| === salem_ is now known as _salem | |
| === chriadam|away is now known as chriadam | |
| === chihchun_afk is now known as chihchun | |
| [05:55] <icosa> qt-creator is for ubuntu apps only now ? | |
| [05:55] <icosa> i had created a project in jan2014 which i need to make some changes to now | |
| [05:55] <icosa> but the whole qt-creator interface has changed so much | |
| === chriadam is now known as chriadam|away | |
| [07:29] <dholbach> good morning | |
| [07:36] <nik90> bzoltan, zbenjamin: Hi, Good Morning. I'd like to raise the priority of bug 1455354 concerning the emulator channels shown in the dialog which are incorrect after the recent channel name changes. I am seeing some reports on google+ where developers are unsure which custom channel to go for. | |
| [07:36] <ubot5> bug 1455354 in qtcreator-plugin-ubuntu (Ubuntu) "Channel names when creating an emulator are wrong" [Undecided,New] https://launchpad.net/bugs/1455354 | |
| [07:51] <bzoltan> zbenjamin: We have a showstopper issue ^ | |
| [07:52] <zbenjamin> hmm | |
| [07:52] <zbenjamin> lets fix it | |
| [07:53] <bzoltan> zbenjamin: the fix is easy but will require a hacl | |
| [07:53] <bzoltan> hack | |
| [07:54] <zbenjamin> a hack? a because the channel names are too long | |
| [07:57] <bzoltan> zbenjamin: Precisely ...we need to add an option selection for bothe the ubuntu-touch and for the channel .. or simple hardcode the ubuntu-touch | |
| [08:11] <nik90> zbenjamin, bzoltan: Thanks for looking into this | |
| [08:12] <nik90> popey: Hey, yesterday nite I reproduced the clock app crash similar to the one we noticed when QA tested the 3.3.270 build few days back after installing OTA-4. Considering that it is not a regression, you think we will be able to persuade QA to push 3.3.270 to the store? | |
| [08:13] <seb128> nik90, is anyone looking at this segfault? | |
| [08:14] <nik90> seb128: not to my knowledge..I tried pinging some, but I cannot find the package at fault without help from suspected package owners | |
| [08:14] <seb128> nik90, did you get a debug backtrace? | |
| [08:15] <nik90> seb128: the issue is that it happens every sporadically and as such at times when I try to get the logs, I cant get clock app to crash. | |
| [08:16] <nik90> I have been trying for a couple of days now | |
| [08:16] <didrocks> no .crash file in /var/crash? | |
| [08:16] <didrocks> that would be a start and you don't have to monitor for getting one | |
| [08:16] * nik90 checks now | |
| [08:18] <nik90> didrocks: I see a few crash files in /var/crash http://paste.ubuntu.com/11723745/ .. however none related to clock app. | |
| [08:19] <didrocks> nik90: isn't clock app started by qmlscene? | |
| [08:19] <nik90> it could be that qmlscene itself is crashing, but this was one of our original suspicion for which it was included in the bug report already. | |
| [08:20] <didrocks> well, if your clock app is crashing, then, the parent process is the one matched here | |
| [08:20] <didrocks> which is qmlscene | |
| [08:20] <didrocks> (but it's actually clock app crashing) | |
| [08:20] <nik90> didrocks: indeed it is and popey uploaded a qmlscene log of it in the bug report at https://bugs.launchpad.net/ubuntu-clock-app/+bug/1463430 | |
| [08:20] <ubot5> Launchpad bug 1463430 in Ubuntu Clock App "Clock app crashes when renabling an alarm" [High,Confirmed] | |
| [08:20] <didrocks> ah nice, with the stacktrace link | |
| [08:20] <nik90> didrocks: on looking through it unfortunately I couldn't make head or tail of it which leads me back to square one where I need help from upstream devs to help debug it. | |
| [08:21] <nik90> I have added qtubuntu and qtbase-opensource-src to it | |
