=== olli_ is now known as olli [05:54] Saviq: something fishy in Unity8 compilation are actually dependencies installation https://bugs.launchpad.net/unity8/+bug/1212944 [05:54] Launchpad bug 1212944 in Unity 8 "unity 8 FTBFS (during package installation)" [Critical,New] [06:57] Mirv, hmm, looks like usermetricsservice configuration failed [06:59] Mirv, "Setting up usermetricsservice:i386 (1.1.1+13.10.20130815.2-0ubuntu1) ... [06:59] Failed to open connection to "system" message bus: Failed to connect to socket /var/run/dbus/system_bus_socket: No such file or directory" [07:00] Mirv, but that shouldn't be a build dependency anyway [07:03] Mirv, it seems http://bazaar.launchpad.net/~unity-team/libusermetrics/trunk/revision/125 is the culprit [07:03] Mirv, it has a .postinst that requires the system bus to be there [07:03] Saviq: yeah, it shouldn't fail installing in case it's not [07:04] Mirv, yup [07:13] hi sil2100 [07:15] Mirv, are you doing something about it or are we waiting for pete-woods? [07:21] Saviq: not doing at the moment and haven't filed a bug yet as I've other urgencies ongoing [07:21] Mirv, I've added affects: libusermetrics to the other bug [07:21] Mirv, so I'll own that, then [07:21] maybe sil2100 can look at it now that he's here, after he catches up :) (thankfully he's back to help!) [07:21] :) [07:22] thank you [07:27] Saviq, btw could we get someone to help with the running apps renderer? it doesn't support regular scope results model... when i was doing the category overrides branch i had a quick look at it, but it was getting too much into the ui bits for my taste [07:28] Saviq, touching ResponsiveFlowGrid... eek :P [07:29] mhr3, not sure we want to spend time on it when the Mir-backed recent apps scope should be working soon [07:30] Saviq, but that's why we need it, it needs to support the regular scope model [07:30] mhr3, ah, got it now [07:31] Saviq, not super high on the list, but once the mir bits are in, we'll want to use it [07:31] mhr3, yeah, will do [07:34] Saviq, off topic what's the current way of getting new image on the phone? phablet-flash ubuntu-system? seems to take too long compared to what was there a week ago :/ [07:34] mhr3, cdimage-touch [07:34] mhr3, otherwise you get read-only / [07:38] at least someone knows these things :) [08:01] Saviq: what do i use to test https://code.launchpad.net/~saviq/unity8/fix-generic-preview/+merge/179932 ? [08:01] tsdgeos, search for something in home [08:01] tsdgeos, and long-press [08:01] tsdgeos, description (if long) should be wrapped and formatted [08:01] tsdgeos, wikipedia is usually useful [08:02] tsdgeos, http://ubuntuone.com/0y3o0lPEaWsVBEP6gdiJdd http://ubuntuone.com/4AGFEchm7D1QqVvDtrbmZb [08:02] tsdgeos, we seem to have triggered an issue in LVWPH :/ [08:02] ah yes [08:02] there was this bug you said the other day in which it stuck in the middle? [08:03] is this that one? [08:03] tsdgeos, yeah [08:04] hmm [08:05] tsdgeos, and then something related to category expansion - go to apps, search for something, clear the search, click the "Dash plugins" category to expand it [08:05] tsdgeos, stuff breaks badly :/ [08:05] tsdgeos, like first of all, the limiting filter gets confused and doesn't actually "pass through" the items [08:06] tsdgeos, and then the whole thing just goes all over the place :/ [08:06] hmmm [08:06] works here? [08:07] what is really weird is that i can't vertically move stuff in the home scope until i expand files & folders [08:08] which kind of makes sense but doesn't at the same time :D [08:08] tsdgeos, i think the limiting proxy doesn't handle few cases - for example removal of one of the visible items, which pushes one of the invisible items to become visible [08:09] and vice versa [08:09] may be [08:09] pushing a new item in the first position, which should remove the last visible one [08:10] sup everyone? scopes still getting you down? [08:11] dednick, it's all your fault! [08:11] dednick, hope you feel down as well now :P [08:12] um. what did i do? [08:12] pete-woods, https://bugs.launchpad.net/unity8/+bug/1212944/comments/1 [08:12] Launchpad bug 1212944 in Unity 8 "unity 8 FTBFS (during package installation)" [Critical,In progress] [08:12] dednick, nothing in particular, we just want you to feel bad too, so you don't feel excluded [08:13] oh. thanks :) [08:13] anytime [08:13] i'm sure i had something to do with it anyway. [08:13] Saviq: thanks for pointing that out! [08:13] pete-woods, well, yeah, you caused FTBFS for unity8 ;) [08:16] Saviq: I don't understand why dbus wouldn't be there - it's a dependency [08:16] but what the hey, I'll make the thing less sensitive [08:16] pete-woods, it's not started, you don't want dbus on the build machines [08:16] Saviq: okay, I suppose that's fair! [08:19] Saviq: there's some weirdness in wrapping+richtext, look for elephant and preview the one in the Info category, see how text gets wrapped too early, are we ok with that [08:19] ? [08:21] tsdgeos, good catch, let me see [08:22] Saviq: but it seems like a qt bug ::/ [08:22] i've added a rectangle on the label and it's covering the whole width [08:22] dednick, i think we broke something quite horribly - http://10.97.0.1:8080/job/autopilot-saucy-daily_release/1027/ [08:23] ouch [08:26] sil2100: hi. What do we need to do next to get mediascanner into the archive? [08:26] tsdgeos, I wonder if it's the line that's causing that [08:26] maybe [08:26] Saviq: anyway it0s an improvement over what we have, so going to aprove it [08:27] and I'll log a bug about it, ok? [08:27] mhr3: told you i did something [08:31] mhr3: yeah, i think i can see what's going on. it is my branch [08:32] ...although [08:34] mhr3: no, dont know what's happening [08:34] tsdgeos, hmm, the interesting thing is... should \n be considered as newlines in richtext anyway? [08:34] dednick, give me a sec, standup [08:35] Saviq: i think not [08:35] that's not how html behaves [08:35] tsdgeos, exactly [08:35] MacSlow, ping [08:36] nic-doffay, yes [08:36] MacSlow, do you have experience with running autopilot tests? Remember you mentioning it in standup a while ago. [08:36] nic-doffay, I do have some... what do you want to run? [08:37] There's an issue with one of my branches and I need to sort it out. [08:37] MacSlow, https://jenkins.qa.ubuntu.com/job/generic-mediumtests-saucy/2286/? [08:37] tsdgeos, ah, I think it's caused by the fact that font size is set [08:37] It's related to the mouse pointer now that the OptionSelector has been added. [08:37] Saviq: weird [08:38] tsdgeos, aactually! [08:38] tsdgeos, check this out http://paste.ubuntu.com/5992013/ [08:39] tsdgeos, it doesn't get re-laid-out if you make the window bigger [08:39] brr [08:39] jamesh: hi! I'll be poking someone to sponsor it today into the archive [08:39] nic-doffay, so you want to only run that "test_can_select_listview" locally for debugging the issue?! [08:39] jamesh: since the packaging is more or less ready now [08:39] sil2100: thanks! [08:40] sil2100: any chance you could take a quick look at lp:unity-scope-mediascanner too? It is a lot simpler, and I think I've got most things right [08:40] mhr3, pstolowski so... richtext... can we nl2br the text? will unity7 be fine with
s? [08:40] Saviq: only if you make it smaller :/ [08:40] tsdgeos, exactly [08:40] Saviq: same with 5.1 actually [08:40] * tsdgeos goes on crying [08:41] MacSlow, ideally yeah. [08:42] nic-doffay, you just cd to unity8/tests/autopilot ; autopilot list unity8 ... and then look for the test you're interested in... [08:42] nic-doffay, then just "autopilot run " [08:42] nic-doffay, will be some looooong name [08:42] Saviq: do you report a bug to Qt? [08:43] tsdgeos, yeah, will do, trying for a workaround now [08:43] Saviq: you can replace the Label with Text to drop Ubuntu.Components [08:43] nic-doffay, sorry... make that toolkit not unity8 :) [08:43] tsdgeos, doesn't matter [08:44] Saviq: i mean for the bug report in Qt [08:44] tsdgeos, right, of course [08:44] MacSlow, great thanks! [08:45] jamesh: sure thing! [08:45] nic-doffay, should be... cd ubuntu-ui-toolkit/tests/autopilot ; autopilot run