=== cwayne is now known as cwayne-afk === olli_ is now known as olli [06:06] hey there [06:07] so I just updated my krillin to current rtm, and the unity dash is empty [06:07] like apps, music, videos, all empty/without icons [06:07] is there any debug info that would be useful? [06:07] Saviq ^ [07:38] hum, no unity8 hackers this morning? [07:38] I sort of need to use the phone but I would like to get debug infos before I reboot it [07:50] seb128: what' up? [07:51] tsdgeos, so I just updated my krillin to current rtm, and the unity dash is empty [07:51] like apps, music, videos, all empty/without icons [07:51] is there any debug info that would be useful? [07:51] seb128: does pulling down to refresh help? [07:51] I can swipe between scopes or use the launcher, start apps, etc [07:51] I didn't try, I wanted to keep it in this state in case that was useful [07:51] do you want me to try? [07:51] yeah [07:51] though it seems more like if the qml is borked [07:51] yes, fixed it [07:51] and not being loaded [07:51] it did? [07:52] yes [07:52] wow [07:52] only on apps [07:52] music still empty [07:52] sure [07:52] pull down only refreshes the current scope [07:52] k [07:52] seb128: sounds more like a backend scopes bug to me [07:52] pstolowski: ↑↑↑ [07:53] the unity8-dash.log is spammed with those [07:53] "file:///usr/share/unity8/Dash/CardTool.qml:179:38: Unable to assign [undefined] to bool" [07:53] not sure that's the issue though [07:53] man, that log is verbose [07:54] nah that's not horrible [07:54] Pre-populating scope "musicaggregator" [07:54] Pre-populating scope "videoaggregator" [07:54] Caught an error from create_query(): unity::scopes::TimeoutException: Request ti [07:54] med out after 500 milliseconds [07:57] that may be [07:58] it's still backend [08:03] seb128, hey, anything interesting in scopes-registry.log? [08:03] pstolowski, hey [08:05] marcustomlinson, pete-woods ^ [08:05] marcustomlinson, pete-woods may still we been the issue with location data? [08:05] pstolowski, [08:05] RegistryObject::ScopeProcess::on_process_death(): Process for scope: "clickscope" exited [08:05] not sure that's "interesting" [08:06] same for music, video, etc [08:07] seb128, no, this is perfectly fine [08:07] k [08:07] nothing looking weird then [08:15] seb128, it's most likely https://bugs.launchpad.net/ubuntu/+source/unity-scopes-api/+bug/1374481 [08:15] Ubuntu bug 1374481 in The Savilerow project "Nearby and News scope do not load on reboot" [Critical,Confirmed] [08:16] pstolowski, so, any info I can get? I refresh apps which made it populated [08:16] music and video are still empty [08:16] wikipedia and my photos have content [08:16] seb128: music and video are not populated on refresh either? [08:18] pstolowski: can you merge feeds with trunk of scopes-shell? also do you have an eta to make the apps scope movable? [08:22] tsdgeos, I didn't try to refresh those, keeping some buggy state for debug [08:22] ok [08:22] sorry that was unclear from what I wrote [08:23] tsdgeos, it has to be movable? [08:23] pstolowski: yes, see mail form James i CC'ed you [08:23] damnit [08:24] maybe i didn't cc you [08:24] Pawel Stolowski <1281602@bugs.launchpad.net> [08:24] ?¿ [08:24] huh? [08:25] yeah no idea how that ended up there :d [08:25] but obviously not in your mailbox [08:25] i'll forward them [08:29] tsdgeos, anyway, that's trivial to do [08:29] pstolowski: cool, can you do it soon-ish along with the merge to trunk? [08:29] tsdgeos, and yeah, i'll merge in a moment [08:29] cool [08:35] pstolowski, so, anything useful I could provide while some scopes are still empty? [08:38] bah [08:38] unity8 segfaulted again [08:38] on r109 [08:38] well, unity8-dash [08:40] do you have a backtrace? [08:40] or way to reproduce? [08:41] no :-/ [08:42] dunno why apport doesn't collect the dump [08:42] trying to reproduce [08:42] it happened when I pressed the top left "back" arrow to go from the click store back to the app lens [08:45] shrug [08:45] rebooted the phone, on rtm 109, no app icons