/srv/irclogs.ubuntu.com/2018/03/28/#ubuntu-desktop.txt

RAOFI gotta say, I'm a *big* fan of the way Mutter “application not responding” dialogs sitting on the top of a background window eat all mouse clicks for the foreground application.05:14
jameshis that wayland or X?05:22
RAOFX05:22
=== maclin1 is now known as maclin
dufluRAOF: Sounds like the foreground window has reverted to being a non-reactive actor. So clicks pass through it. We've fixed one similar bug already and have others unresolved06:18
RAOFduflu: Entirely possible.06:19
didrocksgood morning06:30
jibelSalut didrocks06:30
didrockssalut jibel !06:31
dufluMorning didrocks, jibel06:32
jibeldidrocks, I'm reviewing jbicha's branch Ubiquity, then I think we can do a release that include the new page and the telemetry stuff06:32
jibelbranch of*06:32
jibelAfternoon duflu06:32
dufluCorrect06:32
didrockshey duflu06:39
didrocksjibel: excellent! is the FFe approved?06:39
oSoMoNgood morning desktoppers06:40
oSoMoNhey duflu, jibel, didrocks06:41
didrockssalut oSoMoN06:41
dufluHi oSoMoN06:41
didrocksseems not, but the release team hasn't been suscribed it seems06:42
jibelBonjour oSoMoN06:43
didrocksjibel: added info and paperwork on bug #175545606:47
ubot5bug 1755456 in ubiquity (Ubuntu) "[ffe] Optional recording/sending of installer&system details to help improving Ubuntu" [Undecided,New] https://launchpad.net/bugs/175545606:47
jibeldidrocks, thanks06:50
jibeldidrocks, this bug needs more documentation according to https://wiki.ubuntu.com/FreezeExceptionProcess#FeatureFreeze_for_new_upstream_versions06:54
=== pstolowski|afk is now known as pstolowski
dufluAh crap. Instead of profiling my memory diff I had a look at the memory total for gnome-shell. And 70% of it is from PulseAudio (!?)07:14
dufluUntil you use gnome-shell for a bit. Then the graphics memory usage becomes the bigger issue07:16
didrocksjibel: seb128 filed it, I think it's how he was doing when it's not a new upstream (as another upstream than us) version, let's see with him07:17
mptjbicha, done07:43
duflukoza, got the meeting notification 1 hour early... regardless, are we meeting?07:51
willcookeduflu, koza - anything to talk Bluetooth today?  Do we need to meet?08:00
willcookealso morning08:00
kozaduflu, oh, another change of time08:00
didrocksmorning willcooke08:00
kozawillcooke: we could skip08:00
kozawillcooke: only thing i have is that the bloody speaker has arrived :)08:00
kozaand has some issues08:01
willcookekoza, you think it's that speaker?08:01
kozanoidea yet08:01
willcookekk08:01
willcookeI dont have anything08:01
kozalets skip then08:01
* willcooke gets on with the slide show08:01
dufluAnd good morning willcooke, koza08:02
Laneyyo08:03
didrockshey hey Laney08:03
duflusup Laney08:04
willcookeallo allo08:04
seb128good morning desktopers08:10
willcookehi seb12808:10
=== pstolowski is now known as pstolowski|denti
didrockshey seb12808:10
seb128willcooke, duflu, koza, sorry got delayed ... did you do the bluetooth meeting? is it over? anything interesting?08:10
seb128hey willcooke didrocks08:10
dufluseb128, skipped it08:11
seb128good, I don't think there is much bluetooth changes/work ongoing atm08:11
dufluAlthough it should be 50 minutes from now in theory08:11
seb128we are pretty much settled for bionic08:11
=== pstolowski|denti is now known as pstolowski|bbl
seb128DST fun right?08:11
willcookeoSoMoN, https://chromium-review.googlesource.com/c/chromium/src/+/53229408:14
willcookeoSoMoN, Am I reading that right - did they merge it at last?08:14
oSoMoNgood morning seb12808:17
dufluI think the status is still merge conflict08:17
oSoMoNwillcooke, not quite, they didn't merge it yet08:18
oSoMoNwillcooke, it's an updated patch set, but what's really interesting is that it comes from a google engineer08:18
oSoMoNwhereas the original patch was by someone from intel08:18
oSoMoNso it looks like google is showing some interest at last08:18
dufluHurrah08:18
dufluwillcooke, worlds collide: bug 175949708:19
