/srv/irclogs.ubuntu.com/2012/04/25/#ubuntu-unity.txt

=== andyrock is now known as andyrock|zzz
=== jalcine- is now known as Jacky
=== Jacky is now known as Guest58047
=== Guest58047 is now known as Jacky
thomibschaefer: pong from earlier... although it's a holiday for me :)03:56
bschaeferyeah I looked it up :)03:56
bschaeferthomi, no worries, I just got done with class.03:57
bschaeferI can talk to you tomorrow if you want a day off ;)03:57
thomiit's cool - I'm about to start hacking on some personal projects03:57
bschaefercool!03:57
thomiSo, Ctrl+Space isn't activating ibus in the VM huh?03:58
thomithat's odd03:58
bschaeferIm excited to start one up soon03:58
bschaeferyeah03:58
bschaeferif you look at the error messge03:58
* bschaefer goes to get it...03:58
thomibschaefer: heh, if you're looking for something to do I can suggest several projects for you :P03:58
thomibschaefer: I wonder if we can get a test system that's closer to the one that jenkins runs on - maybe we can reproduce some of these intermittently failing tests as well.03:59
bschaeferthomi, well I might be writing a compiler this summer, or using nux to make a maze solver/generator03:59
thomiahh, cool03:59
bschaeferthomi, AssertionError: After 10 seconds test on SearchBar.im_active failed: True != dbus.Boolean(False, variant_level=1)03:59
bschaeferso03:59
bschaeferfails on this line AssertionError: After 10 seconds test on SearchBar.im_active failed: True != dbus.Boolean(False, variant_level=1)03:59
bschaeferopps03:59
bschaeferdidn't copy haha03:59
bschaefer  self.assertThat(widget.im_active, Eventually(Equals(True)))04:00
thomiyeah, so either ibus isn't being activated, or we're not getting the signal04:00
bschaeferyeah04:00
thomiI wonder if there's anything we can do to narrow down the cause?04:00
bschaeferI was thinking maybe a key is getting held down but I couldn't find a key that you could hold downt o cause that04:00
bschaeferyeah I was wondering that too04:00
thomiyou should be able to see the timings in the test log04:00
bschaeferso it fails on waiting for ibus to be true04:01
thomiit logs a line every time it presses and releases a key04:01
bschaeferhmm04:01
thomimaybe the delay is too long or something?04:01
bschaeferwell I wanted to add some debuging messages in but I guess you need to merge it to trunk for that04:01
bschaeferthomi, but it works locally04:01
bschaeferwell04:01
bschaeferlet me try a super long wait04:01
thomigood idea04:01
bschaeferooo thomi hold down ctrl+space04:02
bschaeferjust hold it down and look at the indicator04:02
thomiI don't get an indicator... hang on04:02
bschaeferhmm well it toggles on/off04:03
bschaeferand I got a random crash doing that...04:03
thomiyeah, I see04:03
bschaefervery odd that it doesn't fail on the local build though04:03
bschaeferI tried it in my VM today also04:04
thomican you do some subtraction from the trimestamps in the test log and see if the timings are unreasonable?04:04
bschaeferlowering the ram to like 300 mb04:04
bschaeferwell it hold it for 0.204:04
bschaeferwhich I think is like 200 ms?04:04
thomiyeah04:04
bschaeferthat seems fast to me04:05
thomithat doesn't seem unreasonable to me04:05
bschaeferyeah04:05
bschaeferI wonder04:05
thomiheh, I was going to say it's on the slow side :)04:05
bschaeferif the VM is just slow04:05
bschaeferthough04:05
thomithat's 1/5 of a second.04:05
bschaeferyeah haha, but I/O is always slow haha04:05
bschaeferhmm04:06
bschaeferIll see what happen is I just set it to 0.004:06
bschaeferor maybe 0.05 haha04:06
bschaeferthomi, we should have a dx-autopilot-test-jenkins04:07
bschaefernot being able to debug the VM is not very nice04:07
thomibschaefer: if you want to set it up I'm sure we can make it happen04:08
thomi... or even set it up on our own jenkins instance.04:08
bschaeferthomi, Ill have to look into that but that would be awesome04:08
bschaeferI also need to show you how I set up a VM for unity 3d04:08
bschaeferthomi, also do you know when the last time the VM was reset?04:08
thomiI have a jenkins instance for my own projects. I'll talk to people and see how hard it'll be to get running on the PA jenkins machine04:09
bschaeferlike just rebooted04:09
thomiit gets reset each run04:09
thomithe dx-autopilot-build-nfsroot job reboots the machine and boots it with a clean image04:09
thomi...but that's not the case if we just trigger the dx-autopilot-run job directly04:09
bschaeferhmm good04:10
bschaeferI just want to eliminate as many possible problems as possible04:10
bschaeferthomi, hmm it seems to work at 0.0504:11
bschaeferill push it 0.004:11
thomi...assuming the issue is that it's too slow. Worth a shot though04:12
bschaeferyeah04:13
bschaeferand 0.0 works fine locally04:13
