafallojust need to wait for packages to test then :-)08:58
Nafalloright. I need to go collect medicine and get the bus home. bbl.08:59
seb128Laney, Trevinho, you managed to land the gdm fix/refactoring for the refcount issue from what I see, well done! :)09:10
Laneythx09:11
Trevinhoseb128: yeah... Who cherry picks it? I can do it later if you want or I just point the commits if in rush09:12
Laneydon't worry about it, I can handle it09:13
Laneyquite a lot of pings about this one...09:14
seb128thx Laney, sorry about the pings09:21
Laneyno worries09:22
Nafallowillcooke: bug #1730765 still on your radar, right? :-)10:29
ubot5bug 1730765 in gnome-shell-extension-ubuntu-dock (Ubuntu) "Mounted external devices do not appear in dock" [Wishlist,Confirmed] https://launchpad.net/bugs/173076510:29
willcookeNafallo, it'll get reviewed when we go through the cc list.  So once 18.04.1 is out the door, or thereabouts10:30
willcookeNo promises though, we've already got a full schedule.10:30
Nafallowillcooke: hmm. I was hoping for 18.04.1 actually ;-)10:31
willcookeWorth seeing if anyone upstream and/or in the community wants to work on it.  I've zero clue about how much work it would be, but I think it's probably non-trivial10:31
Nafallohmm. let me check the release schedule. I might be able to find time for working on it as long as I get some handholding from time to time.10:33
willcookewoot!10:33
Nafallowell, we need it for 18.04.1 I think :-)10:34
Nafallowould make sense to get time allocated.10:34
Nafallowow. 18.04.1 week 30 and 16.04.5 week 31 :-P10:35
=== pstolowski is now known as pstolowski|lunch
didrocksThere is some work being done on it upstream12:08
didrocksI don't think it's realistic for 18.04.112:08
didrocksNafallo: sounds like you volunteer to help them? :)12:08
didrocksas you feel that's crucial (knowing that we didn't huge feedback about the necessity on 17.10…)12:09
didrocksNafallo: willcooke: https://github.com/micheleg/dash-to-dock/pull/677 FYI. Note that we need some nautilus API as well, this is why I pointed them to talk to upstream about it12:10
ubot5-ngmicheleg bug (Pull request) 677 in dash-to-dock "Trash and Removable Device Icons" (comments: 12) [Open]12:10
Nafallodidrocks: as soon as I can land my new role, I've already negotiated doing upstream work :-)12:11
didrocksnice!12:11
NafalloLaney: should be fine running the cosmic proposed builds on bionic still, right? :-)12:12
Nafallodidrocks: I like that the patches Philip did takes the trash can into consideration as well. I've got a separate extension for that at the moment :-P12:26
* didrocks really never use the trash that much…12:32
didrocksremember that dnd isn't supported12:32
didrocksthough12:32
Nafallodidrocks: yeah, me neither. no idea if my users do either :-)12:36
=== pstolowski|lunch is now known as pstolowski
oSoMoNLaney, the chromium-browser autopkgtest run for armhf appears to be stuck (http://autopkgtest.ubuntu.com/running#pkg-chromium-browser , been running for >5hrs), is there a way to cancel and re-run?13:04
LaneyoSoMoN: all I can do is kill the process, no user facing way13:09
Laneyit will time out eventually13:10
Laneyand the container is still running fine, smells like the test itself has hung to me13:10
LaneyNafallo: dunno, sounds like you get to keep both halves if it breaks13:11
NafalloLaney: :-D13:12
NafalloLaney: I'll try it out in a bit.13:12
Laneyif it were me I'd build the package myself13:12
* Nafallo notice Laney is trying to get him to write an ansible role for setting up a schroot environment ;-)13:13
* Laney sounds the sysadmin siren and runs away13:14
Nafalloawesome... lastpass-cli just became completely unusable because Lastpass changed certificate and upstream of the client requires new upstream versions for that... :-(13:14
Nafalloehrm. upstream is Lastpass themselves...13:16
Nafallogah13:16
andyrockgood morning all!13:25
Nafallomorning andyrock :-)13:26
willcookehey andyrock13:26
oSoMoNLaney, ack, thanks13:30
oSoMoNtoo much load on my machine, it almost grinded to a halt, I hoped to restore it by jumping to a tty and killing a few processes, but even that was too much, had to do a hard reboot :/13:31
didrockshey andyrock13:35
kenvandinedidrocks, any idea why i can't request a review from oSoMoN for the helpers?13:40
didrockskenvandine: good question, there is no special team or anything, maybe he needs to start a first review to get in the review list then?13:42
oSoMoNI just went through every possible setting at https://github.com/settings/, and nothing appears to be explicitly disallowing review assignment13:43
oSoMoNI already reviewed some of Ken's changes13:44
didrocksas part of the team?13:44
