/srv/irclogs.ubuntu.com/2017/07/06/#ubuntu-desktop.txt

=== jamesh_ is now known as jamesh
jibelmorning05:14
oSoMoNgood morning desktoppers05:54
jibelbonjour oSoMoN06:09
didrocksgood morning07:00
dufluMorning jibel, oSoMoN, didrocks07:02
jibelhi duflu07:02
didrockshey duflu, jibel07:03
seb128good morning desktopers07:20
seb128jibel, oSoMoN, didrocks, lut les frenchies07:20
seb128hey duflu07:20
seb128how is everyone today?07:20
didrocksrebonjour seb128!07:21
jameshhi seb12807:23
dufluseb128, going OK (modulo a serious sinus headache)07:23
dufluHow are you seb128?07:23
jibelhi seb128, doing well. how are you?07:24
seb128hey jamesh07:33
seb128jibel, I'm good thanks07:33
willcookemorning all08:04
didrockshey willcooke08:07
willcookemorning didrocks, how's it going?08:07
willcookechrisccoulson, more suggestions of thunder storms tonight!  fingers crossed08:08
didrockswillcooke: good good, but getting hot again :/ yourself?08:08
seb128hey willcooke, how are you?08:10
willcookedidrocks, much cooler today, and cloudy.  Yesterday was hot, hopefully today will be a bit cooler.08:11
willcookehey seb128, good!  The kids all slept right through the night, but I still feel tired :D08:11
willcookenot as tired as you I expect08:11
seb128lucky you :p08:11
seb128I got pulled out of bed at 6am today08:12
seb128one of those days...08:12
willcookeheh08:12
willcookeonly another 18 years to go08:12
seb128haha08:12
seb128on the positive side I was done with emails backlog&such at 708:13
willcookehah08:13
willcookeseb128, let me fix that for you... in 3,2,1..08:14
* seb128 puts the finger over the delete key08:15
seb128k, I'm reading, send away!08:15
seb128ready*08:15
willcookesec08:15
chrisccoulsonwillcooke, yeah, I hope so :)08:16
willcookechrisccoulson, before your time?  http://www.bbc.co.uk/news/entertainment-arts-4051675408:21
willcookeseb128, didrocks - did you see:  https://bugs.launchpad.net/ubuntu/+source/gdm3/+bug/168639308:31
ubot5Ubuntu bug 1686393 in gdm3 (Ubuntu) "[MIR] gdm3" [High,Confirmed]08:32
willcookewe're unblocked08:32
didrocksnice! do you know if kenvandine[m][m] did the seemless transition from lightdm to gdm?08:32
didrocks(not really clear on the trello card)08:33
seb128he mentioned being working on him when I talked to him on friday08:33
seb128or monday it was08:33
seb128then he was off for 4th of july08:33
seb128him->it08:33
didrocksok, so let's wait for him. I don't think we should seed it with the prompt before this is and the session transition is done08:34
didrockshopefully, he's already done on that, as it could be done in parallel than the MIR :)08:34
didrocks(I can give a hand if needed on this)08:35
* didrocks tries to go for a run before it's getting to warm. Couldn't run at all this week08:35
seb128didrocks, enjoy08:36
seb128didrocks, he mentioned that he had things mostly working but was having issues with debconf08:36
didrocksseb128: 27°C, will be 31°C very soon. I'll "try" to enjoy :)08:36
seb128so let's see when he gets online08:36
willcookeI think he got it done, one sec08:36
didrocksyeah ;)08:36
didrocksthere are two things:08:36
didrocks- silent debconf08:36
didrocks- session mapping (even for encrypted ones)08:36
didrocksdrive*08:37
willcookemeh, cant find it - but I've a feeling it's either done or nearly done08:38
didrocksgreat! Will be nice to switch it today to see the result on tomorrow's iso08:39
didrocksand continue on demotion/unseeding08:39
* didrocks really goes now08:39
seb128brb, restarting IRC with vpn08:39
seb128still a bit sad to see lightdm being replaced08:39
seb128oh well08:39
seb128didrocks, enjoy08:39
didrocksthx08:39
=== maclin1 is now known as maclin
seb128oSoMoN, so that kernel issue is mostly understood and in the hands of the kernel team now? or do they still want a testcase?08:41
oSoMoNseb128, it’s in their hands, and it appears there's already some work upstream to fix that08:41
