| === jhodapp is now known as jhodapp|afk | ||
| jamesh | Saviq: re. bug 1306343, installing unity8-fake-env gets rid of the immediate problem. | 07:01 |
|---|---|---|
| ubot5 | bug 1306343 in unity8 (Ubuntu) "unity-scope-tool segfaults with message "(dlopen: cannot load any more object with static TLS"" [Undecided,In progress] https://launchpad.net/bugs/1306343 | 07:01 |
| RAOF | Man, there's nothing as awesome as TLS and the ELF ABI for same. | 07:10 |
| === zbenjamin_ is now known as zbenjamin | ||
| jamesh | RAOF: things work fine if you don't fiddle with the defaults: -fPIC defaults to -ftls-model=global-dynamic, which should be fine to dlopen | 07:19 |
| jamesh | that is admitedly slower than -ftls-model=initial-exec, but you only get a few slots for initial-exec .so's from dlopened libraries. | 07:20 |
| RAOF | jamesh: But, but, but mesa! That's a whole extra layer of indirection eating into your sweet, sweet GL performance! | 07:20 |
| jamesh | RAOF: presumably it is fine if you directly link to libGL though? | 07:20 |
| RAOF | Yeah. | 07:20 |
| RAOF | You *should* be able to dlopen libGL, though; many things do. Possibly not libGL *and* libEGL *and* libGLES, though. | 07:21 |
| jamesh | from what I read, there are a few free static-tls slots reserved for dlopen libraries | 07:22 |
| jamesh | once you use them up though, you don't get any new ones | 07:22 |
| RAOF | Right; you can't get new ones, because they're in memory allocated at binary load time. | 07:26 |
| jamesh | do we really need this "dlopen" function? :) | 07:28 |
| jamesh | all it seems to do is cause problems | 07:28 |
| pete-woods | :D | 08:00 |
| pete-woods | tsdgeos: hi, so this CPU usage thing in HUD, the one you're showing me can only be triggered when you actually search for stuff | 08:01 |
| pete-woods | tsdgeos: what things are you searching for in Kontact? I've tried giving the app a good attack with HUD, and not managed to re-create the bug yet | 08:02 |
| tsdgeos | pete-woods: don't use the hud myself | 08:02 |
| tsdgeos | i don't remember hwo to even trigger it tbh | 08:02 |
| tsdgeos | since i think i moved it from alt to somewhere els | 08:02 |
| pete-woods | tsdgeos: dammit, then how is the search being triggered :( | 08:02 |
| tsdgeos | pete-woods: ahh, unity8? | 08:02 |
| tsdgeos | that maybe triggers an "" or QSTring() search on start? | 08:03 |
| pete-woods | tsdgeos: it could be that, yes | 08:03 |
| tsdgeos | i don't search in there either | 08:03 |
| tsdgeos | but it may be doing some searches by itself | 08:03 |
| pete-woods | tsdgeos: but your plugin has a long-lived search that holds open a search for "" | 08:03 |
| tsdgeos | think so yes | 08:03 |
| tsdgeos | jamesh: i can't repro that problem with unity-scope-tool :/ | 08:03 |
| pete-woods | hmm, it's possible that's why you're seeing it.. hmm | 08:04 |
| tsdgeos | jamesh: you just run it and it outputs that? | 08:04 |
| jamesh | tsdgeos: I didn't have unity8-fake-env installed | 08:04 |
| jamesh | tsdgeos: without that package, unity-scope-tool reliably crashed. With it installed, it functioned as expected | 08:04 |
| tsdgeos | it needs the fake application manager | 08:05 |
| tsdgeos | yes | 08:05 |
| tsdgeos | there's no dependency i gather | 08:05 |
