[01:25] mterry: ping === salem_ is now known as _salem === zniavre__ is now known as zniavre [09:34] JohnLea: Hi, sil2100 suggested me to contact you about a bug. ping me when you have a bit of time [09:36] s/a bit of/some/ [09:41] Trevinho: I see Windows() has been changed between 5.0 and 6.0, Re second comment - I build _window_menu_items in AddMenuItemsWindowList and when this method finishes, then I should merge _window_menu_items into results in GetMenus? [09:45] won't be able to get those fixed until next week though :-/ going to fosdem tomorrow morning ... are you guys going? [10:06] luv: unfortunately not... [10:06] luv: no problem, do that when you've time [10:12] rperier; hyia, what's the issue? [10:14] JohnLea: hi, this is the bug 1108956 [10:14] bug 1108956 in Unity "The dash closes when trying to switch from Command lens to home lens" [Undecided,New] https://launchpad.net/bugs/1108956 [10:15] I solved bug which closed the dash when switching from the dash to the command lens. This bug is the oppositive [10:15] and I don't know if this is a real bug or an expected behaviour [10:16] so I need to ask the design team :) [10:20] rperier; ahh, the reason for this issue there is that the ALT F2 command mode was never intended to be part of the Dash. I just tested it now and it brings up the Dash, complete with the bottom Dash nav bar!! however this was never the original intention. Someone must have changed it during the 12.04 development cycle and we never noticed... But given that this change has happened, your suggested fix sounds sensible, so +1 from me, [10:20] I am updating the bug with a 'design ok'. [10:22] awesome ! thanks ! [10:22] So, we should be able to switch to any lens without closing the dash ? right ? [10:38] duflu: Hi :) I fixed everything here: https://code.launchpad.net/~mc-return/compiz/compiz.merge-fix966099-add-unmaximize-or-minimize-window-key/+merge/145464 [10:39] btw does the merge proposal means ayatana team is happy with the idea? [10:39] duflu: Would be nice if you could approve it, then we could merge the Unity branch as well... [10:39] MCR1: Ok thanks. Wait for Brandon or someone to review it. I'm not working on Compiz right now [10:39] duflu: Oh no - we need you ! [10:44] luv: I guess so - Fix Committed in the Ayatana Design task usually means 'ok' to design [10:44] luv: but you can poke JohnLea directly to be sure - but since I indeed saw his merge comment, once these things get fixed it seems OK to be merged into unity raring ;) [10:45] oh, that's great! :-) [10:46] * rye thinks that dynamic workspaces is not such a bad idea. I need to switch context completely and have a new set of apps opened. But I need to do this once a week or even less. And i don't need all other workspaces at the other times... [10:50] not a big fun of dynamic workspaces myself, but I really like the new polished gtk theme in raring [10:53] fan [11:38] bregma: Hi :) This: http://bazaar.launchpad.net/~compiz-team/compiz/0.9.9/revision/3584 seems to have caused a regression. [11:39] bregma: Compiz is supposed to load self-compiled plugins from ~/.compiz-1/plugins - they should override standard installation plugins - Compiz is not doing that anymore [11:41] * MCR1 was wondering why his fixes in ~/.compiz-1/plugins suddenly were not working anymore... [11:51] sil2100: should we perhaps ask for removal of the bamf:s in the p/q queues? the fix we got in there is only 1/2 of the real fix, and it's not worth that much to only publish the partial fix [11:52] sil2100: Trevinho has now committed the full fix to all our branches, so I'd suggest 1. ask for removal, 2. cherry-pick on top of our p/q bamf:s, 3. push again [11:52] (bug #1108380) [11:52] bug 1108380 in BAMF 0.3 "Java Web Start windows launched from browser are still grouped even if different" [Medium,Fix committed] https://launchpad.net/bugs/1108380 [11:52] Mirv: agreedo [11:56] hi, i have a weird issue with openoffice in raring - the global menu somehow gets disconnected from the underlying application - i can click Print Preview and a tick will appear in the menu but nothing else will happen - dbus-monitor says string "No such interface `org.gtk.Actions' on object at path /window/106954808" [11:56] === _salem is now known as salem_ [12:02] bug #1085169, nevermind [12:02] bug 1085169 in libreoffice (Ubuntu) "LibreOffice Menus Stop Working even with libreoffice>=1:3.6.2~rc2-0ubuntu4 and indicator-appmenu>=12.10.3-0ubuntu2.1" [Undecided,Confirmed] https://launchpad.net/bugs/1085169 [12:26] If someone has time to approve those, I would be pleased :) : [12:27] 1. https://code.launchpad.net/~mc-return/compiz/compiz.merge-fix1099100-thumbnail-title-text-issues.0/+merge/144954 [12:27] 2. https://code.launchpad.net/~mc-return/compiz/compiz.merge-fix966099-add-unmaximize-or-minimize-window-key/+merge/145464 [12:27] 3. https://code.launchpad.net/~mc-return/unity/unity.merge-fix966099-shortcut-fails-to-minimize-just-restores/+merge/145474 [12:45] MCR1, http://bazaar.launchpad.net/~compiz-team/compiz/0.9.9/revision/3584 caused an ABI break, you will need to rebuild your self-compiled plugins [12:53] bregma: Ah, yeah - that might be it - thanks :) === MacSlow is now known as MacSlow|lunch [13:09] Mirv: hm, sounds good to me, let me check the fix then [13:11] Mirv: I could take care of the quantal version then, but we'll have to probably re-do basic integration testing for those before we push [13:14] sil2100: yes [13:14] sil2100: I'll ask for the removals [13:19] that was fast [13:21] Removed? [13:21] Probably an one-click action [13:36] yes [13:59] hi all, i have ubuntu 12.10 unity 6.2 , do you know how to show windows of the app when hovering over the icon in the launcher??? [14:00] i know it can be done with double click , but it is really slow to make 3 clicks to access a window [14:00] pokkerface: would this https://bugs.launchpad.net/unity/+bug/1107866 help? [14:01] Launchpad bug 1107866 in unity (Ubuntu) "Show window list when right clicking an icon in launcher - enables quick window switching" [Wishlist,Triaged] [14:01] pokkerface: hi, you mean seeing what windows of an application are open? One click is enough on the launcher btw. [14:02] one click opens the last you had active [14:03] hi, https://bugs.launchpad.net/ubuntu/+source/unity/+bug/806248 [14:03] Launchpad bug 806248 in unity (Ubuntu Precise) "unity::TimeUtil::TimeDelta returns an int value which overflows after 24 days of uptime" [High,In progress] [14:03] need this pushed to precise [14:03] luv: yes that is what I meant [14:03] ritz: I think Mirv was taking care of pushing it to precise [14:03] Mirv: ^ [14:04] sil2100 Mirv thanks :) [14:04] this is from a premium customer [14:04] pokkerface: cool, I started working on that recently - needs some more polishing before unity devs are happy to merge though. I am happy the feature is in demand :-) [14:05] luv, excellent ! thanks [14:06] seb128: could you push that ^ to queue? merge request at https://code.launchpad.net/~timo-jyrinki/unity/ubuntu.lp806248-fix-int32-buffer-overflow-5.0/+merge/145178 [14:06] bregma: the fix has been accepted by the design team (it was you who had asked me to assign this bug to the design team, right ? don't remember...) [14:06] luv, what languages are used for programming the unity desktop? [14:07] I meant, the bug not the fix :) [14:07] Mirv, looking [14:08] Mirv thanks :) [14:08] ritz: no prob :) [14:12] Mirv++ [14:16] seb128: thank you for the upload [14:16] pokkerface: apt-src install unity and see for yourself :-) you can then make your changes and run dpkg-buildpackage to get nice deb pkgs [14:16] Mirv, yw, thanks for the fix ;-) [14:17] luv, ok ;) [14:28] luv, as a suggestion it would be good to have a small preview of the opened windows, to move the mouse less distance [14:31] pokkerface: there is a concept being worked on for getting it even better, here's the bug for that one: [14:31] https://bugs.launchpad.net/unity/+bug/1081843 [14:31] Launchpad bug 1081843 in unity (Ubuntu) "Launcher, Window Management - More Effective window switching for apps with multiple windows using the Launcher" [High,In progress] [14:32] pokkerface: this approach uses the mouse wheel as well, which is fast for mouse users [14:37] mterry, today's ps-indicators is showing an X crash on nvidia and 1 test failure on intel [14:38] mterry, ps-unity-autopilot had 52 total failures, near the normal amount [14:38] fginther, OK, we had 2 test failures on intel yesterday, so we're getting better! :) [14:40] mterry, do we need a clean indicator run before releasing? [14:41] sil2100, that implementation would be great! [14:41] sil2100, do you know what http://10.97.0.1:8080/job/ps-indicators-autopilot-release-testing/label=autopilot-intel/lastCompletedBuild/testReport/unity.tests.test_panel/PanelIndicatorEntryTests/test_menu_opens_on_click_Single_Monitor_/ might be? [14:43] fginther, I'm looking at failures now, trying to see how bad they are [14:43] mterry, ok [14:43] Ah, forgot about the failures [14:43] Looking now [14:44] mterry: ah, this one... it only happens on intel I noticed [14:44] mterry: could not reproduce that one yesterday, I assumed it was a single failure, but I see it happens more frequently ;/ [14:47] mterry: also, sometimes not really autopilot is at fault, but unity just has those random quirks [14:47] sil2100, is