bschaeferthomi, there are also some debugging messages in nux I could uncomment....04:13
bschaeferbut I don't want to push just debugging messages to nux trunk04:13
thomibschaefer: hmmmm. It'd be nice if there was a copy of the jenkins job we could customise...04:14
bschaeferyeah I know...all I want is to be able to reproduce this problem04:14
bschaeferALL of the test are failing04:14
bschaeferbut I cant get 1 to produce anything jenkins says...04:14
bschaeferand it is very constant04:14
thomihmmm04:15
bschaeferim just trying to think of anything that could cause that04:15
thomiif you were feeling brave you could recreate the jenkins config on your local machine04:15
bschaefersure, I can set up a VM haha04:15
bschaeferI actually have a  VM I can toast04:15
thomido you have access to the private jenkins instance?04:15
bschaeferno :(04:15
bschaeferI could walk you through how to setup a VM sometime soon04:16
bschaeferthough I use windows for my VM04:16
thomiOK, I can grab the job details and paste them somewhere for you04:16
bschaeferok cool04:16
bschaeferthomi, ugg my schools internet is crap04:23
thomi:(04:23
bschaeferjust randomly kicked off, anyway did I miss anything?04:24
thominope04:24
bschaefercool04:24
thomiI made fun of you, but that was it04:24
bschaeferIm happy I missed that then04:24
bschaeferhmm should I push these ibus test with the time press set to 0.0?04:25
bschaeferim also wondering if I could use nux to time the inbetween im_active gets active04:25
bschaeferand deactive04:25
bschaeferbut that might not even be the problem :(, you would think it would work at lease once if it was a timing issue04:26
thomibschaefer: I think ideally we'd be able to reproduce the issue before pushing random changes to trunk.04:27
thomibschaefer: maybe push 0.1 or 0.05 instead? seems like a more sensible change04:27
bschaeferthomi, and guess what I did :)04:28
bschaeferI set the time to 1.004:28
bschaeferand it seem to mimic those that problem04:28
thomioooh, that's interesting04:28
bschaeferlet me run the Ignore test to see if that works04:28
bschaefer    self.assertThat(widget.im_active, Eventually(Equals(True)))04:29
bschaeferAssertionError: After 10 seconds test on SearchBar.im_active failed: True != dbus.Boolean(False, variant_level=1)04:29
bschaefer:)04:29
bschaeferboth failed04:29
thomicool!04:30
bschaeferok04:30
thomiwould be interesting to see where the cutoff point is04:30
bschaeferso, why does 0.2 work for us but not jenkins?04:30
bschaeferyeah04:30
bschaeferIll start going closer to 0.204:30
thomibschaefer: probably because the jenkins instance is slow, so it's closer to 0.4 or somthing04:30
bschaeferthomi, ooo you were sooo close04:32
bschaeferit is 0.504:32
bschaeferthat is start failing fo re04:32
bschaeferfor me*04:32
bschaefercould be 0.4 on a different machine04:32
bschaeferthomi, but to be honest that seems like a bug in nux to me04:33
bschaeferI don't think it should toggle like that...04:33
thomibschaefer: probably to do with the tap detection code04:33
bschaeferyeah, or lack of04:33
bschaeferim not sure if there is one04:33
thomibetter not say that too loud, the compiz guys will ragequit04:33
bschaeferhaha04:33
thomibschaefer: ok, so maybe push a branch that sets it to 0.05 or so04:33
bschaeferalright, hopefully that helps!04:34
thomiping me when you have a MP to review04:34
bschaeferok04:34
bschaeferthomi, https://code.launchpad.net/~brandontschaefer/unity/ap-ibus-shorten-timer/+merge/10340704:40
bschaeferalso what were some other side projects were you doing?04:41
thomibschaefer: approved04:41
thomibschaefer: right now I'm working on http://launchpad.net/sloecodfe04:41
bschaefercool, this should fix 32 failures!!04:41
thomi*sloecode04:41
bschaeferthomi, that looks awesome04:42
bschaeferI actually need to learn something about networking04:42
thomibschaefer: we have a todo list that's insanely long :)04:43
thomiso uhh... "patches welcome" :)04:43
bschaeferthomi, I might take a look :)04:43
thomiwe hang out in #sloecode04:43
bschaeferI also might have to wait until I graduate04:43
bschaeferI do to many things atm haha, being a TA/Tutor takes  a lot of time along with classes and working here04:44
thomiyeha04:45
thomi*yeah04:45
bschaeferIll stop by that irc room when I get a chance04:46
bschaeferit is getting a bit late here and I should go eat04:47
bschaeferthomi, do you know when the next jenkins test should run?04:47
bschaeferI suck at waiting..04:47
thomibschaefer: nope, shouldn't be too long though04:47
thomiif you're running indicator-jenkins you can get a helpful popup :)04:48
bschaefercool, Ill check after dinner04:48
bschaeferoo04:48
bschaeferi should install that04:48
bschaeferwell im off to relocate back home, have fun programming :)04:49