ubuntuuitoolkit.tests.gallery.test_gallery.GenericTests.test_can_select_listview [08:46] wtf is up with libusermetricoutput? it keeps on conflicting on my system. [08:47] dednick: the thing that gives data for the welcome screen infographic [08:47] every time i do an upgrade, it buggers up the install and i get a half upgraded system that is totally borked. [08:47] something with the conf file. [08:47] weird [08:48] works fine here [08:50] dednick, maybe you have some ppa enabled that you should not? [08:51] Saviq: http://ppa.launchpad.net/ubuntu-unity/next/ubuntu ? [08:51] dednick, yeah, not needed / wanted [08:51] dednick, for unity8 you don't need any PPAs now [08:51] dednick, so anything you have, you need to know that you want it :) [08:51] Saviq: cool. thanks [08:58] tsdgeos, ugh, can you edit the description in bugreports? [08:59] not sure [08:59] Saviq: why one? [08:59] "Provide more info" [08:59] of course! [08:59] tsdgeos, one what? [08:59] which one i mean [08:59] t [08:59] mhr3: hm. everything looks in terms of searching in the videos... [09:00] tsdgeos, https://bugreports.qt-project.org/browse/QTBUG-33020?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel [09:00] tsdgeos, fixed already [09:00] ah ok [09:01] mhr3: i wonder if something is going wrong with counting the number of items in the categories... [09:02] tsdgeos, pushed a workaround [09:03] oh wow :D [09:03] tsdgeos, actually, need to do one more [09:03] tsdgeos, yah! [09:04] tsdgeos, I knew you'd love it ;) [09:06] tsdgeos, pushed s/\n/
/g [09:06] +\ [09:06] oka [09:07] Saviq, yea, you better do that in the shell, it's coming from the scopes that way [09:07] mhr3, ok, un-TODOing, then [09:07] (and we don't want to fix each and every scope) [09:08] mhr3, yeah, what I though [09:08] t [09:08] tsdgeos, need to re-approve, won't land otherwise [09:08] yep [09:08] let me test it works :D [09:09] tsdgeos, but then it won't land anyway... [09:09] pete-woods, can we expect a fix for libusermetrics sometime soon? :) [09:09] dednick, i guess the category isn't "fully" removed and ap picks up an old one that has the same name and is empty now? [09:09] sil2100, is libusermetrics part of unity8's stack? (it should be, I think...) [09:09] Saviq: it's committed already [09:09] pete-woods, ah [09:09] pete-woods, thanks [09:09] np [09:10] sil2100, Mirv can we release libusermetrics? [09:10] mhr3: hm. maybe. i guess it would be set to non-visible! dang. didnt think about that [09:10] stupid categories [09:11] dednick, but there's a call to ClearCategories() shouldn't that handle it? [09:11] mhr3: it should [09:11] Saviq, sil2100: running indicators stack, can you sil2100 look it goes through? [09:11] Mirv: ok [09:12] well that's a bugger. builddeb failed because of one test... sigh [09:12] and yes for some reason it's in the indicators stack [09:13] dednick, let's corner Trevinho next week and make him fix it, he's in london [09:13] mhr3: hehe. fix our bug! [09:13] will probably need to lock him on the toilet with his laptop [09:13] oh wait... [09:15] mhr3: yeah, it's multiple categories. i can reproduce [09:15] mhr3: just removed SetVisible(false) ... [09:16] dednick, so why aren't they cleared? [09:16] :/ [09:16] i will find out promptly [09:17] oh... the initial setup is called before all the categories are added, isn't it? [09:17] so it adds them twice the first time [09:18] it should clear them every time it switches model. [09:18] unless the bind isnt working correctly.... [09:30] mhr3: right. so apparently sigc::bind doesnt work with sigc::hide. [09:31] mhr3: and http://pastebin.ubuntu.com/5992130/ [09:31] i'm getting 2 application categories... [09:31] dednick, that's fine [09:32] mhr3: why would that be fine? [09:32] dednick, one has id "global", the other "apps" [09:32] you're printing the name, it's the same [09:33] as you can see the index is different [09:35] mhr3: yeah, i thought it was just adding 2 apps at different indexes [09:35] cool. it's my bad then [09:36] it's adding duplicates because the bind for "bool add" isnt working [09:36] rather