[09:53] Saviq: that "just remove it" it's not a good enough solution tbh [09:53] i want to have both unity8 and unity7 installed [09:53] why do i need to have two network indicators in unity7? [09:53] tsdgeos, because indicator-network doesn't yet implement everything nm-applet does [09:54] Saviq: sure, then indicator-network should be blacklisted for the desktop [09:54] or whatever mechanism makes me not have a ugly unity7 experience [09:54] Saviq: sent you the email with the crash-tester btw [09:54] tsdgeos, thanks [09:55] tsdgeos, we could potentially disable the network indicator in the desktop profile, but then unity8 on the desktop would lose it, too... I'm not sure nm-applet is good enough with unity8 [09:55] tsdgeos, but anyway [09:55] sure, not my biggest problem :D [09:56] tsdgeos, yes, my replies were more to the "I don't know where that came from" [09:56] was just giving my opinion [09:56] rather than this is the right solution [09:56] tsdgeos, Saviq: you can edit /usr/share/unity/indicators/com.canonical.indicator.network and drop the desktop section, we should probably do that by default seeing the number of people complaining about it [09:57] then people who want to test the indicator on the desktop would need to edit to add back [09:57] seb128, yeah, but as said above, is nm-applet working good enough under the desktop unity8 session? [09:57] but opt-in probably makes sense, most people get the indicator through other depends, not because they want to use it [09:58] yeah sure [09:58] I've not tried it [09:58] but nm-applet is indicator-application [09:58] does that work under unity8? [10:01] seb128, it does, somewhat [10:02] hmm it actually seems to work fine [10:02] can't enable cellular, though [10:03] hmm or maybe that's actually indicator-network in desktop profile, not nm-applet [10:03] btw, the list of things that get removed: [10:04] with indicator-network: [10:04] account-plugin-ubuntuone* indicator-network* ubuntu-system-settings* ubuntu-system-settings-online-accounts* unity-scope-click* [10:04] not great [10:07] Saviq, just drop the sections from /usr/share/unity/indicators/com.canonical.indicator.network to make sure it's not it [10:07] seb128, yeah, no nm-applet under unity8, but that's probably fine (i.e. if you want unity8 and network, you need indicator-network), as long as ↑ those deps get fixed - like account-plugin-ubuntuone depending on the indicator network? what gives? [10:09] Saviq, it doesn't depends on it, it depends on ubuntu-system-settings-online-accounts [10:09] which depends on u-s-s [10:10] which depends on i-n since it's the backend used for the wifi panel [10:12] seb128, why does it depend on u-s-s, don't we have a desktop version of that? [10:13] mardy, ^ [10:13] mardy, well I guess we could maybe have a | gnome-control-center-signon [10:13] ups [10:13] Saviq, ^ [10:13] but I'm unsure, I didn't look at potential differences [10:14] seb128, right [10:15] Saviq: removing ubuntu-system-settings-online-accounts on the desktop is fine [10:15] Saviq: not sure about account-plugin-ubuntuone [10:15] mardy, yeah, I didn't think that one would be fine [10:17] yeah, "online accounts" → no Ubuntu One account (I can actually add one (???), but that just shows a grey screen - probably plugins cached somewhere and not reloaded on removal) [10:17] mardy, ↑ bug, btw? [10:18] Saviq: IIRC, plugins are not cached [10:18] Saviq, mardy: if "removing ubuntu-system-settings-online-accounts on the desktop is fine" we should probably have account-plugin-ubuntuone depends on "u-s-s-o-a | g-c-c-s" [10:18] so yeah, we need to solve the deps somehow, otherwise people will end up with indicator-network on their unity8-less desktops [10:19] seb128: I think so, unless they have a different package for the U1 plugin on the desktop [10:19] mardy, I mean "cached" in signond or somewhere - in memory basically [10:20] Saviq: I don't think so, but the OA UI is coming from the online-accounts-ui D-Bus service, which exits after a few seconds of inactivity [10:20] Saviq: if you just briefly closed and re-opened it, it may be that the old process was reused [10:20] hmm on second look it seems that online-accounts-ui doesn't even work on the desktop [10:20] seb128, ↑ [10:20] Saviq, you mean? [10:20] for U1 [10:21] seb128, I mean I can't add a U1 account through g-c-c [10:21] http://pastebin.ubuntu.com/7032274/ [10:21] right, I noticed that recently as well [10:21] it works through ubuntuone-control-panel-qt though [10:22] seb128, Saviq: actually, IIRC we never had a U1 account plugin for the desktop [10:22] mardy, but it's showing in the list of "protocols" in the u-c-c UI [10:22] seb128, mardy, yeah, so in fact it's probably fine for the plugin to not be there either [10:22] mardy, which is confusing [10:22] (there == on the desktop) [10:22] it is [10:23] seb128: it definitely shouldn't be listed, but I think it will go away if you remove the plugin [10:24] mardy, not sure how I got the plugin installed, I guess that's because of touch stuff [10:25] but yeah, still a bug [10:25] seb128: yep [10:25] mardy, actually I can't get it to disappear from g-c-c... [10:25] removed he plugin, killed signon* [10:25] it's still there... [10:25] * Saviq files bugs [10:26] Saviq, do you have ubuntuone-credentials-common installed? [10:26] Saviq, that ships /usr/share/accounts/providers/ubuntuone.provider [10:26] seb128, yeah [10:27] if I move that file away it stops being listed [10:27] seb128, +1 [10:27] * Saviq not sure where to file bug... the provider shouldn't be there if the plugin isn't there, should it... [10:27] and it shouldn't be listed in g-c-c anyway, as it's incompatible with it apparently [10:28] Saviq, open on gnome-control-center-signon and let mardy reassign if needed I guess? ;-) [10:29] seb128, ;) [10:34] seb128, mardy, bug #1287640 [10:34] bug 1287640 in gnome-control-center-signon (Ubuntu) "UbuntuOne account plugin does not work" [Undecided,New] https://launchpad.net/bugs/1287640 [10:34] Saviq, thanks [10:34] Saviq: thanks [11:03] Saviq: in here https://code.launchpad.net/~mzanetti/unity-api/new-screenshot-and-focusing-api/+merge/199810 [11:03] Saviq: line 157, is this enough? [11:04] mzanetti, should be [11:04] thanks [11:04] mzanetti, the resulting .pc file has it as the version? [11:05] mzanetti, unity-mir needs an update of Provides, and unity8 needs and update of Depends (and unity8-fake-env of Provides:, too) [11:05] yep, its in there [11:05] right... [11:28] Saviq: ok... I think we can start reviewing the right edge stuff... there's quite a bit stuff to fix for sure, but I think feedback from reviewers would help me a lot in this stage [11:29] mzanetti, any dep or conflict with new-scopes? [11:29] Saviq: yes. it'll conflict with new-scopes as is right now... but as I don't know how long new-scopes will still take I based everything on trunk, hoping that the right edge would land first [11:30] mzanetti, ok, let's see that race ;) [11:31] Saviq: we also could agree on an order... but I'd need an ETA for new-scopes then [11:31] and last time I asked you weren't sure about that => I went for trunk [11:32] in any case. the stuff that conflicts is 90% dropping the new-scopes stuff and using the right-edge stuff [11:32] and I can obviously do/help with the merge, regardless who comes first [11:33] mzanetti, yup, well we'd like new-scopes to land asap, they basically need a review soon, too [11:33] mzanetti, but probably some cleanup first [11:35] mzanetti, could you look into dropping the HUD (just the integration part in Shell.qml for now) and skip the hud tests in autopilot? [11:35] Saviq: ack. [11:39] mzanetti, btw, https://wiki.ubuntu.com/Process/Merges/Checklists/Unity-Mir [11:39] Saviq: ? [11:40] mzanetti, on your merge for unity-mir, please add the checklists [11:40] -s === MacSlow is now known as MacSlow|lunch [11:40] huh? isn't it there? /me checks [11:40] mzanetti, ah, I meant -api [11:40] mzanetti, and yeah, it's there [11:40] mzanetti, sorry for the noise [11:41] np [11:41] so yeah, you start from here https://code.launchpad.net/~mzanetti/unity8/right-edge-2/+merge/204798 and all the related ones are listed in the checklist [11:43] mhr3_, what do we do about: [11:43] 24 -Recommends: ${unity-default-masterscopes}, [11:43] 25 + unity-scope-scopes, [11:43] 26 + unity-scope-onlinemusic, [11:43] 27 + unity-scope-mediascanner2, [11:43] 28 + unity-scope-click, === mhr3__ is now known as mhr3 [11:44] ETOOMANYmhr3s [11:44] mhr3_, what do we do about: [11:44] 24 -Recommends: ${unity-default-masterscopes}, [11:44] 25 + unity-scope-scopes, [11:44] 26 + unity-scope-onlinemusic, [11:44] 27 + unity-scope-mediascanner2, [11:44] 28 + unity-scope-click, [11:45] Saviq, hmmm [11:45] we don't really have desktop vs phone now [11:45] so if we keep it as hard recommends it'd be fine imo [11:45] for now anyway [11:46] mhr3k [11:53] hmmm [11:53] don't know what i did but my unity8 indicators area is now empt [11:53] Saviq: when you said disable the hud, you really meant to get rid of the hud completely? as in: if you drag from the bottom edge the hud button would not appear any more? [11:53] any idea why that may be happening? [11:53] tsdgeos: same here [11:53] oh really? [11:54] tsdgeos: I merged with tnrunk and they appeared again [11:54] that's really unfortunate when i'm trying to fix a crash in indicators :D [11:54] :D [11:54] mzanetti: hmmm, i'm up to date with trunk [11:54] * tsdgeos does a clean build [11:54] tsdgeos: well, I did a run_on_device on my right-edge-stuff and they were gone as of today, I did some merging, rebooting, rebuilding and they appeard again [11:56] ok [11:56] this is on the desktop fwiw [11:56] nope, nothing [11:57] what, i'm getting the fake scopes [11:57] * tsdgeos puzzled [11:58] i must have broken something with the ppa purge :-/ [11:59] mzanetti, yes, get rid of it [11:59] my pleasure, sir [11:59] mzanetti, it's going "somewhere else", but we don't know where yet [12:00] and they are back [12:00] after installing unity8 [12:00] that brought lots of packages that build -c didn't [12:00] weird [12:00] tsdgeos, build -c doesn't [12:00] tsdgeos, build -s does [12:00] ahhhh righto [12:01] Saviq: that pCell stuff is crazy (the cool way) :) [12:18] mzanetti, indeed! [12:27] Saviq: is this how you imagined it? https://code.launchpad.net/~mzanetti/unity8/disable-hud/+merge/209226 [12:27] mzanetti, nah, just remove it from Shell altogether [12:27] mzanetti, so that we don't even instantiate it [12:27] ok [12:27] mzanetti, we can easily bring it back with bzr [12:28] /food [12:28] ok [12:30] tsdgeos: the BottomBar is only there for revealing the hud, right? [12:30] mzanetti: yes [12:30] ok. /me removes the BottomBar too === alan_g is now known as alan_g|lunch === MacSlow|lunch is now known as MacSlow [13:53] dednick, what can we do about the "unity8 exiting kills indicators" thing? ;) it's getting annoying ;D [13:54] saviq, thoughts about adding an activityindicator to a scope view when there are no results visible and the one in the search bar isn't visible either? ie for surfacing [13:54] Saviq: um. check if it's being run on desktop i guess. [13:55] mhr3, I think it could fit somewhere in the new header [13:55] mhr3, on a slightly related topic, could we have somehting like "pull to reload", to update surfacing? (eg. news scope) [13:55] mhr3, like replace the looking glass icon [13:55] davidcalle, nothing to do with mhr3 ;D [13:55] saviq, so basically needs design :) [13:55] davidcalle, ^^ [13:55] yup [13:56] i'll talk with mike [13:56] but tbh i'd like the pull to reload too [13:56] would mean caching == solved :) [13:57] Saviq: i'll take a look [13:57] dednick, don't worry, was just a I-got-annoyed-by-this-right-now issue ;) [13:58] saviq, same for my indicator issue btw ^ :) [14:01] mhr3, ENOUNDERSTOOD [14:01] saviq, no activityindicator -> also I-got-annoyed-by-this-right-now [14:01] mhr3, ah ;) === alan_g|lunch is now known as alan_g [14:06] MacSlow, hey, can I steal you for some investigation on notification ap tests? [14:06] MacSlow, i.e. http://ci.ubuntu.com/smokeng/trusty/touch/mako/219:20140304:20140304/6967/unity8/848175/ [14:06] Saviq, looking... [14:07] MacSlow, somehow the assertion fails, even though I can see the notification correctly and the icon is there [14:07] so iconSource is definitely != "" [14:07] MacSlow, so it might be something with ap [14:08] Saviq, most likely [14:10] Saviq, I'm still wondering, if those could somehow be moved to pure qmltests... what keeps me away from porting them, is the missing interaction with the real backend [14:11] MacSlow, right, these are of the kind that should stay this way - they're integration tests [14:11] MacSlow, *maybe* less extensive [14:13] Saviq, I don't know enough of AP's internals to provide a better (non-string based) test there... which might be more robust [14:15] saviq, btw qtry_compare sucks - i replaced it with qsignalspy and shaved off almost 3seconds from make test [14:18] Saviq, these AP-tests for snap-decisions "failing" have held back jenkins-approvals many times [14:20] MacSlow, yeah, but not that one, I'm getting 100% fail on that [14:20] MacSlow, where before it was just flaky [14:20] MacSlow, and the only failures that we look over currently are unity8 crashes, which