=== Jacky is now known as [Jacky]
=== [Jacky] is now known as Jacky
tsdgeosdidrocks: ping-y07:12
didrockstsdgeos: hey hey07:20
tsdgeosdidrocks: was wondering, might it be that the merger thingy gets confused when a branch name is a subset of another?07:21
tsdgeosyesterday we merged https://code.launchpad.net/~aacid/unity-2d/load_24bit_pixbuf_icons_207:21
tsdgeosand https://code.launchpad.net/~aacid/unity-2d/load_24bit_pixbuf_icons was incorrectly set to merged, Gerry had to put it back to needs review07:21
didrockstsdgeos: it's not the merger it's launchpad AFAIK07:21
tsdgeosah07:22
didrocksif it sees commits that are in trunk, it triggered the status to "merged"07:22
tsdgeosok :-/07:24
=== Guest15469 is now known as Jens
=== Jens is now known as Guest28666
=== Guest28666 is now known as _jens
=== _jens is now known as jbache
=== jalcine- is now known as Jacky
=== andyrock|zzz is now known as andyrock
nik90hi, I am reading a bug report where a user's application and music daemons are not running09:07
nik90how do I ask him to start them?09:07
nik90so basically unity-applications-daemon and unity-music-daemon not running as background processes09:07
davidcallemhr3, hey09:17
mhr3davidcalle, hey, awesome you here, wanted to ask you something :)09:17
davidcallemhr3, ok, you first :)09:18
mhr3davidcalle, https://bugs.launchpad.net/bugs/95387109:18
ubot5Ubuntu bug 953871 in unity (Ubuntu) "[regression] Unity dash quicklist shows Utilities scope" [Low,New]09:18
mhr3is the lens doing something it shouldn't?09:18
mhr3or is that us?09:18
mhr3or perhaps it's not doing something it should ;)09:18
davidcallemhr3, should be invisible... I don't remember, is there an entry in the .lens for this? Because it's not set, so that may be the issue here.09:20
mhr3davidcalle, yea, it should have Visible=false09:20
davidcallemhr3, ok, fixed. People will have it in a few hours.09:21
mhr3ultimately we should be clever there, but we ain't so lenses need to provide this info :)09:21
davidcallemhr3, I feel like it's the third time I'm fixing this o_o Must be forgetting to push it each time.09:22
mhr3heh :)09:22
davidcallemhr3, is it possible to reload only one filter, without changing the state of the others?09:22
mhr3it *should*09:23
mhr3last time i was talking about it with ken, he said that unity doesn't react to the changes09:23
davidcallemhr3, I see.09:23
mhr3and i don't remember anyone fixing it on unity side09:23
mhr3davidcalle, try removing the filter and then re-add it09:24
davidcallemhr3, ok.09:24
mhr3although that will change the order i guess :/09:24
mhr3davidcalle, still, i think it should be possible to even remove all the filters and re-add them (if you keep the original references the state wouldn't change iirc)09:26
davidcallemhr3, yeah, that's what I was going to do.09:27
nik90mhr, can you please look at https://bugs.launchpad.net/bugs/98768909:43
ubot5Ubuntu bug 987689 in unity "dash: application tab disappeared" [Undecided,Incomplete]09:43
nik90its regarding the application and music lens daemon not running09:43
davidcallenik90, if the icons are missing, the lens have probably been uninstalled for some reason. Do you have unity-lens-music installed?09:45
nik90davidcalle, apparently the user states that both are installed but the daemons are not running for some reason09:45
nik90davidcalle: he checked the existence of the package using apt-cache policy unity-lens-application09:46
davidcallemhr3, what do you think ? ^09:48
nik90maybe I should ask him to just try unity --reset and see if it solves his problem??09:49
mhr3starting the lens daemons manually will probably spit some error09:50
mhr3would be nice to know what kind of09:51
mhr3nik90, ^^09:51
nik90mhr, how do you start the daemon manuall?09:52
nik90manually*09:52
mhr3/usr/lib/unity-lens-applications/unity-applications-daemon09:54
davidcalle/usr/lib/unity-lens-music/unity-music-daemon09:54
nik90mhr, davidcalle..nice synchronisation :)09:55
nik90I'll inform the user09:55
nik90thnx09:55
=== MacSlow is now known as MacSlow|lunch
=== greyback is now known as greyback|lunch
=== greyback|lunch is now known as greyback
=== _salem is now known as salem_
=== MacSlow|lunch is now known as MacSlow
=== salem_ is now known as _salem
mhall119more Hello Unity updates: http://mhall119.com/2012/04/hello-unity-episode-3-4/13:55
* mfisch just gave a presentation on Singlet14:10
mfischit was well received14:10
mhall119mfisch: awesome@14:10
mhall119!14:10
mfischmhall119: they wanted to know if singlet was here for the long haul, and I said yes, but of course have no proof14:10
mfischmhall119: the best part is that I presented it as "Singlet" and "The Old Way"14:11
mhall119mfisch: it's here at least as long as I am, which I certainly hope is long haul :)14:13