| [08:21] <didrocks> ProcCwd | |
| [08:21] <didrocks> /opt/click.ubuntu.com/com.ubuntu.clock/3.3.272 | |
| [08:21] <didrocks> clearly matching at least one of clock crash | |
| [08:21] <didrocks> yeah, Mirv, how do you deal with qtbase crashes? ^ | |
| [08:22] <seb128> didrocks, nik90, have a debug bt would be useful | |
| [08:22] <didrocks> https://errors.ubuntu.com/bucket/?id=/usr/lib/arm-linux-gnueabihf/qt5/bin/qmlscene%3A11%3AChange%3AQVector%3Aresize%3Apop%3AQAbstractItemModel%3A%3AendMoveRows | |
| [08:22] <didrocks> seb128: ^ | |
| [08:22] <didrocks> seems seb128 wants to debug Qt now :) | |
| [08:23] <nik90> zsombi: ping (need your help debugging a clock app crash related to alarms..it is blocking its release) | |
| [08:23] <seb128> could be an uitk issue | |
| [08:23] <seb128> adapters/alarmsadapter_organizer.cpp is from there | |
| [08:23] <didrocks> yeah, seems to be the AlarmManager | |
| [08:24] <nik90> seb128: how do you arrive at adapters/alarmsadapter_organizer.cpp by looking at the stacktrace? (just curious) | |
| [08:24] <seb128> nik90, look the e.u.c url didrocks just gave | |
| [08:24] <zsombi> nik90: the crash happens due to the QtOrgabizer not being built with the same Qt release :/ I guess | |
| [08:24] <didrocks> nik90: frame 8 | |
| [08:24] <seb128> #8 0xaf04196e in AlarmsAdapter::updateAlarm (this=this@entry=0x18907b0, id=...) at adapters/alarmsadapter_organizer.cpp:728 | |
| [08:24] <didrocks> #8 0xaf04196e in AlarmsAdapter::updateAlarm (this=this@entry=0x18907b0, id=...) at adapters/alarmsadapter_organizer.cpp:728 | |
| [08:25] <didrocks> zsombi: ABI break? | |
| [08:25] <nik90> seb128, didrocks: ah thnx | |
| [08:25] <zsombi> didrocks: perhaps... QtOrganizer package sais 5.0... We are at 5.4.1 | |
| [08:26] <didrocks> zsombi: this is somewhat worrying we don't even have force rebuilding with the same version if there is no upstream garantee | |
| [08:26] <didrocks> Mirv: do you mind giving a shot at rebuilding QtOrganizer then? | |
| [08:26] <didrocks> I think we can stage that to a silo, and get people running it for a couple of days | |
| [08:27] <didrocks> and see if they continue getting the same clock app segfault | |
| [08:27] <zsombi> didrocks: the problem is that - inm - Qt5.4 brought some deps we couldn't compile it with... | |
| [08:28] <zsombi> didrocks: though this crash seems to be elsewhere now... | |
| [08:29] <didrocks> last = -1358657671 seems suspicious | |
| [08:30] <Mirv> didrocks: find if the crash is in Qt itself or if it's happening actually in some of our component | |
| [08:31] <didrocks> Mirv: mind backloging a little bit in the discussion? There is the backtrace | |
| [08:31] <Mirv> didrocks: zsombi: that's incorrect deduction, qtorganizer is a git snapshot module that upstream has never released, but we do rebuild it against correct Qt every time Qt is updated (otherwise it'd break at packaging level) | |
| [08:31] <nerochiaro> oSoMoN: i fixed the failing test for the keyboard MR. It was verfying the addressbar focus while the test was about something completely different, so i removed these checks for now. they are done elsewhere anyway | |
| [08:31] <didrocks> ok, so not the ABI breakage | |
| [08:31] <didrocks> I guess it's now in your hands zsombi then ;) | |
| [08:32] <zsombi> Mirv: ok, so only the package name staid the same? | |
| [08:32] <didrocks> Mirv: ok, you added the same trick to force the rebuild for those, nice! | |
| [08:33] <Mirv> zsombi: package names stays the same and version stays at 5.0 as it's not actually released | |
| [08:34] <zsombi> Mirv: ok... remember once we had a crash in QOrganizerManager::availableManagers().contains() function? | |