| tsdgeos | we'd need one | 08:05 |
| jamesh | tsdgeos: it does seem worrying that we'd dlopening a set of libraries that triggers that error though | 08:05 |
| tsdgeos | jamesh: why? | 08:05 |
| * tsdgeos kicks email | 08:07 | |
| tsdgeos | why i have only one email about this scope-tool thing | 08:07 |
| tsdgeos | and there have been 7 comments | 08:07 |
| tsdgeos | :/ | 08:07 |
| jamesh | tsdgeos: initial-exec mode TLS isn't guaranteed to work with dlopen, but a few slots are reserved so you can open a few libraries configured that way. If we're hitting this limit in an edge case, it makes me wonder if it will come back to bite later on | 08:08 |
| tsdgeos | i honestly think you have a red-herring there, it just crashes because we make the app exit if you use the wrong applicationManager in the wrong app | 08:10 |
| tsdgeos | but i may be wrong of course :) | 08:11 |
| jamesh | perhaps the bug just leads us to load up a combination of DF_STATIC_TLS libs that pushes us over the limit, when all Mir or all X11 would be fine | 08:12 |
| tsdgeos | Cimi: wait is gone | 08:32 |
| Cimi | tsdgeos, what? | 08:33 |
| tsdgeos | https://code.launchpad.net/~aacid/unity8/carouselLastItemClick/+merge/214230 | 08:33 |
| Cimi | tsdgeos, I've been working for the wizard the whole week | 08:35 |
| Cimi | tsdgeos, I started having a look last niight | 08:35 |
| tsdgeos | i know, you commented in it | 08:35 |
| tsdgeos | and i am answering your comment :) | 08:35 |
| Cimi | tsdgeos, ahahahahah | 08:38 |
| Cimi | tsdgeos, I thought you were like "I've waited enough your review" :D | 08:38 |
| Cimi | I forgot I reviewed a bit last night | 08:39 |
| tsdgeos | .D | 08:42 |
| Cimi | tsdgeos, also, not sure trycompare is enough :) | 08:51 |
| Cimi | tsdgeos, how long it last? | 08:52 |
| Cimi | tsdgeos, can we have a flick that lasts more than 5 seconds? | 08:52 |
| tsdgeos | here or in teory? | 08:52 |
| Cimi | here | 08:53 |
| Cimi | on jenkins | 08:53 |
| tsdgeos | i don't see how | 08:53 |
| tsdgeos | given we have a few similar tryCompares | 08:58 |
| === om26er is now known as om26er|lunch | ||
| === om26er|lunch is now known as om26er | ||
| Saviq | jamesh, right, so you must've had libunity-mir1 installed before installing scope tool, and so -fake-env didn't get in... | 10:14 |
| Saviq | hopefully the dash will soon not need the app plugin any more... | 10:14 |
| Saviq | s/hopefully/fortunately/ | 10:14 |
| tsdgeos | guys, do autopilot tests work for you on the desktop? | 10:22 |
| tsdgeos | i can't get it to unlock the greeter | 10:22 |
| tsdgeos | damnit, and it works now :S | 10:23 |
| tsdgeos | come on, that's not funny anymore :D | 10:34 |
| * tsdgeos tries to pin point what's killing its X | 10:34 | |
| tsdgeos | hmmmm, one by one they don't do it | 10:37 |
| * tsdgeos tries again all in a row | 10:37 | |
| tsdgeos | and there you go, killed again :S | 10:40 |
| tsdgeos | i guess i'll run them on the phone ^_^ | 10:44 |
| tsdgeos | Saviq: any clue why running the autopilot tests may be killing my X session back to login manager? | 10:44 |
| Cimi | tsdgeos, do you have debug? :) | 10:53 |
| tsdgeos | Cimi: what do you mean if i have debug? | 10:53 |
| Cimi | tsdgeos, .xsession-errors or so | 10:54 |
