=== duflu_ is now known as duflu === zequence_ is now known as zequence [08:18] tsdgeos, o/ [08:18] Saviq: \o [08:40] tsdgeos: I found a note about dropping 'Don-t-use-ebx-in-the-x86-register-allocator.patch'.. I'd be replacing it with 'V4-JIT-save-restore-ebx-GOT-ptr-on-linux-x86.patch' + 'V4-fix-range-sorting.patch' + 'V4-regalloc-fix-interval-splitting-when-register-pre.patch', does that sound sane? [08:41] Mirv: looks correct yes [08:44] thank you [09:09] mzanetti, top-ack https://code.launchpad.net/~unity-team/unity8/new-scopes-clean-to-trunk/+merge/209459 ? [09:09] mzanetti, I'd like you to take another stab at testing landing-013 [09:10] didrocks, could we have packaging review of ↑↑? [09:10] Saviq: yeah, but not now, like in an hour or so :) [09:10] didrocks, sure, thanks [09:10] * didrocks sees everything fallind down [09:10] falling* [09:11] didrocks, yeah, I saw Larry's and Paul's emails... [09:11] Saviq: and it's only part of the story :p [09:11] didrocks, I can imagine [09:19] Saviq: on the plus side I've locally run all AP:s for #256 and it seems like everything is still passing with exception of gallery-app failing 1 test for me, so it's not like disaster would have happened after #250 [09:19] Mirv, cool [09:19] I just hope my local run is not giving more positive results than what the automated tester would give [10:24] dednick, ping [10:24] Cimi: yo [10:24] Saviq, so how do we look on landing scopes? [10:24] dednick, ciao [10:25] dednick, I'm reusing the unity menu model stack for welcome wizard [10:25] and I arrived at this menuitemfactory [10:25] mhr3, want to, today [10:26] mhr3, we're fixing a flaky ap test, found one small UX issue [10:26] dednick, is there a way to get only the list of the access points? [10:26] Saviq, very well, i'll land some extra raciness fixes meanwhile [10:27] Cimi: hm. not sure other than maybe using a filter model. [10:29] dednick, what's a filter model? [10:29] dednick, something that doesn't pass items in the model? [10:30] Cimi: i mean a qt filter model. like when you filter out items in a model based on a property [10:31] ok [10:31] Cimi: It's possible there should be a different indicator profile for the welcome wizard. [10:31] dednick, can you help me on that? [10:31] dednick, or guide me [10:31] dednick, where can I create this profile? [10:32] Cimi: the profile is in the indicator itself in which case tedg would be your man to ask (not sure if it's the thing to do really). [10:33] Cimi: otherwise you should probably be using the "phone_wifi_settings" profile [10:34] I'm using that for now [10:49] Saviq: answered on the MR [10:49] didrocks, /me too [10:50] didrocks, only LP didn't refresh its INBOX yet... [10:51] ah :p [10:51] didrocks, now [10:51] didrocks, anyway, new scopes have nothing to do with libunity any more, and there's not a desktop version of those dependencies [10:51] Saviq: shouldn't they be recommends only though? [10:52] Saviq: and we seed the new scopes [10:52] so that, on desktop, people are not forced to installed onlinemusic for instance [10:52] and we give more granularity [10:52] didrocks, they are [10:53] hum? [10:53] 39-Recommends: ${unity-default-masterscopes}, [10:53] hmm [10:53] 40+ unity-scope-scopes, [10:53] so Recommends [10:53] no* [10:53] didrocks, you're right [10:53] \o/ [10:53] Saviq: just reminds me to check they are all seeded when this branch lands (once you downgraded them as recommends) [10:54] didrocks, so what do you think, click and scopes depends, but mediascanner2 and onlinemusic recommends? [10:54] Saviq: why click? I would say scope-scopes depends only [10:54] didrocks, yeah, I pung ogra this morning, didn't respond yet, though [10:54] Saviq: I can do the seeding if needed [10:54] mhr3, opinion ↑↑↑? [10:55] didrocks, ok [10:55] scope-scopes is the scope to enable/disable, right? [10:55] didrocks, ultimately, yes [10:55] so if you want to depends on anything, that one, the rest is surely just recommends [10:55] and we seed them [10:55] ok [10:55] Cimi: dont think adding a new profile is the way to go on this one. maybe just use the settings profile. Think it's a bit of overkill just for something we're really only displaying first time you use the phone. [10:56] mhr3, only -scope-scopes as Depends, rest as Recommends? [10:56] Cimi: so I'd just go for filter model if you can. [10:56] dednick, yes but it contains other menuitems, no? [10:56] damn, i broke something in my machine [10:56] Saviq, no, click is deps too [10:56] "file:///home/tsdgeos_work/phablet/unity8/new-scopes-clean-to-trunk/builddir/plugins/typelist.qml:2:1: plugin cannot be loaded for module "AccountsService": The file '/home/tsdgeos_work/phablet/unity8/new-scopes-clean-to-trunk/builddir/plugins/AccountsService/libAccountsService-qml.so' is not a valid Qt plugin." [10:56] Saviq, we're all the time saying that scopes and apps are two required components [10:56] mhr3, right [10:56] didrocks, ↑ [10:57] does anyone get this? [10:57] Cimi: ./builddir/src/Panel/Indicators/client/indicators-client -profile phone_wifi_settings [10:57] tsdgeos, no [10:57] tsdgeos, ./build -c ? [10:57] Saviq: i nuked builddir, same thing [10:57] Cimi: open up the "indicator-menu" section and deselect visual representation [10:58] mhr3: Saviq: is click available on the desktop? I think it's not [10:58] mhr3: Saviq: and we have that preview session where we don't want to promote click on the desktop for now [10:58] didrocks, sure it is, with the unity8 preview session [10:58] didrocks, uhm hum [10:59] why not, the preview session does mir [10:59] Cimi: you should be able to set up a filter for type "unity.widgets.systemsettings.tablet.accesspoint" [10:59] tsdgeos, that on new scopes? [10:59] mhr3: there is still the xorg one [11:00] mhr3: will it crash without the click scope? [11:00] Saviq: yes [11:00] didrocks, of course it won't, but will you be able to launch anything? no [11:00] Cimi: using a Utils.SortFilterProxyModel [11:01] mhr3: so gallery depends on click scope? [11:01] and so on? [11:01] like unity-scope-mediascanner2 depends on unity-scope-click? [11:01] didrocks, click surfaces apps [11:01] right [11:02] but as of today unity-lens-applications is a recommends from unity7 [11:02] as every other scopes [11:02] any reason to not be the same for click scopes? [11:05] didrocks, wasn't it recommends until now? [11:05] didrocks, you can't really launch apps other than through the click scope on unity8 (except for those on the launcher) [11:05] mhr3: yeah, it was [11:05] mhr3: that's why I wonder why we are going to change this to depends [11:05] didrocks, then why did you ask to change it in the first place? :) [11:06] mhr3: did I? I didn't, I'm arguing to keep it as recommends [11:07] well, pks wise it's really an optional dep, yes unity is useless without it, but you can write your own apps scope to make it useful [11:07] s/pks/pkg/ [11:07] yeah, so keeping it as recommends + seeding it makes sense to me [11:07] so i'm fine with keeping everything recommends + seeding [11:07] ok, thanks mhr3 :) [11:07] Saviq: ^ [11:07] Cimi: something like this: http://pastebin.ubuntu.com/7145753/ [11:08] dednick, trying [11:08] didrocks, mhr3, ack [11:08] Cimi: except with the sound replaced with network [11:08] didrocks, mhr3, so depends on scopes, recommends+seed everything else? [11:09] Saviq, recommends on everything, no? [11:09] mhr3: scopes-scope isn't mandatory? [11:09] didrocks, no [11:09] same way apps is [11:09] ok, all recommends as it was [11:09] didrocks, not code-wise [11:10] unity is useless, but it works [11:10] and we'll seed everything [11:10] mhr3: come on, there is the launcher! don't bring back everything to scopes :) [11:10] so then, we can land the new scope infra? [11:10] as thostr_ told on the ML that we were only blocking on landing? [11:11] * greyback hates Mondays, accidentally reverted code written on Friday [11:11] didrocks, launcher doesn't list everything :P [11:11] Saviq: ok, it seems i had something weird exported in my shell, i switched to a new terminal and now it works :S [11:12] didrocks, Saviq is finishing a fix for some flaky tests, should be able to land it in a bit [11:12] tsdgeos, :) [11:12] didrocks, pushed the change [11:13] greyback, no .~1~ file? [11:13] Saviq: *mwah* totally forgot those [11:14] * Saviq wipes [11:14] * Saviq goes for a shower [11:14] * Saviq goes to an auto-wash [11:27] mzanetti, top-ack https://code.launchpad.net/~unity-team/unity8/new-scopes-clean-to-trunk/+merge/209459 ? [11:27] mzanetti, I'd like you to take another stab at testing landing-013 [11:27] mzanetti, actually, wait a bit, will rebuild with the latest changes [11:27] Saviq: ok [11:28] didrocks, can you upload the updated seed to landing-013? [11:28] Saviq: this is now everything merged together? [11:28] didrocks, we need unity-scope-{scopes,click,mediascanner2,onlinemusic}, the rest can go away [11:28] mzanetti, yes [11:28] Saviq: sureeee [11:28] Saviq: uno momento [11:28] didrocks, thanks [11:29] Saviq: mhr3: dropping libunity-scopes-json-def-phone, right? [11:30] in addition to the rest of scope [11:30] didrocks, ack [11:30] didrocks, not sure why it was there in the first place [11:30] unity-scope-video-remote is handled by mediascanner2? [11:30] yes [11:30] waow, all was spread out all over the seeds [11:30] let me clean this [11:30] didrocks, well, in that case I should probably drop that, and the perl thingy to parse it? [11:30] didrocks, yeah [11:30] didrocks, from unity8's debian/control? [11:30] Saviq: yeah [11:31] * Saviq does [11:31] didrocks, also -scope-onlinemusic should go byebye [11:31] Saviq: snif, my perllllll := [11:31] :) [11:31] ;) [11:31] it's a unicorn! [11:31] heh, I'll blame mhr3 for this Friday evening hack :p [11:31] mhr3: and -home [11:31] didrocks, and you loved