| === Malsasa_ is now known as Malsasa | ||
| dednick | greyback: morning. | 08:59 |
|---|---|---|
| dednick | greyback: what is the difference between processFailed & processStopped from upstart? would an application which exits cleanly generate a processFailed notification from upstart, since it wasn't expecting it to close? | 09:00 |
| dednick | tedg: ^ ? | 09:00 |
| dednick | that's the ubuntu_app_launch_observer_add_app_failed vs ubuntu_app_launch_observer_add_app_stop notifications | 09:02 |
| greyback | dednick: I believe process fails if it exits with a non-zero return code | 09:05 |
| === vrruiz_ is now known as rvr | ||
| dednick | greyback: ah. ok, thought it might be that | 09:09 |
| dednick | greyback: hm. wouldn't expect that from sending a close request to surface would we? | 09:10 |
| greyback | dednick: no, I wouldn't. Perhaps I'm wrong | 09:11 |
| greyback | in which case, ual is being a bit paranoid | 09:11 |
| dednick | i've looked at it and it seems to only generate on RESULT=failed. so... | 09:19 |
| dednick | unless the close request is causing the app to close with error, dunno what's happening. | 09:19 |
| dednick | greyback, tedg: looks like legacy apps get an ApplicationFailed dbus notification from UAL when exiting. | 09:41 |
| dednick | click seems fine | 09:41 |
| dednick | or it's just random :/ | 09:43 |
| dednick | maybe it's just the impl | 09:43 |
| dednick | dialer does | 09:45 |
| greyback | dednick: dialer does cause a fail notification on quit? | 09:46 |
| dednick | greyback: ya | 09:46 |
| greyback | dednick: ok, that's a bit of a problem | 09:47 |
| dednick | greyback: heh. it segfaults actually | 09:47 |
| greyback | hah! so it was correct | 09:47 |
| dednick | just started manually | 09:47 |
| dednick | yeah | 09:47 |
| dednick | nevermind! | 09:48 |
| dednick | :) | 09:48 |
| greyback | legacy apps still have that problem tho, no? | 09:48 |
| greyback | that's a bug to solve | 09:48 |
| dednick | no, it seems ok. i thought i saw another but i've just tried | 09:48 |
| dednick | again | 09:48 |
| dednick | greyback: erg. clock app just generated one. but i think it's probably correct. it seems to be random so it was probably a crash | 09:50 |
| greyback | dednick: this might open up a crop of bugs, as apps are not accustomed to being politely asked to shut down! | 09:51 |
| dednick | greyback: true! :) | 09:52 |
| dednick | greyback: yeah, it's qmlscene crashing | 09:52 |
| greyback | would be worth grabbing a stacktrace or two, to check if something shared is hte cause of the crash (qtubuntu maybe) | 09:52 |
| dednick | we're not even supposed to use qmlscene... not supprised | 09:53 |
| dednick | "The qmlscene utility is meant to be used for testing your QML applications, and not as a launcher in a production environment. To launch a QML application in a production environment, develop a custom C++ application or bundle the QML file in a module. See Deploying QML applications for more information." | 09:53 |
| dednick | from qt. | 09:53 |
| dednick | boom | 09:54 |
| greyback | well that's not an idea everyone agrees with | 09:54 |
| greyback | by rights, should be using 'qml' | 09:55 |
| greyback | but as that's a more recent appearance, people got used to qmlscene. qmlscene just adds some extra dev flags which a user wouldn't care for | 09:55 |
| === MacSlow is now known as MacSlow|lunch | ||
| === alan_g is now known as alan_g|lunch | ||
| === mzanetti is now known as mzanetti|run | ||
| ChrisTownsend | dpm: Hey, I heard you have been having issues with unity8-lxc. | 12:25 |