ubot5bug 1759497 in PulseAudio "80% (512MB) of gnome-shell's memory mappings at start up are due to PulseAudio" [Medium,Confirmed] https://launchpad.net/bugs/175949708:19
seb128lut oSoMoN08:19
seb128duflu, that seems excessive indeed08:20
willcookeduflu, whaaa?!?!  odd08:21
dufluIt could just be address space and not real memory08:22
dufluBut still, lots of address space08:22
willcookeCould this be related to us running pulse under GDM as well for a11y?  Oh, no hold on, I think I've got that backwards, we *don't* start PA & BlueZ in GDM.08:23
seb128no, that's part of the gnome-shell user process mapping08:25
seb128if I understand correctly08:25
dufluwillcooke, gnome-shell uses libcanberra for sound(?) which uses pulseaudio08:25
dufluIt's an upstream thing, hardcoded in there08:25
dufluOn a related note, what is goa-daemon and why does it use 96GB of address space?08:27
seb128welcome to webkit08:29
seb128it's the daemon handling online accounts08:29
seb128duflu, webkitgtk upstream said that when we were discussing the webkit/ulimit issue impacting deja-dup08:31
seb128<mcatanzaro>seb128: Just address space08:31
seb128<mcatanzaro>It's not actually just a spectre mitigation, it's also a general security feature08:31
seb128<mcatanzaro>Looks like it's allocating 144 GiB of address space. If I've done my division right.08:31
willcookegetting faster as this now: https://imgur.com/a/BXgpf08:32
dufluThat's wacky, but potentially safe if there's no real memory needed behind most of it08:33
seb128right08:34
dufluEven Chrome only needs 1.2GB08:35
dufluper process :P08:35
duflu_My DSL is broken. Oddly at 5pm two days in a row09:13
duflu_Did I miss anything?09:13
=== duflu_ is now known as duflu
willcookeduflu, nothing09:17
=== pstolowski|bbl is now known as pstolowski
seb128duflu, should bug #1748450 be marked as invalid for gnome-shell since that's an xwayland issue?10:22
ubot5bug 1748450 in gnome-shell (Ubuntu) "gnome-shell crashed with SIGTRAP in _g_log_abort() from g_log_default_handler() from default_log_handler(message="Connection to xwayland lost") from g_logv() from g_log() from <bug 1505409>" [High,Triaged] https://launchpad.net/bugs/174845010:22
dufluseb128, no. Xwayland and gnome-shell are co-dependent so when one crashes the other does too. It's not always easy to tell who crashed first10:23
dufluthat's what the apport fix will help with10:23
seb128oh ok, I though that it was clear from the error that gnome-shell was hitting the sigtrap because it losts the connection to xwayland10:23
seb128which meant xwayland went away first10:24
dufluseb128, but with built-in crash handling you could bounce between each more than once. So it's not obvious who crashed first, or how many times10:25
dufluPerhaps Xwayland dies because gnome-shell was in the process of dying already10:25
seb128k, I see10:25
seb128let's see if we get better reports now that the apport changes landed then10:25
seb128well, if gnome-shell was already down, would it still sigtrap due to xwayland going away?10:26
seb128anyway let's hope we get better reports that tell us more10:26
dufluseb128, yes a process can crash multiple times in its lifetime if it caught its own the first time10:26
dufluI think we might need to check and remove any self-crash handling10:27
dufluIt's always a dumb idea to have such things10:27
seb128right10:27
dufluXorg is the bigger offender10:27
dufluAnd then there's the other big problem of X errors being delayed and you never know where they originated10:28
darkxstduflu, you can obviously force xorg errors to be synchronous, but that doesnt work in a apport world etc10:32
dufludarkxst, yeah. A good system is one that is always-on in production10:32
dufluAlthough I don't think the origin of any X errors should matter in this case. We should always expect X errors, and we should be able to survive them10:36
* duflu runs away10:38
darkxstduflu, just to be clear this is the greeter (gdm) session crashing?10:38
seb128he left, and I think we don't know, it's bug #174845010:43
