/srv/irclogs.ubuntu.com/2014/03/28/#ubuntu-desktop.txt

=== infinity_ is now known as infinity
=== Tokyo is now known as Zachary_DuBois
MirvLaney: not really, normal desktop apps probably won't hit what we do in Touch, and Debian does not (yet) have much mobile apps or heavy Qt 5 users there. I should initiate discussion though probably that others would take a look at the Ubuntu's cherry-picks05:21
MirvLaney: I've not planned a new qtdeclarative upload (just did one a couple of days ago), but those updates are probably best through CI Train manual upload so that AP tests can be run to make sure there are no regressions.05:22
Mirvthey'll be at Qt 5.3.0 though probably before any KDE5 bits hit Debian05:23
=== maclin_ is now known as maclin
seb128good morning desktopers!09:00
Laneyhey there09:03
Laneyhappy friday09:03
LaneyMirv: Fair enough, I usually commit but don't upload for worthy cherry-picks09:03
seb128Laney, happy friday indeed!09:03
LaneyEven though they are important for us they're still usually actual bug fixes09:03
Laneypeople seem to like the beta changes :-)09:12
seb128what beta changes?09:14
seb128you mean trusty changes compared to saucy?09:14
Laneyyeah09:15
Laneyspeaking of, we should get the gtk scale settings for u-c-c in09:15
seb128oh indeed09:16
seb128let me test it locally09:16
seb128or did you do that already?09:16
Laneynope, lemme do that09:17
smbmlankhorst, Are you already aware about the glitches I tried to summarize in bug 1298517?09:36
ubot2Launchpad bug 1298517 in unity (Ubuntu) "Rendering issues in unity with xserver-modeset" [Undecided,New] https://launchpad.net/bugs/129851709:36
seb128shrug09:37
seb128update-manager segfaults on start for me now :/09:37
seb128hum, apport as well09:38
seb128pitti, do you know about those?09:38
pittibonjour seb12809:38
seb128pitti, guten tag! happy friday!09:38
pittiseb128: yes, it's bug 129882409:38
ubot2Launchpad bug 1298824 in libffi (Ubuntu Trusty) "libffi 3.1~rc1 regression: crashes on i386; python3.4 crashed with SIGSEGV in g_callable_info_free_closure()" [Critical,Confirmed] https://launchpad.net/bugs/129882409:38
pittiseb128: it kept me busy in my morning :/09:38
seb128oh, doko, who else09:38
pittiseb128: hereux vendredi à toi aussi !09:39
seb128how come none of the autopktest caught it?09:39
pittiseb128: yeah, big fail of britney-adt :(09:39
seb128:-(09:39
pittiseb128: the point is, the autopkgtest did catch it09:39
pittiseb128: but it wasn't held back09:39
seb128is that still the same issue of "let things through despite failing tests"?09:39
pittiseb128: jibel is looking into the logs to try and find out why09:39
seb128it's not the first time :/09:40
seb128pitti, ok, I'm going to let you work then, thanks for the update!09:40
pittiseb128: yes, which we haven't managed to understand yet; I thought I reproduced it in a britney test, but that was just a bug in the test09:40
pittiseb128: yeah, libgcc1 a few weeks ago09:40
seb128right09:40
pittiseb128: I prepared a revert for that, it's waiting in unapproved09:40
seb128pitti, did some nudging on release09:43
pittiseb128: thanks; I pinged doko this morning, but no response yet; I think we can declare this a timeout, given how critical this bug is09:44
seb128pitti, yeah, especially just when we released beta, it means users upgrading are going to be without a working update-manager to get the fixed version09:44
pittiseb128: ah, got accepted09:46
seb128pitti, ;-)09:46
pittiseb128: looking forward to a flood of "jenkins fixed" emails :)09:47
seb128yeah09:47
seb128but more important, looking forward having users without an hosed system, it's not only tests :/09:47
seb128wth doko, did he even test that update?09:47
pittiseb128: it works fine on amd64, and the upstream tests pass on i386 as well09:48
pittibut it seems to have an ABI break on i38609:48
seb128oh ok09:48
pittiseb128: I kept some notes in the bug09:48
seb128it's you guys hating on us i386 users :p09:48
seb128I got the message, I should migrate to 64 bits one day ;-)09:48
pittiit's whispering "go 64 bit!", don't you hear it?09:48
* seb128 hugs pitti09:48
* pitti hugs you back09:49
pittiseb128: it's actually good that some folks test i38609:49
pittiseb128: it's already bad enough that most developers have Lenovo or Dell laptops these days :)09:49
seb128indeed, it's one of the reason I'm keeping that machine i38609:49