mhall119"The Old Way" isn't quite accurate14:13
mfischmhall119: true14:14
mfischmhall119: the "traditional way"14:14
mhall119it's also the only way for non-Python lenses and scopes, or anything more complex14:15
seb128mhr3, hey, do you know if bug #915203 is a lens side issue? a design decision?14:44
ubot5Launchpad bug 915203 in unity-lens-music (Ubuntu) "Unity 5.0: Not all music shows up in Music lens" [Undecided,Confirmed] https://launchpad.net/bugs/91520314:44
seb128mhr3, i.e having limited number of songs and albums listed14:44
mhr3seb128, i'm sure would like to put 20 thousand items in the lens, people who implemented the lens didn't14:45
mhr3..design would like...14:45
seb128mhr3, ok, so unity-lens-music is the right component? i.e it's coming from the backend side?14:46
mhr3yes, its implemented to list only up to 100 items14:46
seb128mhr3, thanks14:47
=== _salem is now known as salem_
=== yofel_ is now known as yofel
hyperairany lightdm devs around?15:15
hyperairhow does lightdm find out what wallpaper the user is using?15:15
mdeslaurhyperair: it asks accountsservice IIRC15:17
hyperairmdeslaur: i see. is it possible to get lightdm to pull the entire wallpaper image through this as well?15:17
hyperairmdeslaur: if you use a custom wallpaper in your $HOME (i don't see where else you would put it) and chmod o-r $HOME, lightdm refuses to log in.15:18
hyperairnot getting the wallpaper is one thing, but not being able to log in is ridiculous15:18
hyperairhttps://bugs.launchpad.net/ubuntu/+source/lightdm/+bug/98761415:18
ubot5Ubuntu bug 987614 in lightdm (Ubuntu) "Stuck at login screen with status "Logging in..."" [Undecided,New]15:18
hyperairwe have this bug here15:18
mdeslaurhyperair: well, that's a bug. It should be able to obtain the wallpaper so it doesn't bypass permissions, but shouldn't fail either15:18
mdeslaurs/should/shouldn't/15:18
hyperairit looks like background.vala:138 loads it directly in the unity greeter process15:19
mdeslaurhyperair: I'm not a lightdm developer, can't help you any further15:19
hyperairah, okay thanks anyway15:20
=== salem_ is now known as _salem
mfischmhall119: do you have a simple example of something too complex for singlet?15:51
mhall119mfisch: the video lens would be one15:58
mfischmhall119: whats the reason?15:58
mfischmhall119: trying to think of an example to tell people and explain why16:02
mhall119mfisch: mostly just because Singlet only covers simple cases16:04
mhall119you can do more complex stuff using Singlet, but you don't get any real benefit from using Singlet at that point16:04
mhall119it's like Quickly, good for small fast development, but once you reach a certain size it doesn't do you much good anymore16:05
mfischgot it16:21
=== elopio_ is now known as elopio
=== _salem is now known as salem_
=== Guest85884 is now known as Jacky
=== Jacky is now known as Guest70035
=== Guest70035 is now known as Jacky
=== Jacky is now known as Guest53736
=== Guest53736 is now known as Jacky
bschaeferlamalex, hey, so thomi and I might have figured out why the ibus test don't work18:19
lamalexnice!18:20
bschaeferI pushed some changes last night, but it didn't seem to fix it18:20
bschaeferit seems when you hold ctrl+space down, ibus toggles on/off18:20
lamalexok18:20
lamalexare we turning it off?18:21
bschaeferI think it is how fast we detect tabing18:21
bschaeferfor the hud18:21
bschaeferso events come through faster, so holding it down18:21
bschaeferwill push that event through nux multiple times18:21
bschaeferI have a different solution but Im not sure how much I like it :(18:23
bschaeferlamalex, o yeah to reproduce the problem on a local machine. Go into test_ibus.py and under def activate_ibus change the press/release time to 0.518:24
lamalexso it's purely timing issue?18:25
lamalexyarrr18:26
bschaeferthat is what it looks like...18:26
bschaeferbut that is the only way I can reproduce it on a local machine...18:26
bschaeferI changed the time to 0.05 last night, but it still fails so I think a fix is to iterate from 0.2 to 0.9 and try and press/release ctrl+space18:27
bschaeferif it activates its the go, else it will drop out and fail on an assert18:28
lamalexbschaefer, after doing that- now it fails without setting the time for me18:28
bschaeferhmm really?18:28
bschaeferwhat test are you running?18:29
lamalexautopilot.tests.test_ibus.IBusTestsPinyin.test_simple_input_dash18:29
bschaeferok, ill see if I can get the same18:29
bschaeferlamalex, hmm mine works after setting it back to default18:33
lamalexodd18:33
bschaeferyeah...18:33
bschaeferI have a branch that should find the correct time for presssing ctrl+space...18:34
bschaeferwhich should find a correct time, if the timing is the issue18:34