every bit of that friday evening :P [11:32] mhr3: soooo much [11:33] didrocks, pushed [11:33] didrocks, do we have a process for landing a thing risky enough like the new scopes? [11:33] seed updated [11:33] like additional QA? [11:33] Saviq: yeah, it's just asking for some manual dogfooding from our team [11:33] didrocks, ok [11:34] in addition to all AP tests that you run + all tests that you are doing as usual of course [11:34] ok, generating the metapackage [11:34] didrocks, you're dputting the seed to landing-013? [11:34] didrocks, ah ok [11:34] Saviq: yeah, once the package is generated [11:34] didrocks, yeah, of course (re: testing) [11:34] (it's a 2 step process) [11:34] didrocks, right, remember now [11:35] * didrocks just relooked at his perl script and cries on all the $ [11:39] Saviq: mhr3: looking good to you? http://paste.ubuntu.com/7145879/ -click and -onlinemusic were already in the seed, I just moved them) [11:40] didrocks, +1 [11:40] * didrocks dput to landing-013 [11:40] will need a reconfigure then [11:41] didrocks, will I be able to, or do you need to? [11:41] Saviq: you can't, as it's about adding on more component [11:41] didrocks, yup, what I thought [11:42] hum, it's building [11:42] Saviq: should I just stop the monitoring and relaunch it? [11:42] didrocks, sure, go for it [11:45] Saviq: uploaded, configure and restarted with watch only [11:45] didrocks, yup, thanks [11:45] yw! [11:45] oh, miss the reject [11:45] let me see why [11:45] (from my dpugt) [11:47] didrocks, no onlinemusic [11:47] didrocks, it's handled by mediascanner2, same as onlinevideos [11:48] 12:28:33 Saviq | didrocks, we need unity-scope-{scopes,click,mediascanner2,onlinemusic}, the rest can go away [11:48] hum? [11:48] ok, I'll do a second upload then :) [11:48] oups [11:48] Saviq: you're mean! :) [11:48] didrocks, no, we still have it in unity8, too [11:48] didrocks, sorry [11:48] no worry ;) [11:49] ok, refreshing with only [11:49] pushed [11:49] * unity-scope-scopes [11:49] * unity-scope-click [11:49] * unity-scope-mediascanner2 [11:52] mhr3: do a search do the online music and videos show up then? [11:54] * Removed unity-scope-onlinemusic from touch [11:56] davmor2, yes, just provided by different pkgs [11:56] didrocks: does this mean I have to keep my eyes out for a landing of the new scopes? [11:56] davmor2: yeah, this is going to happen very soon :) [11:56] Saviq: I see QT devs have replied to your bug :( [11:56] davmor2, yeah, tvoss did, too [11:56] davmor2, but yeah, I was expecting that [11:57] davmor2, they have a point, and we need to work out a compromise of sorts [11:57] Saviq: Meh! [11:58] someone needs to budge ;) [11:58] Saviq: I blame it all on popey for finding the bug [11:58] davmor2, :) [11:59] * popey loads is gun and looks for messengers to shoot [11:59] *his [11:59] * davmor2 takes popey 's gun off him before he shots himself ;) [12:04] Saviq: seems it's still picking up 116, waiting a publisher cycle to get 117 [12:04] didrocks, k [12:14] Saviq: all build, tell us if that's enough once the testing is done (or if you need to rebuild and so on) [12:14] Saviq: but all on you now :) [12:14] didrocks, yup [12:14] mzanetti, ↑ can you do a round of testing of landing-013? [12:15] Saviq: I didn't rebuild unity8 though, not sure if your latest packaging changes are there [12:15] Saviq: sure. but I just set up my phone for autopilot testing the right edge stuff before you pinged me before. can I finish that first or is this so high priority that I need to get started immediately? [12:16] didrocks, they don't really matter for dogfooding, but yeah, /me does [12:16] mzanetti, sure, it's fine === MacSlow is now known as MacSlow|lunch [12:21] Saviq: have you noticed a crash after applying silo 13? [12:22] thostr_, no, any more data? [12:22] Saviq: let me check if apport has written anything [12:29] Saviq: crash file is there, not sure how helpful though [12:29] thostr_, please do /usr/share/apport/whoopsie-upload-all [12:30] thostr_, and after that's completed, send the .crash somewhere [12:30] will do [12:44] 140 [12:44] bah === _salem is now known as salem_ [12:47] Saviq, mhr3: guys, what about the transition to the new scopes? How's it proceeding? [12:47] sil2100, almost there, tweaking the last flaky test and we'll be in for dogfooding soon [12:48] Saviq: awesome ;) [12:49] sil2100, silo for 31 for a prereq for that pls [12:49] sil2100, 002 still not freed :P [12:51] mhr3: pfff ;p [12:52] Nitpicks ;p [12:52] Saviq is faster than I at the unassign war :) [12:53] ;) [12:53] hey all: when I launch a lot (but not all) applications in the Unity8 desktop preview session they abort during startup with a failure to create an instance, and Unity8 logs the message 'ApplicationManager REJECTED connection from app with pid 8208 as no desktop_file_hint specified' .. anyone have any idea what might be going on with that? [12:53] these