| [08:35] <zsombi> then we realized that the QtOrganizer has been built with an older Qt... | |
| [08:35] <Mirv> didrocks: that backtrace I'd give to zsombi and then if he thinks a Qt bug in qvector or something I'd bring it back to qtbase | |
| [08:35] <seb128> Mirv, then? do we have anyone dealing with qtbase issues? | |
| [08:35] <Mirv> zsombi: hm, I don't remember. qtpim has been within the rebuilds at least since the beginning of 2014. | |
| [08:35] <seb128> it doesn't do much good to do re-assign games if there is no owner of the issue at the end | |
| [08:35] <zsombi> ok | |
| [08:36] <Mirv> seb128: no, we don't have anyone fixing Qt bugs on average | |
| [08:36] <zsombi> nik90: is there an easy way to repro this? | |
| [08:36] <zsombi> didrocks: seb128: ^ | |
| [08:36] <Mirv> tsdgeos is on holidays, he might. and of course I'll try to search upstream gits for something that sounds like a fix. | |
| [08:36] <seb128> zsombi, bug #1463430 | |
| [08:36] * didrocks didn't get it, just looked at the current discussion | |
| [08:36] <ubot5> bug 1463430 in Ubuntu Clock App "Clock app crashes when renabling an alarm" [High,Confirmed] https://launchpad.net/bugs/1463430 | |
| [08:36] <nik90> zsombi: usually when you try reenabling a single time alarm that has been disabled automatically by indicator-dt after it rang | |
| [08:36] <seb128> zsombi, basically on the current ota4 image try to re-enable an event | |
| [08:37] <zsombi> nik90: hmm... I did that thi snight... and it worked... | |
| [08:37] <seb128> zsombi, you are on ota4/vivid? | |
| [08:37] <nik90> zsombi: but the crash happens rarely | |
| [08:37] <nik90> zsombi: and only on ota4/vivid | |
| [08:37] <seb128> zsombi, try a few times, eventually you hit it | |
| [08:37] <zsombi> seb128: I have not flashed my device for few weeks, but keep getting dev-proposed images, so I'm at r243 | |
| [08:38] <seb128> zsombi, unsure what channel/device that is | |
| [08:38] <seb128> what qt version do you have on there? | |
| [08:38] <zsombi> seb128: 5.4.1 | |
| [08:38] <seb128> k, so that should do it | |
| [08:38] <seb128> try to do what nik90 said, a few times | |
| [08:39] <zsombi> seb128: it's an 15.10 | |
| [08:39] <zsombi> ok, lemme try | |
| [08:39] <zsombi> nik90: do I have to have the app running? i.e. not sleeping etc | |
| [08:40] <nik90> zsombi: doesn't matter..when you reenable the alarm it crashes | |
| [08:40] <nik90> zsombi: I have noticed that it occured more frequently when trying to fix https://code.launchpad.net/~themeles/ubuntu-clock-app/bugfix-1442518/+merge/261910 | |
| [08:41] <nik90> zsombi: try following the steps in that MP without applying the patch. | |
| [08:44] <zsombi> nik90: it works for me... | |
| [08:45] <nik90> zsombi: how many tries? its rare making this hard to debug and fix :/ | |
| [08:46] <zsombi> nik90: don't know.... like 10? I got 10 alarms ringing, then re-enabled them, and all work | |
| [08:55] <zsombi> nik90: so what were the changes you had in the latest release? | |
| [08:55] <nik90> zsombi: nothing alarm related :/ here is the changelog http://pad.ubuntu.com/clock-app-changelog | |
| [08:55] <zsombi> nik90: toolkit hasn't had either... | |
| [08:56] <zsombi> wth??! | |
| [08:56] <zsombi> Qt is the same as well afaik | |
| [08:56] <nik90> zsombi: well not true..there was this huge uitk change in the alarms backend that we did | |
| [08:56] <nik90> and that was for vivid afaik | |
| [08:56] <nik90> zsombi: https://bugs.launchpad.net/ubuntu-clock-app/+bug/1429273 | |
| [08:56] <zsombi> nik90: you mean the backend change landed just now? | |
| [08:56] <ubot5> Launchpad bug 1429273 in Ubuntu Clock App "Clock App crashes when editing an alarm on vivid" [Critical,Fix committed] | |