that one isn't [14:21] Saviq, it's not new notification-code at fault I bet, as I'm still waiting on some reviews :) [14:22] MacSlow, yeah I know ;) nothing ours changed recently - while autopilot did [14:22] MacSlow, so ok, I'm looking into it further [14:23] Saviq, is there maybe a "state" that AP stores, which is not in sync with the real notification? [14:29] MacSlow, not sure, but just confirmed old autopilot passes this test, upgrading 1 by 1 now [14:29] Saviq, bisecting the other way ;) [14:31] @unity: standup? [14:48] mzanetti, bug #1287689 is already marked as dupe of the non-rotating shell [14:48] bug 1210199 in unity8 (Ubuntu) "duplicate for #1287689 [Shell] support rotation" [High,Triaged] https://launchpad.net/bugs/1210199 [14:48] mzanetti, or maybe we just multiplexed - I was clearing it up with Jamie in #ubuntu-touch [14:48] Saviq: ok. sorry. missed that [14:49] mzanetti, don't be :) [14:49] but in any case, Jamie's report seems a dupe of some bug :) [14:49] mzanetti, also [14:49] plugins/Utils/easingcurve.h UNKNOWN *No copyright* [14:49] plugins/Utils/easingcurve.cpp UNKNOWN *No copyright* [14:49] qml/Stages/SwitchingApplicationImage.qml UNKNOWN *No copyright* [14:49] qml/Stages/SpreadDelegate.qml UNKNOWN *No copyright* [14:49] qml/Stages/TransformedSpreadDelegate.qml UNKNOWN *No copyright* [14:49] meh... [14:49] thanks. I'll fix [14:50] Saviq: doing a hangout on air now for an hour. so if possible, ping me again after 5pm :) [14:50] mzanetti, will do [16:08] mterry, qt is crashing on startup some 5-10% of the runs [16:08] Saviq, :( that makes tests hard to rely on [16:09] mterry, is fixed with 5.2 [16:09] Saviq, ah! OK, I saw you all talking about that in emails [16:09] mterry, yup [16:10] Saviq, well, that branch is ready, I think [16:11] mterry, please chase reviewers, then :) [16:12] Mirv: we need https://codereview.qt-project.org/#change,79857 in 5.2.1 [16:16] fixes bug #v [16:16] 1277206 [16:16] bug #1277206 grr [16:16] bug 1277206 in unity8 (Ubuntu) "QT5.2: Disabling bluetooth crashes unity8" [Critical,In progress] https://launchpad.net/bugs/1277206 [16:32] saviq, how do we want to deal with gotoScope / openScope that wants to define the search string / filter/department state? [16:32] saviq, should i be passing that somehow to the signals, or try to deal with it internally? [16:33] saviq, basically i'm pretty sure that if i set a .searchString on the scope, you're going to invalidate it as soon as you create the visual component [16:34] or am i wrong? [17:07] tsdgeos: Saviq: luckily weekly meeting now, pushing a build so that I check it compiled fine in the morning and copy the landing PPA === robru is now known as robru-sick === alan_g is now known as alan_g|EOD === jhodapp is now known as jhodapp|dr === WebbyIT is now known as rpadovani === jhodapp|dr is now known as jhodapp === veebers_ is now known as veebers [22:18] mterry, is there a command line way to start usc? [22:19] kdub, not really, it heavily relies on its two-way communication with lightdm [22:19] kdub, if you want to insert yourself [22:19] kdub, edit /usr/share/ubuntu-touch-session/usc-wrapper [22:20] mterry, alright, thanks [22:31] mterry, So, it seems we did similar work... [22:31] tedg, on volume/mute? bummer [22:31] mterry, I've got a bunch of stuff going into account service. [22:31] mterry, But not volume/mute [22:31] tedg, the source package? [22:31] mterry, But I set up nice proxies and stuff :-) [22:32] mterry, https://code.launchpad.net/~ted/indicator-sound/account-service-support/+merge/205891 [22:35] tedg, interesting. Shouldn't conflict with my stuff though [22:36] mterry, Yes, but the volume/mute should probably go into sound settings. [22:36] tedg, also, these policykit permissions keep being duplicated around. We might want to consolidate on com.ubuntu.AccountService.GreeterReadAny and GreeterModifyAny [22:36] tedg, settings? [22:36] +1 [22:36] mterry, The account service schema [22:37] tedg, like this? https://code.launchpad.net/~mterry/gsettings-ubuntu-touch-schemas/volume/+merge/209158 [22:38] tedg, that also can provide the generic polkit actions [22:38] mterry, It should go in the one installed by indicator-sound [22:38] It seems they're more about indicator-sound than touch. [22:38] tedg, why necessarily? That package above already contains lots of sound related settings [22:38] tedg, they are trying to remove the touch namespace wherever