seb128my next laptop is going to be 64bit install, but I keep that one on 32bits09:50
jibelpitti, seb128 I found at least 2 problems: libffi has been copied before britney received all the results of packages depending on it 2. Some results are marked as PASS while the result file contains a FAIL.09:50
seb128pitti, do you know if anything was built against the new libffi that needs to be rebuilt with the revert?09:50
pittiseb128: no, I don't; we need to traverse through build logs09:51
pittii. e. all libffi rdeps which got uploaded yesterday09:51
seb128pitti, ok, having a look09:51
pittiseb128: I know I synced g-i and pygobject, but they obviously got built against the old one09:51
pittiseb128: and everything which has an autopkgtest (ruby-ffi and friends) will be automatically covered09:52
pittiseb128: so I suspect there's not that many packages to check manually09:52
pittijibel: eek; so for 1. britney didn't wait until all tests got done, but accepted on RUNNING?09:52
seb128pitti, \o/, working python guis again, thanks ;-) (I got the debs from proposed)10:04
pittigreat10:05
seb128pitti, you synced gnome-icon-theme-extras for Jackson, was that a bug report?10:05
seb128https://bugs.launchpad.net/ubuntu/+source/gnome-icon-theme-extras/+bug/1297645 found it10:05
ubot2Launchpad bug 1297645 in gnome-icon-theme-extras (Ubuntu) "Sync gnome-icon-theme-extras 3.12.0-1 (universe) from Debian unstable (main)" [Wishlist,Fix released]10:05
pittiseb128: yes; it's used nowhere except ubuntu-gnome, and only has a new icon, so it seems fine?10:06
seb128pitti, britney says "gnome-icon-theme-extras/i386 unsatisfiable Depends: gnome-icon-theme (>= 3.12) "10:06
seb128pitti, I commented on the bug/reopened/assigned to Jackson10:06
seb128pitti, our g-i-t is not new enough10:06
pittiseb128: erk, sorry; thanks; I should check this myself for noskcaj10:06
seb128yeah, he's motivated but not very careful sometimes10:07
darkxstpitti, any idea what is happening with the gjs tests?10:25
darkxstI am lacking internet at this point ;(10:26
seb128not sure if you noticed but new gjs is quite buggy10:30
seb128well at least it has errors ranked high on e.u.c that started in 1.39 versions10:30
seb128g-s-d is also quite high ranked10:30
darkxstseb128, there are some re-entry issues with the GC.10:35
darkxstbut I can't do much right now. stuck on phone ;(10:35
Laneyricotz: https://bugs.launchpad.net/ubuntu/+source/libosinfo/+bug/1297957 please more details in future10:36
ubot2Launchpad bug 1297957 in libosinfo (Ubuntu) "Sync libosinfo 0.2.9-1 (universe) from Debian unstable (main)" [Undecided,New]10:37
darkxstmoving in a few weeks, but when I booked it in, my stupid telco decided to disconnect everything straight away10:37
Laneytether ;-)10:37
darkxstI am10:37
darkxstbut over here its super expensive10:38
darkxstlike $20 per 700MB10:38
Laneygosh10:38
darkxstLaney, yup! I should just move to europe, but need to finsih phd first10:40
margaHi, can someone re-open https://bugs.launchpad.net/ubuntu/+source/unity/+bug/1051921 for Precise?  It was incorrectly marked as Fix Released.  I can still reproduce with 5.20.0-0ubuntu311:04
ubot2Launchpad bug 1051921 in Unity 5.0 "lens-bar-keynavigation periodically writes to /tmp/wut.png" [Medium,Fix released]11:04
seb128marga, done11:05
margaseb128, tnx11:05
LaneyTrevinho: Just checked out the combo11:08
LaneyThere's a problem with sizing11:08
LaneyIf you go to a big scale size and then back down to a smaller one then the combos in the right column keep the larger width11:09
Laneyhaha also if you change the size while a notify-osd bubble is displayed it gets messed up11:11
Laneyoh MAN I finally lost the ubuntu one context menu in nautilus11:12
LaneyTrevinho: http://ubuntuone.com/783mzxI7O0dTLUZTiLd2gG11:13
LaneyAlso isn't it weird that some things are instant apply given the 'Apply' button?11:14
pittidarkxst: another fallout of libffi; the revert should land soon now11:15
pittidarkxst: yep, fixed again11:15
Laneydarkxst: lucky you ;)11:17
Laneylibffi fallout is tricky11:17
darkxstLaney, why me? I have nothing to do with libffi!11:18
Laneybecause you don't have to fix anything11:19
darkxstLaney, oh you are forgetting, I can't fix anything anyway!11:20
LaneyI don't think anyone was going to fix gjs for you :P11:20