| [08:58] <nik90> zsombi: the fix for the above bug was necessary because of the big backend end change that landed in UITK 1.2 final release..and since then we have no new clock app release | |
| [08:58] <zsombi> hmm... | |
| [08:58] <zsombi> crap! | |
| [08:59] <nik90> atleast that's my suspicion .. I have been pushing for a new clock app release for ages..but AP tests came in the way of becoming a release blocker few months back | |
| [09:00] <zsombi> nik90: whoah... I got a crash if I kept the clock app open while I changed the date/time, and then re-enabled the alarm | |
| [09:01] <nik90> zsombi: yup as I mentioned in the MP steps I asked you to test with | |
| [09:01] <zsombi> nik90: well, that sais I shoudl kill the app... if I do that it works, but if I keep it open while I change the date, then it crashes | |
| [09:02] <nik90> zsombi: I suspect that when a fault alarm.date is provided (before the current date for example) it crashes.. | |
| [09:03] <nik90> zsombi: I have *never* had clock app crash when editing the alarm and then saving it again.. | |
| [09:03] <nik90> but only when reenabling it...and the alarm switch code is rather simple..it deals with alarm.date and alarm.daysOfTheWeek | |
| [09:04] <nik90> s/fault/faulty | |
| [09:04] <zsombi> nik90: do you happen to have a property binding there? like having any property of the alarm bount to any component? | |
| [09:05] <nik90> well the listitem shows a label where label.text: model.date ... but not necessary alarm.date | |
| [09:05] <karni> Where can I learn more about 'New ListItem with fancy features' as mentioned in ota-4.changelog ? | |
| [09:06] <zsombi> karni: in the docs :) | |
| [09:06] <nik90> zsombi: https://bazaar.launchpad.net/~ubuntu-clock-dev/ubuntu-clock-app/utopic-3.0/view/head:/app/alarm/AlarmDelegate.qml | |
| [09:06] <karni> they already up to date? cheers | |
| [09:06] <zsombi> karni: New ListItem landed in 1.2 I guess those should be out there | |
| [09:07] <zsombi> karni: https://developer.ubuntu.com/api/apps/qml/sdk-15.04/Ubuntu.Components.ListItem/ | |
| [09:07] <karni> thank you, zsombi | |
| [09:07] <zsombi> karni: yw | |
| [09:07] <zsombi> nik90: that binding is OK, it doesn't cause further changes | |
| [09:16] <zsombi> nik90: from the logs seems the QAbstractItemModel is the one which somehow screws us up... | |
| [09:17] <nik90> yes | |
| [09:17] <zsombi> how the heck it gets a negative number for allocating the item? | |
| [09:26] <karni> I just had a look at Ubuntu.DownloadManager.DownloadManager and .SingleDownload. I know mandel worked on this, not sure who's on it now. I'd like to ask - 1) where are the files downloaded to/how do I get handle of the file, since the documentation does not mention it 2) trottling says 'This property can be used to limit the bandwidth used for the download.' and is of type long. What is supposed to be the value of this, bytes per second? | |
| [09:27] <nik90> karni: I might be able to help regarding where the files are downloaded to since we make use of this in Podbird to download Podcasts | |
| [09:27] <karni> I just saw an example usage here https://developer.ubuntu.com/api/apps/qml/sdk-15.04/Ubuntu.DownloadManager.index/ | |
| [09:27] <nik90> karni: the downloaded files are stored in the app directory .local/share/app-name/ | |
| [09:27] <karni> and still it doesn't tell where they're stored | |
| [09:28] <karni> nik90: in the top directory, or some 'downloads/Downloads' dir? (thanks! :) ) | |
| [09:28] <nik90> karni: just the app root directory...no downloads/Downloads dir and so on afaik | |
| [09:29] <karni> nik90: thank you bro | |