seb128oSoMoN, great news :-)08:42
oSoMoNso hopefully within a few days we should have an updated kernel in proposed that unblocks us08:42
ricotzseb128, oSoMoN, hey :), note this fixed kernel needs to be in builder vm08:45
seb128of course08:46
oSoMoNricotz, yes indeed, but once it’s in proposed we will be able to test it and validate whether it fixes the crash08:46
ricotzright, I assumed this kernel needs to be transitioned from proposed to be picked up by the builder-vm08:47
oSoMoNand it’s not just the build that fails, it’s also the app itself that crashes on x86 under certain circumstances when the JVM is enabled08:47
ricotzI know08:47
ricotzfingers crossed this get fixed soon :)08:47
oSoMoNyup08:47
chrisccoulsonwillcooke, I'm old enough to remember grotbags :)08:50
willcooke:)08:53
ricotzchrisccoulson, hey :), https://bugs.launchpad.net/ubuntu/+source/rustc/+bug/170155609:05
ubot5Ubuntu bug 1701556 in rustc (Ubuntu) "Backport rustc 1.17 and cargo 0.18 to 14.04/16.04 and 17.04" [Undecided,New]09:05
seb128ricotz, he's off for the rest of the week09:07
ricotzseb128, hi, I see09:08
chrisccoulsonricotz, yeah, I'm aware of that. I just need somewhere I can prepare the transition (it needs to be updated to 1.16 / 0.17 first, and then eventually binary copied to the security PPA)09:09
ricotzchrisccoulson, ok, I already monkey-updated/patched cargo to 1.19 to get nightly builds on artful09:12
oSoMoNseb128, chrisccoulson: FYI the chromium 59 update is crashing at startup for a number of users, details in bug #170240709:28
ubot5bug 1702407 in chromium-browser (Ubuntu) "Startup crash after upgrading to 59.0.3071.109 on xenial" [Critical,Confirmed] https://launchpad.net/bugs/170240709:28
seb128urg09:28
oSoMoNI’m actively looking into it09:29
seb128thanks oSoMoN09:29
seb128bah09:32
seb128jbicha, could you push your most recent gnome-control-center revision to the vcs?09:32
seb128brb09:33
Trevinhoandyrock: good luck https://code.launchpad.net/~unity-team/unity/x-sru5/+merge/326916 :)09:43
Trevinhoandyrock: and... i'd need some help to update the bugs to be SRU friendly too09:44
seb128hey andyrock Trevinho09:51
Trevinhohi seb12809:51
=== davmor2_ is now known as davmor2
* Trevinho feels dirty: just used some browser macros to do launchpad actions instead of using APIs :|10:10
andyrockTrevinho: kk10:12
andyrockand hey10:12
dufluAwesome. In figuring out initialization failures I seem to have also found the cause of video corruption elsewhere.10:24
dufluTime for a celebratory meal10:24
Trevinhodupondje: yeah10:25
Trevinhoerr sorruy, duflu*10:26
GunnarHjdidrocks: Hi Didier! Thanks for approving 'your' part of the MP. As regards usb-langsupport, and even if you tried to explain it on the MP, I still don't understand the purpose well enough to formulate a proper question. Hence it would be good if you could post to the -devel ML about it.11:53
didrocksGunnarHj: yw! yeah, I'll check first on next team meeting if anyone has any use of it. I'm almost sure even QA don't try it11:57
didrockswe'll can kick it off anytime on the -devel ML, I can handle that next week. But yeah, let's not block on that11:58
GunnarHjdidrocks: Ok, thanks!11:58
didrocksyw, thanks for working on this! :)11:58
GunnarHjjbicha: Hi Jeremy! Ready to approve the rest of the MP, including the sync of fonts with pkg_depends, and merge it?12:01
GunnarHjhttps://code.launchpad.net/~gunnarhj/ubuntu-seeds/ubuntu.artful_lang-support/+merge/32626112:01
andyrockTrevinho: do you have a ppa for the sru? on bileto?12:12
andyrockfound it sorry12:12
Trevinhoandyrock: yeah. it's there thanks12:28
andyrockstill building unity12:28
Trevinhoandyrock: yep12:31
Trevinhoas i forgot to do it before lunchj12:31
Trevinhoandyrock: I've packages but.... I guess you can wait a sec12:31
andyrockyeah otherwise I can build it too12:32