retardedly [09:37] Mirv, hey [09:37] dednick, but your log says "false", doesn't that mean it works? [09:37] mhr3: after fixing yes [09:37] ok [09:37] mhr3: was always true [09:38] * mhr3 loves when compiler / library doesn't work [09:39] jamesh: one thing about mediascanner - would you mind if I change the MEDIASCANNER_VERSION variable in CMakeLists.txt to the proper version? [09:39] sil2100: that sounds fine. [09:43] mhr3: https://code.launchpad.net/~nick-dedekind/unity/duplicate.categories/+merge/180511 [09:44] * dednick endevours to actually test code he merges [09:44] om26er: hello [09:45] Saviq: shall we get the indicators show on the road? https://code.launchpad.net/~nick-dedekind/unity8/remove.battery.time.indicators/+merge/180375 [09:45] dednick, not yet, we still can't release unity8... [09:45] Mirv, the qtorganizer5-eds project that you setup yesterday didn't have a source code branch target, but I just saw someone added that [09:45] dednick, need to wait for unity to release first [09:46] Mirv, thanks anyway :) [09:46] om26er: yes, that "someone" was me as well :) [09:46] after I noticed there was some branch now to set as such [09:46] dednick, at least libusermetrics should not be killing us anymore [09:46] dednick, how about using standard sigc::mem_func(this, ClearCategories), and do the forced add at the end? will be more readable [09:46] Mirv, cool, thanks [09:46] sil2100, Indicators finished, yellow though (packaging change) [09:46] om26er: so it's now just lp:qtorganizer5-eds like it should be [09:46] Saviq: needs +1 from core-dev [09:47] Saviq: if you see any around, please get, for http://pastebin.ubuntu.com/5992152/ [09:47] Mirv, can you +1 this https://code.launchpad.net/~om26er/cupstream2distro-config/ubuntu_keyboard_release_daily/+merge/180442 ? [09:47] Mirv, who's "core-dev"? [09:47] mhr3: yeah, dont know why still doing func for a single line... [09:47] Saviq: https://launchpad.net/~ubuntu-core-dev/+members [09:47] Saviq: yes, we're poking around about that [09:47] Mirv, its to enable daily release for ubuntu-keyboard, I got the packaging reviewed from mterry yesterday, he made a few improvements [09:47] Saviq: it's the deal we have, need an ack from one always [09:47] sil2100, ok thanks [09:48] Mirv, understood [09:48] preferably in 12 minutes before the next cycle starts :) I wonder how long ogra's meeting is [09:48] :D [09:48] dednick, if i had this comment yesterday we'd save ourselves quite some trouble :) [09:48] om26er: ok, looking [09:51] Saviq, we need vim in the default touch image, i can't stand vi :P [09:51] mhr3, https://code.launchpad.net/~ubuntu-core-dev/ubuntu-seeds/ubuntu-touch.saucy [09:51] ;) [09:51] need to apt-get it everytime... terrible user experience [09:51] mhr3, and ssh [09:51] +100 :) [09:51] :| [09:51] Saviq: got this http://paste.ubuntu.com/5992173/ [09:52] just starting and unlocking [09:52] Without didrocks and seb128 it's so much more complicated to get things released :| [09:52] tsdgeos, https://code.launchpad.net/~mhr3/unity/only-valid-filters [09:52] At least until ken and mathieu pop up later [09:52] OH ZE FRENCH! [09:52] Saviq: but that is merged already [09:52] and i just got it [09:52] VER AR THEY WHEN YOU NEED THEM? [09:52] tsdgeos, not released [09:52] ah wait [09:52] it's on unity [09:52] not unity8 [09:52] tsdgeos, yes [09:52] true-ish [09:52] tsdgeos, and not released [09:53] Saviq: they are enjoying 15th august holidays? :D [09:53] tsdgeos, sounds like it indeed ;) [09:53] mhr3, what's the deal with the Unity stack? any ETA? [09:54] Saviq, yea, we have a fix now, should land soonish [09:54] Saviq, so... 4hours? [09:55] mhr3, daily starts in 5 mins, will it land by then? ;) [09:55] 5 minutes leeeft! [09:55] erm quaternourly [09:55] Mirv: did you find anyone?! [09:55] no [09:55] :/ [09:55] maybe if you allow dednick to push the fix directly [09:55] sil2100, ^? [09:56] sil2100: yes, now [09:59] Mirv: \o/ [10:00] Saviq: so, indicators published, so the daily-release run that will start in a minute should already have the fixed