lamalexodd that this is the only place where key press timing is an issue18:37
lamalexso you think you understand the root cause of this?18:37
bschaeferI think it has to do with ctrl+space getting pushing into ibus multiple times in nux18:37
bschaeferyeah, the problem is Ctrl+Space can change to anything...18:38
bschaeferalso gnome seems to toggle it18:39
bschaeferbut just not as much18:39
bschaeferwell not on pinyin...it really is up to ibus engine then18:40
bschaeferlamalex, this is one possible fix...https://code.launchpad.net/~brandontschaefer/unity/ibus-ap-test-range-time/+merge/10354418:42
lamalexbschaefer, this is such a hack :\19:09
bschaeferlamalex, I know....Ill spend today looking into fixing this in nux19:10
lamalexthomi, review my branch when you're back! https://code.launchpad.net/~alexlauni/unity/unity.ap-switcher-emu/+merge/10333819:58
gemahey , wondering if unity is supposed to show the launcher bar and the top bar on all of my screens (3 of them)19:59
gemaand also if the mouse is supposed to get stuck when moving from one to the next (which I find rather annoying)20:00
seb128gema, yes and yes, and both can be changed in the system settings20:00
gemaseb128: you are a lifesaver20:00
seb128gema, well the top bar can't, but the display configuration dialog has a combo to select if the launcher should on one screen, which one in that case, or on all screens20:01
seb128gema, it has a config to turn off the stickyness between screens as well (which is there so you can easily get to a scrollbar against the border)20:01
gemaseb128: already much more usable for me20:02
gemaseb128: thanks20:02
seb128gema, yw!20:06
bschaeferthomi, ping20:10
thomibschaefer: pong20:13
thomilamalex: ok20:13
bschaeferhey20:13
bschaefersoo that fix didn't work :I(20:14
bschaefer:(20:14
lamalexwah20:14
bschaeferthomi, ill see if this is a possible fix in nux. The problem is there are multiple keys to activate/deactivate ibus and you can add/subtract from that list20:15
thomibschaefer: I saw - odd huh?20:15
bschaeferyeaah20:15
thomilamalex: is there any way to run an identical environment to the one the jenkins instance uses locally?20:15
bschaeferso it must have a min slow time, ie it is always 0.5 or higher20:15
bschaeferif it is a timing issue20:15
seb128thomi, hey20:16
lamalexthomi, i mean it should be possible to run the nfsbuild root script stuff locally20:16
lamalexbut im not sure how you'd boot the overlayfs image. you'd need a second machine to pxe boot20:16
thomilamalex: hmmm20:16
seb128thomi, lamalex, bschaefer: is there any chance one of you help writing a test for https://code.launchpad.net/~mvo/unity/usc-launcher-fix-982921/+merge/10225720:16
lamalexi guess really you need a jenkins master and slave set up in your apt20:16
bschaeferVirtualBox has pxe20:17
seb128it will not make into the first SRU otherwise20:17
seb128i.e mvo is maintaining software-center but is not familiar with the unity test framework and he's busy with the precise release (he's looking at i.e upgrade bugs as well)20:17
thomilamalex: ok, another idea - do you think we could duplicate the autopilot-run job into another jenkins job that we fire off manualkly? We could point it at a custom unity branch, so we can try things out and run the tests in an identical environment without having to merge with trunk.20:18
lamalexthomi, yah for sure20:18
lamalexjust make a new job and copy from dx-autopilot run and change the branch ha20:18
thomilamalex: we can just go ahead and do that, or do we need to ask someone?20:18
lamalexthomi, on it20:18
thomilamalex: thanks!20:18
thomibschaefer: now we can try some wacky fixes and not have to merge with trunk :)20:19
bschaefersweet!20:19
thomiseb128: looking into it...20:19
seb128thomi, thanks ;-)20:19
bschaeferI have a bad hack to check if it is a timing issue20:19
bschaeferthomi, https://code.launchpad.net/~brandontschaefer/unity/ibus-ap-test-range-time/+merge/10354420:20
bschaeferwe should be able to use that to check if/what is the timing problem20:20
thomibschaefer: you want to at least sleep before that if statement. It will take a while for unity to update20:21
thomiyou could use wait_for, but then you'd be sleeping 10 seconds each time it didn't work20:21
thomialso, the more I think about this, the less I think it's a timing issue20:21
bschaeferhmm yeah,20:21
thomiThe times seem similar to what works for me locally20:21
thomiand we're seeing 100% failure rate20:21
thomiso I suspect it's something a bit more basic.20:22
bschaeferyeah20:22
thomibschaefer: a vlue might be to look at what changed when they all started failing?20:22
bschaeferim hoping...20:22
bschaeferIll look through the logs20:22