Trevinhoit's already at the test stage12:32
andyrockbut I didn't want to build compiz+nux+etc.12:32
Trevinhowill finish in 5 minutes amx12:32
Trevinhomax*12:32
Trevinhoyeah..12:32
* Trevinho prepares a unity7 snap :)12:32
* Trevinho jokes, but not so much12:32
Trevinhowould perfect for classic12:33
didrocksTrevinho: you can read my "trip" to classic snap with ubuntu make :)12:34
didrocks(part 1 published on planet ubuntu, you can read the rest on github until published if you want)12:34
Trevinhodidrocks: I will, although I already have some experiencve with it, but interesting indeed12:34
didrocksthere are some interesting thinking I guess on 14.04 & 17.04 snaps ;)12:35
didrocks(part 3)12:36
Trevinhoandyrock: ok packages are there https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu/2841/+build/1285889712:36
Trevinhodidrocks: where's the rest of the story in gh?12:38
didrocksTrevinho: https://github.com/didrocks/website/tree/master/content/blog12:43
didrocks(static websites ftw!)12:44
jibeldidrocks, did you do the change to migrate u-c-c to g-c-c?13:08
jibeldidrocks, the key is unity-control-center.desktop not ubuntu-control-center.desktop13:11
didrocksjibel: argh, probably typoed, sorry :/13:13
jibelnp13:13
Trevinhodidrocks: thanks :)13:13
didrocksjibel: doing it right now13:13
jibelthanks13:14
didrocksjibel: ofc, same scripts, so people who already migrated won't see that change13:14
kenvandinecyphermox_, the security team ack'd bug 1686393 so i reassigned it back to you13:15
ubot5bug 1686393 in gdm3 (Ubuntu) "[MIR] gdm3" [High,Confirmed] https://launchpad.net/bugs/168639313:15
Trevinhodidrocks: it looks like I'm able also to run snapcraft to build classic inside a scrhoot now... but I had to bind mount manually the core, once I bind mounted the snap dir from host13:15
didrockskenvandine: why is it reassigned back to cyphermox_, he did +1 if security was happy about it, isn't it? Isn't that just a question13:15
didrocksof us promoting it?13:15
kenvandineoh, maybe he did13:15
didrocksthat's the normal procedure13:16
didrocksso, now, it's up to us to seed it if the migration is handled :)13:16
kenvandinedidrocks, i'd still prefer cyphermox_  mark it fix committed :)13:17
kenvandinei'll go ahead and update the seed13:17
didrockskenvandine: well, technically I can as he's on holidays :)13:17
didrockskenvandine: all the transition is done?13:17
kenvandineyeah13:18
didrocksI think it will be good to get an image tomorrow with that done so that we can move it forward13:18
kenvandinenot uploaded yet13:18
didrocksnice!13:18
kenvandinei can do that today13:18
didrocksjibel: uploaded13:18
didrocks(sorry again)13:18
didrockskenvandine: just being curious, how did you handle the "I install gdm after the fact and have lightdm, but want a debconf prompt"?13:18
* didrocks is interested :)13:18
kenvandinedidrocks, it's a combination of postint in ubuntu-session and gdm13:19
didrocksinteresting :)13:19
kenvandineubuntu-session checks to see if any user on the system uses the ubuntu session13:19
didrocksI guess if current user session is ubuntu-session as we discussed…13:19
kenvandinenot just current, any user13:19
didrocksyep, what I did propose on the card IIRC13:19
kenvandineyup13:20
didrocksstill some false positive13:20
kenvandineif any use it, we touch a file13:20
didrocksbut couldn't come up with anything else13:20
didrocksand then debconf script check that file13:20
kenvandinei think it was laney that suggested we do it in ubuntu-session13:20
didrocksand don't show the question, correct?13:20
kenvandineyeah13:20
didrocksyeah, I didn't suggest package, I still wonder about order, did you force anything on gdm?13:20
didrocksto see "we need ubuntu-session postinst before gdm"13:21
kenvandinegdm will prompt in the case ubuntu-session didn't flag it13:21
kenvandineyes13:21
* didrocks is eager to test :)13:21