| dpm | hi ChrisTownsend, thanks for following up! Last evening I gave it a go on a vivid host, and I encountered 3 issues: 1) on vivid, unity8-lxc-setup fails to run, which can be worked around by installing the wily .deb from the PPA (it seems the vivid one is outdated) 2) At some point the LXC container did not start, so I was always thrown back to a disabled lightdm screen. I worked around this by manually starting the LXC container 3) Once I had done all t | 12:28 |
| dpm | his, I could enter the unity8 session, but then I ended up with a black screen (but working mouse pointer) | 12:28 |
| ChrisTownsend | dpm: First, it all works fine for me on my Vivid machine, so let's try to figure out why your machine is having issues. | 12:29 |
| ChrisTownsend | dpm: Issue 1) is strange because the wily and vivid versions in the PPA are the same. | 12:30 |
| ChrisTownsend | dpm: Both are built from version 72 of lp:unity8-preview-lxc/development. | 12:31 |
| ChrisTownsend | dpm: What graphics do you have in your machine? | 12:32 |
| ChrisTownsend | dpm: And let's start from a fresh baseline. Please do 'sudo unity8-lxc-setup --rebuild-all --redownload". | 12:33 |
| dpm | ChrisTownsend, oh, I might have gotten confused by the different versions in the _archive_ not in the PPA: http://pastebin.ubuntu.com/11724543/ | 12:34 |
| dpm | ChrisTownsend, intel graphics | 12:35 |
| ChrisTownsend | dpm: Yeah, the wily archive version is greater than what's in the PPA (for now), but the archive version is just cherry picks of what it is in the PPA except for the debian/changelog. | 12:36 |
| ChrisTownsend | dpm: Ok, Intel, that should definitely work unless it's super old. | 12:36 |
| dpm | ChrisTownsend, shouldn't have more than a year | 12:38 |
| dpm | ChrisTownsend, in any case, regarding 1) I reported bug 1465383 | 12:38 |
| ubot5 | bug 1465383 in Unity8 Desktop Preview in LXC "Vivid PPA build out of date: failure to finish setup" [Undecided,New] https://launchpad.net/bugs/1465383 | 12:38 |
| ChrisTownsend | dpm: Yeah, that's fine. | 12:39 |
| ChrisTownsend | dpm: Yeah, saw that bug. Not sure what's going on with that as I don't see that issue, but it looks like an apt problem. | 12:39 |
| dpm | ChrisTownsend, I will not have much time to test this until the evening, but at least I can try once to log into the session again. So I finished doing the setup: http://pastebin.ubuntu.com/11724558/ | 12:40 |
| ChrisTownsend | dpm: And it's very strange you don't see it using the wily version as the wily version and vivid version do the same exact thing. | 12:40 |
| ChrisTownsend | dpm: Ok. Except for the locale issues, that looks good. | 12:41 |
| dpm | ChrisTownsend, yeah, I think the locale issues are harmless. So I guess next step is to reboot, try to login and report any errors? | 12:42 |
| ChrisTownsend | dpm: Right, next step is reboot and try logging in and see what happens. | 12:42 |
| dpm | ok, off I go | 12:42 |
| === mzanetti|run is now known as mzanetti | ||
| dpm | ChrisTownsend, so I ended up with the black screen again | 12:49 |
| ChrisTownsend | dpm: Ok, could you paste the contents of /var/log/lightdm/unity-system-compositor.log? | 12:49 |
| dpm | ChrisTownsend, http://pastebin.ubuntu.com/11724597/ - looks better than yesterday, which showed a cryptic mir error | 12:52 |
| dpm | I quit after ~2 minutes of black screen, I guess that was waiting time enough? | 12:53 |
| ChrisTownsend | dpm: Hmm, that looks fine. I think the container is not starting for some reason. What happens if you do 'sudo lxc-start -n unity8-lxc'? | 12:53 |
| ChrisTownsend | dpm: It should definitely start within 2 minutes. More like 5-10 seconds at worst. | 12:53 |
| dpm | ChrisTownsend, in fact, the container hadn't started, so I started it manually before logging into the unity8 session | 12:54 |
