/srv/irclogs.ubuntu.com/2019/10/30/#ubuntu-desktop.txt

hellsworthricotz, yep it looks like gnome-calculator-master builds with vala 0.40.8 since that's what the valac deb is in 18.0402:47
hellsworthwe're working some kinks out with the extensions newer than gnome-3-28. Once gnome-3-34 extension is ready to be widely used, gnome-calculator will build with 0.4602:50
hellsworthmarcustomlinson: kenvandine , i built the build snap with the invalid-free-crash branch (to test https://gitlab.freedesktop.org/cairo/cairo/issues/382#note_278108), put it in edge, rebuilt the platform snap, put it in stable, and rebuilt evince. it still segfaults when it opens a pdf.03:11
=== _thumper_ is now known as thumper
hellsworthmarcustomlinson: kenvandine the build snap in edge and platform snap in stable were built with the invalid-free-crash. i'm going to leave these in the store and it would be interesting if one of you could test evince or drawing or something with them to see if you still see a segfault03:46
hellsworthafterwards, we can build the build snap with cairo source-tag: 1.16.0, and platform snap, and put those in the store03:47
hellsworthi didn't comment on https://gitlab.freedesktop.org/cairo/cairo/issues/382#note_278108 because i'm having an issue registering an account on gitlab.freedesktop.org03:52
marcustomlinsonhellsworth: thanks, I’ll give it a try and leave a comment on the bug when I wake up in a couple hours :P06:06
oSoMoNgood morning desktoppers!07:18
dufluMorning oSoMoN (and marcustomlinson and didrocks?)07:24
oSoMoNhey duflu07:25
oSoMoNsalut didrocks, hey marcustomlinson07:26
didrocksgood morning duflu, oSoMoN, marcustomlinson!07:32
didrocksforgot to press "enter" on my good morning ;)07:33
RikMillsmorning07:46
RikMillsah, wrong channel but still valid :P07:47
didrocksgood morning RikMills07:50
=== pstolowski|afk is now known as pstolowski
marcustomlinsonmorning oSoMoN duflu didrocks08:08
seb128gooood morning desktopers08:09
marcustomlinsonhey seb12808:11
seb128hey marcustomlinson, how are you today? you are online earlier than usual?08:11
dufluHi seb128, how goes?08:11
seb128hey duflu, good! and you?08:11
seb128I see you are quite busy with bug triage recently08:12
marcustomlinsonseb128: I'm alright thanks, was eager to test something :) Hw you?08:12
dufluseb128, OK... Yeah, I'm trying to estimate if it's a busy release or not08:12
seb128seems low traffice/no much high profile bugs atm from what I saw08:12
seb128do you have the same impression?08:12
dufluseb128, yes for sure about the lack of showstoppers08:14
dufluBugs like the HDMI audio switching are among the busiest08:14
dufluHmm, there's a simple way to check and I haven't yet08:15
seb128that one is annoying, I wonder if we should just put back the code to ignore HDMI in the dynamic switcher08:15
didrockshey seb12808:15
seb128lut didrocks, en forme ?08:15
dufluseb128, I was considering reverting the commit that upstream say caused it, but currently am more likely to wait for them to decide08:16
didrocksseb128: ça va, et toi ?08:16
seb128duflu, right08:18
dufluseb128, Ha! Yes my guesses were right - https://bugs.launchpad.net/ubuntu/eoan/+bugs?orderby=-heat&start=008:18
seb128didrocks, ça va !08:18
oSoMoNsalut seb12808:45
seb128oSoMoN, lut, en forme?08:46
oSoMoNseb128, ça va, et toi?08:47
seb128oSoMoN, ça va bien !08:47
didrockssil2100: RAOF: hey! Just confirmed the 2 use cases are fixed for bug #1850052 (new and previous install) with zsys in -proposed. As previous version prevents people from log in, what do you think about bypassing the retention period?09:01
ubot5bug 1850052 in zsys (Ubuntu Eoan) "installing zsys breaks desktop login" [High,Triaged] https://launchpad.net/bugs/185005209:01
Laneysup09:03
RAOFdidrocks: 20:00 here, so I'll let sil2100 take that!09:03
didrocksRAOF: no worry, have a good evening :)09:04
didrockshey Laney09:04
seb128hey Laney09:04
didrocksRAOF: and thanks for acking it in -propsoed09:04
RAOFI'm always happy to deal with SRU pings!09:05
didrocks:)09:05
sil2100RAOF: o/09:08
sil2100didrocks: hey, oh! Ok, let me take a look09:08
didrocksthx ;)09:09
didrockssil2100: thanks a bunch! :)09:20
sil2100didrocks: yw!09:22
dufluAnd I killed gitlab09:23
dufluor something did09:23
marcustomlinsonsil2100: now that you're here ;) could I please ask you to have a look at libreoffice in the disco queue?09:33
sil2100marcustomlinson: hey! Will try to get to that, but since today is not my SRU shift I might not be able to find enough cycles10:07
jibelLaney, how did you update the translations of ubiquity to add new strings?10:21