kenvandinemade gdm depend on ubuntu-session13:21
didrockssounds exciting13:21
kenvandine:)13:22
didrockskenvandine: want me to promote gdm so that it's done, and you can upload/update the seeds?13:22
kenvandinei'm going to do one more round of tests in a VM today before i upload it13:22
kenvandinedidrocks, please do13:22
didrockslet's gooooooooooo13:22
kenvandinewoot13:22
willcooke\o/13:23
didrocks*DONE* next publisher cycle would have it13:25
* kenvandine merges gnome-session branch again :)13:26
willcookenice, exciting times ahead13:26
didrockskenvandine: I guess as you are going to test that will be fine, but wait for rmadison to show gdm3 in main before uploading, or there is a race and we'll need to repromote it twice (once in proposed, once in the release pocket)13:26
kenvandinedidrocks, ok13:26
didrockskenvandine: who did just push for a u-c-c rename? I wonder :)13:26
kenvandinei wonder13:27
Trevinhodidrocks: on Fixing by using snap internal dependencies, actually the best solution is just to use snapcraft-preload so far IMHO :)13:42
Trevinhoas you don't have to worry about all the env variables13:42
didrocksTrevinho: got a load of issues with it some months ago13:42
didrocksI didn't really retry since then13:42
Trevinhodidrocks: oh, like? I'd like to address those13:42
didrocksTrevinho: gettext not looking at the right paths and such, I guess that were uncovered syscalls13:43
didrockskenvandine: I promoted at the right moment! Already in main :)13:44
Trevinhodidrocks: might be... for qt apps for example it worked like a charm https://github.com/nextcloud/client_theming/blob/master/linux/snap/snapcraft.yaml13:44
kenvandinei saw that :)13:44
Trevinhodidrocks: it used not to work in X some months ago though, but I've fixed it13:45
didrockskenvandine: can't wait to see the remaining indicators off the live :)13:45
didrocksTrevinho: probably at that time, TBH, I didn't retry since13:46
Trevinhogive that a look, it helps a lot.13:46
didrocksTrevinho: if we are pretty confident about it, why not having the desktop helper removing some variables override and using this?13:46
Trevinhoactually I would love to do a snapcraft part using that for desktop, but...13:46
Trevinhodidrocks: I wanted to, but seb128 isn't much confident on that since it increases the syscall's amount.13:46
didrocksTrevinho: I guess you are looking at a relative path, and if not found, the system one?13:46
Trevinhowhich is true, but in my benchmarks it wasn't really something noticeable13:47
didrockswe already have a lot of -1 open() calls when stracing… (even due to those env variables)13:47
didrocksTrevinho: could we add 2 paths, like relative snap one, then another one (like the platform snap one) before it defaults to system ones?13:48
Trevinhodidrocks: yep.. . That's the plan13:48
Trevinhodidrocks: but there is a bug in apparmor which gives some false positive13:48
Trevinhodidrocks: see https://bugs.launchpad.net/apparmor/+bug/165543513:49
ubot5Ubuntu bug 1655435 in AppArmor "stat() unconditionally allowed via apparmor_inode_getattr()" [Undecided,Triaged]13:49
Trevinhodidrocks: so due to that in some cases we can't be as efficent as we could be, as we get that snapcraft can stat something that it can't actually open13:49
Trevinhodidrocks: but in any case I think that the overhead of that tool isn't really much, and I've rewritten it to C++(14) also so it's really easy to maintain avoiding issues13:50
didrocksoh yeah, will be good to get it fixed before switching it on, indeed13:50
seb128I really dislike that preload hack13:51
seb128hope that doesn't become default or standard13:51
seb128it's such a hack13:51
didrocksnot as much as env variables?13:51
didrocksquite some hack level to me, no?13:51
didrockssame*13:51
seb128I feel like intercepting syscalls is lower level and has more potential for real bugs13:52