| dpm | otherwise, from yesterday's experience I would have got stuck at the inactive lightdm screen | 12:54 |
| ChrisTownsend | dpm: Hmm, there is definitely something up with the container then. Let's try to get some debugging output. | 12:55 |
| dpm | ChrisTownsend, is the container supposed to be set as autostart? http://pastebin.ubuntu.com/11724608/ | 12:56 |
| dpm | or does the login into the unity8 session triggers the start? | 12:56 |
| === alan_g|lunch is now known as alan_g | ||
| ChrisTownsend | dpm: The log in to the unity8 lxc session is supposed to start it. | 12:56 |
| dpm | ah, ok. So there was no need for me to manually start it | 12:57 |
| ChrisTownsend | dpm: No, that is an issue. | 12:57 |
| ChrisTownsend | dpm: So please do this: 'sudo lxc-start -n unity8-lxc -o unity8.log -l DEBUG' and please paste the contents of the file. | 12:58 |
| === MacSlow|lunch is now known as MacSlow | ||
| dpm | ChrisTownsend, http://paste.ubuntu.com/11724615/ | 12:59 |
| === dandrader_ is now known as dandrader | ||
| dpm | ChrisTownsend, I'm about to start some calls now, but I'll be around again in a few hours in my evening | 12:59 |
| ChrisTownsend | dpm: Ok, it starts fine from command line, so next step is to figure out why it doesn't start when logging in. | 13:00 |
| ChrisTownsend | dpm: Sure, just ping me later. | 13:00 |
| tedg | pstolowski, pete-woods1, did we ever get an API call for "is this a scope" that takes an appid? | 13:19 |
| tedg | I remember putting it on the backlog, but I lost track of it. | 13:19 |
| pete-woods1 | tedg: I don't remember working on anything like that | 13:19 |
| === pete-woods1 is now known as pete-woods | ||
| tedg | pete-woods, Seems like it'd be a "check to see if a file is in these directories with this name" | 13:21 |
| pete-woods | tedg: well, more likely a "talk to the scope registry", but yes, should still be simple | 13:23 |
| tedg | Oh, figured checking the filesystem would be faster. | 13:23 |
| pete-woods | tedg: I'm sure if you log a feature request bug explaining what you need against unity-scopes-api, then michi would be able to cook this function up pretty quickly | 13:26 |
| tedg | pete-woods, done. bug 1465675 | 13:36 |
| ubot5 | bug 1465675 in unity-scopes-api (Ubuntu) "API needed to determine if App ID is a scope" [Undecided,New] https://launchpad.net/bugs/1465675 | 13:36 |
| mhall119 | mzanetti: I spent all weekend demoing the phone to people who got confused when shell didn't rotate, and the very next day you land it :/ | 14:01 |
| mhall119 | It's pretty awesome though :) | 14:02 |
| balloons | mhall119, LOL | 14:06 |
| mzanetti | mhall119, on the phone, will come back to you in a bit | 14:25 |
| om26er | mzanetti, do you know how I can verify https://code.launchpad.net/~aacid/unity8/borderSourceRegression/+merge/258883 fix ? | 14:29 |
| mzanetti | kgunn, FYI, https://docs.google.com/document/d/1XBSKpFu86KSWSeNI92osoGHACY1CJ7G0n0K9bvRK1wQ/edit | 14:30 |
| mzanetti | om26er, looking. gimme a minute | 14:31 |
| mhall119 | mzanetti: unless you have a witty retort (which I don't doubt) there's nothing you need to come back to me about :) | 14:32 |
| mzanetti | om26er, I think you won't notice it in the ui... It was causing warning messages in the test setup | 14:33 |
| om26er | mzanetti, aah, ok. thanks | 14:33 |
| om26er | mzanetti, can you help with this https://code.launchpad.net/~aacid/unity8/noTemplateInGeneratedCardCreatorCode/+merge/258980 ? | 14:34 |
| mzanetti | om26er, I just tried to see if those warning messages also appeared in the real world thing, but couldn't find any... | 14:34 |
| mzanetti | mhall119, I didn't even read your question before... :D | 14:35 |
| mzanetti | mhall119, nice, thanks | 14:35 |