again [08:45] pstolowski, ^ [08:45] that's twice in a row [08:45] on the rtm image [08:45] that's suboptimal :/ [08:45] music and video have content this time, only apps are missing [08:46] seb128, icons missing is not a scopes / backend issue [08:46] pstolowski, sorry, when I mean "icon", it's the app page is empty [08:46] no content [08:46] seb128, aah [08:47] seb128, anything in logs this time? [08:48] no :/ [08:49] seb128, and if you swipe down to refresh, it works? [08:49] pstolowski, yes [08:51] seb128: can you pastebin the scope-registry.log [08:54] marcustomlinson, http://people.canonical.com/~seb128/scope-registry.log [08:58] marcustomlinson, e.g the run from 16 07:43:26 was buggy [08:59] tsdgeos, ok, it's easy to segfault [08:59] that's bad :D [08:59] by clicking through click store and back to dash [08:59] (gdb) bt [08:59] #0 0xadac0626 in unity::scopes::Category::query() const () [08:59] from /usr/lib/arm-linux-gnueabihf/libunity-scopes.so.3 [08:59] #1 0xadb69f5a in scopes_ng::Categories::data(QModelIndex const&, int) const () [08:59] from /usr/lib/arm-linux-gnueabihf/unity8/qml/Unity/libUnity-qml.so [08:59] #2 0xb64ebe4a in QSortFilterProxyModel::data(QModelIndex const&, int) const () [08:59] from /usr/lib/arm-linux-gnueabihf/libQt5Core.so.5 [08:59] * seb128 installs dbg [09:00] pstolowski: ↑↑↑ this points at you guys [09:02] jamesh: the discussion about scopes crashing is here ^ [09:18] seb128: when did all this start happening? any way to roll back. This wasn't happening yesterday right? [09:18] seb128: something broke recently [09:19] marcustomlinson, what issue? the segfault or the empty dash? [09:19] the empty dash, I got it for the first time today [09:19] and got it twice in 3 reboots [09:23] Saviq: so we just landed into rtm and not trunk because of time, right? [09:23] tsdgeos, and because not everything got reviewed, ye [09:23] s [09:24] ok [09:24] tsdgeos, but there's a silo to land everything that went into rtm (and more) into trunk already, yes [09:24] ok :) [09:24] but awaiting reviews [09:24] tsdgeos, and if I get any say in it, we'll just sync trunk to rtm soon after that [09:25] ok [09:26] tsdgeos, i'be pushed feeds [09:26] tx [09:43] jamesh, marcustomlinson, pstolowski, hit that one again [09:43] #0 unity::scopes::Category::title (this=0x30) [09:43] at /build/buildd/unity-scopes-api-0.6.7+14.10.20141010.1/src/scopes/Category.cpp:56 [09:43] No locals. [09:43] #1 0xadd9de02 in scopes_ng::Categories::data(QModelIndex const&, int) const () [09:43] from /usr/lib/arm-linux-gnueabihf/unity8/qml/Unity/libUnity-qml.so [09:43] No symbol table info available. [09:43] #2 0xb6531e4a in QSortFilterProxyModel::data(QModelIndex const&, int) const () [09:43] from /usr/lib/arm-linux-gnueabihf/libQt5Core.so.5 [09:43] segfault [09:46] jamesh, marcustomlinson, pstolowski, opened https://bugs.launchpad.net/ubuntu/+source/unity-scopes-api/+bug/1381970 about it [09:46] Ubuntu bug 1381970 in unity-scopes-api (Ubuntu) "unity8-dash segfaults when using click store" [Undecided,New] [09:46] seb128, i'm reflashing to try to repro [09:47] pstolowski, it's not trivial to trigger, but I had a bunch of times since this morning [09:47] just tapping through click store results and back for a while [09:50] seb128, this is odd.. we haven't been touching this area for very long time, not sure why that started happening just now [09:50] pstolowski, not sure it did, the empty scope is new [09:51] pstolowski, unity8-dash segfaulting sometime is not new, but there are enough issues that I never bothered trying to spend time looking at the issue [09:51] I assumed that people were working on those issues [09:51] but now we are hitting rtm [09:51] and I hit like 3 segfaults yesterday, which is a bit embarassing [09:52] so I decided to try to see if I can help provide info/see if they are being worked on [09:53] lunch time [09:53] bbiab [09:57] tsdgeos, did we enable qml caching recently? [09:58] pstolowski: we did [09:58] and i think we disabled it [09:58] shortly