=== ljp is now known as lpotter === vrruiz_ is now known as rvr [11:54] Saviq: hey! You mentioned that you guys were looking into the unity8 security-bugfix and checking if the deployed fix is enough - all good in regards to that? [11:54] No follow up fix needed? [12:12] sil2100: there's a follow up fix needed [12:12] tsdgeos: hmm... this won't make it for OTA-9 then, we'll have to get that into OTA-9.5 [12:12] sil2100: saviq is on usa still, visiting friends, not sure if he said he'd be working or not though [12:13] sil2100: yes, 9.5 is like "next week" anyway, right? [12:13] Yeah [12:13] ok [13:27] tsdgeos, hey, yeah, I'm working [13:28] oki [13:28] morning [13:41] Saviq: was wondering, the few last MRs failed on xenial jenkass [13:42] would it make sense like "wait 5 min and try again before failing" [13:42] if the apt-get step fails? [13:43] tsdgeos, assuming I'm getting a meaningful exit code out of it, I might do that, yeah [13:47] tsdgeos, better than that I'll try to switch to the internal archive cache, which should have less trouble like that [13:47] oki [14:01] Saviq: where's the autopkg test run in https://requests.ci-train.ubuntu.com/#/ticket/877 ? can't see it [14:01] tsdgeos, they only run when it's approved by the lander [14:01] ah [14:02] tsdgeos, and since it got rebuilt, they got cleared [14:02] * Saviq rebuilds [14:56] Saviq: does API this work for you? https://code.launchpad.net/~alan-griffiths/mir/add-mir_surface_spec_set_shell_chrome/+merge/283818 === maclin1 is now known as maclin [14:58] alan_g, looks good here, dednick, can you have a look ↑ [15:03] tsdgeos, added that test. was harder that I thought. but should make us more bulletproof for the future [15:04] mzanetti: cool, checking [15:04] alan_g, Saviq: is that dynamic? [15:05] IIUC, yes [15:05] Saviq: or doesnt it need to be? [15:05] it does need to be [15:05] i'm not familiar with the surface spec, but it doesnt look like it to me. [15:05] dednick: yes. It can be set on creation, or applied later [15:05] dednick, the second test is "apply..." [15:06] ah. ok. mir::options = cool. [15:06] then looks all good. [15:07] ahh. theres mir_surface_apply_spec. got it. [15:11] hey, is https://launchpad.net/bugs/1537782 a known issue? [15:11] Launchpad bug 1537782 in qtubuntu (Ubuntu) "Modifier ignored when pressing a key if TextInput has active focus" [Undecided,Confirmed] [15:14] dandrader, does that ring a bell ↑? [15:21] Saviq, yeah... it might be possible that the modifiers get lost in translation in the "mir server interface -> unity8 qml scnene -> mir server-client wire procotol -> qtubuntu" road [15:24] dandrader, think you could have a look? [15:26] Saviq, once I get back home preferably. no bluetooth kbd or test machine to play with [15:27] dandrader, ack [15:31] @unity: whoever is not jetlagged, standup :) [15:39] mzanetti, wrong irc server :P [15:39] ? [15:39] cimi, ^ [15:39] mzanetti, we usually ping for standup in canonical irc [15:40] ah... didn't know that was intentionally [15:40] mzanetti, was taking piss that you are probably jetlagged too :) [15:41] I kinda am, yes [15:41] feels like 4am, rather than 4pm right now [15:44] :D [15:44] mzanetti: is there a bug for https://code.launchpad.net/~mzanetti/unity8/edgebarrier-click-transparent/+merge/283735 ? [15:48] tsdgeos, don't think so, but it's easy to repro and quite annoying when typing messages [15:48] * mzanetti searches [15:48] mzanetti: i'm just trying to know what i need to do to test it :D [15:48] tsdgeos, type something with the OSK. the q key is quite nasty to press [15:49] depending on the layout, the a and shift key too [15:49] the ones that touch the left edge [15:50] tsdgeos, you'll notice the difference immediately if you try to press the osk very close to the edge. [15:50] tsdgeos, also have to ensure that the edge push to show spread & launcher are no affected [15:50] *not [15:50] if you just type a normal message, you'd notice the q key often just doesn't react [15:50] dandrader: you mean with the mouse? [15:51] tsdgeos, yes [15:51] k [16:13] mzanetti: dandrader: about the push to show launcher, can you guys confirm that if you exit the launcher form the unity icon it won't autohide? [16:16] tsdgeos, confirm if it happens or if it's the expected behavior? [16:16] dandrader: happens [16:16] i guess it's not the expected behaviour [16:17] but if it is, that also helps :D [16:17] tsdgeos, by clicking with a mouse? [16:18] tsdgeos, I don't think it should autohide while the mouse is still hovering over it.... [16:18] dandrader: "exit the launcher" [16:19] i have a typo [16:19] form -> from [16:19] tsdgeos, ah, so you mean the mouse if hovering over the dash icon and then you move it away from the launcher? [16:20] dandrader: yes [16:20] tsdgeos, I confirm it happens and that's a bug in trunk it seems [16:22] tsdgeos, doesn't happen all the time though... [16:23] dandrader: ok, so do i file a bug, correct? [16:24] tsdgeos, yes.to safely reproduce it you have to perform the edge push already at the y position where the dash button will show up [16:24] that may be, yes [16:25] hmm, haven't managed to hit it yet [16:25] tsdgeos, if you do the edge push from elsewhere, move the pointer to the dash button, then exit the launcher. it will still autohide [16:25] ok [16:25] so it's not hide from [16:25] ah now it did [16:25] but show and exit from [16:25] yeah, bug [16:25] tsdgeos, yes [16:25] nice catch :D [16:26] I hope we can catch this with a tst_Shell.qml test... [16:28] dandrader: mzanetti: https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1537817 