| mzanetti | om26er, that second branch is only code readability/performance improvements, shouldn't change anything visible in the ui. | 14:39 |
| mzanetti | om26er, basically just make sure the scope still look as you'd expect them | 14:39 |
| om26er | mzanetti, crashes ? | 14:39 |
| om26er | :p | 14:39 |
| mzanetti | it crashes? | 14:39 |
| === davidcalle_ is now known as davidcalle | ||
| om26er | mzanetti, no I mean does it have a potential to crash | 14:39 |
| om26er | ? | 14:39 |
| mzanetti | I had a quick read through and it looks like it moves the same code from inside the cards to a more generic place outside... so that it is only evaluated once. | 14:41 |
| mzanetti | so in theory nothing should be more prone to crashes than before... | 14:41 |
| mzanetti | om26er, ^ | 14:41 |
| om26er | mzanetti, ack, thanks | 14:43 |
| om26er | mzanetti, the launcer is partially visible in the edge tutorial, is that new behavior ? | 14:49 |
| mzanetti | om26er, hmm... don't think so, no | 14:53 |
| mzanetti | om26er, there was a bug before that the launcher could be dragged too far | 14:53 |
| mzanetti | that should be fixed now, but no intentional behavior changes | 14:54 |
| om26er | mzanetti, http://i.imgur.com/BNbcjvc.jpg | 14:58 |
| mzanetti | om26er, that's odd... will check it out | 14:58 |
| om26er | mzanetti, I was testing the silo, so probably its related to that. | 14:59 |
| mzanetti | om26er, yeah... very likely. Thanks for catching this | 14:59 |
| om26er | mzanetti, I am not sure if the silo should land with that. What do you think ? | 15:00 |
| mzanetti | om26er, probably not, no | 15:00 |
| mzanetti | om26er, trying to verify it right now, I have a branch I suspect is the culprit | 15:01 |
| mzanetti | will drop that branch from the silo | 15:01 |
| om26er | mzanetti, btw its not only the edge tutorial, launcher is a bit out in its normal condition | 15:04 |
| mzanetti | yeah... just saw it too | 15:04 |
| mzanetti | om26er, well, it only happens if the tutorial ran before | 15:05 |
| om26er | hmmm | 15:05 |
| mzanetti | om26er, and it depends on when exactly you drag the launcher in the tutorial while it's bounching | 15:05 |
| mzanetti | om26er, but anyways, this can't land like this... will drop that branch from the silo and rebuild it | 15:05 |
| om26er | guess I caught a race. *pats self on back* | 15:05 |
| mzanetti | heh | 15:06 |
| om26er | mzanetti, Following practice, I will fail this silo and when the new package is built, will pick the new card again. | 15:07 |
| mzanetti | om26er, ok. thanks again for catching it | 15:07 |
| mzanetti | dandrader, hey, given mterry is not with us any more, mind taking a look at this when you have some time? https://code.launchpad.net/~mterry/unity8/tutorial-launcher-gap/+merge/259127 | 15:09 |
| dandrader | mzanetti, ok | 15:13 |
| === cwayne1 is now known as cwayne | ||
| === alan_g is now known as alan_g|EOD | ||
| dpm | ChrisTownsend, I've got some time now if you're around. I've done a couple of test logins, and I can consistently log into the session, but I get the black screen with the mouse cursor. I've got a dual output intel card, and I'm getting the same result with either one monitor (i.e. the other unplugged) or two (cloned black screen on both monitors). Not sure about the next steps to debug | 17:35 |
| ChrisTownsend | dpm: Hey, I'm here now. | 17:49 |
| ChrisTownsend | dpm: Is there a ~/.cache/upstart/unity8.log file and if so, can you post the contents? | 17:49 |
| dpm | ChrisTownsend, thanks, looking | 17:54 |
| ChrisTownsend | dpm: sure | 17:54 |
| dpm | ChrisTownsend, on the host, I've got unity8.log.1.gz type of files (1, 2, 3, 4), but not unity8.log | 17:56 |
| ChrisTownsend | dpm: Sure, it gzip's up pretty fast. Unzip the .1.gz one let's see what is has. | 17:56 |
