[08:48] tsdgeos: oh, is there a 'floor' problem in qmluitests too? https://objectstorage.prodstack4-5.canonical.com/v1/AUTH_77e2ada1e7a84929a74ba3b87153c0ac/autopkgtest-yakkety/yakkety/amd64/u/unity8/20160727_024846@/log.gz [08:48] or is that test running some funky version of unity8, that's also possible [08:49] no, that was fixed, so yes unfortunately autopkgtests do the wrong thing [08:49] Mirv: seems very much what i fixed, no? [08:49] is it compiling some old version or something? [08:50] tsdgeos: yes, it's compiling old version [08:50] tsdgeos: then elsewhere with new version, two failing tests: https://objectstorage.prodstack4-5.canonical.com/v1/AUTH_77e2ada1e7a84929a74ba3b87153c0ac/autopkgtest-yakkety/yakkety/amd64/u/unity8/20160727_034401@/log.gz [08:51] that could be real Qt 5.6.1 breakage [08:52] FAIL! : qmltestrunner::PhoneStage::test_selectAppFromSpread(App 5) Uncaught exception: Type error [08:52] that looks bad-ish [09:06] Mirv: if you file a bug maybe i can have a look later today/tomorrow if noone beats me to it? [09:25] tsdgeos: bug #1606835 [09:25] bug 1606835 in unity8 (Ubuntu) "qmluitests: test_quicklist_positioning fails with Qt 5.6.1" [Undecided,New] https://launchpad.net/bugs/1606835 [09:25] oh right there were others [09:27] updated === dpm is now known as dpm-lunch [11:48] mzanetti, Hello [11:48] om26er, hi [11:48] mzanetti, mind taking a look at https://code.launchpad.net/~canonical-platform-qa/unity8/add_object_name_for_setup_wizard/+merge/300380 [11:49] we need to land that for some testing needs in ubuntu-system-tests, it just adds object names [11:52] Saviq: is dandrada around or vac? [11:52] Saviq: or can you help with validating silo https://requests.ci-train.ubuntu.com/#/ticket/1715 (the relative mouse events part) [11:52] Mirv, he'll be around soon (timezone) [11:53] yeah I was thinking timezone but then thinking Brazil should be up already :) [11:54] Mirv, http://www.timeanddate.com/worldclock/brazil/joinville almost ;) [12:38] dandrader: hey! could you test the relative mouse events part of https://requests.ci-train.ubuntu.com/#/ticket/1715 on vivid and report on the ticket? [12:39] dandrader: and if you have ideas for the test plan feel free to improve [12:39] Mirv, ok [12:40] Mirv, from qt's perspective, it's about hover events carrying timestamps [12:41] dandrader: yeah, the QA would just like to know what to look for and regression potential === dpm-lunch is now known as dpm [13:27] Mirv: something is weird with those type error failures [13:27] it fails when doing [13:27] tile.mapToItem(spreadView) [13:27] with tile being TransformedSpreadDelegate_QMLTYPE_40_QML_54(0x560f74f440a0, "spreadDelegate_32") [13:27] and spreadView QQuickFlickable_QML_50(0x560f721a7de0, "spreadView") [13:27] it "should work" [13:35] * Saviq wonders if Popescu uses yakkety proposed, he mentioned this error in a bug [13:36] in bug #1606501 [13:36] bug 1606501 in unity8 (Ubuntu) "initial app surface size wrong" [High,Confirmed] https://launchpad.net/bugs/1606501 [13:39] mterry, FYI: https://code.launchpad.net/~saviq/ubuntu-system-settings/no-test-powerpc/+merge/301271 but also bug #1606501 [13:39] bug 1606501 in unity8 (Ubuntu) "initial app surface size wrong" [High,Confirmed] https://launchpad.net/bugs/1606501 [13:40] noo [13:40] bug #1606835 [13:40] bug 1606835 in unity8 (Ubuntu) "qmluitests fails with Qt 5.6.1" [Undecided,Confirmed] https://launchpad.net/bugs/1606835 [13:40] so even when we "fix" uss/powerpc, we'll be fooked [13:41] and mapToItem failing isn't even funny [13:41] Saviq, this seems like a case where we back out 5.6 and let them fix their mess without blocking all of ubuntu [13:41] mterry, tell that to Mirv... [13:43] qmluitests was supposed to be fine for 5.6 :( [13:43] Mirv, we in unity8/ubuntu-system-settings/maliit-framework are hitting some real test problems with qt5.6, is that widespread or isolated to these three? [13:43] anyway, Kubuntu are already doing their uploads to yakkety since they've been holding off their cycle work until 5.6 is in [13:44] mterry: it seems isolated to those three, although likely Qt on powerpc is not perfect (and never supported by upstream in the first place) [13:44] mterry: Debian