didrockswell, org group member?13:44
seb128hey andyrock13:45
didrocksoh13:45
didrocksI know I think13:45
didrocksyou are in private mode13:45
didrockslet me see if I can switch you to public13:45
didrocksoh no, you are public…13:45
didrockslet me grand you explicitely access to that repo13:46
didrocksoSoMoN: you probably received an email13:47
oSoMoNdidrocks, yep, and I accepted the invitation13:48
oSoMoNgithub says "You now have push access to the ubuntu/snapcraft-desktop-helpers repository."13:48
oSoMoNkenvandine, can you assign me a review now?13:48
didrocksand you are on the list!13:48
didrocksoSoMoN: you couldn't push before, just being part of the org?13:48
oSoMoNdidrocks, dunno, I didn't try in that timeframe13:48
didrocksmaybe it's better to add directly people to repos thus…13:49
oSoMoNI was added to the org yesterday or the day before, can't remember exactly13:49
didrocksI thought I added you to the org last week13:49
oSoMoNit could be last week, my memory is failing me13:52
oSoMoNaccording to my inbox, it was the 14th13:53
oSoMoNso neither yesterday, nor the day before, nor last week :)13:53
didrockswe were both wrong!13:55
didrockswhat, it's not Friday? Felt already like a week :p13:56
oSoMoNyou wish :)13:57
kenvandineoSoMoN, yes14:03
kenvandineoSoMoN, what did you do?14:03
didrocksI added him to the repo explicitely14:04
kenvandineah14:04
kenvandineok14:04
kenvandinethx14:04
didrocksyw14:05
andyrockseb128: livepatch guys say we just need to check if we're in a LTS or not14:07
andyrockseb128: in order to understand if we need to show the UI or not14:08
didrocksandyrock: what's their answer to people installing the snap on non LTS?14:09
seb128what didrocks said14:09
didrocksconfusing for them to install non working services…14:09
andyrockdidrocks: well if you install it you still need to enable it14:10
andyrockonce you enable it you get an error message14:10
didrocksstill seems unnice while we have an existing mecanism to prevent this14:11
didrocksany reason for them to not use per-release track?14:11
andyrockbut the per-release track will not prevent people to install the snap14:12
didrocksI don't know if it's only a naming convention or something backed in into snapd, would need to reach out the snapd people14:13
didrocksat least, it would give you a clear hint that's purposed for one release14:13
seb128andyrock, checking if we are on a LTS is probably going to give us the right user experience, but yeah it would be nice if they handled that better...14:14
seb128we should still communicate that to them14:15
seb128but I think it's fine to do the lts check meanwhile on our side14:15
andyrockthe question is how we do the lts check? 😂14:15
andyrockseb128: python has some nice libraries to do that14:16
seb128what is the livepatch snap doing?14:16
seb128can you see their code?14:16
andyrockthey're checking the kernel14:16
seb128you mean?14:16
andyrockthe kernel version14:17
seb128shouldn't we do the same thing then?14:17
seb128but do they have a mapping table?14:17
seb128or do they ask the server if it's supported?14:17
andyrocklet me check14:17
seb128andyrock, otherwise you could parse /etc/os-release and see if VERSION contains "LTS"14:18
andyrockI would go with the /etc/os-release thing14:18
seb128or use the ID and see if it's an <even number>.0414:18
* didrocks would think "LTS" is safer14:18
andyrockit feels less likely to change in the future14:18
didrocksbtw, with .1 .2 release, do anyone of you know if the ID change as well?14:19
* didrocks was wondering that for ubuntu-report14:19
didrocks(VERSION_ID)14:19
kenvandinewillcooke, seb128: oSoMoN and I are talking about how important is it to build chromium snaps for armhf/arm6414:19
kenvandinethoughts?14:20
seb128not?14:20
kenvandinesince we don't have arm desktop images14:20
kenvandinedoesn't seem useful14:20
seb128didrocks,14:20
seb128VERSION="16.04.4 LTS (Xenial Xerus)"14:20
seb128VERSION_ID="16.04"14:20
seb128seems not14:20
didrocksseb128: thx for confirming :)14:20
seb128np!14:20
oSoMoNnot sure if there's a way of getting snap usage statistics per architecture?14:20
didrocksand for having old vm/install around ;)14:21
willcookekenvandine, ah - flavours do have, e.g. mate - but not sure if that means we are obliged or not.14:21
seb128:)14:21
seb128willcooke, do they care enough to contribute some work to help with it?14:22
seb128I would say it depends of the cost14:22
seb128but it's not trivial and just to help a flavor so I think it's fair to ask for shared cost14:22
oSoMoNthe problem is not so much an extra maintenance cost (armhf is unsupported upstream so it requires a patch to keep building every now and then, but it's not too bad), but rather that it takes forever to build chromium on arm (typically more than 20 hours)14:22