the video instructive? [14:48] sil2100, we're trying to get rid of these random quirks! [14:48] :) [14:48] mterry: it's as it was before - the Calculator doesn't export its menu to the panel, maybe because of some previously executed test? I'll try checking that [14:49] mterry: yes ;p But we need to fix those quirks in unity! [14:49] :) [14:51] fginther, I'm seeing "/tmp/hudson1736751589422415328.sh: line 4: ./fail_if_interesting_crashes.sh: No such file or directory" in unity's autopilot test console output [14:52] mterry, that was caused by a typo from veebers adding the "fail on crash" step [14:52] as it has been validated by the design team, could I get inputs on bug 1108956 ? (when you have some time) [14:52] bug 1108956 in unity (Ubuntu) "The dash closes when trying to switch from Command lens to home lens" [Undecided,New] https://launchpad.net/bugs/1108956 [14:52] mterry, If not for that error, the build would have finished unstable [14:52] fginther, is a fix on the way / deployed? [14:53] mterry, he claims to have fixed it in the job, should work on the next run [14:53] fginther, (I'm still getting familiar with verbiage) and by unstable, you mean tests failed, but not enough to cause the build to fail? [14:53] fginther, i.e. a "normal" amount of failure? [14:54] mterry, yes, the build is unstable because there were some test failures [14:56] Who's looking after misc stack? cyphermox ? [14:56] yeah [14:59] cyphermox, here's a chance for you to practice manually publishing if ya like. The stack failed for packaging changes [14:59] right [14:59] I haven't had time to create an account and all of that [14:59] mterry: could you please take care of it? [14:59] cyphermox, sure [14:59] I'm still working on nexus7 stuff [15:00] assuming of course the packaging changes make sense :) [15:00] cyphermox, I can't seem to get into nexus7 stuff :) [15:00] really? [15:00] cyphermox, they do :) [15:00] cyphermox, I have done a bit, just not as much as I wanted [15:00] aye [15:00] well... the not as much as I wanted is still true for me [15:00] it's crazy slow to get stuff done [15:01] rperier: I think you might assume that the best fix to this issue is being able to switch back to the dash when Super is pressed ;) [15:08] sil2100: this is what I suggested in the fix, it should switch to the home.lens (the default len when you press Super) without closing the dash :) [15:08] no? [15:08] rperier: yes, I think that's the right way to go [15:09] rperier: once you submit a merge request, it will be additionally validated by design and told when something's wrong ;) [15:09] I think that the code fix might be acceptable, however I am not sure at all about the autopilot test [15:09] sil2100: ok [15:09] noted [15:10] rperier: I can be of help with the autopilot test if anything [15:11] rperier: but it should be simply 'open the command lens, tap keybinding for dash home and check if the dash is visible and the home lens selected' [15:12] fginther, did you get around to adding the result of "ubuntu-bug XXX.crash --save=XXX.logs" to the artifacts? I don't see such a bug-bundle [15:12] yeah, I agree, I meant, I am not familiar yet with autopilot API ;) [15:12] sil2100: https://code.launchpad.net/~rperier/unity/command-lens-switching/+merge/145882 [15:13] mterry, checking to see if veebers got to it... [15:13] (feel free to do another task if you have something with a higher priority, there's no hurry) [15:13] rperier: will test and review in a moment! [15:13] ok [15:13] np [15:14] as I said, there's no hurry :) [15:18] mterry, sadly no, looks like only the extra -dbg packages made it. My irc logs indicate he was working on the bug collection but may have run into issues [15:19] mterry, I'll follow up with him today and see if we can get it in today [15:21] mterry, do you know if ubuntu-bug requires network access? [15:23] fginther, I don't believe so if you use it with --save. I think it just collects logs and such [15:24] mterry, thanks [15:24] fginther, so the problem with the nvidia crash is that the proprietary drivers are going to make it hard to get a useful stacktrace (and that is born out by no good stacktrace this time, even with -dbg). So logs are going to be most helpful here [15:28] mterry: it seems that the test_menu_opens_on_click failure is a regression that appears seldomly, will try to find the reason for that one once I'm done with something [15:28] mterry: the funny thing is that it only seems to happen on intel in jenkins, right ;) ? [15:29] sil2100, I'm not sure; I haven't noticed its history [15:29] sil2100, only on intel this last run for sure === MacSlow|lunch is now known as MacSlow [15:30] mterry: for instance build 96 has a failure in test_menu_closes_on_click_outside because of the same bug, since the menu didn't get exported [15:30] And it's also intel only [15:31] sil2100, maybe a race condition and only the intel machine is slow/fast enough to hit it occasionally? [15:31] sil2100, or uninitialized var [15:32] mterry: maybe? I'll be looking for that, looks like another uninitialized var, but since it's in the panel it might be a bit more to find it [15:32] mterry: andyrock said it also happens for him too [15:32] andyrock: what machine do you have ;) ? [15:33] sil2100, ivy bridge i7 [15:33] sil2100, I have an intel too, but haven't run those tests to reproduce yet (in middle of something else, can swing around in a bit) [15:34] andyrock: makes sense - thanks! [15:34] mterry: thanks :) No hurry, I'll browse through the code in the meantime [15:35] sil2100, what ap test is failing? [15:35] are you sure that unity just fails to display the menus [15:36] i mean the menus are correctly exported [15:36] andyrock: tests from the test_menu_closes_on_click_* suite [15:36] unity just fails to draw them [15:37] brb (coffee time) [15:37] andyrock: hm, well, maybe? But AP polls the panel emulator for menu entries with no luck [15:37] andyrock: ok === mmrazik is now known as mmrazik|afk [15:57] sil2100, so self.panel.menus.get_entries() fails right? [15:59] andyrock: yes [15:59] andyrock: no results returned, and the videos show no menu on the panel [16:17] sil2100, andyrock : I can get some errors in mouse_open_indicator() calls because the previous window's menus are on the panel instead of Calculator's [16:17] (on my local intel machine) [16:18] 3 instances in one test run of unity.tests.test_panel.PanelIndicatorEntryTests [16:18] hm? That's strange indeed [16:19] mterry: you have menus but wrong ones yes? This might be the same problem even, hmm [16:19] mterry: how reproducible is it? [16:20] sil2100, rather. 3, 1, and 2 failures when running that suite of 4 tests [16:20] sil2100, I get a character printed to the console during the test [16:21] sil2100, ^[ during the test, and [A or [B on the prompt after.. [16:21] which is weird, because joined, isn't that arrow keys? [16:21] hm, heh [16:21] Yes, maybe it's for workspace switching [16:21] mterry: you have one workspace, right? [16:21] sil2100, I have workspaces off [16:22] So maybe the key combinations have no effect and now compiz simply doesn't fetch the keys [16:22] And they get through into the terminal [16:22] sil2100, but to your earlier question, yes. menus, but the wrong ones [16:22] Sp it shouldn't be a problem [16:24] sil2100, hmm... I'm not running staging. So I might be a bit behind trunk. Let me update and rerun these tests [16:28] I'll help you fix importants bugs next week and write tests [16:28] :) [16:31] sil2100, something strange there are two other tests in that suite which go through open_app_and_get_menu_entry and pass [16:32] fginther: yes, since it's broken only sometimes [16:32] It's not completely reproducible [16:33] Just it seems sometimes when an application is opened, the menu *might* be broken [16:37] maybe open_app_and_get_menu_entry needs to assert that the calculator has focus before trying to grab the menus [16:37] sil2100, ^ [16:38] hmm, maybe not, the refresh_fn 'should' take care of that [16:38] sil2100, harder to reproduce when I'm running staging, but still possible. 1 failure in 5 test runs [16:39] sil2100, should not be related but on trunk panel menus don't respect Fitt's law [16:39] fginther: but it does have focus, since the videos show that it does [16:39] not always [16:40] Now this is interesting... test_menu_opens_on_click has shown me two failures: one with the previous menu, and one with the Calculator menu that never opened [16:40] Oh! [16:40] When it never opened, did you see the calculator menu appear? [16:41] sil2100, define appear. It showed on the panel. The mouse was hovering over it. But it never opened [16:41] mterry: ah, the mouse was hovering over it? hmmm [16:43] I'll try a bit to get it reproduced here, wait one moment [16:44] sil2100, export i=0; while autopilot run unity.tests.test_panel.PanelIndicatorEntryTests.test_menu_opens_on_click; do export i=$((i + 1)); echo $i; done [16:44] sil2100, that seems to get it within 10 [16:44] sil2100, usually the previous-menu failure [16:46] Strange thing [16:50] mterry: ran it 21 times and could not reproduce - what machine are you running? [16:52] mterry: anyway, I think this needs creating a bug about it anyway, I'll fill it in anyway if you don't mind [16:52] sil2100, a dell sputnik. amd64, intel core i7 [16:52] sil2100, sure [16:52] Since I didn't