| Cimi | when it kills | 10:54 |
| Cimi | log | 10:54 |
| tsdgeos | ah | 10:54 |
| Cimi | or just run gdb from a tty and run the autopilot test | 10:54 |
| tsdgeos | not really | 10:54 |
| tsdgeos | gdb on what? | 10:54 |
| tsdgeos | on compiz? | 10:54 |
| Cimi | tsdgeos, whatever crashes | 10:58 |
| Cimi | tsdgeos, don't think compiz | 10:58 |
| Cimi | tsdgeos, if compiz goes down you don't go back to login | 10:58 |
| Cimi | seb128, tsdgeos runs autopilot tests and it hangs and he is logged out | 10:59 |
| Cimi | seb128, what process shall he check with gdb? | 10:59 |
| seb128 | check the logs first | 10:59 |
| Cimi | seb128, any idea which one? | 11:00 |
| seb128 | gnome-session or Xorg I would say | 11:00 |
| Cimi | tsdgeos, ^ | 11:00 |
| seb128 | .cache/upstart/gnome-session-Unity.log and /var/log/Xorg.0.log.old | 11:00 |
| tsdgeos | nothing obvious i can see there | 11:04 |
| tsdgeos | anyhow, running on phone it's ok | 11:04 |
| === MacSlow is now known as MacSlow|lunch | ||
| === jhodapp|afk is now known as jhodapp | ||
| Saviq | tsdgeos, sorry, no :| | 12:16 |
| Chipaca | hello all. Who should I pester about https://bugs.launchpad.net/ubuntu/+source/unity/+bug/1304265 ? | 12:21 |
| ubot5 | Ubuntu bug 1304265 in unity (Ubuntu) "Unity bugs out when changing screen size" [Undecided,New] | 12:21 |
| Cimi | Saviq, import "file:///usr/share/unity8/Notifications" as Notifications ok? | 12:33 |
| Saviq | dednick, Cimi, can you please look into bug #1306499 and bug #1301463 (assign yourself as applicable) | 12:37 |
| ubot5 | bug 1306499 in unity8 (Ubuntu) "Sound indicator is laggy #287 on mako" [Undecided,New] https://launchpad.net/bugs/1306499 | 12:37 |
| ubot5 | bug 1301463 in unity8 (Ubuntu) "Send button when responding to texts is grayed out and non-functional" [Undecided,New] https://launchpad.net/bugs/1301463 | 12:37 |
| Saviq | Cimi, with a huge FIXME, ok | 12:37 |
| dednick | Saviq: sure | 12:37 |
| Saviq | Cimi, and make sure to depend correctly on unity8 and whatnot | 12:37 |
| Cimi | Saviq, OK | 12:37 |
| Saviq | Cimi, maybe it'd make sense to add notifications to https://code.launchpad.net/~saviq/unity8/split-common/+merge/214907 too in that case | 12:41 |
| Cimi | Saviq, yup | 12:42 |
| Saviq | Cimi, please comment on the MP so that I don't forget | 12:42 |
| Cimi | Saviq, would be just for welcome wizard though | 12:43 |
| Cimi | or something else? | 12:43 |
| Saviq | Cimi, greeter, too | 12:43 |
| Saviq | Cimi, let's see | 12:43 |
| Cimi | Saviq, I'd be happy to have inputfilterarea if it's not already as a module (maybe) | 12:43 |
| Saviq | Cimi, that's from unity-mir, isn't it? | 12:43 |
| Cimi | Saviq, maybe indeed | 12:43 |
| Cimi | don't remember | 12:44 |
| === _salem is now known as salem_ | ||
| Saviq | mardy, hey, you might now... I'm trying to build new empathy, it's complaining about missing symbols - do you know where I could get them http://paste.ubuntu.com/7234231/ (bottom)? | 12:51 |
| tsdgeos | Saviq: in https://code.launchpad.net/~mzanetti/unity8/support-short-appid-in-gsettings/+merge/209071 you mention something in ApplicationManager ? | 13:00 |
| Saviq | tsdgeos, see description | 13:00 |
| tsdgeos | oh, so not really useful then | 13:01 |