ubot5bug 1748450 in gnome-shell (Ubuntu) "gnome-shell crashed with SIGTRAP in _g_log_abort() from g_log_default_handler() from default_log_handler(message="Connection to xwayland lost") from g_logv() from g_log() from <bug 1505409>" [High,Triaged] https://launchpad.net/bugs/174845010:43
seb128GunnarHj, hey11:08
GunnarHjHi seb12811:08
seb128GunnarHj, I rechanged e-d-s/did an upload, that worked, .po are being imported11:08
seb128GunnarHj, see https://translations.launchpad.net/ubuntu/bionic/+source/evolution-data-server/+imports11:08
seb128GunnarHj, also "make template.update-pot" is not a strandard target in the inltool world afaik11:09
GunnarHjseb128: Nice. Hope that the upstream link does not get re-enabled again..11:09
seb128yeah, I don't know why/how that happened11:10
seb128are you sure your changed worked? like launchpad didn't timeout or something?11:10
GunnarHjseb128: That has nothing to do with intltool. It's the command used when intltool is not made use of.11:11
seb128GunnarHj, sorry that I didn't comment on bug #1751820 earlier, I was on vac some days and things have been crazy here otherwise, I'm doing that now11:11
ubot5bug 1751820 in pkgbinarymangler (Ubuntu) "Teach dh_translations how to deal with meson" [Medium,In progress] https://launchpad.net/bugs/175182011:11
seb128GunnarHj, I'm not sure I understand the question then11:11
seb128GunnarHj, what I meant is11:12
seb128 LC_ALL=C make -C eog-3.18.2/po eog.pot-update11:12
seb128make: Entering directory '/tmp/eog-3.18.2/po'11:12
seb128make: *** No rule to make target 'eog.pot-update'.  Stop.11:12
seb128$ LC_ALL=C make -C eog-3.18.2/po eog.pot11:13
seb128make: Entering directory '/tmp/eog-3.18.2/po'11:13
seb128INTLTOOL_EXTRACT="/usr/bin/intltool-extract" XGETTEXT="/usr/bin/xgettext" srcdir=. /usr/bin/intltool-update --gettext-package eog --pot11:13
darkxstseb128, yeh that is gdm again11:13
seb128GunnarHj, was your question specific to meson projetcs?11:13
seb128projects11:13
darkxstseb128, aapart from the apport noise are there any real reports of issues?11:14
seb128darkxst, not that I know but often we don't get user feedback about segfaults out of apport/whoopsie reports11:14
GunnarHjseb128: No, my question was specific to good old build system which makes use of po/Makefile instead of intltool. But it looks like I'd better test before talking more about that topic. ;)11:15
darkxstseb128, that has been ongoing for a few cycles before you guys switched back to GNOME, I never managed to find the cause though11:18
seb128:/11:19
seb128I don't even understand why we need xwayland under gdm11:19
darkxstmutter is entirely dependant on Xwayland, upstream were working to decouple that but that work seems to have stagnated11:20
seb128k :/11:23
darkxstI have next week off work, so might try and look into a bit more, if there are any other nasty bugs let me know11:24
seb128GunnarHj, commented on that dh_translations bug, I don't know if we can figure out how to get the domain from meson introspect in a robust way, if we can't maybe we should add a --domain to dh_translations and just specify the domain by hand in the rules for those projects, would be cleaner that the overrides copied around11:24
seb128darkxst, http://reqorts.qa.ubuntu.com/reports/rls-mgr/rls-bb-incoming-bug-tasks.html and http://reqorts.qa.ubuntu.com/reports/rls-mgr/rls-bb-tracking-bug-tasks.html are lists worth looking at if you are interested in flagged issues11:25
GunnarHjseb128: Yeah... The other problem is figuring out the build dir. I suppose that also that would need to be passed to dh_translations.11:26
GunnarHjseb128: But it feels like a post release thing.11:26
seb128GunnarHj, I don't think we would have an issue adding that --domain option before release, it's not changing an existing behaviour11:27
seb128then we probably don't need to replace the hacks now11:27
seb128next time we update those packages/merge11:27
seb128GunnarHj, the builddir we could use the obj-arch by default and override manually if needed yes11:28