libusermetrics, so unity8 should pass now [10:00] Saviq, getting a weird error assigning a deligate to a ListView, mind giving your thoughts? [10:00] sil2100, probably won't, as we still need the unity fix [10:00] sil2100, and unity won't publish before https://code.launchpad.net/~nick-dedekind/unity/duplicate.categories/+merge/180511 [10:00] nic-doffay, details, please [10:01] Saviq, https://pastebin.canonical.com/95961/ [10:01] First line is the error. [10:01] You'll see the model is just an int for now. [10:01] And the delegate is the option selector. [10:01] nic-doffay, you need to wrap in Component { } [10:01] nic-doffay, if you don't assign to "delegate" directly [10:01] nic-doffay, but really [10:02] nic-doffay, in that case you don't need "delegate: " at all [10:02] nic-doffay, IIRC the child of ListView is the delegate by default [10:03] nic-doffay, or maybe not, and anyway it's better to be explicit [10:03] nic-doffay, also it's recommended that you put the Component { OptionSelector { } } outside of ListView { } [10:03] nic-doffay, see http://qt-project.org/doc/qt-5.0/qtquick/qml-qtquick2-listview.html#details [10:05] Saviq, cool. Thing is it's still only displaying one selector now. [10:05] Even though the model is clearly set as 3. [10:05] nic-doffay, Component { id: foo OptionSelector { } } ListView { model: 3; delegate: foo } [10:05] Mirv, is that change enough for daily release ? [10:06] Saviq, yeah that's what it looks like atm. [10:06] nic-doffay, that creates 3 selectors for sure [10:07] nic-doffay, code, please? [10:07] Saviq, https://pastebin.canonical.com/95962/ [10:08] The cos recommended not manipulating contentY/X directly so I'm redoing the scrolling. [10:08] there's something broken in the filter grids too :-/ [10:08] Saviq, that's the reason for redoing it with a model is correct because it can be provided from the backend. [10:09] i can easily reproduce this [10:09] FilterGrid Model 26 [10:09] FilterGridLimited Model 6 [10:09] FilterGridLimited Items 13 [10:09] Saviq, at least that's how I feel atm. [10:09] why the view has 13 items if the model only has 6? [10:09] eh! tellme! why! [10:09] mhr3: again. https://code.launchpad.net/~nick-dedekind/unity/duplicate.categories/+merge/180511 [10:10] woops. guess i missed daily [10:11] Saviq: any idea when unity8 release will be? [10:11] nic-doffay, your ListView doesn't have enough height [10:11] nic-doffay, to show more than one [10:12] om26er: not yet, the stack config needs to be 'redeployed', and it can't be done while the current run is running. me or sil2100 will do that in eg. two hours or so. [10:12] nic-doffay, height: contentHeight [10:12] Mirv, that's fine [10:12] Mirv, last review https://code.launchpad.net/~om26er/cupstream2distro-config/ci_for_qt-eds/+merge/180515 [10:12] dednick, thx, approved [10:13] Mirv, adds CI for qtorganizer5-eds [10:13] nic-doffay, you should avoid manipulating ListView.contentY directly, btw [10:13] om26er: ok, will check [10:14] nic-doffay, if possible, use positionViewAtIndex() [10:14] Saviq, when I add that there are errors about a binding loop. [10:14] Saviq, yeah that's what I'm experimenting with atm. [10:14] Or positionViewAtEnd [10:14] I was planning to see what worked best. [10:14] nic-doffay, AtEnd will position at end of the whole ListView [10:14] nic-doffay, so AtIndex is what you most probably want [10:15] Saviq, then I'll just position it at the last index using the other function. [10:15] Saviq, but back to the height issue. [10:15] nic-doffay, but anyway, just anchor your ListView's bottom to the parent, that's what you need anyway in the end, no? [10:15] Saviq, I'm not sure what you mean. [10:16] nic-doffay, somehow you need the ListView to have a height [10:16] nic-doffay, it doesn't assume a height by default [10:16] Saviq, you mean use the anchor instead of setting it to the contentHeight? [10:16] nic-doffay, contentHeight is the height of all the items in the ListView [10:17] nic-doffay, it's rarely the right thing to use [10:17] nic-doffay, it was just a quick'n'dirty approach