afterwards [09:58] since it made everything crash like crazy [09:58] Saviq: ↑↑↑ [09:59] tsdgeos, pstolowski it's enabled in rtm [09:59] it doesn't exist in utopic yet [09:59] tsdgeos, hmm ok. that bt seb128 pasted looks like a row out of valid range is requested in ::data() [09:59] tsdgeos, which is weird [09:59] pstolowski: not that weird tbh, you have to protect against that :D [10:00] tsdgeos, and then it crashes trying to access invalid object [10:00] tsdgeos, yes, that's true === cwayne-afk is now known as cwayne [10:15] Saviq: what exactly is expected of indicator-sound with regards to the volume bubble? (I saw you reassign a couple of bugs to it) [10:15] Saviq: or is tedg on top of it? [10:15] larsu, that it doesn't appear when you have a slider on screen (from indicator or settings) [10:17] Saviq: how should indictor-sound know that there's a slider on screen? [10:17] larsu, it needs to be told by the indicator and settings ;) [10:18] larsu, basically if indicator-sound.open == true [10:18] larsu, and that action will be set by the shell when the indicator is open, and by settings app when its sound page is open [10:19] larsu, I'm worried, though, that people will want to drag the bar in the bubble, which would be a little more involve [10:19] d [10:20] Saviq: hm, right now it only seems to show a notification in the scroll action, which isn't even used on the phone [10:20] Saviq: clearly this is getting out of hand. We need a proper volume control module in unity8 [10:21] Hola! [10:22] larsu, we actually don't, we need to pass the vol+/vol- to the indicator and let it do with it as it pleases [10:25] Saviq: except for handling user feedback, because clearly it doesn't have enough information [10:25] Saviq: I'm really starting to come around to the idea that indicators need to become smarter in the ui layer [10:25] larsu, right now it does, it's easy to add an "open" action [10:25] larsu, but yeah, the latency of it is not great [10:25] Saviq: which doesn't help in the wizard case, does it? [10:26] larsu, well, the wizard case is different, it's triggering volume changes for no apparent reason [10:26] larsu, the problem isn't that the bubble appears, rather that it's triggered in the first place [10:27] Saviq: we'll have system components that change the volume at some point [10:27] Saviq: and then we add another exception to i-sound? [10:28] the default case should be _not_ to show a notification [10:35] larsu, I don't really care, if you have a better plan, fine by me [10:36] Saviq: I do, but nobody has time === MacSlow is now known as MacSlow|lunch === _salem is now known as salem_ === alan_g is now known as alan_g|lunch [12:34] larsu, could you help with bug #1381871, looks like indicator-sound triggers a bubble on start [12:34] bug 1381871 in ubuntu-system-settings (Ubuntu) "[wizard] volume notification showing up during wizard (with volume as 0)" [Critical,Confirmed] https://launchpad.net/bugs/1381871 [12:34] and it's restarted because of language change in the wizard [12:35] Saviq: the restarting makes sense, but I wonder why on earth it would trigger a notification on startup [12:39] larsu, I know even less [12:39] tsdgeos, pstolowski, what do we know about the "apps scope empty on boot" issue? [12:39] Saviq: it doesn't do that for me. My bet is on unity8 requesting a state change... [12:40] * larsu builds unity, again [12:40] larsu, bear in mind it's not landed in trunks, only in rtm [12:40] larsu, and unity8 isn't running at that point [12:40] Saviq: can you point me to the branch that'S in there? [12:40] larsu, https://code.launchpad.net/~indicator-applet-developers/indicator-sound/trunk.rtm-14.09 [12:42] Saviq: ya, this one shows a notification on every volume change. wtf? [12:42] in the wrong place, too [12:42] larsu, that's on desktop is it? [12:42] larsu, yeah, that's https://code.launchpad.net/~ted/indicator-sound/synchronous-notification/+merge/238252 [12:43] larsu, on the phone it behaves ok [12:43] larsu, it's approved into trunk as