| mzanetti | tsdgeos: re, no. not useful yet | 13:07 |
| * Saviq missed his transfer in FRA ;| | 13:07 | |
| tsdgeos | Saviq: what?! | 13:09 |
| tsdgeos | :/ | 13:09 |
| tsdgeos | Saviq: so what now? | 13:09 |
| tsdgeos | elopio: ping | 13:09 |
| Saviq | tsdgeos, flying to KTW in an hour, bus back home | 13:09 |
| Saviq | and yeah... /my fault, got into the lounge and fired up IRC... missed boarding by some 5 mins | 13:09 |
| mzanetti | Saviq: :/ but one hour seems still ok | 13:10 |
| Saviq | mzanetti, well, would have been home 1430, will be (hopefully) 2130 | 13:10 |
| * mzanetti missed the highway exit and only realized when passing the sign: Welcome to Bavaria. | 13:10 | |
| Saviq | ;) | 13:10 |
| mzanetti | Saviq: oh... then its quite a difference indeed :/ | 13:10 |
| Saviq | anyway, I got what was coming to me... crazy thing, though, is I helped a Polish guy get his Lufthansa-broken ticket (yes, to Wrocław) sorted out | 13:11 |
| syst3mw0rm | Where can I read more about appindicator? | 13:11 |
| Saviq | older guy is going back home after 6mths in India | 13:11 |
| mzanetti | :) | 13:11 |
| Saviq | and the flight he had the ticket for didn't exist any more, 6 months after booking | 13:12 |
| tsdgeos | ouch | 13:12 |
| Saviq | and he doesn't speak languages much, so it must've been fate ;) | 13:12 |
| tsdgeos | yeah that's karma for you | 13:12 |
| tsdgeos | help dude, get stranded | 13:12 |
| Saviq | other way | 13:12 |
| Saviq | get stranded, help dude | 13:13 |
| tsdgeos | ah | 13:13 |
| tsdgeos | ok, then it works | 13:13 |
| Saviq | only met him at the service counter | 13:13 |
| tsdgeos | had the same when i got stranded in london last time | 13:13 |
| tsdgeos | spanish guy flying to brazil with noone speaking spanish in heatrow (that was weird) and he knowing 0 english | 13:13 |
| * mzanetti waits for his uncle to get lost in bilbao in 2 weeks :) | 13:14 | |
| tsdgeos | i'm not flying there to help him, sorry but it's far man | 13:15 |
| mzanetti | wouldn't help anyways. neither spanish nor english | 13:16 |
| mardy | Saviq: those symbols are from libsignon-glib and libaccounts-glib | 13:19 |
| tsdgeos | Cimi: ping | 13:20 |
| Cimi | tsdgeos, pong | 13:20 |
| tsdgeos | Cimi: commit message and checklist missing in https://code.launchpad.net/~cimi/unity8/lp1301309/+merge/214810 | 13:20 |
| tsdgeos | Saviq: what's https://code.launchpad.net/~saviq/unity8/split-common/+merge/214907 for? | 13:21 |
| tsdgeos | for wizard? | 13:21 |
| tsdgeos | or? | 13:21 |
| Saviq | tsdgeos, scope tool, see bug | 13:28 |
| Saviq | tsdgeos, but yeah, wizard, too, now, for notifications | 13:28 |
| tsdgeos | i see | 13:29 |
| tsdgeos | anyway i guess with didrocks and Cimi doing the review i don't need to jump on board :) | 13:29 |
| * Saviq goes | 13:32 | |
| cwayne | Saviq, ping | 13:33 |
| tsdgeos | Cimi: how i test that fix? | 13:36 |
| Cimi | tsdgeos, previews or scope tool? | 13:37 |
| tsdgeos | Cimi: previews | 13:37 |
| Cimi | tsdgeos, I did not test it | 13:37 |
| Cimi | tsdgeos, but works for sure | 13:37 |
| tsdgeos | Cimi: then stop lying in the checklist :D | 13:38 |
| Cimi | tsdgeos, everybody lies :D | 13:38 |
| tsdgeos | Cimi: i'll let a lier to approve it then | 13:41 |
| Cimi | tsdgeos, I can approve! | 13:42 |