seb128when env can't not really be harmful13:52
Trevinhomh, yeah... but also very controllable.13:52
seb128also env is well defined13:52
TrevinhoI mean we know what's happening13:52
Trevinhothe problem with env's is that there are too many13:52
Trevinhoand out of our control13:52
seb128where when you start derailing syscalls you might impact things you didn't think about13:52
Trevinhofixing 1 instead of 100013:52
seb128I'm reading to bet that you are going to hit weird cases you didn't think about with the syscall hacking13:53
TrevinhoIt's true... It's just that it's a different path we can follow, allowing both the cases.. I'm not in the idea of setting is as default, but adding something like desktop-gtk3-preload as option13:56
Trevinhothen we see13:56
didrockshum, adding a part will add confusion13:56
didrocksit's either we replace or we don't IMHO13:56
didrocksbut don't give too much options that developers won't care/just get confused with IMHO13:57
seb128I guess my main issue is that it's an hack on low level stack and those might bite you back one day since they are doing undocumented things13:57
* didrocks would like /usr to be really /usr in snaps13:57
seb128where the env thing is an hack13:57
seb128but those env variables are documented13:57
seb128and there is little risk that they fire back13:58
seb128don't mess up with libc level also :p13:58
seb128yeah, /usr to be at the correct place would make things easier13:58
seb128Trevinho is laaaate14:02
didrocksbut he was preload*ed* :)14:03
jbichadidrocks: can I go ahead and replace unity-greeter/lightdm with gdm3 in ubuntu-desktop?14:22
jbichaI just merged Gunnar's language MP14:23
seb128jbicha, no14:25
didrocksjbicha: kenvandine is doing a last round of testing and doing it14:26
seb128jbicha, doesn't seem polite to hijack the transition from somebody who is doing it and testing to upload instead of them14:26
kenvandinejbicha, i need to upload gnome-session and gdm first14:26
seb128also Ken said he was testing and would upload, see backlog14:27
kenvandinei guess it doesn't have to be first...14:27
kenvandinebut i was going to14:27
jbichasorry, I didn't see the backlog14:28
jbichadidrocks: did we want to do Gunnar's language changes now or should we just do one thing at a time?14:28
didrocksjbicha: hum, I would like to have an image with gdm first if possible14:29
didrocksand that one on the next one14:29
didrocksso that we can see additional packages and such14:30
didrocksmaking sense?14:30
seb128jbicha, in case you didn't read backlog, can you push your g-c-c changes to the vcs14:30
jbichadidrocks: that's fine, I'll revert Gunnar's changes for now then14:30
didrocksthx!14:30
jbichaseb128: done, thanks14:34
seb128jbicha, thank you14:34
Trevinhoseb128, didrocks : so... ./plugins/xsettings/gsd-xsettings-manager.c... by migrating first and patching this code to use a different key we can fix things I think14:35
Trevinhoit's thing I should have had in the past actually... but it should work14:36
Trevinhoi'll check it later14:37
seb128thanks14:37
didrocksTrevinho: excellent! Please once you do it, add me to the card so that I handle the pre-migration stuff14:37
didrockshum, that will be interesting with the wayland session…14:38
didrocksas the migration script can run twice, we'll see¿14:38
Trevinhodidrocks: nope, as that only uses xsettings to inform gtk14:38
didrocksyeah, but I mean, migration-wise14:38
Trevinhoah14:38
Trevinhook14:38
didrocksthe migation is per session14:38
didrocksI'll deal with it14:38
jbichaseb128: your g-c-c upload is a bit broken14:41
seb128jbicha, because of the missing C/R/P?14:41
seb128or something else?14:41
jbichayes and it's missing the final 's' from gnome-control-center-face.install14:41
jbichafaces14:41
seb128?14:42
seb128facess?14:42
jbichathe package is gnome-control-center-faces but you have a gnome-control-center-face.install14:42