doesn't have packages they'd have problems with it since they've carried Qt 5.6 since June [13:44] Mirv, qmluitests isn't a powerpc issue, but sure, the other two are [13:44] Mirv, got it [13:45] right, the build phase only knows of unit tests [13:45] Mirv, but presumably we've also ran qt5.6 :) [13:45] and not come up with problems? [13:45] mterry: yeah, I've run it on yakkety Unity 8 + Plasma + LxQT desktop and on xenial phone. [13:46] Mirv, OK. Seems to be just a lot of problems in our immediate space. We'll work around it [13:46] Saviq, I do like the amd64 re-enabling in that MP... [13:46] arm64 rather [13:46] Keep getting those confused [13:47] mterry: yes, I wouldn't care about powerpc much (unliked s390x which got fixed before upload), but I would have blocked on qmluitests had I known there is trouble - albert checked it in May/June and it was fine back then but this is possibly due to changes in bug fix release 5.6.1 [13:48] mterry, yeah let's see how that works, I had a look at the bug and it said fixed, so... [13:49] Saviq, I see you added to silo, cool [13:49] except that won't help with britney on yakkety :/ [13:53] inevitable Qt upgrades will happen in development Ubuntu versions from now on too, (Qt 5.7 in yakkety+1) unless we start to make noise about wanting to stay in the upstream's LTS (5.6). other Qt users however expect 5.7, 5.8... [13:57] Mirv: nice that you don't see https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1582144 anymore :) [13:57] Launchpad bug 1582144 in unity8 (Ubuntu) "Unity 8 scopes possible to get vertically squeezed with Qt 5.6" [High,Incomplete] [13:58] Mirv, yeah I expect Qt upgrades, just seemed like the problems were more widespread than they were (lots clustered in our stuff) [13:59] Saviq, yeah :-/ [14:00] mterry: yeah it'd help if we would have time (other than me) to test new releases, I've been filing bugs and offering PPA since March [14:00] and not that I'd have too much time in general either [14:06] Mirv: we will get Qt 5.6 only in yakkety right? not on vivid overlay [14:07] tsdgeos: in yakkety, and xenial overlay as soon as everyone agrees. we should be switching to xenial soon anyway. [14:07] SOON (TM) [14:07] :d [14:07] Half Life 3 confirmed! [14:07] tsdgeos: or at least I've understood 5.6 is wanted by "everyone" [14:07] SCNR [14:08] tsdgeos: and it's upstream LTS so it makes sense to combine with our LTS [14:08] total sense [14:08] tsdgeos: I'd mostly like zsombor back and look at UITK + Qt 5.6. but that's a couple of weeks. and obviously Unity 8 will need these qmluitests fixed. [14:09] on it [14:16] Mirv, added a comment to https://requests.ci-train.ubuntu.com/#/ticket/1715 . Thanks! === JanC is now known as Guest24709 === JanC_ is now known as JanC [14:23] ok, first problem found [14:23] seems mapToItem now requires the 3 arguments [14:23] not only item [14:23] luckily we were only [ab]using it in 1 place [14:24] mzanetti: Saviq: ↑↑↑ [14:24] ah [14:24] it's just that the guy that added the check may have added a warning [14:25] would have been much easier to undersatnd [14:25] what issue is this? [14:25] tests failing in Qt 5.6.1 [14:25] ah [14:25] and for good reason [14:26] weird thing is, I've fixed for sure like 5 occurances of that mapToItem argument in the last year [14:26] (after making that very same mistake quite ofter before that time) [14:26] tsdgeos: great [14:26] whenever I debug thigns it seems to totally not work, but still strange that things don't break more because of wrong usage of this [14:27] Mirv: now i'll tackle the other, but that other one may "just" be that we have unstable tests [14:27] tsdgeos: the other did happen on both amd64 and i386 so at least fairly often [14:45] mzanetti: will you review https://code.launchpad.net/~aacid/unity8/mapToItem3Args/+merge/301283 ? [14:46] tsdgeos, ack [14:47] tsdgeos, in order to test it I'd require Qt 5.6, no? [14:47] kind of D: [14:47] or you can just say "ok it works now" [14:47] i mean [14:47] "ok it still works" [14:47] yeah, I know what you mean [14:47] and read the code of mapToItem [14:47] where it checks for args to be 3 or 5 [14:49] mzanetti: actually give me a minute, there's a wrong mapFromItem call too