| === dpm is now known as dpm-afk | ||
| didrocks | Saviq: -common should depends on misc | 14:01 |
| tsdgeos | didrocks: please comment on the MR, he's on a plane atm | 14:03 |
| didrocks | tsdgeos: yeah, I've done it | 14:04 |
| tsdgeos | cwayne: he's on a plane, can we help? | 14:04 |
| didrocks | thanks | 14:04 |
| Cimi | MacSlow, I am reusing code from the shell for the input filter area | 14:14 |
| Cimi | MacSlow, but weirdly it's not working | 14:14 |
| Cimi | MacSlow, height: parent.contentHeight | 14:15 |
| Cimi | onHeightChanged: console.log(height) | 14:15 |
| Cimi | blockInput: height > 0 | 14:15 |
| MacSlow | Cimi, the wifi-wizard specific notifications?! | 14:15 |
| Cimi | MacSlow, height is 0 | 14:15 |
| Cimi | but still blocks input | 14:15 |
| MacSlow | Cimi, sure you don't touch (alter) blockInput elsewhere by accident... | 14:16 |
| Cimi | MacSlow, nope.. | 14:16 |
| MacSlow | Cimi, just guess here atm... would need to look at all the code... | 14:16 |
| MacSlow | Cimi, branch? | 14:16 |
| Cimi | MacSlow, http://paste.ubuntu.com/7235083/ | 14:16 |
| Cimi | MacSlow, it's weird | 14:17 |
| MacSlow | Cimi, set block explicitly to false and test that first... | 14:18 |
| cwayne | tsdgeos, was just wondering if there were plans for differentiating the ubuntu button at the bottom of the launcher | 14:19 |
| Cimi | MacSlow, I did and works | 14:19 |
| MacSlow | Cimi, ok... | 14:19 |
| cwayne | i'd seen some design mockups, but was unsure if it was planned | 14:19 |
| Cimi | MacSlow, it's weird | 14:19 |
| tsdgeos | cwayne: don't know much about that i'm afraid, you'll have to wait for him | 14:19 |
| MacSlow | Cimi, set height to 0 | 14:19 |
| Cimi | MacSlow, blockinput turns false | 14:19 |
| MacSlow | Cimi, good | 14:19 |
| Cimi | http://paste.ubuntu.com/7235101/ | 14:20 |
| MacSlow | Cimi, so parent.contentHeight should then be what you neend to look after | 14:20 |
| Cimi | onHeightChanged: console.log(height) | 14:20 |
| Cimi | blockInput: height > 0 | 14:20 |
| Cimi | onBlockInputChanged: console.log(blockInput) | 14:20 |
| Cimi | BUT | 14:20 |
| Cimi | still blocks input | 14:20 |
| MacSlow | hm... | 14:21 |
| Cimi | I know it's super weird | 14:22 |
| MacSlow | Cimi, in the console.log() spit out height and parent.contentHeight | 14:22 |
| Cimi | you see height | 14:22 |
| Cimi | http://paste.ubuntu.com/7235101/ | 14:23 |
| Cimi | so ehight is 0 | 14:23 |
| Cimi | and blockInput is FALSE | 14:23 |
| MacSlow | weeee... my GoogleDocs-page is in Japanese! | 14:38 |
| MacSlow | Cimi, so does the work-around with visible: work? | 14:38 |
| Cimi | MacSlow, trying now | 14:38 |
| MacSlow | what's the recommended doc (or source) to learn about input in mir? | 14:39 |
| Cimi | no idea :D | 14:40 |
| Cimi | MacSlow, looks like input hangs or so | 14:40 |
| MacSlow | Cimi, how about in a Timer toggling the inputfilterarea's blockInput directly | 14:45 |
| MacSlow | Cimi, just to get closer to the real issue | 14:45 |
| Cimi | MacSlow, but the blockInput is off | 14:45 |
| Cimi | MacSlow, I'll do without it | 14:47 |
| Cimi | MacSlow, it's used only in the wifi page | 14:47 |
| MacSlow | Cimi, still an odd bug | 14:47 |
| Cimi | it definitely is | 14:48 |