lamalexthomi, where do we want this branch20:23
lamalexlp:~unity-team/unity/ibus-tests20:23
lamalexsound good?20:23
thomiyup20:23
thomilamalex: we can always tweak the jenkins job config to point it elsewhere, right?20:23
lamalexyah20:23
lamalexalso going to change the ap run command to just run ibus tests20:24
lamalexsave time20:24
bschaeferlamalex, awesome, thanks!20:24
lamalex(will do a test run to make sure we're still seeing same fails this way)20:24
bschaeferthomi, man, im looking at these test and it has been failing to activate ibus for a while...20:25
thomibschaefer: yeah20:25
lamalexok should be set to build on every commit20:28
bschaeferthomi, the first build to not activate ibus correctly20:35
bschaeferthough that is for pinyin only20:35
bschaeferhttps://jenkins.qa.ubuntu.com/job/dx-autopilot-run/695/#showFailuresLink20:35
thomibschaefer: ok, so what changed in that build? can you see the bzr revno?20:36
thomilook in the console output for the default configuration20:36
bschaeferyeah, I started looking for where they ALL started failing20:36
bschaefergoes to do that20:36
bschaeferthomi Now on revision 2206.20:37
* bschaefer goes to look at that rev20:37
thomiuh oh - dx-autopilot-run build just failed.20:37
lamalexi might have cancelled it20:37
lamalexi forgot to change the downstream job to the ibus job20:38
lamalexs/might have//20:38
bschaeferthomi, http://bazaar.launchpad.net/~unity-team/unity/trunk/changes/220620:39
bschaeferim just going to dig until if find anything useful haha20:40
bschaeferthomi, ugg so a problem, a problem happen between rev 2120 - 2206...since the jenkins test fail for a while20:44
bschaeferunity rev 2120-220620:44
thomihmmmm20:45
thomibschaefer: it could also be a change in the autopilot infrastructure branch that caused it20:45
bschaeferIll see if I can find where ALL start failing as that doesn't help at all20:45
thomi...or a change oin nux.20:45
bschaeferyeah20:45
thomilamalex: I take it you didn't want to go the extra step and use the Eventually() matcher in your switcher refactor branch?20:48
thomiI guess it makes sense to do that in a separate branch20:48
lamalexthomi, where? i thought i did20:49
thomioh, ok, there are some, I didn't get to them yet20:49
thomiI'll add a note in my review20:49
lamalexthere are some places where I couldn't use it- for instance the mode property isn't from introspection20:49
lamalexi added a card to extend wait_for so that you can use it on properties not added by introspection20:50
bschaeferthomi, wow even more odd things. What happen to jenkins dx ap builds 723-763?20:51
bschaeferbecause after that is were all start to fail20:51
* thomi looks20:52
bschaeferand it could have happened in unity form 2228-227520:52
bschaefer:(20:52
lamalexerm well, good news- all the tests fail20:52
bschaeferhaha20:52
lamalexgood news in that we get the same results as in the normal AP env with only running ibus tests20:52
thomi\o/20:52
bschaeferyeah20:52
lamalexbut you know, still bad news20:53
lamalex:P20:53
thomiit'd be funny if they all passed :)20:53
bschaeferI would .... idk20:53
bschaeferhaha20:53
bschaeferthat would be very inconsistent20:53
thomibschaefer: let me finish this review for alex, do the standup, then I'll investigate - will you be around still in 20 minutes?20:53
bschaeferyup!20:53
bschaeferIll dig some more20:53
lamalexbschaefer, thomi i would just say screw it and have a seperate ibus job20:54
thomiibus job?20:55
bschaeferI ment to dig through nux and the ibus api those logs aren't going to be very helpful :(20:55
lamalexthomi, as in we'd have two AP test jobs, one for most of the tests one for the ibus tests20:56
lamalexof course that's moot becuase they failed20:57
lamalexbut had they passed...20:57
bschaeferor if we can get them to pass :)20:57
thomilamalex: https://code.launchpad.net/~alexlauni/unity/unity.ap-switcher-emu/+merge/103338/comments/223103  just a few small tweaks needed :)20:59
thomilooke awesome otherwise20:59
lamalexsick20:59
lamalexwill fix now20:59
lamalexdamn i thought i changed that docstring about the tuple21:01
lamalexwth21:01
gemacan you guys help me add a console based app to the launcher?21:03
gemanevermind, got it already21:10
NotALamerhey how does unity figure out what icon and title to use for alt-tab21:10
NotALamerhaving trouble getting an application to show up properly21:10
NotALamerunity is calling it win0 and giving it a generic icon21:11
NotALamerbut xwininfo gives the correct title21:11
thomibschaefer: still around?21:14
bschaeferthomi, yeah21:17
thomi...just looking for clues as to the ibus failures...21:18
bschaeferthomi, was just pulling/compiling that unity-team/ibus-tests branch21:18
bschaeferyeah, the logs don't seems to help besides a general time line21:18