darkxstseb128, hmm I never seen those before, are they new(ish)11:28
GunnarHjseb128: But dh_translations is a perl script. Does it understand that variable?11:28
seb128darkxst, no, some years old at least11:29
seb128GunnarHj, well, I'm sure there is a way to query the arch tripplet from perl :)11:29
GunnarHjseb128: 15 years ago I might have known how. Now I don't. :(11:30
seb128GunnarHj, well, I'm not saying that you are the one that need to do that work11:33
seb128I'm going to try to help having a look11:33
seb128adding the --domain would be a first step and probably easier11:33
GunnarHjseb128: I'll be happy to help if I can, but you or somebody else needs to help with the approach.11:35
seb128k11:36
Mirvjbicha: right, I guess I could upload 0.28.1 as well with lighter testing, as the only other 0.28.1 changes are two translation updates and adding "static" to one function11:37
jbichaMirv: please :)11:37
Mirvwill do11:38
darkxstseb128, some years is indistinct, keeping I mind I didnt have internet for 18months of the last 2 years!11:38
seb128darkxst, I don't remember but my emails suggest it was started in 201111:41
Laneythis glib test failure sucks11:45
LaneyAND I forgot my lunch at home11:46
seb128:/11:51
darkxsthmm i had internet in 201111:51
willcookeI'm trying to build GNOME Software with a patch I got from Robert for the slideshow.  I've git cloned it, applied the patch, now... how the heck do I build this thing?11:55
willcookeah, meson11:56
jibeljbicha, Hi, can you update lp:~jbicha/ubiquity/update-updates-page following latest changes from m_pt? also there are periods missing at the end of the label of the radio buttons to perfectly match the spec.11:56
jibelother than that it's ready to merge11:57
jbichajibel: yes, I'm working on that now11:57
seb128willcooke, meson builddir && cd builddir && ninja ?11:57
jibeljbicha, thanks11:57
willcookeseb128, ta, I'll try that.  Looks like I need to build in 180411:58
seb128willcooke, you are trying to build a binary or a package?11:58
willcookemy bloody 1804 machine keeps going to sleep11:58
willcookeseb128, just a binary will be fine for now11:58
seb128willcooke, I would rather start from the source package and add the patch11:59
seb128k, then meson might be what you need11:59
willcookeI'll try the source package approach, that might be easier in the long run11:59
seb128willcooke, https://paste.ubuntu.com/p/DjSrN98Y4r/ is what the package does, if you want to build with the same options12:01
seb128you might set some plugin to false for local testing12:01
xnoxI like the new picker!12:03
xnoxlooks awesome, and nice icons12:04
willcookethanks seb12812:06
seb128np12:06
seb128xnox, picker?12:06
jbichaacheronuk: ping12:07
xnoxseb128, the whatever that "google-chrome" shows when one tries to attach file to a gmail e.g. "open/browse a file" dialog12:07
acheronukjbicha: pong12:07
seb128xnox, ah, that sounds like a chrome thing :)12:07
xnoxbut it's a gnome thing12:07
xnoxlet me get a screenshot, or trigger it elsewhere12:08
jbichaacheronuk: can I change your "Minimal Install" to "Minimal installation" to match the updated gtk string?12:08
willcookeseb128, what's that metapackage that brings in all the deb tools, like fakeroot etc?12:09
xnoxseb128, Super -> search for evince; launch; Ctrl-O -> see the dialog that pops up with loads of small icons column, then subdirs as a sidebar12:09
xnoxoooh, and it has some orange highlits12:09
jbichaI am rebasing the ubiquity merge proposal so it will have fewer commits and changes to be easier to read12:09
acheronukjbicha: please do. I prefer that, but decided not to change while it was being debated12:09
* xnox ponders if i have a weird ass nautilus extension and/or customizations12:09
willcookeseb128, ignore, it's building12:10
seb128willcooke, I don't know of one, you can apt-get build-dep gnome-software though12:10
seb128if you have a deb-src enable12:10
willcookeyeah build-essential didnt bring fakeroot in12:10
willcookebut that's all I was missing12:11