| MacSlow | tsdgeos, am I imagining things or is jenkins especially slow today? | 14:48 |
| tsdgeos | MacSlow: it's a bit random with slowness tbh | 14:48 |
| tsdgeos | sometimes it's ultra fast and others its slooooooow | 14:49 |
| tsdgeos | MacSlow: is it even running? | 14:49 |
| MacSlow | tsdgeos, oh it certainly is... http://s-jenkins.ubuntu-ci:8080/job/unity8-ci/2802 | 14:50 |
| MacSlow | but slow | 14:50 |
| tsdgeos | yeah | 14:52 |
| tsdgeos | there you go | 14:52 |
| tsdgeos | it just finished, no? | 14:52 |
| MacSlow | tsdgeos, hm.. everything failed -> https://jenkins.qa.ubuntu.com/job/autopilot-testrunner-otto-trusty/4142/?#showFailuresLink | 14:54 |
| tsdgeos | that's otto | 14:54 |
| tsdgeos | ignore him | 14:54 |
| tsdgeos | focus on the phone :D | 14:55 |
| MacSlow | tsdgeos, ok :) | 14:55 |
| tsdgeos | i have no clue what's wrong with otto, but oh well, someone that is not me will have to worry about it | 14:55 |
| MacSlow | we should call the servers after fictional characters... | 14:55 |
| MacSlow | tsdgeos, :) | 14:55 |
| Cimi | dednick, ubuntu settings components doesn't work with sbuild | 15:21 |
| Cimi | error http://paste.ubuntu.com/7235366/ | 15:22 |
| josharenson | I'm trying to run a clean build of trunk and I get a symbol lookup error. Any pointers? | 15:30 |
| Cimi | josharenson, paste the rror | 15:33 |
| josharenson | cimi: http://pastebin.ubuntu.com/7235408/ | 15:34 |
| Cimi | josharenson, looks like you need to recompile something, seb128 might tell you what | 15:35 |
| seb128 | josharenson, do you have libgsettings-qt1 installed and which version? | 15:37 |
| Cimi | dpkg -l qtdeclarative5-gsettings1.0 | 15:37 |
| Cimi | 0.1+14.04.20 is what I have | 15:37 |
| dednick | Cimi: hm. never tried it with sbuild before. | 15:38 |
| seb128 | Cimi, no, the lib | 15:38 |
| dednick | Cimi: working with dpkg for you? | 15:38 |
| josharenson | qtdeclarative5-gsett 0.1+14.04.20140 amd64 | 15:39 |
| josharenson | same | 15:39 |
| Cimi | josharenson, recompile unity? | 15:47 |
| josharenson | cimi: working on it | 15:47 |
| seb128 | josharenson, what about "libgsettings-qt1" | 15:47 |
| josharenson | same version | 15:48 |
| josharenson | seb128 cimi, works now | 15:53 |
| elopio | tsdgeos: pong. | 15:53 |
| tsdgeos | elopio: see comment at https://code.launchpad.net/~elopio/unity8/use_fake_instead_of_messaging/+merge/214615 | 15:54 |
| elopio | tsdgeos: done. Thanks for reviewing! | 16:03 |
| tsdgeos | elopio: yw | 16:06 |
| dednick | larsu: hey. re https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1306499 . It's not the UI | 16:16 |
| ubot5 | Ubuntu bug 1306499 in unity8 (Ubuntu) "Sound indicator is laggy #287 on mako" [Undecided,New] | 16:16 |
| dednick | unity8 does the same after a dist-upgrade | 16:16 |
| dednick | larsu: think it's account services | 16:17 |
| Cimi | vaffanculo | 16:45 |
| Cimi | vaffanculo | 16:45 |
| === gatox is now known as gatox_lunch | ||
| === dpm-afk is now known as dpm | ||
| === gatox_lunch is now known as gatox | ||
| medikus | Hey, how would I change the shortcuts for moving and resizing windows from alt to super? | 19:01 |
| === salem_ is now known as _salem | ||
| === jhodapp is now known as jhodapp|afk | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!