are core apps, there are desktop files [12:53] bregma, sounds like unity8 and upstart have issues communicating that the apps are being launched [12:54] greyback, can you support bregma please ↑? [12:54] ok [12:54] things like Sudoku and browser-app all work OK [12:54] bregma: hey, how are you launching these apps? Does it go through upstart-app-launch? [12:54] /food [12:55] greyback, I would imagine, this is stock Unity8 (although, as usual, it may need further tweaking for the desktop) [12:56] damn you, Dropping Letters, I only want to see if you launch [12:56] bregma: can I install the PPA to try it? [12:57] greyback, https://launchpad.net/~unity8-desktop-session-team/+archive/custom [12:57] bregma: usually that message due to an app starting up which shell wasn't expecting. upstart-app-launch should always notify shell that the app is launching. So either u-a-l or unity-mir is broken [12:58] bregma: ok, installing now [12:58] you need the qtubuntu from the PPA for a workaround in a Qt5.2, the rest of the things in there are for (a) mouose support and (2) session logount support [12:59] greyback, if you could take a look at qtubuntu bug #1295309 that would also be swell [13:00] bug 1295309 in qtubuntu "qtubuntu-desktop fails using Qt 5.2 on Mir" [Critical,In progress] https://launchpad.net/bugs/1295309 [13:02] bregma: sure, will review [13:02] heh "feature" [13:05] mhr3: silo for your prereq [13:05] sil2100, ty === alan_g is now known as alan_g|lunch === MacSlow|lunch is now known as MacSlow [13:22] dednick, can I use that plugin from the ubuntu system settings? [13:23] Cimi: hm. not the Utils one. [13:23] dednick, so probably I have to copy/paste [13:24] Cimi: yup. unless qt 5.2 has offered a qml filter model now. [13:34] Saviq, finally noticed that we no longer (since January!) have the "Approved by Foo" line added to commit messages. Do you know why? And, secondly, should we try to get it back somehow? [13:35] I find them useful as the reviewer/approver shares part of the responsibility over a commit (one more guy you can poke about it) [13:37] bregma: how does one log out of a unity8 session? [13:37] didrocks, ↑ [13:38] please file a feature req. Not sure when I can safely add that though [13:38] greyback, if you have everything from the PPA installed, pull down one of the indicators and switch to indicator-session (it doesn't have an icon in the top bar for some reason) [13:38] then click on logout [13:40] bregma: on unity8-x11, mouse gestures aren't working (I can't unlock the greeter with my mouse). We've a "-mousetouch" switch for unity8 that should fix that. [13:42] greyback, yeah, that package is pretty much unsupported at the moment and should probably be removed [13:42] bregma: ah ok [13:42] it needs to have a window manager added for most featuures to work anyway [13:43] yeah I expected it to be a bit useless :) [13:43] ain't nobody got time fo' dat [13:43] word [13:45] bregma: ok, trying the unity8-mir, I just get blank screen. Where do I look for logs? [13:45] greyback: ~/.cache/upstart -- that problem sounds like the QT5.2 problem [13:46] bregma: ibEGL warning: unsupported platform (null) [13:46] terminate called after throwing an instance of 'boost::exception_detail::clone_impl >' [13:46] what(): MesaNativePlatform::create_internal_client is not implemented yet! [13:46] interesting, I wonder what I broke [13:48] what version of Mir do you have installed? [13:49] that should have been fixed by 0.1.7 [13:49] you might need to upgrade [13:52] I have updated all, but now I can't even start mir_demo_server_shell [13:53] elopio: hi === alan_g|lunch is now known as alan_g [14:00] MacSlow, if you do work on the boot animation, please feel free to do it in ~unity-team/unity-system-compositor/new-gl-screen [14:00] greyback, are you using proprietary drivers by any chance? [14:00] mterry, ok [14:01] MacSlow, last week I already made it use a black background and no text [14:01] MacSlow, so it should mostly be an asset swap and maybe some small tweaks on your end [14:01] bregma: crisis averted [14:01] mterry, still busy in notification-land... might be much later today I'll be able to switch to the loading-screen [14:02] MacSlow, no rush [14:04] bregma: have you tried this on nouveau hardware? I have some text characters missing, maybe driver bug? [14:05] it's not exactly buttery smoother either [14:07] greyback, no, all my test hardware is Intel [14:07] and graphics corruption and crash.... [14:10] bregma: well anyway I've approved the qtubuntu qt5.2 fix [14:11] and once I get used to the invisible cursor, I'll launch some apps [14:12] greyback, it's the 21st century, get a HDPI touchscreen [14:15] one day someone should investigate this errors we're getting [14:15] QtDBus: cannot relay signals from parent QObject(0xabb0e330 "") unless they are emitted in the object's thread QThread(0x15a9c68 ""). Current thread