seb128xnox, oh, that's a theme update from Trevinho :)12:11
jbichaacheronuk: gtk uses "What apps would you like to install to start with?" qt uses "Which applications would you like to start with?" is that intentional?12:11
willcookeseb128, woot, it work12:16
willcookeseb128, I used the deb-buildpackage12:16
seb128:)12:16
seb128does the patch work as well? ;)12:17
willcookeit does!12:17
seb128nice12:17
acheronukjbicha: yes, KDE is much more applications still than apps12:17
jbicha🙄12:18
seb128tkamppeter, hey, I don't know if you noticed but the cups update isn't migrating because the autopkgtest are failing, see12:20
seb128https://objectstorage.prodstack4-5.canonical.com/v1/AUTH_77e2ada1e7a84929a74ba3b87153c0ac/autopkgtest-bionic/bionic/amd64/c/cups/20180328_101808_e6118@/log.gz12:20
seb128"                 FAIL stderr: lpadmin: Raw queues are deprecated and will stop working in a future version of CUPS."12:20
seb128tkamppeter, those warnings are writting on stderr which makes the test fail, either you need to write on stdout, or filter out, or allow writting on stderr12:21
oSoMoNkenvandine, I've prepared a LO 5.4.6 build for artful (SRU: bug #1759404), would you be able to upload it for me? https://people.canonical.com/~osomon/libreoffice-5.4.6/12:42
ubot5bug 1759404 in libreoffice-l10n (Ubuntu) "[SRU] libreoffice 5.4.6 for artful" [Undecided,New] https://launchpad.net/bugs/175940412:42
oSoMoNI've built it in a PPA, verified that autopkgtests pass (on amd64) and successfully ran the manual test plan12:42
oSoMoNbtw, is there a way to request an autopkgtest run for a package in a PPA? that would be super useful12:54
oSoMoNLaney, you probably know ^12:54
seb128oSoMoN, https://wiki.ubuntu.com/ProposedMigration#Testing_against_a_PPA12:56
jbichaacheronuk: I updated my merge proposal and new PPA packages (bionic3) are building https://launchpad.net/~jbicha/+archive/ubuntu/temp-ubiquity/+packages12:58
oSoMoNseb128, perfect, thanks!13:05
seb128oSoMoN, yw!13:05
oSoMoNit would work if I were allowed to upload libreoffice… seb128, mind running this for me? https://autopkgtest.ubuntu.com/request.cgi?release=artful&arch=amd64&package=libreoffice&ppa=osomon/lo-test&trigger=libreoffice/1:5.4.6-0ubuntu0.17.10.1~ppa113:08
oSoMoN(I will request PPU for libreoffice someday, it's on my to-do list)13:08
seb128oSoMoN, done13:09
oSoMoNcheers13:09
kenvandineoSoMoN, sure13:12
oSoMoNthanks kenvandine13:20
jbichaacheronuk: new PPA packages are published now13:22
didrockskenvandine: hey, mind pushing your g-c-c changes to the vcs? I had to revert my commits13:22
kenvandinedidrocks, sorry!13:22
kenvandinewill do13:22
didrockskenvandine: no worry, I was wondering why I got a reject :)13:23
seb128that's annoying when it happens :/13:24
kenvandinedidrocks, done13:26
didrocksthx!13:26
kenvandineoSoMoN, sponsored13:28
oSoMoNcheers!13:29
xnoxseb128, re: https://bugs.launchpad.net/ubuntu/+source/systemd/+bug/1756006 -> if and when it is approved, i can do the systemd upload with relevant changes. But that FFe is not yet approved, right?13:29
ubot5Ubuntu bug 1756006 in systemd (Ubuntu) "FFe: Support suspend-to-hibernate" [Undecided,New]13:29
seb128xnox, right, looks like Lukaz wanted and UIFe, maybe GunnarHj can help you with that13:30
seb128well I think the systemd change doesn't impact the UI13:30
seb128then g-s-d ... dunno if adding an enum is user visibile/reflecting on g-c-c choices (but then we need a string added there I guess)13:31
jbichajibel: I added an updated screenshot (and merge proposal) to LP: #175808213:33
ubot5Launchpad bug 1758082 in ubiquity (Ubuntu) "[UIFe] Update ubiquity's Minimal Install page to match the spec" [Undecided,In progress] https://launchpad.net/bugs/175808213:33
jbichaseb128: gnome-control-center hard-codes the list of Power options, so there won't be a new string there unless we add it13:34
seb128that's what I though13:35
willcookejbicha, we've got the OK from product management to drop the Amazon web app from the minimal session.13:51