| dpm | ok | 17:57 |
| dpm | ChrisTownsend, http://paste.ubuntu.com/11726217 - perhaps the locale warnings were more important that we thought... | 17:58 |
| ChrisTownsend | dpm: Yep, your locale is not correct. I had another user with similar issues and I guess he fixed it. | 17:59 |
| dpm | ChrisTownsend, can I just regenerate the locale in the container? | 18:00 |
| ChrisTownsend | dpm: It's the host locale that's it's messing up on. | 18:01 |
| dpm | ChrisTownsend, well, the host locale is definitely correct | 18:01 |
| ChrisTownsend | dpm: The container gets it from the host, so... | 18:01 |
| ChrisTownsend | dpm: This is kind of beyond my expertise, so I'm not sure what to tell you. | 18:03 |
| dpm | ChrisTownsend, somehow the container doesn't recognize the host locale as a valid one, but it is valid: http://pastebin.ubuntu.com/11726263/ | 18:03 |
| ChrisTownsend | dpm: Does booting off of a Live session work? | 18:04 |
| dpm | ChrisTownsend, not sure I understand what you mean | 18:06 |
| === dandrader is now known as dandrader|lunch | ||
| ChrisTownsend | dpm: Download the Ubuntu Desktop Next ISO and make a USB stick and boot to the Live session and see if that all works. | 18:06 |
| ChrisTownsend | dpm: Or use /var/lib/lxc/unity8-lxc/ubuntu-next.iso | 18:07 |
| dpm | ChrisTownsend, ah, I see, I can try that, yes. But I think the issue is this: http://pastebin.ubuntu.com/11726290/ | 18:07 |
| dpm | the container gets the locale messed up | 18:07 |
| dpm | "POSIX" is not a valid locale | 18:08 |
| ChrisTownsend | dpm: Errr, yeah, that looks messed up. I'll see what mine shows. | 18:08 |
| dpm | ChrisTownsend, I think that might have happened when running unity8-lxc-setup, when the locale warnings appeared | 18:09 |
| ChrisTownsend | dpm: Yeah, mine shows the same locale as the hosts. | 18:09 |
| ChrisTownsend | dpm: See if fixing up the locale in the container fixes the issue. I really don't know why it got messed up. | 18:10 |
| ChrisTownsend | dpm: You're the first person to have this. The other user had an issue of mixed locales. | 18:11 |
| dpm | ChrisTownsend, I'd be tempted to run locale-gen on the container, but I noticed that it tells me a) I cannot sudo inside the container (my user is not in the sudoers file) and then I don't quite know which things done in the container affect the host | 18:11 |
| dpm | e.g., I still do not quite grasp how the unity8.log is in the host and not the container | 18:11 |
| dpm | is it safe for me to add myself to sudoers in the container to fix the locale? | 18:12 |
| ChrisTownsend | dpm: /home is mounted inside the container. | 18:12 |
| ChrisTownsend | dpm: If you do 'sudo lxc-attach -n unity8-lxc', it will be a root console. | 18:12 |
| dpm | ah, cool | 18:12 |
| ChrisTownsend | dpm: It shouldn't mess up the host's locale, but I'm not 100% positive of that without researching that further. | 18:13 |
| dpm | ChrisTownsend, this looks promising: http://pastebin.ubuntu.com/11726330/ | 18:14 |
| ChrisTownsend | dpm: Yep. Try it again and let's see if it gets further this time. | 18:15 |
| ChrisTownsend | dpm: It meaning logging into the Unity8 LXC session. | 18:15 |
| dpm | indeed. Hopefully if I'm not back in a few minutes it's because I'll be playing with the unity8 session ;) | 18:16 |
| ChrisTownsend | dpm: Ok, good luck! | 18:16 |
| dpm | thanks! | 18:16 |
| dpm | ChrisTownsend, success! https://plus.google.com/+DavidPlanella/posts/C9x8yqdM2KY :-) | 18:40 |
| ChrisTownsend | dpm: Sweet! You were gone for a while, so I thought it was good news! | 18:41 |
| dpm | indeed | 18:42 |
| dpm | ChrisTownsend, thanks a lot for the help, and great work! | 18:43 |