that i'll add to the branch [14:51] mzanetti: done [14:51] ack [14:52] ok. now I need to actually test it on my device [14:53] he he [14:53] sorry [14:53] Mirv: so https://code.launchpad.net/~aacid/unity8/mapToItem3Args/+merge/301283 should fix the two test failures, is it easy for you to add to that auto testing or we need to land it? [14:54] tsdgeos: it is easy for me to push to yakkety and update your trunk manually, if that's what you're asking [14:54] tsdgeos: similar to what I did when landing the floor include fix [14:55] yes/no i mean if you can test it without our trunk being changed [14:55] tsdgeos: no, I can't test it easily. and I guess the change should be #ifdef for < 5.6 ? [14:55] Mirv: no [14:55] mterry, hey ho. running your branch still and rebooted for the first time. seems you broke the SIM PIN entry [14:56] mzanetti, ah crud. I thought I kept Lockscreen compatibility [14:56] mzanetti, will look into that, thanks [14:56] mterry, I looked at the diff, didn't seem you removed it [14:56] mterry, but maybe some wiring died === dandrader is now known as dandrader|afk [15:58] mterry, grrrrrr https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu/landing-077/+sourcepub/6755846/+listing-archive-extra [15:59] wtf, I tried it on a porter box and it does the right hting [15:59] Saviq, huh... [15:59] Saviq, and your change was small -- the support already seemed there? (unless we've been running tests in arm64 unbeknownst to us?) [15:59] which wouldn't surprise me, if we've been working there for a while [16:00] mterry, easy to check [16:00] hah it was testing on arm64 https://launchpadlibrarian.net/274491605/buildlog_ubuntu-yakkety-arm64.ubuntu-system-settings_0.4+16.10.20160722.1-0ubuntu1_BUILDING.txt.gz [16:01] *but* I tried on a porter box and it did the right thing! [16:01] the previous version didn't test, though https://launchpadlibrarian.net/271833732/buildlog_ubuntu-yakkety-arm64.ubuntu-system-settings_0.4+16.10.20160708-0ubuntu1_BUILDING.txt.gz [16:03] wth [16:12] :-/ [16:12] Saviq, check phases of the moon === dandrader|afk is now known as dandrader [16:21] mterry, I added some debug prints to debian/rules, let's see... [16:51] ChrisTownsend, Hi! [16:51] om26er: Hey [16:51] ChrisTownsend, whats the recommended way to install libertine on desktop ? and which tech should be used container or chroot ? [16:52] om26er: apt-get install libertine [16:52] om26er: And LXC for sure [16:52] om26er, trust the tools to choose the right default [16:52] om26er: Yes, also what bregma says. Don't specify the container type and let it choose it for you. [16:55] ChrisTownsend, what would an example to create a container look like ? [16:56] om26er: If using the cli, 'libertine-container-manager create -i test -n "Test"' Of course, you can call -i and -n whatever you want. [16:56] om26er: If you use the UI, well, it's just clicking the right things:) [16:57] om26er: Also, the manpage and help is your friend:wink: [16:58] ChrisTownsend, ok, seems to have failed for me :/ http://paste.ubuntu.com/21161321/ [16:59] om26er: Hrmm, lxc issue. This is the pertinent line: [16:59] lxc 20160727215736.439 ERROR lxc_utils - utils.c:safe_mount:1692 - Invalid argument - Failed to mount /sys/kernel/debug onto /usr/lib/x86_64-linux-gnu/lxc/sys/kernel/debug [17:00] om26er: Are you using yakkety or xenial? [17:00] we need lxc guys in this channel [17:00] ChrisTownsend, that's yakkety [17:01] om26er: They hang out in #lxcontainers [17:01] Need ubuntu specific channel as well :) [17:02] ChrisTownsend, shall I just try in a VM and see how it works out of the box in 16.04.1 ? [17:04] om26er: It's working for me on my Yakkety system. [17:05] om26er: I guess you can try in a VM. I've never tried it, so I'm not sure what your experience will be like. === dandrader is now known as dandrader|afk === dandrader|afk is now known as dandrader === dandrader is now known as dandrader|afk === dandrader|afk is now known as dandrader [20:58] mzanetti, you said sim unlock screen didn't work for you? it seems to be working for me using phonesim with silo 23 (which granted has more than just greeter-no-lockscreen in it... maybe that branch broke something that another branch fixed...)