willcookecongrats!13:51
willcooke;)13:52
jbicha🎉13:52
didrocksand committed! :)13:53
willcookeoSoMoN, there seem to be no scroll bars in LO Impress when you shrink the window from full screen.  Have you seen that one?14:49
willcookeoSoMoN, they're back once I restarted it14:53
GunnarHjseb128: For now I take it that we await the LP update of gettext instead of playing with a hack.14:53
seb128GunnarHj, we can do that and revisit in a week14:54
GunnarHjseb128: Yeah - suppose you by "do that" mean wait. ;)14:55
seb128right14:55
GunnarHjAck.14:55
willcookexnox, I've got the slideshow ready, but I updated the text slightly on the Software slide - do I need to do anything about that?15:02
xnoxwillcooke, last time i touched this, it was eons ago. If you run the previewer from the branch, if you see what you are expecting to see, it should be all just fine.15:03
xnoxwillcooke, i can review the branch, if you want....15:03
willcookethanks xnox15:06
willcookeoh, my bad, it was cyphermox who pinged about it yesterday ^15:06
cyphermoxas xnox said15:09
oSoMoNwillcooke, weird, if you manage to reproduce please file a bug15:09
cyphermoxthe previewer helps a lot to make sure things look the way they should, sometimes spacing can be a little odd15:10
Trevinhoxnox: yeah is the file sector sidebar, even in nautilus... Based in ambiance-rw. There might be some improvements to do, but after you got used to, it's nice15:30
xnoxTrevinho, i liked from the first time I saw it!15:32
TrevinhoGood :-)15:33
willcookecyphermox, https://bugs.launchpad.net/ubuntu-themes/+bug/1759601  I checked with the testing tool, and it all looks good to my eyes.15:40
ubot5Ubuntu bug 1759601 in ubuntu-themes (Ubuntu) "18.04 LTS Bionic Beaver Mascot" [Undecided,New]15:40
seb128jibel, the gnome-menus update issue, do you get steps where you can still trigger it without hitting the plymouth issue that is blocking L_aney to investigate?16:08
jibelseb128, I can try to find steps. I'd like to finish the review of ubiquity today but planned to work on upgrades tomorrow and fraiday. I'll check if I can find steps that do not hit the plymouth issue.16:11
cyphermoxwillcooke: cool, will review after lunch16:20
willcookecyphermox, merci!16:20
didrockshave a good evening everyone16:48
=== pstolowski is now known as pstolowski|afk
seb128jibel, thanks, thursday or friday sounds good17:10
jibeljbicha, I merged the "update & software" page. Thanks for your work!18:00
jibeljbicha, can you drive the uife18:00
jibel?18:00
jbichaum, I thought it was provisionally accepted based on fixing Kubuntu18:00
jibelI see nothing in the bug report and ubuntu-release is not subscribed18:01
jibelotherwise i'll do it tomorrow morning18:01
jbichaI'm interpreting comments #4 and #1418:02
jbichaLaney: do you want to reconfirm the UIFE for LP: #1758082 ?18:02
ubot5Launchpad bug 1758082 in ubiquity (Ubuntu) "[UIFe] Update ubiquity's Minimal Install page to match the spec" [Undecided,In progress] https://launchpad.net/bugs/175808218:02
willcookegnight all18:05
jibeljbicha, ah sorry, I missed #4. It's all good then. thanks18:10
* jibel -> EOD18:10
jbichacool, thanks for reviewing :)18:10
seb128good work guys!18:13
oSoMoNEOD, have a good evening everyone18:59
=== amano_ is now known as amano
jbichaTrevinho: do you want to close LP: #1689239 now?20:29
ubot5Launchpad bug 1689239 in gnome-software (Ubuntu) "Headerbar used as toolbars in unity are missing proper css classes" [High,In progress] https://launchpad.net/bugs/168923920:29
Trevinhojbicha: correct20:52
Trevinhoc-lobrano: hey, just to know... Will you do something for https://github.com/CanonicalLtd/desktop-design/issues/28 or should I?22:52
ubot5-ngCanonicalLtd bug 28 in desktop-design "Nautilus path bar buttons are inconsistent" (comments: 13) [Community Team, Open]22:52
ubot5bug 28 in Launchpad itself "Outgoing email configuration ready for production - gogo!" [Medium,Fix released] https://launchpad.net/bugs/2822:52

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