bschaeferdo you want to push a test in to see if we can find a press/release time that works?21:18
bschaeferworst case is we can show it is'nt a timing issue21:19
bschaeferthomi, what if we hold Ctrl, then tab space very fast21:20
bschaeferinstead of having it be "Ctrl+Space"21:21
thomibschaefer: worth a shot, although it shouldn't make any difference.21:21
bschaeferi know :(21:21
thomiI suspect this is an environmental thing.21:21
bschaeferok21:21
bschaeferdam I wonder if ibus ever gets activated at all21:22
bschaeferwe should start with push some debugging changes I suppose21:22
bschaeferthomi, can we print out 'ps aux' on that VM?21:25
bschaeferps aux | grep ibus21:25
thomibschaefer: sure, if you use subprocess and log it in a test21:26
thominow that we have a junk branch to push to that's easy :)21:26
thomigood idea BTW :)21:26
thomibschaefer: you want me to set that up?21:26
bschaeferhaha that is true, sure!21:26
thomiok, I'm on it21:26
bschaeferI can look it up, but Ill start trying to figure out how this could happen w/o a timing issue21:26
bschaefervery odd21:27
bschaeferand thanks, I like being able to debug things manually haha21:27
bschaeferthomi, sweet also found a good spot in put a print statement in SearchBar.cpp to see if im active gets set21:30
thomicool21:31
bschaeferdid you want to throw it in your branch when you get a chance?21:31
thomibschaefer: I just pushed a new revision of that testing branch21:33
thomibschaefer: also, your print statemnt won't work, since we're not compiling unity from the branch, only running the autopilot tests21:34
bschaeferoo21:34
bschaeferok21:34
* bschaefer goes to look up some python ibus api21:34
bschaefermaybe we can add some ibus stuff in the emulator for debugging21:35
thomisure21:35
thomiok, i just kicked off a new run of the debugging jenkins job21:35
bschaefercool21:36
bschaeferalso got that jenkins indicator it is awesome21:37
thomithanks :)21:37
bschaefero you made it? haha21:38
thomiyeah,. one of my side projects21:38
thomithere's several things that I need to fix in it, but it's useful at least21:39
bschaeferawesome, didn't check who pushed it to launchpad, just pulled it and ran it21:39
bschaeferyeah!21:39
thomibschaefer: there's a PPA as well21:39
bschaeferO, well it worked from launchpad haha21:40
=== salem_ is now known as _salem
tiredGeekOn Pangolin, desktop is blank when trying to use vncserver for a remote session.  Can Unity be redirected to another X display?21:46
thomihmmm, the ibus test jenkins jobs eems to be hanging21:52
bschaeferyeah, Im trying to see if we can push the trigger signal through ibus21:53
bschaeferinstead of pressing ctrl+space...21:54
bschaeferI know how to get the list of shortcuts but need to look more21:54
thomibschaefer: hmmm, tht's odd. My debug test fails in a strange way on the server21:59
* bschaefer looking21:59
bschaeferthomi, so hud doesn't want to show now?22:00
thomiok, I made it even simpler22:00
bschaeferodd22:00
thomipushing a version that just does the 'ps' command without trying to show or hide the hud22:01
thomire-running it now22:02
bschaeferalright, also I wonder what would happen if you ran "ibus-setup", isn't it taking a video the whole test?22:02
bschaeferso we could in a sense have a video of the test, screen shots of ibus-set if anything is wrong with it?22:02
bschaefersetup*22:02
thomisure, except you'd only see the first tab.22:02
bschaeferyeah22:03
thomilet's see what this produces first22:03
bschaeferagree22:03
thomibrb22:03
bschaeferok22:04
mhall119mhr3: are global search strings still stored in .props.search_string?22:04
mhall119mhr3: I'm trying to help with http://askubuntu.com/questions/125121/unity-lens-not-included-in-global-search/125186#12518622:04
mhr3mhall119, i dont think there was such property... ever :)22:06
mhall119uh...22:07
mhr3mhall119, not on scopes anyway22:08
mhr3oh... it is on the LensSearch instance of course22:09
mhall119yeah22:09
* mhr3 needs time to process stuff this late :P22:09
mhall119which gets passed to the function listening to search-changed signals, right?22:09
mhr3yep22:10
mhall119mhr3: so, Singlet sets search_in_global=False all the time, is it possible that it would still get a search_changed signal on a gloabl search, but with an empty search string?22:14
mhr3mhall119, are we talking strictly 12.04?22:15
mhall119I assume, since Singlet only works at all in 12.0422:16
mhr3cause there was a race in unity4 were that could happen22:16
mhr3but afaict i fixed it :)22:16
thomibschaefer: OK, did you see the results?22:16
bschaeferlooking at it right now22:17
bschaeferand comparing22:17
thomilooks good to me22:17
bschaeferthomi, I have ibus-daemon --xim22:17
bschaeferand that log has22:17