LaneyI forgot already10:23
Laneydebconf-updatepo was it?10:23
jibelit does something, not sure it's the right thing10:24
didrocksI guess that was it, indeed10:25
WimpressMonring desktoppers o/10:44
seb128hey Wimpress, how are you today?10:45
didrockshey Wimpress10:47
tseliotLaney, seb128 can we have somebody with systemd knowledge have a look at LP: #1845801, please? (see my comment in the bug report)11:43
ubot5Launchpad bug 1845801 in gnome-shell (Ubuntu Eoan) "[nvidia] Automatic login fails and then all subsequent logins fail. Killing gnome-session-binary fixes it, or just not using automatic login." [Undecided,Confirmed] https://launchpad.net/bugs/184580111:43
Laneytseliot: ah, wtf, good find11:53
seb128tseliot, thx12:06
tseliotyw12:06
Laneyrbalint: don't suppose you have any idea about the above?12:09
LaneyI don't know what paused for a drm device node means12:10
rbalintLaney, i took a look but it seems nvidia driver related12:12
Laney:/12:12
* Laney wibbles12:12
rbalintLaney, not that i don't have other problems with my nvidia card, but they are not related :-)12:13
Laneyright12:16
Laneynot entirely sure how best to move forward12:16
rbalintLaney, suggested trying systemd 243 from a ppa i already have, but to triage i need to assemble a system from the nvidia card - or fix gpu passthrough which freezes my systemd atm (bug to be filed)12:29
Laney:O12:30
Laneyxnox had that working at the release sprint12:30
Laneythanks for the suggestion there12:30
rbalintLaney, what worked? gpu passthrough?12:31
Laneyyeah12:31
rbalintLaney, in my case it is external tb3 gpu case on bionic host and disco guest12:31
rbalintLaney, i imagine xnox had it working on eoan host12:32
Laneyindeed12:33
rbalintLaney, and maybe without tb212:33
rbalinttb312:33
Laneyit was prime iirc12:33
rbalintLaney, that's also cool, but a different way12:35
Laneynod12:36
Wimpressseb128 didrocks Head down trying to ctach up on stuff today. Sorry for not replying ealier.12:53
didrocksnw!13:01
seb128Wimpress, no worry, busy is good :)13:03
kenvandinemarcustomlinson: i see the same weirdness in the headerbar when using cairo 1.17 with that crash fix13:23
kenvandinemarcustomlinson: in drawing13:23
kenvandinedefinately better in 1.1613:23
marcustomlinsonkenvandine: alright, so that's definitely a separate issue then13:23
kenvandineyeah13:23
marcustomlinsonthanks13:23
kenvandinethat PR does fix the crash though13:24
kenvandinemarcustomlinson: i also don't see the window controls13:24
marcustomlinsonkenvandine: perhaps we call off upgrading to 1.17 until possibly the next sdk pair13:30
kenvandineyeah13:30
kenvandinemarcustomlinson: i suspect nobody has updated to that yet :)13:30
kenvandinethis bug seems to effect everything!13:30
kenvandinethe crasher that is... and the PR with the fix has been sitting for 8 months13:31
marcustomlinsonyeah that's why I thought that guy could use the help with some confirmation from us13:31
kenvandineyup13:32
marcustomlinsonkenvandine: it's all bryce's fault ;)13:33
kenvandine:)13:33
kenvandinein this case it is :)13:33
kenvandinemarcustomlinson: and looking at that PR, I bet this is the cause for those weird issues in the headerbar13:35
kenvandinerelated to clipping boxes13:36
kenvandineand what i saw on the headerbar looks like boxes that overrun a bit13:37
marcustomlinsonkenvandine: ah so probably not unrelated, I'd not looked at the PR admittedly13:37
kenvandinemarcustomlinson: yeah, it's a code path that wasn't being handled before13:38
kenvandineso he fixed that13:38
kenvandinebut something's not right13:38
marcustomlinsonwell I brought it up, we'll see if he has something to say about it13:39
kenvandineclobrano: hey, i'm seeing a theme issue in the drawing snap.  Specifically the headerbar when using yaru.  yaru-dark and yaru-light are both fine13:39
kenvandinemarcustomlinson: thanks13:39
kenvandineclobrano: https://usercontent.irccloud-cdn.com/file/7MHfIVC8/drawing_theme_issue.png13:39
kenvandineclobrano: the drawing snap is being built with the latest gtk and other bits13:40
kenvandineclobrano: if you look really close at that screenshot there is a subtle gear in the background of the headerbar towards the right hand side13:40
Laneythat bryce bug is still there?????13:50
Laneyis something actually wanting 1.17?13:52
Laneythe archive still doesn't have it13:52
marcustomlinsonLaney: nobody's requesting it no, we were just giving it a go in the gnome platform snap13:56
Laneynod13:57
LaneyI think that's why we never bothered upgrading13:57