is QThread(0x158c8c0 ""). === greyback is now known as greyback|lunch === dandrader is now known as dandrader|afk [14:39] mterry, doh... imap... I'll wait half a day before considering to reply to anything :) [14:40] kgunn, if you have some time today, I'd appreciate you removing telephony-service branches from the split silo. And adding in the lp:~mterry/lightdm/fake-no-change-mp branch -- this should be enough for people to play with functionality [14:40] MacSlow, I know, right! Everyone in that thread is doing it ;) [14:40] kgunn, (we can remove the telephony stuff because it landed in trusty) === pete-woods is now known as pete-woods-lunch [14:44] mterry, I can do, so we don't want telephony-service in the silo any more? [14:45] Saviq, right. And we want to add lightdm via lp:~mterry/lightdm/fake-no-change-mp [14:45] mterry, yup, doing [14:45] mterry, although the service won't disappear from the silo, does it need to? looks like it does? [14:46] i.e. there's not a newer version in distro [14:46] Saviq, we had a version land today in trusty I thought === dandrader|afk is now known as dandrader [14:47] Saviq, ah, it entered with an older date than the silo has [14:47] Saviq, yeah, we can delete the one in the silo [14:47] mterry, yeah, ok, we'll have to delete [14:47] sil2100, can you please delete telephony-service from landing-004 ppa? [14:48] Saviq: just the package from the PPA, right? Sure [14:48] sil2100, right [14:49] sil2100, yup [14:49] mterry, so only lightdm needs building? [14:49] Saviq, mterry: it's removing [14:49] thanks! [14:50] mterry, ok, lightdm is gonna build soon [14:50] Saviq, for the nonce yeah [14:50] thanks..i was majorly distracted :) [14:52] Saviq: ok, i know what's going on with that AP test that fails [14:52] tsdgeos, good [14:52] mterry, please add commit message [14:52] now i only need to fix and add a test :D [14:53] tsdgeos, looking forward to it! :) [14:53] Saviq, whoops, done [14:53] mterry, cheers [14:54] didrocks, uh oh [14:54] unity-scopes-master-default [14:55] didrocks, we need that gone, too [14:56] actually it's not there maybe? /me no get it [14:56] but upgrading ubuntu-touch didn't say the old scopes can be removed... [15:00] mhr3, is it expected that we still see songs instead of albums in the local music carousel? [15:01] Saviq, yes [15:01] Saviq, the design has a separate category for albums, and we have a branch that does that, but music-app needs to support launching those first [15:02] mhr3, well, songs should be random or something, no? so that we don't see the same cover all the time? [15:02] Saviq, not really specified anywhere afaict [15:03] * Saviq doesn't think there should be "songs" at all in the dash, not for surfacing [15:03] convince design :) [15:03] mzanetti: hello. [15:03] elopio: hi. I'm struggling with getting one of the autopilot tests working again. Would you have time to help me? [15:04] mzanetti: I do. Which one and from what branch? [15:04] elopio: I'll ping you again in a few minutes, will prepare instructions for you to reproduce [15:05] that's ok [15:06] mhr3: http://91.189.92.89/api/scopes/ubuntu-14.04/ === greyback|lunch is now known as greyback [15:11] mhall119, wooow, looks.. consistent :) [15:11] MacSlow, hey, bfiller reported https://bugs.launchpad.net/unity8/+bug/1296777, can you please have a look when you have a chance? [15:12] Ubuntu bug 1296777 in Unity 8 "avatar displayed incorrectly in notification" [High,New] [15:12] mhall119, so mono font here and there would be nice though ;) [15:12] mhr3: where and where? [15:13] Saviq, ah ok [15:14] mhall119, the code snippets for example [15:17] ah, yes, I can do that [15:17] Saviq: https://code.launchpad.net/~aacid/unity8/LVWPHcullLostItems/+merge/212457 [15:18] elopio: I think I've solved it. The issue seems to be that QML *sometimes* optimizes the AppPreview to be a DashPreview, sometimes doesn't. [15:18] mhall119, would be also good if the categorization you did was connected with the docs, i guess right now you did it manually? [15:18] elopio: I've removed the select_single by classname and only fetch it via objectName now [15:18] mzanetti: hum, that might be a problem. [15:19] why? [15:19] tsdgeos, on it [15:19] mzanetti: because we have to define some helpers in AppPreview that don't apply to DashPreview, like install() [15:20] tsdgeos, TODO somewhere for a test? [15:20] tsdgeos, if you think not → not [15:20] mzanetti: but that will be a problem later, when I have to update the click scope tests. [15:20] elopio: but I have the impression that this select by classname doesn't really work well with QML anyways [15:20] Saviq: honestly i would not know how to test it, but i can add a TODO saying "think how to do this" [15:21] elopio: as just changing one property can cause it to change classname in the qml context [15:21] mzanetti: that could be even worse, we rely on it too much. [15:21] mzanetti: oh, yes, I hate that. I had some problems before with CheckBoxes and Switches because of that. [15:21] tsdgeos, your call [15:22] Saviq: i'll add it, someone invests some time and decides if to do the test or remove the todo [15:22] tsdgeos, yup [15:22] mzanetti: they are working on an alternate way to match the helpers with a QML class. [15:22] elopio: yeah, had to change some in my branch now, just because I somewhat touched a file, but yeah, the real problem is that this optimization might even change at runtime === alan_g is now known as alan_g|tea [15:22] I think that will make it a little better. [15:22] elopio: what's wrong with objectNAme? [15:23] elopio: that one will stay the same unless you really change the code in way that you have to update tests anyways [15:23] mzanetti: objectName is perfect. The thing is that autopilot currently needs the name of the python class to match the name of the QML class. [15:23] elopio: no it doesn't [15:23] and that's sometimes painful, not just your example. [15:23] Saviq: pushed [15:23] select_single(objectName="dashPreview", isActive=True) works fine [15:24] tsdgeos, thanks [15:24] mzanetti: yes, that works, but it returns an instance of the python class that matches the QML class. [15:26] which sounds correct to me... [15:26] mzanetti: you will see some tests on test_emulators.py with self.assertIsInstance [15:27] those check that we return the right instance. If they pass for you, things are ok. [15:27] elopio: yeah, changed that to just check for GenericScopeView instead of DashApps [15:28] given that the line above checks if scope.id == apps.scope, I don't think its necessary to check for the isInstanceOf too [15:28] mzanetti: then that's not ok. [15:28] mzanetti: go to unity8/shell/emulators/dash.py [15:28] ok [15:30] you will that the DashApps class has an extra method, get_applications. [15:30] that one will not be available if the instance returned is GenericScopeView. [15:30] elopio: yeah, moved that into GenericScopeView [15:30] we need a way to solve that. [15:30] mzanetti: ok, that's one way :) [15:31] not the nicest one, because not all GenericScopeViews should have a get_applications method [15:31] yeah... [15:31] but due to the current autopilot implementation, there' [15:31] s not much we can do about it. [15:31] maybe raise an exception if we call it with a scope that doesn't show apps. [15:32] mhr3: yeah, manually for now, because I didn't see any way to derive that information from the docs === alan_g|tea is now known as alan_g [15:32] mzanetti: I have a meeting, I'll be back in 1 hour, but it seems you have it properly covered. [15:33] just one detail, your changes might conflict with what tsdgeos is doing for new scopes. [15:33] elopio: same here... [15:33] noOoOoOoOoOoO [15:33] elopio: well, in that case we'll merge it [15:33] mhall119, can't doxygen do some tags? [15:34] mhall119, otherwise it's just asking to get stuff out-of-sync [15:34] mhr3: I'm not really a doxygen expert === pete-woods-lunch is now known as pete-woods [15:50] Saviq, this is the 1st step of changing how jenkins unlocks: https://code.launchpad.net/~mterry/unity8/unlock-script/+merge/212170 You seemed interested last time we talked [15:50] mterry, yeah, I'll look through, but we definitely need ACK from QA and CI folks [15:51] Saviq: just one more query on 5.2 render loop update issue, so do we need to put another thread in unity-mir for the screen off state to make qt happy ? [15:51] mterry, a) ship it in unity8-autopilot, not unity8 [15:51] Saviq, sure... read the description though, this is a very incremental small change and does what you want -- encapsulates more info inside the component. This change itself shouldn't need QA signoff, but the next step where I change the jenkins scripts will [15:52] kgunn, it's not that [15:52] kgunn, there's two threads in QML threaded rendering: GUI and render [15:52] kgunn, GUI processes almost everything, render just does the scenegraph thing [15:53] kgunn, on 5.0, they acted more or less independently [15:53] kgunn, now they do a sync once every frame [15:53] Saviq, done [15:53] arenson, o/ [15:53] \o [15:53] hello [15:53] arenson, if you can use one nick across IRC, that's helpful :) [15:54] kgunn, and if that sync doesn't return, the GUI thread is ultimately blocked [15:54] kgunn, so we basically can't block the render thread, unless we tell Qt to stop rendering via an exposeEvent [15:54] Saviq: oh wow, so yeah that seems weird [15:54] for some reason it says josharenson is taken on this server, perhaps I'll just make it arenson across servers [15:54] kgunn, read https://bugreports.qt-project.org/browse/QTBUG-37677 [15:55] Saviq: thanks...i'll read up [15:55] arenson, must be "josharenson" was registered by someone maybe, not here at this time, though [15:55] still, based on your description...seems like a funny decision [15:55] maybe