seb128Laney, Trevinho: it's also a bit weird, the dialog resize horizontal when selecting screens because of the different names length11:21
Laneyah, I don't see that11:22
seb128Laney, it might depends of the locale, for me things file the horizontal space11:22
darkxstLaney, right, I will just go sulk in my internet-less corner not much else to do ;)11:23
seb128so longer labels strech the dialog11:23
seb128Laney, Trevinho: btw the "instant apply" is also an issue, it makes easy to screw your session, like if you move to 3, the slider/buttons get offscreen and you are screwed for good11:23
Laneyyou need to know that you can hold alt to move it up11:24
seb128I know11:24
Laneywhich is bad, but it'd also be bad if you did that with non-instant apply11:24
seb128non tech users likely not11:24
Laneyunless it does the timeout thing11:24
seb128well, the resolution change has a "revert after 30s if not confirmed"11:24
seb128Laney, is there any of this issues that you consider a blocker to land the current mp? or should we just land that and consider tweaks as bugfixes for another iteration?11:25
LaneyI don't think so, but I'd like to hear that the work can be on the list for trusty11:26
seb128right11:26
Laneybut feel free to get it in a silo if you want and the code review is ok11:27
seb128well, no hurry to get it in a silo11:29
seb128I didn't review the code yet, just did some local testing11:29
Laneyshrug, would like to get it in sooner rather than later11:31
seb128k, let me finish what I'm doing and do a code review/put u-c-c in a silo in 10minutes or so11:31
Laneyno über rush since Trevinho seems afk anyway11:32
seb128right, he has weird work hours, like he seems to work at night often and then sleep through the morning ;-)11:32
* seb128 used to do that at school11:33
Laneyyeah me too, but these days I like being free at the same time as other people :p11:33
seb128yeah, same here ;-)11:33
seb128Laney, do you see an issue landing https://code.launchpad.net/~ted/indicator-keyboard/startup-cleanup/+merge/207768 ?11:36
Laneythe data/upstart desktop file has a different OnlyShowIn11:37
seb128I wonder if we should include it for trusty or not (I'm putting a landing for another fix)11:37
LaneyI think those things shouldn't have OnlyShowIn at all11:37
seb128hum, yeah11:37
Laneybut the effect of that as it is is that you get XDG autostarted in gnome (!gnome3) sessions11:38
seb128ok, let's skip over that one, I really care about the bugfix11:38
ricotzLaney, thank you12:16
Laneyyw12:16
=== MacSlow is now known as MacSlow|lunch
nturner_Is anyone else seeing https://bugs.launchpad.net/ubuntu/+source/unity/+bug/1281806 ? What package should this really be assigned to?13:42
ubot2Launchpad bug 1281806 in unity (Ubuntu) "Alt-Tab switcher regularly goes into "flicker loop of death" or just hangs" [Undecided,New]13:42
seb128nturner_, well, as said on #ubuntu-devel, no, nobody else mentioned that/saw that before afaik, you might want to check with bregma, he might have seen it mentioned13:43
ritzhappy weekend folks13:55
=== sil2100_ is now known as sil2100
=== MacSlow|lunch is now known as MacSlow
=== m_conley_away is now known as m_conley
TrevinhoLaney, seb128: ok, I'm back... And unfortunately I wasn't sleeping this time, but I had to go to the doctor to renew my driver license...14:37
* Trevinho is getting old14:37
Trevinhobut yeah, I do prefer hacking by night :D14:37
seb128Trevinho, good afternoon ;-)14:37
seb128bregma, btw I took over your unity landing ask to get a silo/kick a build, I hope it's ok14:38
Trevinhoanyway... About the resizing, unfortunately it's not easily controllable14:38
bregmaseb128, I saw that, I'm currently running the usual AP tests14:38
seb128good14:38
seb128the update seems to run fine for me14:39
Trevinhoalso switching monitor selection can change it, but this depends on the grid we use... And I didn't find any valid workaround to avoid that, without breakiing the grid in some cases..14:39
Trevinhobut if you've suggestions...14:39
Trevinhoin theory all the elements are set not to expand or fill... so they should resize14:40
Trevinhobut theory might not match reality14:40
LaneyTrevinho: hmm, not sure, maybe ask in #gtk+14:59
seb128Laney, Trevinho: I've u-c-c in a landing silo, did you have any reason to not land it? as said we can do another landing next week for tweaks and fixes if needed15:00
Trevinhoseb128: I'm happy if it lands :)15:01