| [09:29] <nik90> karni: it seems there is a variable called "path" that has the location of the downloaded file | |
| [09:29] <nik90> (after looking through podbird.qml) | |
| [09:30] <karni> nik90: thanks, I'll file a bug it's not documented | |
| [09:30] <nik90> karni: so basically onFinished: { console.log(path} } | |
| [09:30] <karni> right :) | |
| [09:30] <nik90> karni: yeah I am not sure how Elleo figured it out..just glad that he did though :) | |
| [09:30] <karni> lol hahah | |
| [09:30] <karni> probably browsed the DM source | |
| [09:31] <nik90> ;) | |
| [09:33] <karni> nik90: you may want to confirm it https://bugs.launchpad.net/ubuntu-download-manager/+bug/1465568 | |
| [09:33] <ubot5> Launchpad bug 1465568 in ubuntu-download-manager "Download path is not documented" [Undecided,New] | |
| [09:35] <nik90> karni: done..hey btw do you still use Ubuntu.Connectivity to determine if the phone is connected to 3G or Wifi? Does it work on krillin? | |
| [09:36] <karni> nik90: Yes we do, and it seems it works fine. | |
| [09:36] <karni> karni@zen:~/src/canonical/telegram/app-dev$ grep -ir ubuntu.connectivity | |
| [09:36] <karni> components/TelegramPage.qml:import Ubuntu.Connectivity 1.0 | |
| [09:36] <karni> :D | |
| [09:36] <karni> I had to grep as it's been a while I touched that part. | |
| [09:37] <nik90> karni: I remember adding that code ;) but also remember QA saying that it didnt work on krillin..and I had the same issue when using it in Podbird as well. | |
| [09:37] <karni> nik90: I believe there were some bugs with it earlier, but perhaps they've already been sorted. | |
| [09:37] <nik90> karni: yeah I will test it out again | |
| [09:37] <nik90> karni: people apparently don't like downloading podcasts in 3G :P | |
| [09:37] <karni> I haven't paid much attention to testing Tg offline, TBH | |
| [09:37] <karni> haha | |
| [09:48] <nik90> zsombi: hey, are the trailingaction widths dependent on the listitem height somehow? I am hitting this bug https://imgur.com/a/Tx3K5#0 | |
| [09:50] <zsombi> nik90: yes it is... | |
| [09:51] <nik90> zsombi: hmm perhaps I should show the description in a dialog to avoid a big listitem in that case | |
| [09:52] <zsombi> nik90: but what's wrong there? | |
| [09:52] <nik90> zsombi: in the 3rd image, my entire list item is hidden by the trailing actions and also the 3rd trailing action "info" has no background | |
| [09:53] <zsombi> nik90: hmm! that's bad, seems we have to adjust the width calculation | |
| [09:54] <nik90> zsombi: yeah ... perhaps add a Math.max() to ensure it doesnt exceed a design set max value | |
| [09:54] <zsombi> nik90: I think we have a clamp there... | |
| [09:56] <zsombi> nik90: yes, we have, but the minimum value is the height of the listitem, and that's not good | |
| [09:56] <nik90> zsombi: ah ok.. | |
| [09:56] * nik90 reports a bug | |
| [09:59] <nik90> zsombi: https://bugs.launchpad.net/ubuntu/+source/ubuntu-ui-toolkit/+bug/1465582 | |
| [09:59] <ubot5> Launchpad bug 1465582 in ubuntu-ui-toolkit (Ubuntu) "TrailingAction widths needs to be clamped" [Undecided,New] | |
| [10:15] <zsombi> nik90: when I'm trying to run the clock-app on desktop, I'm gettin' this: ubuntu-clock-app/app/clock/ClockPage.qml:247, unknown) - file:///home/zsombor/projects/canonical/core-apps/ubuntu-clock-app/app/clock/ClockPage.qml:247:20: Unable to assign [undefined] to QColor | |
| [10:15] <zsombi> Unable to select module, using dummy. | |
| [10:15] <zsombi> Loading module: 'libubuntu_application_api_test.so.2.9.0' | |
| [10:18] <nik90> zsombi: on wily? | |
| [10:18] <zsombi> nik90: vivid | |
| [10:18] <zsombi> nik90: desktop | |
| [10:18] <nik90> zsombi: does it run after those messages? | |