jibeldidrocks, when you have some time later this week https://code.launchpad.net/~jibel/ubiquity/+git/ubiquity-1/+merge/37492014:39
jibelit's the new design for advanced features14:39
didrocksjibel: thx! I'll have a look (probably tomorrow)14:47
hellsworthmorning folks15:33
hellsworthmarcustomlinson: what is the state of the build and platform snaps in the store?15:33
hellsworthwere they built with the invalid-free-crash or 1.16?15:34
hellsworthoh i can go look up the commits. nvm15:34
marcustomlinsonhellsworth: for 3-32?15:34
hellsworthoh i meant for 3-3415:34
marcustomlinsonhellsworth: I'm not sure the state of that15:35
hellsworthmarcustomlinson: it looks like gnome-3-34-1804-sdk in edge is built with cairo source-tag: 1.16.0 (based on the commit)15:35
hellsworthcan't tell on the platform snap though15:35
hellsworthi think the evince crash is different from the drawing/epiphany crash15:36
hellsworthmarcustomlinson: is it worth it for me to build/push the invalid-free-crash to the store to test epiphany? it looks like you and kenvandine have already come to the same conclusion..15:37
marcustomlinsonhellsworth: no don't bother, we've investigated enough there15:38
marcustomlinsonthanks15:38
oSoMoNgood morning hellsworth15:38
hellsworthok that's what i thought15:38
hellsworthgood morning oSoMoN !15:38
marcustomlinsonwe're gonna hold off upgrading to 1.17 until possibly the next round of sdk snaps15:38
hellsworthyes that sounds like a solid plan15:38
hellsworthi'll assume that both gnome-3-34-1804-sdk and gnome-3-34-1804 in the store are built with 1.16, and retry evince in case it's related with cairo. honestly, evince could be crashing due to any of the build snap updates.15:40
Laneyforget 1.17 imho15:44
seb128Laney, do you know offhand of an easy wait to make gsd displays g_debug entries? hacking the systemd unit to add --debug?15:51
seb128or is there an easier way?15:51
Laneyif it were me I'd do15:52
Laneysystemctl --user edit gsd-whatever.service15:52
Laney[Service]15:52
LaneyEnvirionment=G_MESSAGES_DEBUG=all15:52
Laney(with correct spelling)15:52
Laney(--runtime if you don't want it to persist)15:53
Laneythen restart the unit15:53
seb128Laney, thx, I didn't know about systemctl edit, handy :)15:55
Laneyyeah, no need to hack units in /usr most of the time15:56
Laneyif you do systemctl --user edit --full whatever.unit then you get the whole thing to edit too15:56
seb128right, I'm reading about it now, it surprised first because I was expecting edit to give a copy you can edit :)15:57
Laney:>15:58
Laneyit's like an overlay15:58
Laneysystemd calls them drop-ins15:58
* Laney frowns16:54
LaneyI had an old todo to fix up a merge request16:55
Laneyand now I can't remember what I wanted to fix :(16:55
marcustomlinsonfix it all!16:59
didrocksunless you didn't even note what MP to fix… ;)17:01
Laneyheh17:01
Laneynah I have the commits I wrote back then17:01
Laneybut now they look finished to me ...17:01
jdstrandkenvandine: if you recall I was asking about theming issue with firefox on wayland after eoan upgrade. it wasn't wayland specific. firefox forgot Custom/Title Bar for that profile. so feel free to forget this whole conversation :)17:34
kenvandinejdstrand: ok :)17:40
hellsworthkenvandine: glimpse snap support was just merged! so should i put this in the store under my name and then we move it to being from ubuntu-desktop?17:41
kenvandinehellsworth: sure17:41
kenvandinehellsworth: even better would be if upstream published it17:41
hellsworthhmm ok then i'll see about coaching them on that17:43
kenvandinepopey could probably help17:44
kenvandinehellsworth: since the yaml is upstream they could use build.snapcraft.io17:44
* hellsworth looks for steps on the docs about build.snapcraft.io17:46
popeyIs it on GitHub?17:46
popeyGlimpse that is17:47
hellsworthyes17:47
hellsworthhttps://github.com/glimpse-editor/Glimpse/blob/dev-g210/snap/snapcraft.yaml17:47
popeyNeat17:47
popeySomeone from the upstream team should sign up for a store account and then goto build to hook it all up17:48
popeydiddledan: ^17:49
hellsworthupstream doesn't want to maintain the snap17:51
hellsworththey are happy to carry the snap/snapcraft.yaml though17:51
kenvandineok, then probably snapcrafters17:52
kenvandinepopey ^^ right?17:52
hellsworthi'm checking to see if the contributor of the snap packaging would be willing to submit it to the store17:52
hellsworthif not then we can go with snapcrafters. what is the proces for that?17:52
hellsworthalso if this convo should be moed to #snap-advocacy, we can do that..17:53