willcookeoSoMoN, what's the knock on effect of slow building?  Like, can you kick it off and just leave it until its done?14:24
willcookeor do you need to keep checking back on it14:24
oSoMoNwillcooke, yeah, I can totally kick it off and wait for it to be done, it's just very asynchronous14:27
willcookeoSoMoN, My 0.02USD, if it's not a big burden to keep it building, it will be useful to people, but when that changes I think we reconsider14:27
oSoMoNit's not even something that I can kick off before going to bed and expect to have it done when I wake up, so it makes for very slow iterations, but I can live with it14:27
willcookeoSoMoN, now, if we can get the snap building, then we can get an idea of user numbers14:28
willcookeand then that makes a case for or against pretty easy14:28
oSoMoNwillcooke, do we get snap stats per architecture?14:28
willcookeoh14:28
willcookehm14:28
willcookeI just assumed we did14:28
willcookeupload it as a new snap if not ;)14:29
oSoMoNI couldn't figure out how to extract that info from the dashboard, but maybe there's a way14:29
didrocksyou can: https://dashboard.snapcraft.io/docs/api/snap.html#fetch-metrics-for-snaps14:29
didrocksbut you have to poke the API manually14:29
oSoMoNaha14:30
didrocksand you need to request per arch14:30
oSoMoNperfect, thanks didrocks14:30
didrocksoSoMoN: hold on, actually, I'm unsure, as I thought you can filter by revision, and getting the revision from another API call…14:32
didrocks(for each arch)14:32
kenvandineLaney, i have a question about seeded snaps in my desktop-preinstalled image14:32
didrocksbut it seems it doesn't return the revision :/14:32
didrocks(on the metrics filter)14:32
kenvandineLaney, the snaps are in /var/lib/snapd/seed/snaps/14:33
kenvandinebut snap list doesn't show them as installed14:33
Laneydo you get the seed.yaml?14:33
Laneylivecd-rootfs makes that14:33
kenvandineyes14:33
kenvandineand it has them14:33
Laneythen snapd is supposed to do a thing to install them14:33
Laneyon the first boot14:33
didrocksoSoMoN: the only way would be to hardcode to arch in the "Version" field from what I see :/14:34
kenvandinehttp://paste.ubuntu.com/p/NN4Ph3hhr5/14:34
kenvandineLaney, snapd is running :(14:34
kenvandinenothing in /snap14:34
LaneyI don't know how that part is supposed to work, sorry14:35
Laneytry #snappy ?14:35
kenvandineLaney, no worries14:35
kenvandinewill do14:35
Laneyas far as I was concerned our job ended at writing seed.yaml, snapd is supposed to arrange for that to work14:36
kenvandineoh... snapd.seeded service was still marked as "starting"14:38
kenvandinesince yesterday14:38
Laney/o\14:53
NafalloLaney: followed your advice and built them in an lxd guest. debs for amd64 if someone needs to test on bionic is at http://people.ubuntu.com/~nafallo/gdm3/15:06
Laneyphat15:11
TrevinhoLaney: as per reading the email / telegram, it might be already too late, but to get better gdm development experience, and want to run it from jhbuild, I can help...15:44
Laneyoh yeah, show me IRL?15:45
LaneyI was using a testing laptop and making debs15:45
Laneyor ... write a wiki page :-o15:45
TrevinhoLaney: I generally do such things in a VM (things which supprot mesa drivers, as vmware or maybe kvm are better)15:45
Trevinhobut yeah... I wanted to write down something15:46
Laneynice15:47
Trevinhoanyway the basic trick is just create an alternative .service file, stripped of the pre/post-exec and the dbusbusname15:47
Trevinhothen setting the exect to a script that sets proper PATH, LD_LIBRARY_PATH and GI_TYPELIB_PATH and runs the actual gdm... Then you've to symlink something too15:47
Laneyhaha15:48
Laneyeasy!15:48
Trevinhobut that's mostly for xorg stuff as you might not want to run it from your jhbuild (i did, but still)...15:48
Trevinhowell it's quite easy, just need to check a bit the logs for erros, but eventually it works (once you use upstream gdm with my patches)15:48
LaneyI never even ran the shell from jhbuild15:48
Trevinhonoooo?!15:49
TrevinhoHow can you!15:49
Laneysymlink the binaries to /usr/bin15:49
Trevinho-_-15:49
Trevinhoyou can also run funny stuff such as `jhbuild run gdb gnome-shell -batch -ex "set logging on" -ex "run --replace" -ex "bt full" -ex "call gjs_dumpstack()"   -ex quit15:49
Trevinho`15:49
Trevinhowhich bsicallty gives you the traces (gjs and c) when it crashes15:50
* oSoMoN is done for today, have a good evening16:23
=== pstolowski is now known as pstolowski|afk
willcookenight17:50
ricotzchrisccoulson, please push your packaging branches!18:15

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!