| ChrisTownsend | dpm: You're very welcome and thanks! | 18:43 |
| dkessel | dpm: do you have nvidia optimus graphics? | 18:50 |
| === dandrader|lunch is now known as dandrader | ||
| dpm | ChrisTownsend, I've been trying to do something more meaningful with the unity8 session: first thing I tried was to build the phone terminal app for wily and install it into the container. However, after installation it does not appear in the launcher, so I can't start it. Any idea how to make it show up there? | 20:42 |
| dandrader | mterry, so I'm taking over https://code.launchpad.net/~mterry/unity8/tutorial-launcher-gap/+merge/259127 if you don't mind | 20:42 |
| dpm | ChrisTownsend, also, every time I log into the unity8 session I get the welcome wizard. Is there a way do disable it? | 20:42 |
| mterry | dandrader, oh thanks that's swell | 20:43 |
| mterry | I saw the comment go by about a problem with it, but yeah | 20:43 |
| ChrisTownsend | dpm: Do you have a desktop file for the terminal app? | 20:43 |
| dpm | yeah | 20:44 |
| dpm | let me check the contents | 20:44 |
| ChrisTownsend | dpm: You need to put it in ~/.local/share/applications and make sure there is a line in that has "X-Ubuntu-Touch=true". | 20:45 |
| ChrisTownsend | dpm: Regarding Welcome Wizard, do you mean the one that shows you how to swipe or the one that has the password setup? | 20:45 |
| dpm | ChrisTownsend, the swipe one | 20:46 |
| ChrisTownsend | dpm: Yeah, I'm not really sure how to keep that from coming up. I haven't investigated it any. | 20:46 |
| dpm | ChrisTownsend, here's the content and location of the terminal app's desktop file in the .deb build (as opposed from the .click build) http://pastebin.ubuntu.com/11727135/ | 20:46 |
| ChrisTownsend | dpm: Yea, that should work. You built and installed this in the container, right? | 20:48 |
| dpm | ChrisTownsend, I built it on a PPA from a daily recipe for wily, and I then installed it into the container | 20:48 |
| dpm | ChrisTownsend, but I'm not quite sure why it's not showing up in the launcher | 20:49 |
| ChrisTownsend | dpm: It should be in the App Scope only. | 20:49 |
| ChrisTownsend | dpm: I'm not sure how Launcher pinning works in Unity 8. | 20:49 |
| dpm | ChrisTownsend, why in the app scope and not the launcher? And btw, the "Scopes" window is empty | 20:49 |
| dpm | ah, I see what you mean, it'd need to be pinned to appear in the launcher | 20:50 |
| ChrisTownsend | dpm: Hmm, nothing in the Scopes window? | 20:50 |
| ChrisTownsend | dpm: You should have some things in there. | 20:50 |
| dpm | ChrisTownsend, ok, I'll investigate how to do pinning. So I guess the next question is, is the Scopes window supposed to show things? | 20:50 |
| dpm | yeah, no luck, it's a blank window | 20:50 |
| dpm | well, grey | 20:50 |
| ChrisTownsend | dpm: Uggg | 20:50 |
| dpm | if I close it, then it shows a spinner in a black background, and then it's back to the grey (empty) content | 20:51 |
| ChrisTownsend | dpm: Seems maybe the favorites gconf key needs reset. I'll see if I can dig up how the Unity 8 guys told me how to do that. | 20:51 |
| dpm | ChrisTownsend, ok, cool | 20:52 |
| dpm | I'm going afk for a bit, but I'll be idling around | 20:52 |
| ChrisTownsend | dpm: Ok, found it. To see if there are any favorite-scopes there, do "gsettings get com.canonical.Unity.Dash favorite-scopes" and if it's empty, do "gsettings reset com.canonical.Unity.Dash favorite-scopes". | 20:59 |
| ChrisTownsend | dpm: Although I'm not really sure how to do this from an LXC session. | 20:59 |
| ChrisTownsend | dpm: I need to go in a couple of minutes, so we can pick this up tomorrow if you're still having issues. | 21:00 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!