bregmaseb128, I've marked the Unity landing (landing-001) as "tests passed", shall I ask for a publish?15:02
seb128bregma, I can publish for you15:02
bregmacool15:03
Laneyseb128: yeah, do it15:03
seb128bregma, note that Trevinho had a review comment on the compiz change, but I guess that can be addressed in another mr later on, I don't feel like respinning the whole update for that15:03
seb128bregma, published15:04
seb128Laney, thanks15:04
seb128Laney, do you have any comment on https://code.launchpad.net/~seb128/sessioninstaller/current-gstreamer-tweaks/+merge/213163 ?15:05
LaneyNeed to look at it, hang on15:06
seb128Laney, thanks15:07
seb128Laney, the version split is probably not ideal15:07
seb128but I didn't feel like changing the gstreamer logic/rebuilding all the gstreamer plugins15:07
seb128like the code considers 1.0 as the serie, but we flag the packages 1.215:08
seb128in practice we have only 0.10 and 1. anyway, so split['.'][0] gives 0 or 1 and that's going enough for the current situation15:08
hikikoseb128, and Laney15:37
hikikohere?15:37
seb128hikiko, hey15:37
hikikohey :)15:37
hikikoI am going to add a dialog in the u-c-c when the scaling is so large or so small that the u-c-c window exceeds the screen boundaries, a dialog will appear asking the user to confirm the setting15:39
LaneyHey15:40
hikikoso that he can revert in extreme scaling but doesnt have to press confirm each time he moves the slider15:40
Laneywhy not make it consistent with the resolution settings?15:40
Laneynot instant apply, revert if not confirmed15:40
hikikoLaney, because the resolution settings15:40
hikikobecause the slider is interactive15:40
hikikoand the resolution settings can revert on apply15:41
hikikoand we revert the whole desktop settings15:41
hikikothis dialog will only appear in extreme cases15:41
hikikowhen you use a scale factor 2 for example15:41
hikikoand your text is huge and you cant scroll down15:41
hikikoor when the scale factor is too close to 0.515:42
hikikoand the win goes that small you cant drag it15:42
hikikonot while you move tha slider15:42
hikikoI think it's a good idea isnt it?15:42
seb128Laney, if we have linear scaling it feels like validation-try would be tedious15:42
seb1281.15, no, 1.20, no, 1.25, no15:42
seb1281.30, oh, that looks ok15:42
seb128how was 1.25 again15:42
seb128etc15:42
seb128hikiko, not sure what you want to do is easily doable/reliable though15:44
seb128hikiko, one other option would be to revert if the dialog is closed by the (x) wm control, but I guess that would be weird15:44
hikikoit's just a dialog that sets the gsettings/ui_scale value to the previous one15:44
hikikoI already store the prev value15:44
hikiko:)15:44
hikikoit's easy, the only tricky part is to use the resolution dpi to check if the win is inside the screen boundaries15:45
hikikobut it's not too difficult15:45
hikikoI think it's a good feature15:45
seb128yeah, we need a way to revert anyway15:46
LaneyI don't think it's that tedious, it's not like you do this operation all the time15:46
seb128can you open a bug for tracking?15:46
seb128maybe check with mpt for design recommendations15:46
hikikosure seb12815:46
seb128Laney, no, but it's sort of nice of move the slider and seeing the result to get to the right ration ... but agreed, having to do it by tries wouldn't be the end of the world15:47
seb128Laney, don't forget my few liners codec review please ;-)15:47
LaneyI am going to in a minute15:47
seb128Laney, I'm mostly asking for review because of the split hack, in case you have an opinion, otherwise I can just upload15:47
seb128it's a small/safe change anyway15:48
seb128Laney, I put some context on the channel a bit earlier in case you didn't notice ;-)15:48
seb128thanks15:48
LaneyI would try and find out where struct.version (or whatever it is) comes from15:48
=== Ursinha is now known as Ursinha-afk
Laneybut I think it's fine to only compare the major versions if that is what they use for abi compatibility anyway15:49
seb128it's coming from gstreamer15:49
seb128I think it's the "current serie"15:49
seb1281.2 versions are 1.0 abi serie15:49
seb128so we have a mismatch between the packages and the abi version15:49
seb128one way would be to fix the description flags and rebuild all the plugins15:49
LaneyDepends what Gstreamer-Version is meant to be15:49