well https://code.launchpad.net/~ted/indicator-sound/synchronous-notification/+merge/237666 [12:43] Saviq: didn't you just tell me it doesn't behave ok? [12:43] larsu, with that one exception on restart [12:44] s/re// [12:44] this code has changed quite a bit since I last touched it.. [12:45] larsu, you nice code got tedified? ;-) [12:45] you said that [12:45] ;-) [12:46] seb128, popey, bug #1382039 [12:46] bug 1382039 in unity8 (Ubuntu) "Apps scope empty on boot" [Critical,Triaged] https://launchpad.net/bugs/1382039 [12:46] Saviq, thanks [12:46] thanks! [12:46] seb128, popey, was wizard involved in your case? [12:46] Saviq, no [12:46] Saviq, I had it after an ota update and one after a debug reboot [12:46] both on installed system with no wizard [12:49] Saviq: i *think* it's the scopes timing out, or that seemed what seb128 had [12:50] tsdgeos, yeah, something like that... [12:50] Saviq: I don't see it on the desktop because something about pulse stream restore I don't understand [12:50] Saviq: and also refreshing got them back [12:50] Saviq: however, the code very much looks like it triggers on every volume change [12:50] tsdgeos, yeah yeah, same here [12:51] Saviq: which cannot be what we want, or is it? [12:51] larsu, is it, see spec [12:51] larsu, why not? [12:51] larsu, https://wiki.ubuntu.com/Sound === MacSlow|lunch is now known as MacSlow [12:51] well then the notifiction is correct as well [12:51] volume changes -> notification [12:51] larsu, https://bugs.launchpad.net/bugs/1232633 [12:51] Ubuntu bug 1232633 in unity8 (Ubuntu) "[system] Changing volume doesn't show feedback on the phone" [High,In progress] [12:51] larsu, well, it doesn't change, does it, it's just the indicator that restarts [12:52] larsu, it shouldn't change on indicator restart [12:52] bss, I hate it when you're right [12:52] lol [12:53] sorry :) [12:59] pstolowski, where would I look for info on the "apps scope empty on boot" issue? === alan_g|lunch is now known as alan_g [13:07] tedg, bug #1381871 [13:07] bug 1381871 in indicator-sound (Ubuntu) "[wizard] volume notification showing up during wizard (with volume as 0)" [Undecided,Confirmed] https://launchpad.net/bugs/1381871 [13:08] i notice it shows up on boot on my laptop too [13:08] since a couple of weeks back [13:08] larsu, So are you fixing that or should I? [13:08] Saviq, seb128: got a fix. Feel dirty. [13:08] tedg: just fixed it [13:08] kgunn, I think we fixed that case. [13:08] * larsu commits [13:08] ack [13:09] \o/ [13:10] larsu: thanks! [13:10] larsu, Saviq, tedg: I see we have a separate branch for 14.09 - so we need to release it separately for utopic and ubuntu-rtm, right? [13:11] sil2100, yes, this didn't even land in utopic yet [13:11] sil2100, Correct [13:11] sil2100, yesteday's silo, that is [13:11] Ok, I would love to fast-track it with priority to ubuntu-rtm anyways [13:11] We would like to land that ASAP and kick a new image with that + qtmir [13:12] sil2100, there is also bug #1382039, for which we have no steps to repro yet, but 3 people saw it today [13:12] bug 1382039 in unity8 (Ubuntu) "Apps scope empty on boot" [Critical,Triaged] https://launchpad.net/bugs/1382039 [13:13] sil2100, has silo 12 (?) landed yesterday, one that mentioned the long delay in dash on stratup? [13:13] Saviq: ok, noting - for now management didn't put it on the high prio list, same for QA [13:13] Saviq: you mean the scopes fix? [13:13] sil2100, yes [13:14] Saviq: yeah it landed [13:14] Saviq: yes, it still made it in [13:14] sil2100, that sounds relevant for this new bug [13:14] Saviq: branch is attached. Should I merge propose that onto the rtm branch? [13:14] larsu, yes please [13:14] larsu, you'll need a separate branch, since trunk and rtm diverged [13:15] Saviq: a separate branch for what? [13:15] larsu, ah, you based that off of rtm anyway, right [13:15] tedg, so you might want to incorporate that into your branch for trunk ↑ [13:16] larsu, It needs to be a separate branch but you just need to merge lp:~ted/indicator-sound/rtm-silent-mode-item