it was me at some point.... [15:55] arenson, weird, NickServ doesn't know it, you should be able to [15:55] arenson, type /nick josharenson now? === arenson is now known as josharenson [15:57] cool [15:57] changed back on canonical too... sorry for the confusion [15:57] josharenson, no worries [15:59] Saviq: how is going testing the LVWPH fix for the AP test? [16:00] and does anything else in new-scopes needs me fighitng against it? [16:00] tsdgeos, unity8 just built [16:00] I think [16:00] tsdgeos, so will run it now [16:01] didrocks, I think I reconfigured landing 013 to not include ubuntu-touch-meta, can you add it back? [16:03] sil2100, maybe you're around ↑ [16:03] Saviq: ok [16:03] Saviq: doing [16:03] sil2100, thanks [16:05] mterry, looks like landing-004 is good to go? [16:05] Saviq: tssss :p [16:05] didrocks, yeah :| [16:05] Saviq, define good to go? [16:06] mterry, everything built [16:06] mterry, and ready for testing [16:06] Saviq, yeah, everything built. I'm going to send out an email shortly after I do a double-check test myself [16:06] mterry, yup, cool [16:06] Saviq, thanks btw! [16:06] mterry, cheers [16:08] Saviq: reconfigured [16:08] sil2100, thanks! [16:10] mzanetti, tsdgeos: how does one usually get a models role-types known on the QML-side? [16:11] MacSlow: not sure I understand the question [16:11] MacSlow: read the documentation of the model [16:11] mzanetti, tsdgeos: is some Qt/moc macro needed in the C++-sources? [16:11] or in its absense read the c++ of the model :D [16:11] MacSlow: besides the roleNames function? [16:11] no [16:12] tsdgeos, I need to make a SortFilterProxyModel work [16:12] MacSlow, it takes int as models I'm afraid [16:12] s/models/roles/ [16:13] MacSlow: ok, what's the question? [16:13] :) [16:13] MacSlow, so you just need to know the id of the role [16:13] Saviq, well sure... but I've seem it used in other places in unity8 and there symbolic names were used [16:13] Saviq, I wanted to avoid that :) [16:14] MacSlow, ah then yeah, you need Q_ENUM [16:14] Saviq, argl... so I do need a new branch for unity-notifications :) [16:16] MacSlow, ;) [16:16] MacSlow, good thing is, these land together now [16:16] Saviq, didn't expect your change-request there on my MP to be that involved [16:16] MacSlow, i.e. we can put unity8, unity-notifications in the same silo, and they'll land in the same time [16:17] Saviq: once sentence update on the Qt issue? [16:17] (preparing the landing team meeting) [16:17] so that you don't have to come [16:18] didrocks, Qt said "you're doing it wrong", we're disagreeing to some extent, but looking for a local solution anyway, until we can convince upstream otherwise [16:18] Saviq: thanks :) [16:19] Saviq: you are the one leading it, right? [16:19] didrocks, reverting Qt change is not an option, it's a rewrite of the threded rendering engine [16:19] didrocks, with tvoss taking over from an arch PoV [16:19] didrocks, but for now I'm the PoC still [16:19] Saviq: perfect… and good luck! :) [16:21] Saviq: mterry: so the split branch conflicts a bit with new-scopes-clean-to-trunk, given that AFAIK new-scopes-clean-to-trunk is the "blocker" for commiting anything else would make sense to merge it in already in split? [16:21] Saviq: is this about the one with the events being queued while not rendering? [16:21] tsdgeos, OK. Can do [16:21] tsdgeos, we should land it today [16:21] tsdgeos, so by tomorrow it will hopefully be in trunk [16:22] ok [16:22] Still good to get on top of the merge [16:23] ugh, just got 5 dbus failures on it :| [16:26] woot? [16:29] Saviq, btw once you have a moment - happy with [16:29] https://code.launchpad.net/~mhr3/unity-scopes-shell/expose-attributes/+merge/212213 ? [16:29] Saviq, and we need to talk about filters, how do you want those exposed [16:30] mhr3, looking good, think we're not limiting ourselves with text and optional icon? didn't we have some place where there was no text, actually? [16:31] Saviq, there was always at least a number or something [16:31] Saviq, and noone's saying the value can't be a dict of something [16:31] mhr3, ok [16:32] mhr3, as filters go, I think they way they were done until now was pretty good [16:32] mhr3, and without looking at some visuals... pretty difficult to say otherwise [16:32] ok, will check that out === dandrader is now known as dandrader|lunch [16:55] * tsdgeos EODs [17:11] * Saviq too [17:11] o/ [17:15] Saviq, no longer racy -> https://code.launchpad.net/~macslow/unity8/modal-snap-decisions/+merge/210988 [17:26] mzanetti, if you like Karma... this gives you an easy way to harvest it :) https://code.launchpad.net/~macslow/unity-notifications/modal-snap-decisions/+merge/212483 [17:27] MacSlow: ok. will review [17:27] mzanetti, thx [17:30] MacSlow: I think we require a checklist [17:30] mzanetti, it didn't save the text for that? [17:30] mzanetti, one sec [17:32] mzanetti, but now it is === alan_g is now known as alan_g|afk === arenson is now known as josharenson [17:39] greyback, no, I haven't seen that behavior from USC... [17:40] mterry: would you mind double-checking on your hardware, to see if it's my setup or not? [17:41] greyback, I can. Are you blocking on this right now or can I get back to you in a bit? I'm in the middle of something [17:42] MacSlow: I hate to say it, but IMO that enum should be defined in unity-api and tested for its existance [17:43] mterry: not blocking me, so no rush [17:43] mzanetti, it wasn't missed before... so why now? [17:44] MacSlow: before it was only used internally in the backend. [17:44] now it is used in Unity8 too [17:44] which means we have a inter-repo dependency on that enum [17:44] => unity-api defines it [17:44] mzanetti, what process is to be followed to get that done? [17:45] MacSlow: add the enum to unity-api, remove it from the backend and file the mp as usual [17:45] MacSlow: later the silo takes care for all 3 landing at the same time [17:45] hence the "Are there any related mps..." question in the checklist [17:48] mzanetti, Saviq: I guess that then needs to go into lp:unity-api/include/unity/shell/notifications/Enums.h ?! [17:49] MacSlow: this is really specific to the model. put it inside the model [17:49] the stuff in Enums.h is more generic, appliying to many models/structures [17:52] mhr3: what 'tags' were you referring to earlier? [17:52] I'm digging through the doxygen output trying to find something to use to categorize thigns [17:56] mhall119, nothing specific, was mostly wondering if doxygen has something like that [17:57] not that I can see [17:58] and since all of the classes except testing are in the same namespace, I don't see any alternative to manual mapping [17:58] FWIW, the script defaults to putting things into the "Searching" section unless told otherwise === dandrader_ is now known as dandrader [18:05] mhall119, http://www.stack.nl/~dimitri/doxygen/manual/grouping.html === alan_g|afk is now known as alan_g [18:40] mhr3: those might work, are any of them used in the current docs? [18:42] mhall119, no [19:25] tedg, I saw the greeter-volume changes landed for indicator-sound, awesome! [19:25] tedg, but I thought https://code.launchpad.net/~mterry/gsettings-ubuntu-touch-schemas/volume/+merge/209158 was in the same silo? [19:26] tedg, can you add that to the next indicator silo? Otherwise the volume/mute syncing won't do anything [19:26] mterry, Ah, bummer, I put the volume changes in the player settings silo, I forgot to include that one. [19:26] mterry, Sure, not sure when that'll be, if you get one throw it in there as well. [19:27] tedg, k [19:27] mterry, Told you we should put the setting in the indicator sound package ;-) [19:27] tedg, hrrm [19:28] I think that's all the indicator-sound stuff though, the project is flushed. Which is exciting. [19:28] tedg, :) [19:51] greyback, do you know much about what makes the osk keyboard become "maximized" (unity-mir triggers some of its keyboard code off when the keybaord surface becomes maximized) [19:58] mterry, there's a direct comm channel between qt and maliit, if that's what you're asking [19:58] mterry, that says 'hey, a field got focused' [19:58] mterry, and then there's an API between ubuntu-keyboard and unity8 (over a socket) to tell it what's the keyboard state [19:59] Saviq, I think I'm interested in ubuntu-keyboard more than maliit. Found an issue where it is passing clicks through the keyboard into shell with split branch [19:59] as well as directly back to Qt for the Qt.inputMethod [19:59] mterry, than you want this ↑ [19:59] mterry, OSKController.qml [19:59] Saviq, seems to be because OSKController never sees that the keyboard surface gets maximized [19:59] Saviq, right. It has code to watch the surface state. This state gets triggered without my branch, but doesn't with [20:00] Trying to find code that actually "maximizes" the ubuntu-keyboard surface === dandrader is now known as dandrader|bbl [20:11] greyback, testing your PPA, will be back (hopefully) [20:14] greyback, seems to work fine to me (stop lightdm and restart lightdm both killed USC) [20:14] greyback, I wasn't logged in [20:14] greyback, should I have tried a unity8 session and then stopped lightdm? [20:24] mterry: ok, strange it behaves differently for me. Will have a look tomorrow [20:24] * greyback eod === ajmitch_ is now known as ajmitch === salem_ is now known as _salem === dandrader|bbl is now known as dandrader [23:02] the scopes have landed! [23:23] Saviq, go get some sleep [23:24] eod [23:55] mterry, are we ready to rebuild silo 4 on top of new scopes? [23:56] mterry, they got merged [23:56] Saviq, uh, OK. Let me merge from trunk. I'm also debugging a weird issue with the ubuntu keyboard [23:56] Saviq, that's why I didn't send out the email [23:56] Saviq, but I'll merge now [23:57] mterry, ok cool, let me know