seb128well, nothing tell us that they are not going to change abi at 1.615:50
seb128yeah, I don't know, I didn't find a definition of what Gstreamer-Version is meant to be15:50
seb128I should ping slomo about it when he's around15:50
Laneyyeah I'm pretty sure that will never happen15:50
Laneythey've said many times that 1.x is going to remain compatible15:51
seb128right15:51
LaneyProbably ideally there would be a Gstreamer-ABI field like the -Version one, but in reality the major version is that anyway15:52
Laneyso, do it15:52
seb128Laney, thanks15:55
=== Ursinha-afk is now known as Ursinha
hikikoseb128, Laney https://bugs.launchpad.net/unity-control-center/+bug/129907915:58
ubot2Launchpad bug 1299079 in Unity Control Center "allow UI scale revert" [Undecided,New]15:58
hikikohere it is15:58
seb128hikiko, thanks15:58
seb128mpt, ^ if you have any opinion on the best interaction mode for that we would welcome hearing it :-)15:58
LaneyI replied with the alternative15:59
Laneyhikiko: We usually assign mpt to get a design first, doing that15:59
LaneyMatthew Paul Thomas did not previously have any assigned bugs in Unity Control Center.16:00
hikikook :)16:00
Laneymuhahaha16:00
hikikoI can find him16:01
hikikoto assign it16:01
hikiko:s16:01
Laneyi did it16:01
seb128shrug16:02
seb128we need to close that projects for bug report16:02
hikikompt, could you please get a look so that I start it on monday? :)16:02
Laneyoh yeah, i didn't notice that16:02
seb128reassigned to the ubuntu package16:03
hikikohahaha16:03
mpthikiko, ok16:03
seb128mpt, thanks16:04
* mpt wonders if this is going to be just another case of “undo your last System Settings change”16:04
mptoh, it’s more specialized, okay16:04
LaneyIt's something the spec doesn't talk about16:05
Laneyso you should be happy getting to think about a new thing :P16:05
hikikompt, no it will only appear in extreme scales16:10
hikikonot everytime the user sets the slider16:11
hikikothat would be annoying16:11
TrevinhoLaney, dpm: something else for you :|  https://bugs.launchpad.net/unity/+bug/129908016:14
ubot2Launchpad bug 1299080 in Unity "[UIFe] New style force-quit window dialog" [Medium,Confirmed]16:14
seb128time for some exercice, be back in 1 hour16:18
LaneyTrevinho: ok, you didn't manage to reuse the strings then?16:21
LaneyTrevinho: force *the* application to exit16:22
dpmTrevinho, added comment to the bug16:22
TrevinhoLaney: well, it would look not so good... As the contained the name of the window And this is a repeated thing I think16:22
dpmTrevinho, I also think the *the* or a *this* is needed16:23
Trevinhodpm, Laney: yeah my tyopo16:23
=== gatox is now known as gatox_lunch
dpmok, no worries16:23
Laneyrepeated between the decoration and the dialog you mean?16:24
TrevinhoLaney: yeah16:24
TrevinhoLaney: it also would make it to grow too much if too long16:24
Trevinhoand well, cutting it is not that nice16:24
LaneyOK, I guess you weighed it against breaking translations16:24
seb128ok, really going this time, bbl16:30
tremseb128: \o/16:37
* Sweetshark sees seb128 is still here and tries to thing of a topic lock him down here for a bit longer ...17:03
Sweetsharkhttp://xkcd.com/356/17:03
desrtso you can start by establishing some loose bounds quite easily17:12
desrtthe solution must be between 0.25 and 317:13
desrt0.5 and 3, in fact17:13
* desrt resists the urge to get trapped17:14
=== gatox_lunch is now known as gatox
Laneyhappy weekend!18:02
* Laney phases out18:02
elfyhave a good one18:03
seb128Laney, thanks, you too!18:03
=== FJKong is now known as FJKong_afk
=== Ursinha is now known as Ursinha-afk
hyperairhmm, how do i tell unity not to clobber my circular scrolling settings?19:14
BigWhaleHoly codecs, Batman! There's no faac support in gstreamer1 in 14.04?19:15
=== Ursinha-afk is now known as Ursinha
=== gatox_ultra is now known as gatox
Sweetsharkdesrt: ah, almost got you. May I interest you in a game of http://gabrielecirulli.github.io/2048/ ?22:00
sarnoldit's a trap!22:00
sarnoldyou'll lose your weekend if you start :)22:00
Sweetsharksarnold: shshshsh22:03
sarnoldSweetshark :)22:04
=== moCrazyLe is now known as CrazyLemon
=== darkxst_ is now known as darkxst
=== m_conley is now known as m_conley_away
=== Ursinha is now known as Ursinha-afk
=== Ursinha-afk is now known as Ursinha
BigWhale:q23:57

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