| [10:18] <zsombi> nik90: I'm gettin' a segfault after that | |
| [10:18] <nik90> zsombi: https://bugs.launchpad.net/ubuntu-clock-app/+bug/1464343 | |
| [10:19] <ubot5> Launchpad bug 1464343 in location-service (Ubuntu) "Crash on wily-proposed" [Undecided,New] | |
| [10:19] <nik90> zsombi: the warnings you get are very similar to that reported in the bug | |
| [10:20] <nik90> zsombi: for your testing purposes, just disabled qt position code in https://bazaar.launchpad.net/~ubuntu-clock-dev/ubuntu-clock-app/utopic-3.0/view/head:/app/clock/ClockPage.qml#L56 while testing it locally | |
| [10:21] <nik90> line 56 to line 111 | |
| [10:21] <nik90> zsombi: that should disable them all for you | |
| [10:21] <zsombi> ah, so it is the GEO location... | |
| [10:22] <nik90> zsombi: no .. its QT Location .. GeoLocation C++ Plugin that I wrote is not the cause of the crash | |
| [10:22] <nik90> :) | |
| [10:22] <zsombi> nik90: uhm... yes... sorry | |
| [10:27] <zsombi> nik90: hmm!!! I cannot get this crash with memory manager, but with EDS, I do get it straight! | |
| [10:27] <nik90> zsombi: bah...:/ | |
| [10:28] <zsombi> nik90: and on desktop as well | |
| [10:31] <zsombi> nik90: and I'm no longer getting crashes :/ | |
| [10:32] <nik90> zsombi: how did it just stop crashing if the environment was the same | |
| [10:32] <zsombi> nik90: no idea... | |
| [10:33] <zsombi> nik90: something feels to be racy... like the signal handler in the model when fetches the data from the hash... | |
| [10:36] <zsombi> nik90: but, none of the alarms ring anymore!!!! | |
| [10:36] <nik90> zsombi: while extensive testing I noticed that i-dt did not pick up the alarms I created..but a reboot fixed it. | |
| [10:37] <nik90> i-dt did not pick up alarms after a while* | |
| [10:37] <zsombi> uhh... | |
| [10:37] <nik90> zsombi: but I just ignored it and focused on reproducing the crash | |
| [10:38] <zsombi> nik90: no matter how hard I try I don't get it... | |
| [10:38] <nik90> zsombi: well EDS is the black box here that we're poking too hard at | |
| [10:39] <zsombi> nik90: yeah... agree... and noone has time to get a working backend between toolkit and i-dt | |
| [10:40] <zsombi> nik90: I am wondering why QAbstractItemModel::endMoveRows appears to be called... I do not see anything between the logs | |
| [10:41] <zsombi> nik90: ah, hold on... yes, I am calling it | |
| [10:41] <nik90> zsombi: when you enable/disable alarms the alarms order changes..could it be that? | |
| [10:41] <mcphail> Now OTA4 is released, can we use 15.04 framework as standard? | |
| [10:42] <popey> yes | |
| [10:42] <popey> but anyone on pre-15.04 won't get your app | |
| [10:42] <popey> (ie. those who haven't updated) | |
| [10:44] <mcphail> great | |
| [10:48] <mcphail> of course, will need to read up on qmake... | |
| [10:54] <zsombi> nik90: dude... | |
| [10:54] <zsombi> nik90: when the alarm enabling crashed, wasn't the alarm about to stay in the same position as it was before? | |
| [10:56] <zsombi> nik90: the crash seem not to have any changes in the stack, so that may be an issue here | |
| [10:58] <nik90> zsombi: hmm I think so..many a time I see the string "AlarmPassed" which indicates that the alarm date is incorrect..which in turns results in it staying in the same position as it was before. | |
| [10:58] <nik90> and that's when I have noticed it crash.. | |
| [10:59] <nik90> zsombi: upside is that a community contributor found a fix for the invalid alarm date being calculated by the clock app. But it still doesn't fix the underlying crash issue upstream. | |
| [10:59] <zsombi> nik90: I think I found teh problem: I got a crash at least when there was no more than one alarm, and the update was about to happen on the same item... | |