makes sense? [16:28] Launchpad bug 1537817 in unity8 (Ubuntu) "Launcher autohide fails if opening and exiting the launcher via mouse push in the "unity" icon" [Undecided,New] [16:29] tsdgeos, yes [16:29] tsdgeos, it's also called BFB :) [16:29] but yeah, this works [16:29] and it's the ubuntu icon, not unity :D [16:29] someone was confused about the BFB terminology the other day [16:29] mzanetti, BFB? [16:29] big fat button [16:29] designers call it that way [16:30] mzanetti, was searching for BFB and all kinds of stuff showed up :) === alan_g is now known as alan_g|EOD === DanChapman is now known as DanChapman_ === DanChapman_ is now known as DanChapman [19:30] Hallo [19:32] mzanetti: Oh Fat yeah not the version I heard [19:35] Hello everybody. I am using Ubuntu 15.10 unity. Work areas are not working isolated from each other. I am writing this post to report it. Maybe the solution can be found. [19:35] work areas "workspace" :) [19:36] gnukarabatak, can you please file a bug by running Alt+F2, "apport-bug unity"? [19:37] or via https://bugs.launchpad.net/ubuntu/+source/unity/+filebug, but the former is preferable as it will add some details about your setup [19:38] thanks. [20:23] tedg, I'm debugging why my qtmir branch doesn't work on top of your app-object branch -- it looks like I'm no longer seeing callbacks from ubuntu_app_launch_observer_add_app_starting [20:24] tedg, you said to keep using the old callbacks, right? [20:24] mterry: Hmm, not sure of a good reason that would happen. Yeah, use the old callbacks. [20:24] (rather than the new-style object callbacks) [20:25] mterry: I've tried to keep it as mostly a wrapper on the old code to reduce risk... [20:25] tedg, I could still be doing something dumb... [20:25] mterry: Hmm, okay, you can start a ubuntu-app-watch to see the signals in the CLI. [20:26] mterry: That might help to see if they're getting setn. [20:26] sent [21:43] tedg, ok... so what would cause Application::info() to return a null ptr? Seems I'm getting that for "ubuntu-system-settings" [21:54] tedg, line 61 in application-impl-legacy? [21:54] .cpp [22:14] tedg, and line 60 needs an "applications" in it [22:17] mterry: Fixed and pushed, need to still write tests for that code. [22:18] mterry: Does U8 have tests for all the desktop keys? [22:18] * tedg will have to look [22:18] tedg, I don't think u8 does... qtmir might test some of them [22:18] tedg, but it would mock out u-a-l in that case [22:19] mterry: I mean, from before when it read the desktop file itself. [22:20] tedg, fwiw qtmir has some tests for its desktop file reader I think [22:27] tedg, oh yeah, qtmir had some tests yeah [22:27] tedg, now I'm seeing why click apps aren't launching either :-P but ubuntu-system-settings does! [22:27] * tedg prefers to steal rather than write [22:28] mterry: Heh, are you saying the real world is different than the one that exists inside my head? ;-) [22:30] tedg, who's to say we're not all figments of your imagination? But please tell your imagination to let click apps work [22:40] tedg, ... [22:41] tedg, so if I give parse() "com.ubuntu.camera_camera" what should I expect? [22:41] AppID::parse that is [22:42] mterry: I think you need the full appid today. [22:43] mterry: Yeah, it'll error. [22:43] tedg, hrm. So I have an appid input from an API, there's no easy way to get a proper AppID object... I don't know whether I have package or appname or version or what, so I can't use the discover API [22:45] So we need it to be smart enough to handle legacy AppIDs as well. [22:45] legacy/short/full [22:45] tedg, yeah I hoped parse() would parse what it could and use discover() behind the scenes if it wasn't enough [22:46] tedg, maybe that behavior should be a third call. But it would be a useful api call [22:46] Yeah, that's what I'm thinking. I can't think of a good name though. [22:46] "figure it out" [22:46] fromWTF [22:47] interpet? [22:48] applyHuristics() [22:49] tedg, an overloaded parse call? or another discover call? [22:49] Forgot the "e", knew that didn't look right. [22:49] Yeah, kinda thinking I want it a different name to just say "this might be doing more than you want" [22:50] stringSolver() [22:51] tedg, "discover" already implies that [22:51] tedg, "find" ? [22:52] Ah, I like find(), let's go with that. [22:52] tedg, also I don't think parse() does throw any errors [22:52] tedg, but if you try to use the result in other places, you might get errors yeah [22:52] mterry: No, it returns an empty() AppID. [22:53] tedg, only if the input is empty [22:54] mterry: Sure, otherwise it returns a legacy AppID. [22:54] tedg, otherwise it looks like it will have a valid package/appname but empty version [22:54] I think package is empty as well. [22:54] in the com.ubuntu.camera_camera case [22:54] tedg, that's in the legacy case [22:54] Yeah, so when you're going back to string it's not putting anymore '_' in. [22:54] tedg, but in any case, those aren't errors [22:55] We probably shouldn't let legacy appid's include a '_' — wonder if that'd break anything. [22:55] tedg, I'm not talking about legacy apps right now. com.ubuntu.camera_camera is a click id without a version [22:55] tedg, which AppID::parse correctly parses as an appId without a version [22:56] tedg, but it doesn't give an error. You implied you expected it to [22:57] Yeah, what it's doing is returning the tuple { '', 'com.ubuntu.camera_camera', '' } which is a legacy appid. [22:58] tedg, ah I see. weird behavior, yeah [23:04] mterry: Cool, I need to head out now, but I should get to this tomorrow.