bschaeferibus-daemon -d --xim22:17
bschaeferI need to look up what that -d does22:17
thomi-d = daemonise22:17
bschaeferyeah22:18
bschaeferhmm22:18
mhr3mhall119, so, no it shouldn't, and if it does sound alarms :)22:18
mhall119mhr3: ok, I'll see what responses I get from my current comments, thanks22:19
mhr3mhall119, i feel bad when i chat with you, you don't make mistakes and you put commas where they should be :P22:20
mhr3not to mention the capital I :)22:21
mhall119lol22:22
mhall119my mother was a paralegal, taught me to write on one of those old smash-the-keys typewriters22:23
mhall119I still put two spaces after a period, despite all my subsequent English teachers yelling at me for it22:23
bschaeferthomi, ugg, so im still not sure what could cause this besides that toggling problem.22:24
mhall119I still make typos though, if you're paying attention22:24
bschaeferthomi, when I had ibus set to press/release at 0.5 it was failing consistently for me22:24
thomibschaefer: hmmm22:25
bschaeferthomi, hmmm, should we just go ahead and eliminate a timing issue by seeing if the ibus can get active?22:25
thomibschaefer: if you look at this: https://jenkins.qa.ubuntu.com/job/dx-autopilot-fix-ibus/3/label=1EBEE0FF-DAC9-11DF-BBDA-64A98C34D485/testReport/autopilot.tests.test_ibus/IBusTestsAnthy/test_simple_input_dash_user_commit_j_/?22:26
thomiit looks to me like the tap is taking just less than 200mS22:26
thomi...which seems reasonable?22:26
mhr3mhall119, good, otherwise i'd have to assume you installed some clever autocorrect in your irc client :)22:26
bschaeferthomi, it says 50 ms doesn't it?22:26
bschaeferfor ctrl+space22:26
bschaeferthomi, also how accurate are those times? it seems like it just prints the var you push into the function22:27
thomibschaefer: nope, 201 ms22:28
bschaeferthomi, wait where are you looking?22:28
bschaeferPressing keys 'Ctrl+Space' with delay 0.05000022:28
thomioh nuts - I'm being an idiot22:29
thomiwas looking at the Ctrl+a :(22:29
thomiright, so 23mS for the space key22:29
bschaeferhow are you getting 23ms?22:29
bschaefer23mS?22:29
thomi22:03:54.965 DEBUG X11:194 - Sending press event for key:22:30
thomi22:03:55.017 DEBUG X11:197 - Sending release event for key:22:30
bschaeferoo22:30
bschaeferok yeah duh22:30
thomi:)22:30
bschaeferhmm that is very odd22:30
bschaeferthis would be funny. So you can kill the ibus daemon and the ap test starts a new one when you open an engine22:31
bschaeferso I think a new daemon is getting started each test...nevermind that idea does no where...22:32
=== davidcalle_ is now known as davidcalle
davidcallejono, poing22:32
jonohey davidcalle22:32
thomihmmmm22:32
jonodavidcalle, nice work on the lens!22:33
jonodavidcalle, I responded to your mail on the list22:33
bschaeferthomi, so, what could we be missing that we have but jenkins doesn't22:33
bschaefercould we have*22:33
jonodavidcalle, also, we have #ubuntu-accomplishments if you ever want to hang out there22:33
bschaeferugg that is worded very badly...22:33
thomibschaefer: possibly an engine, but they should be installed, and the tests should skip if it can't find the engine22:33
bschaeferor I just can't read right now haha22:33
bschaeferthomi, yeah, we could print out all engines22:34
davidcallejono, thanks. I've seen your answer, when are you going to push your API changes to the PPA ?22:34
jonodavidcalle, the PPA has now been updated22:34
jonodavidcalle, so you should be good to go22:34
davidcalleOh, cool22:34
jonowhen you have updated the lens I will install the PPA and test too :-)22:35
thomibschaefer: hmmm, I'm going to try some things...22:36
bschaeferthomi, alright...22:36
bschaeferill try to think of something...22:36
bschaeferhaha22:36
* thomi kicks off another jenkins run22:42
bschaeferugg I have to leave in 20 min22:43
bschaeferthomi, what are you testing?22:43
thomiI added a bunch more debugging statements, and made the sleep after we restart the ibus daemon a bit more sensible22:43
bschaefercool!22:44
bschaeferI like the graph, Test Results Trend, all red22:44
thomiheh22:44
bschaeferthomi, I don't think it's an engine problem. I removed hangul and those test just reported back as no test taken22:46
bschaeferso it seem to skip them :(22:46
thomigood22:46
bschaeferwell yeah22:47
bschaeferhaha22:47
thomiI'm getting virtualbox set up to PXE boot the remastered CD, so I should have anidentical setupo22:47
thomi*seup22:47
* thomi waits for CD to download...22:48
bschaeferyeah, I'll end up doing that when I get back form this seminar I have to go too...22:48
bschaeferto*22:48
bschaeferthomi, ugg well I have to go for a bit. Ill be back in a couple hours, good luck!22:52
thomicheers22:54

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