| [11:00] <zsombi> nik90: so, beginMoveRows() may return false if the from==to | |
| [11:00] <zsombi> nik90: in which case there's no change pushed into the QAbstractItemModel | |
| [11:00] <zsombi> nik90: but endMoveRows() doesn't check that | |
| [11:02] <nik90> ah ok | |
| [11:02] <mcphail> Would be good to cull the old frameworks from the SDK. The choices are messy and confusing. Perhaps relocate them to some kind of "legacy" menu? | |
| [11:03] <nik90> mcphail: report a bug against qtcreator-ubuntu-plugin on launchpad. That does sound like a good idea btw ..in a way encouraging developers to go with newer frameworks | |
| [11:04] <davmor2> popey: wow I haven't looked at app stats in a while 183 people have installed Sad Trombone :) | |
| [11:04] <mcphail> nik90: i'll do that later and see if it gathers support :) | |
| === chihchun is now known as chihchun_afk | |
| [11:18] <zsombi> bzoltan: here you have https://bugs.launchpad.net/ubuntu-clock-app/+bug/1463430 | |
| [11:18] <ubot5> Launchpad bug 1463430 in Ubuntu Clock App "Clock app crashes when renabling an alarm" [Critical,Confirmed] | |
| [11:36] <zsombi> nik90: https://code.launchpad.net/~zsombi/ubuntu-ui-toolkit/alarmmodel-segfault/+merge/262074 | |
| [11:38] <nik90> zsombi: were you able to reliably get clock app to crash on vivid desktop before applying the patch? | |
| [11:38] <zsombi> nik90: I did not get any crashes anymore... | |
| [11:39] <nik90> zsombi: I am just thinking of how to reliably test the patch and ensure it fixes the crash. | |
| [11:39] <nik90> ok | |
| [11:39] <zsombi> and I got few more chashes before this | |
| [11:39] <nik90> zsombi: alrite I will test on my vivid desktop with EDS and approve/comment on the MR tonight. | |
| [11:40] <nik90> zsombi: Just wondering if I should postpone clock app release to OTA-5 in this case since UITK wont be updated until then. | |
| [11:41] <zsombi> nik90: unless we land one more landing is possible... but perhaps not... | |
| === MacSlow is now known as MacSlow|lunch | |
| === mzanetti is now known as mzanetti|run | |
| === mzanetti|run is now known as mzanetti | |
| === _salem is now known as salem_ | |
| === MacSlow|lunch is now known as MacSlow | |
| [13:32] <davmor2> sad trombone receives it's first update \o/ 0.2 | |
| [13:36] <DanChapman> \o/ | |
| [13:52] <Mirv> zsombi: do you think the bug is therefore not in Qt, or is that MP just a workaround to a Qt bug? | |
| [13:53] <zsombi> Mirv: it doesn't seem to be a Qt bug, right | |
| [13:55] <Mirv> zsombi: ok, thanks, marking so | |
| [13:56] <zsombi> Mirv: if we still find it to be the guilty, then we're gonna get back to it, but so far I found that the crash comes when the model data doesn't change position in the list => QVector size is 0 when the stack tries to pop a value from it, which asserts/segfaults as there's no data in | |
| [14:00] <Mirv> zsombi: ok, for any next possible Qt problem we now have Qt 5.4.2 in wily (starting today) so it'll be useful to compare whether there's an existing upstream fix | |
| [14:01] <zsombi> Mirv: +1 | |
| === chihchun_afk is now known as chihchun | |
| === davidcalle_ is now known as davidcalle | |
| [15:01] <davmor2> popey: once upon a time did I here you say that the calendar guys were looking at a better sync-er? | |
| [15:02] <popey> davmor2: yes, bfillers team are looking at buteo | |
| [15:02] <davmor2> popey: ah thanks I was sure I heard it but thought I was dreaming ;) | |
| [15:03] <popey> :) | |
| === salem_ is now known as _salem | |
| [17:44] <meles> Hey nik90, bugfix #1442518 seems to work. I had problems that my alarm clock wasn't playing any sounds but that can't be related to the fix. I just hope it changes after the update tdy. | |