from my side to show what's the issue [10:17] nic-doffay, I don't know what's the height you need [10:17] other than != 0 [10:19] Saviq, setting the bottom anchor was def the best way. [10:19] ta [10:19] nic-doffay, told you so ; [10:19] ;) [10:27] Saviq, tsdgeos, mzanetti: Is there a _lean_ way to have a single method (providing the "reply" for the snap-decision dialogs) exposed on DBus using QDBus without the need to deal with xml-files and qdbusxml2cpp? [10:28] MacSlow: well, you don't need the xml stuff [10:28] MacSlow: that's just a helper to generate the stub. If you think you're confident enough with D-Bus and QtDBus, feel free to write it yourself [10:29] MacSlow: see http://bazaar.launchpad.net/~mir-team/unity-mir/trunk/revision/17 on how to expose stuff using qtdbus [10:29] mzanetti, I have what I want working in Python (as proof of concept) and now want to get it into the notification-backend [10:29] tsdgeos, looking... [10:30] MacSlow: what's the problem with writing down the D-Bus interface in xml and then generating a Qt class you can inherit from? [10:31] mzanetti, nothing really... just wanted to avoid touching the build-steps [10:33] MacSlow: you shouldn't call the qdbus2xml in the build steps anyways [10:34] MacSlow: only manually once, and then committing the resulting .cpp/.h files. It's really just a helper to generate the class for you out of a D-Bus interface description [10:36] +1 [10:37] mzanetti, I tend be used to hook such things up into the build-scripts... won't do it [10:37] Saviq, btw... were you ok with my last summary on the extended snap-decisions? [10:38] MacSlow, well, I'm still not sure it will provide us enough flexibility... [10:38] MacSlow, as I mentioned earlier, I don't think sending an "updated" notification with the whole list every time the list of AP changes [10:38] MacSlow, will provide for good UX [10:39] Saviq, I rather want a green light so I don't have to redo it again [10:39] MacSlow, that's why I was opting for the MenuModel approach [10:40] MacSlow, sending *some* things through as hints I'm ok with [10:40] MacSlow, but the list of APs with their signal strength... I'm really not sure that's going to cut it [10:41] MacSlow, and as for "replying" to the notification - a single action in the MenuModel that you'd submit the data to would be good enough, too [10:41] MacSlow, although I remember you wrote that there's issues with invoking an action with parameters / arguments? [10:41] Saviq, correct [10:41] i.e. not supported yet in UnityMenuModel? [10:42] Saviq, if that's not in there yet I'd be stuck [10:43] MacSlow, and since larsu isn't here, I doubt it's "there" [10:43] Saviq, who'd add that? Do I have to patch it then?! [10:43] MacSlow, afaik larsu is working on UMM [10:43] but he's on holidays [10:44] * Saviq checks [10:44] or on debconf [10:44] hmm nothing in cadmin [10:45] Saviq, pressure is pretty high for these extended snap-decisions... so I guess I've to further swim into unfamiliar territory to get it done [10:45] MacSlow, I know [10:45] MacSlow, in absence of the "submit action with parameters" [10:46] MacSlow, we can go with the custom interface you're doing now [10:46] MacSlow, unless you want to dig into UMM [10:46] jamesh: a review for you, last one before getting it into the archive I guess? https://code.launchpad.net/~sil2100/mediascanner/version_bump/+merge/180531 [10:46] MacSlow, but really I'd rather not send the list of APs as hints, but just point at a MenuModel [10:47] MacSlow, and have that MenuModel updated [10:47] Saviq, can we briefly chat about the custom interface (which is really small... just one call) after the stand-up? i want to avoid following a path that's not going to be ack'ed [10:47] MacSlow, I think a(ss) is good enough [10:48] MacSlow, i.e. [('username', 'foo'), ('password', 'bar')] [10:48] [('ap', 'whatever-ap-id-wanted')] [10:48] can someone do quick review (typo fix)? https://code.launchpad.net/~nick-dedekind/unity8/indicators.network.password/+merge/180530 [10:49] dednick, there's a bug for it, right? [10:49]