possible [22:39] tedg, they renamed the binaries, and will rename source [22:39] mterry, === added file 'accountsservice/com.ubuntu.AccountsService.Sound.xml' ? [22:40] tedg, that's because com.ubuntu.touch.AccountsService.Sound.xml with the other sound settings exists, but they haven't migrated them to non-touch namespaces yet [22:40] Who is setting those? [22:40] tedg, telephony-service and ubuntu-system-settings look at the other settings like ringtones and silentmode [22:41] tedg, it's true that indicator-sound is the only one that currently needs to look at volume/mute, but they seemed like system-y settings that indicator-sound was just an implementation detail of [22:42] tedg, but I'd be happy to move them over if you feel like that's where they ought to live [22:42] mterry, Well, I didn't realize that we had an entire repo of random settings… still deciding how I feel about that. [22:43] What I'd really like is that they're stored in ALSA and the greeter and user session both read from there. [22:43] That's probably a minor pipe dream right now. [22:43] tedg, does ALSA have the per-user permission structure to allow that? [22:43] I think that's what logind is setting the permissions of using the ACLs. [22:44] The stuff we're getting around with the audio group. [22:44] tedg, but for storing settings? And we'd need to be able to adjust settings when user isn't logged in by logind [22:44] We'll they're not settings. They're state. [22:45] logind thinks that the greeter is a login session :-) [22:45] tedg, fair. "but for storing state? And we'd need to be able to adjust state when user isn't logged in by logind" [22:45] Yes, it does. It's reading/writing that state to the audio chips. Really it's stored there. [22:46] tedg, how does that work for multiple users? [22:47] mterry, Well, you only have one set of speakers. So the audio chip doesn't really understand multiple users. Logind gives permission to modify the driver based on who as the active session. [22:48] tedg, I get that. But how does volume setting normally work in that environment? Like, if I have two users open, who sets the volume when I switch from one session to another? [22:48] mterry, I believe it's who ever is the active session. So when you switch, you trade control. [22:49] Fast user switching volume fight, go! [22:49] tedg, sure, but which component is changing the volume is my interest [22:49] tedg, because the correct volume to set must be saved somewhere for that user [22:50] mterry, I don't know how Pulse behaves there, but I'd hope it doesn't reset it when you change. You wouldn't want to take your laptop to a coffee shop, mute it, and then switch users to music blaring. [22:51] tedg, I thought volume setting was per-user today. Just not shared with greeter [22:51] mterry, It could be. Each user has a Pulse daemon. [22:52] mterry, I think the issue is that logind isn't giving access to lightdm? [22:52] tedg, ok, so theoretical ideal place for greeter to talk to would be pulse [22:52] tedg, but it doesn't have per-user outside-of-home support for accessing/setting state [22:52] Well, no, pulse is per-user. ALSA. [22:53] tedg, but then when they logged in, pulse would override that state, right? [22:53] either pulse holds the right user volume state or it doesn't [22:53] mterry, I'm not sure on that, but I don't think it should. [22:53] We really need diwic here for this. [22:53] Perhaps tomorrow morning would be a better time :-) [22:54] tedg, anyway, I restarted the design discussion, but the design has been and still is per-user volume state [22:54] tedg, hence keeping it in AS [22:55] mterry, Hmm, where is that discussion? [22:56] tedg, there was some in email, but mostly in long-standing bug 840777 [22:56] bug 840777 in pulseaudio (Ubuntu) "Muting sound indicator in Unity Greeter does not mute sound on login" [Undecided,Triaged] https://launchpad.net/bugs/840777 [22:56] * tedg subscribes [22:57] tedg, the bug description outlines what design (JohnLea) wants. I recently re-confirmed with him that it's correct. But mpt and laney disagree [23:01] We have a bunch of multi-user issues. Need to put design time there. [23:01] tedg, not too much time! [23:02] mterry, We wouldn't want to take away from redesign v10 of the notifications ;-) [23:04] tedg, :) I was more worried about 14.04 [23:04] mterry, We've got at least 6 weeks before final designs are due for that! [23:05] ;-) [23:05] :)