see an existing bug for this strangeness [16:52] sil2100, I'm happy to test solutions [17:08] fginther, looking at the clientlogs/dx-autopilot-nvidia_master.run_2013-01-31_03-01-10.yaml file for the failed nvidia run, I see that apport-retrace fails with a None options.cache [17:08] fginther, looks like that is keeping us from having a retraced crash file [17:10] https://bugs.launchpad.net/unity/+bug/1111623 [17:10] Launchpad bug 1111623 in Unity "Rarely application menus not exported to the panel correctly" [Undecided,New] [17:28] hello, i had a question like this for ages - when I connect my external display to my laptop I need to go to Appearance settings for Gnome/Unity to realize the display is connected. Is that supposed to be so? [17:29] veebers, ^ on the apport-retrace failure [17:54] how to markup the panel clock applet and decrease the font size in ubuntu 12.10 [17:55] mterry, thanks. I'll have a look [17:56] how to markup the panel clock applet and decrease the font size in ubuntu 12.10 [17:57] how to markup the panel clock applet and decrease the font size in ubuntu 12.10 like this http://www.omgubuntu.co.uk/wp-content/uploads/2010/10/Top-Expanded-Edge-Panel_015-copy.png [17:59] fginther: you think it makes sense to re-run the indicator tests? [17:59] anyone???? [18:00] ignerous: hi, I think there was a tool for that, but I can't really help right now [18:00] fginther: I don't want to spam the indicator test results [18:01] anyone???? === salem_ is now known as _salem [18:02] ignerous: if anyone present would know how to help, he would answer [18:02] ignerous: try #ubuntu for Ubuntu support [18:02] ignerous: current implementation does not allow markup [18:03] ignerous: you can set a custom format but not change the way the indicator is rendered [18:04] :( [18:05] i tried #ubuntu but the recommend this channel for help [18:06] ignerous: the format is set as in gsettings set com.canonical.indicator.datetime custom-time-format '%Y-%m-%d' [18:07] ignerous: and to switch to custom format - gsettings set com.canonical.indicator.datetime time-format 'custom' (by default it is locale-default) [18:07] okay [18:09] ohhh === mmrazik|afk is now known as mmrazik [19:05] Hi hi. I see my preferences clobbered back to default every so often. Key combinations are particularly bad to have to set up again. Should I be looking for a storage problem, or are there any circumstances under which unity thinks it's okay to discard preferences? [19:37] qengho: all preferences or some? === AlanChicken is now known as alanbell === alanbell is now known as AlanBell === _salem is now known as salem_ [19:49] JanC: The ones I really care about are key combinations and number of desktops. [19:49] JanC: if others change, i do not notice. [19:49] qengho: there is/was a bug about that [19:49] e.g. keys to switch to particular desktops [19:51] you might want to search LP [19:51] mterry: interesting, that appears to be a different issue that I had originally: https://bugs.launchpad.net/apport/+bug/1103709 [19:51] Launchpad bug 1103709 in Apport "apport-retrace gives an exception retrace when processing a crash file" [Undecided,New] [19:54] veebers, no, looks the same to me [19:55] mterry: oh? Ah yes I see now [19:55] * veebers thinks he needs a coffee :-P [19:55] veebers, do we need to pass -C? [19:56] * veebers checks the manpages [19:57] mterry: I hadn't need to before [19:57] (i.e. it was working in the past) [19:59] veebers, looking at the code, it seems that --sandbox and --cache go hand in hand [20:00] veebers, you either have both or neither. There's a nice check for cache without sandbox, but they forgot the other way [20:18] veebers, well, it seems like we could easily work around this for the moment. I'd like to be able to use those logs to decide how bad this nvidia thing is [20:24] mterry: sorry I was in a meeting [20:24] mterry: oh ok, I'll add that and fire off another run [21:34] bregma, a recent nvidia unity-autopilot run had a lot of new failures: http://10.97.0.1:8080/job/ps-unity-autopilot-release-testing/label=autopilot-nvidia/lastCompletedBuild/testReport/ [21:34] * bregma looks [21:34] bregma, errors like: [21:34] Traceback (most recent call last): [21:34] File "/usr/lib/python2.7/dist-packages/unity/tests/test_dash.py", line 1106, in test_right_click_on_preview_ratings_widget_cancel_preview [21:34] ratings_widget = self.preview_container.current_preview.ratings_widget[0] [21:34] AttributeError: 'NoneType' object has no attribute 'ratings_widget' [21:41] hmm, I see no recent changes in unity that should have that effect [21:41] most of the falures seem related to DBus somehow === salem_ is now known as _salem