to handle the reversion that happened on the rtm branch. [13:16] larsu, Then it can be proposed into trunk [13:18] tedg: the branch is broken. Don't merge it into trunk. [13:18] tedg: for example, scrolling over the indicator on desktop now makes two notifications [13:19] larsu, ? I thought I blocked the second notification by checking to see if we're using role based pulse. That shouldn't happen on desktop. [13:20] tedg: ah sorry, I turned that off in my built to be able to see the problem. It's still a bit weird that we emit notifications from two different places [13:20] tedg: and the role based stuff doesn't really scream "not on desktop" to me (and maybe we'll get it there too at some point?) [13:21] larsu, Yes, so "the idea" was that we'd turn that role stuff on desktop and then not make u-s-d send the notifications. [13:21] larsu, So we wouldn't have two sources of them either. [13:21] V-work [13:25] larsu: do we have an MP for the 14.09 branch? Since we would like to assign a silo for that ;) [13:26] sil2100: ya, https://code.launchpad.net/~larsu/indicator-sound/lp1381871/+merge/238555 [13:26] I linked it to the bug as well [13:26] larsu: thanks! === tvoss is now known as tvoss|food [13:31] tedg, Saviq: can I get anyone of you guys reviewing that branch ^ [13:31] sil2100, Just did :-) [13:31] larsu, tedg, Saviq: landing prepared and silo building as we speak :) [13:31] coolz [13:31] tedg: can you help out with testing as well? [13:31] sil2100, How do you test that without building an image? [13:31] easy to test... `restart indicator-sound` [13:32] Ah, okay. [13:32] tedg, just install the .deb and enable wizard if you want to test the actual path [13:32] phablet-config --help to the rescue [13:33] Usually for me the wizard is just the annoying side effect of --wipe [13:33] Never realized you could turn it back on :-) [13:33] Saviq, scope-registry.log is the main starting point for finding out about any scopes backend issues, but it didn't lead us to anything useful yet [13:33] not for this particular issue [13:34] pstolowski, ok, I'll let you know if I stumble upon it again === karni is now known as karni-lunch [13:36] Saviq, pete-woods has experienced it, grabbed logs etc but what's causing it is still unknown [13:37] pstolowski, :/ [13:37] pstolowski, please use bug #1382039 to record findings [13:37] bug 1382039 in unity8 (Ubuntu) "Apps scope empty on boot" [Critical,Triaged] https://launchpad.net/bugs/1382039 [14:11] sil2100, larsu's fix works for me mako r91 [14:11] sil2100, Tested by restarting and reenabling the wizard [14:13] tedg: mako 91 is ubuntu-rtm, right? [14:13] Let me do a re-confirm as well [14:13] sil2100, Yes [14:38] Saviq, tedg, larsu: confirmed, fixed [14:39] sil2100, awesomes [14:39] sil2100, did you talk to pete-woods about the empty scopes issue? [14:39] nice, thanks [14:39] sil2100, they've a revert ready, everyone and their mother have been hitting it === karni-lunch is now known as karni [14:44] Saviq: didn't poke them about that yet... QA didn't yet mention this as an issue so not sure if we'll be able to get this in [14:44] Saviq: is the revert in a landing already? [14:44] (or you prefer just reverting it in the archive?) [15:03] Saviq: FYI, we just reproduced the empty scopes bug with our reverted branch [15:04] pete-woods, uhm === mhall119 is now known as mhall119|vacatio === gatox is now known as gatox_lunch [15:59] tedg, larsu, Saviq: since I missed that out probably... did we already release the indicator-sound fix for utopic? [15:59] sil2100, We have not yet. [15:59] sil2100, We haven't released any sound notifications for utopic. [16:00] ACK, let's write that down somewhere and try not to forget, don't want to go out of sync too much [16:00] sil2100, Once CI Train becomes unfroze we have a bunch of syncing to do. [16:00] sil2100, Utopic has a bunch of bug fixes that weren't allowed in rtm. === gatox_lunch is now known as gatox === alan_g is now known as alan_g|EOD === dpm is now known as dpm-afk