| [17:58] <nik90> meles: yeah..same had good results with it. | |
| [17:59] <nik90> meles: btw it seems alarms created before OTA-4 stop working after the upgrade..so you will need to disable and then reenable them again. | |
| === _salem is now known as salem_ | |
| [18:06] <meles> nik90: ok, i don't have ota-4 yet. but thanks for the info | |
| === salem_ is now known as _salem | |
| [18:25] <davmor2> nik90: https://bugs.launchpad.net/ubuntu/+source/evolution-data-server/+bug/1465806 I'm blaming eds :) | |
| [18:25] <ubot5> Launchpad bug 1465806 in evolution-data-server (Ubuntu) "Alarms not shown in indicator datetime after upgrade to ota4" [Undecided,New] | |
| [18:29] <nik90> davmor2: our common enemy :P | |
| [18:30] <nik90> davmor2: but how is that bug useful now that everyone is already on OTA-4? | |
| [18:30] <nik90> I suppose we could add it to our manual test-suite but that's the best we can do at the moment | |
| [18:31] <davmor2> nik90: well after looking at it, the clock app does what it is meant to, the indicator does what it is meant to, eds just says there are no alarms/appointments or anything else to the indicator | |
| [18:31] <davmor2> nik90: it is more this is how you reproduce the issue so the devs can fix it and it never happens again | |
| [18:32] <davmor2> nik90: also if you look I added the workaround till there is a fix too :) | |
| [18:33] <nik90> davmor2: ack. just waiting to get of the EDS band wagon. | |
| [18:33] <davmor2> nik90: no you have to suffer it forevererer | |
| [18:34] <nik90> haha. | |
| [18:35] <davmor2> nik90: also shell rotate and clock is not pretty ;) just saying :) Be nice if it showed clock to the left and timezones on the right but I don't know how much work that is :) | |
| [18:36] <nik90> davmor2: shouldn't be too much work really..just column -> row conversion..but I cannot change UI without design approval. At the moment I have a MP ready which locks it to Portrait. | |
| [18:37] <davmor2> nik90: I think the big issue especially on krillin is that because it is so small it is too easy to swipe up the alarms basically :) | |
| [18:38] <nik90> davmor2: you find yourself accidentally swiping the bottom edge while trying to scroll through the world city list? | |
| [18:38] <davmor2> nik90: yeap because the clock takes up the whole of the screen | |
| [18:38] <nik90> I have that issue in the calculator app when trying to press the 0 and = button | |
| === _salem is now known as salem_ | |
| === salem_ is now known as _salem | |
| === _salem is now known as salem_ | |
| [19:42] <balloons> nik90, is it a-ok to make some infrastructure changes to clock AP tests? | |
| [19:43] <balloons> going to add in the cmakepluginparser and remove the emulators language amongst other things | |
| [19:45] <mcphail> nik90: https://bugs.launchpad.net/ubuntu/+source/qtcreator-plugin-ubuntu/+bug/1465825 - as discussed earlier | |
| [19:45] <ubot5> Launchpad bug 1465825 in qtcreator-plugin-ubuntu (Ubuntu) "Outdated frameworks and templates should be moved or removed" [Undecided,New] | |
| === athairus is now known as afkthairus | |
| === afkthairus is now known as athairus | |
| === chihchun is now known as chihchun_afk | |
| [20:54] <nik90> davmor2: why is this bug marked duplicate? https://bugs.launchpad.net/canonical-devices-system-image/+bug/1465806 | |
| [20:54] <ubot5> Launchpad bug 1463430 in ubuntu-ui-toolkit (Ubuntu) "duplicate for #1465806 Clock app crashes when renabling an alarm" [Critical,Confirmed] | |
| [20:54] <nik90> davmor2: the bug it is marked duplicate of is a crasher bug | |