diddledanI imagine the glimpse yaml is largely the same as gimp's which I maintain18:08
hellsworthyep18:09
hellsworthminor differences18:09
popeyhellsworth: yes, we can put it under snapcrafters (like gimp is)18:26
hellsworthpopey: should we encourage the contributor to submit it to snapcrafters?18:39
popeyWell, it's a bit tricky now the yaml is in the upstream project18:39
popeycan we discuss this in the desktop call on monday maybe?18:40
hellsworthyeah i thought the upstream project was willing to support the snap packaging going forward18:40
hellsworththere's a desktop call on monday?18:40
popeyWe might be able to do it with a travis.yml18:40
popeythere was between snap advocacy and will / ken.18:40
popeymaybe it needs to change now the structure has changed a bit18:41
=== pstolowski is now known as pstolowski|afk
popeythere's another couple of snaps which don't have out-of-tree snapcraft.yamls which I need to work on, may be able to re-use that work for glimpse18:47
hellsworthsure. what would the result of that work be? moving the glimpse upstream snap/ to snapcrafters?18:49
popeyno, a travis.yaml in a repo in snapcrafters which periodically grabs the glimpse repo and then runs snapcraft on it18:56
popeyeffectively18:56
popeythe downside is if we need to fix the snapcraft.yaml we have to wait for upstream to land those fixes18:56
popeywhereas traditional out of tree snapcraft yamls we can ninja (and often do) in snapcrafters repo18:57
popeybut i reckon we can cope18:57
=== ondra_ is now known as ondra
popeywe can from within travis, use the snapcraft remote-build option to fire it over to launchpad to build18:58
popeytravis would just be used to manage the job18:58
hellsworthpopey: if we were to move the glimpse snap packaging info over to snapcrafters (and remove it from upstream source), would a new snap be automatically built on upstream master commit?19:05
popeyhm, we can make that happen19:06
hellsworthit sounds like if we leave the snap packaging in upstream glimpse, and add a travis.yaml to snapcrafters, then that just periodically builds/pushes a new snap to the store - not based on upstream commit, but rather more of a cron job19:07
hellsworthwhich is better?19:07
hellsworththe glimpse team is pretty fine with whatever we think is best19:07
popeydepends who'd use the edge channel19:07
popeylikely not many people19:07
popeymost people tend to only care about stable releases19:07
popeyits most useful to just build whenever upstream do a release19:09
popeywhich can be tested in edge, and then migrated to beta/stable as appropriate19:09
hellsworthok glimpse upstream does want to keep snap in the glimpse project, review snapcraft.yaml prs, and maybe maintain the snap in the future. that is nice :)19:20
hellsworthpopey: ^19:20
popeynice one19:21
hellsworthbut they like the idea of snapcrafters having a yaml that will periodically build the snap and push it to the edge channel in teh store19:21
popeyok, so i/we need to make a simple travis.yml to do the builds19:22
hellsworthpopey: is there anything i can do with helping out with that?19:22
popeyif you have some time, that'd be awesome19:22
popeyI'll ping you a mail with the details, okay?19:22
hellsworthi have never worked with travis.yaml before so yeah any details you can give me would help19:22
hellsworthsounds good. thanks so much :)19:22
hellsworthway to advocate19:22
hellsworthok i'm off for the dentist.. I'll be back (in my best terminator voice)19:25
popeyoooh I have a better idea19:26
popeyglad i typed this mail up, now I can throw it away and give you a better idea :D19:27
sarnoldpity your rubber ducky required a whole email :)19:31
popeyya19:38
popeyhellsworth: you have mail19:38
popey</aol>19:38
diddledanI maintain and build a synchronised kernel from the Microsoft WSL2 repo on github which is a complete mirror that I've augmented by adding my own .github/workflows/* action definitions to. maybe that would be useful for making an in-tree snapcraft.yaml work?20:47
diddledanit resynchronises every hour with upstream changes which then get rebuilt if a new tag appears - the build service would take over the equivalent of the build step, but the sync step would be useful20:48
diddledanhere's the workflow/action: https://github.com/diddlesnaps/WSL2-Linux-Kernel/blob/master/.github/workflows/sync-upstream.yml20:49
hellsworththanks popey!21:35
hellsworthdiddledan: no idea if yoru workflow could help but it's good to know aobut in case so thanks for the link :)21:35

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