seb128ah, the filename14:42
seb128doh14:42
seb128thanks for spotting that14:42
seb128I was already commiting a fix for the C/R/P14:42
seb128going to rename as well14:42
didrocksand then, u-c-c off main! :)14:44
didrocksthis is the big package archive gambling this week :)14:44
seb128:-)14:44
jbichaseb128: are you going to have u-c-c-faces be a transitional package?14:45
seb128jbicha, new revision uploaded, hopefully better14:45
seb128no14:45
seb128why?14:45
seb128g-c-c-faces provides it14:45
jbichaI'm not sure when transitional packages are required, but I think it might be better to have it than not14:45
seb128so just going to drop it from u-c-c14:45
seb128Provides should work fine as long as we don't have versioned depends on it14:45
jbichaprovides alone does not ensure the upgrade happens14:45
seb128which I don't think we have14:45
seb128right14:46
seb128but g-c-c recommends g-c-c-faces14:46
seb128that should work14:46
seb128if it turns out to be problematic for some reason we can add back a transitional14:46
seb128but the case is easy enough that I think it should work14:46
seb128it gets trickier when you have several solutions apt can choose to a problem14:47
jbichaok, I was just wondering14:47
seb128but I don't think it's complex enough to be an issue in that case14:47
seb128well, it's always difficult to say14:47
seb128so let's see what upgrade tests say14:47
kenvandinei think dropping unity-greeter will allow demoting all the indicators15:02
seb128that's what we think as well15:03
didrocksand another part of Qt15:03
seb128:-)15:03
didrocks(not all, due to fcitx)15:03
didrocksso yeah, that's why I'm eager to see that effect :)15:04
kenvandine:)15:04
kenvandinedidrocks, seb128, jbicha: seed updated15:10
kenvandineremember, you can't blame me if we don't have images tomorrow :-D15:10
didrocksohhhhh /me removes all his premade jokes in his note15:11
kenvandinelol15:11
seb128\o/15:16
seb128didrocks, oh, and thanks for the binNEW reviews15:16
didrocksyw ;)15:17
seb128jbicha, GunnarHj, didrocks, I rolled artful langpacks today and upgraded now, gnome-shell is correctly translated in french now15:23
didrocksnice!15:26
jbichaseb128: thanks!16:03
seb128yw!16:03
jbichaI guess we need to add X-Ubuntu-Use-Langpack to all the Unity packages or something?16:03
seb128that would be nice16:03
jbichaseb128: please subscribe the Desktop bug teams to https://launchpad.net/ubuntu/+source/spice-vdagent/ for LP: #120029616:44
ubot5Launchpad bug 1200296 in ubuntu-meta (Ubuntu) "[MIR] spice-vdagent" [Wishlist,Confirmed] https://launchpad.net/bugs/120029616:44
seb128 jbicha, done16:46
jbichapitti: did you see that were pinged on https://bugzilla.gnome.org/690866 ?16:50
ubot5Gnome bug 690866 in gnome-session "Integrate better with systemd" [Normal,New]16:50
willcookenight all17:19
ahayzenHi, I've been trying wayland on Ubuntu GNOME 17.04, and I've found that QtCreator crashes when trying to open a file dialog, this appears to be fixed upstream in Qt and in Fedora 25. Any chance this can be patched in Ubuntu as well ? I've reported bug 1702741 to track it.19:16
ubot5bug 1702741 in qtcreator (Ubuntu) "QtCreator crashes when opening a filechooser under wayland" [Undecided,New] https://launchpad.net/bugs/170274119:16
=== Guest15020 is now known as fredp
=== fredp is now known as Guest73021
=== JanC is now known as Guest56451
=== JanC_ is now known as JanC
jbichaahayzen: oh maybe that's what I saw months ago with LP: #165792419:24
ubot5Launchpad bug 1657924 in virtualbox (Ubuntu) "Clicking the file chooser button crashes VirtualBox on Zesty" [Medium,Expired] https://launchpad.net/bugs/165792419:24
ahayzenjbicha, could be related19:25
jbichaVBox is the only Qt app I use19:25
jbichahmm, it was fixed in Debian in January :|19:33
=== charles_ is now known as charles

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