[01:55] robert_ancell: Is there a bug open for the black notification box on login screen in wily? [01:55] Seems to be an alpha-related issue specific to lightdm(?) [01:56] duflu, I haven't seen a bug [01:57] duflu, I guess notify-osd might have started doing something differently? [01:58] robert_ancell: Maybe. So do a log a bug against just notify-osd? lightdm I guess isn't doing any active compositing.. [01:58] duflu, yeah, I'd file against notify-osd + unity-greeter [01:58] I doubt lightdm would be having any influence === duflu_ is now known as duflu [07:49] good morning desktopers [07:53] morning seb128 [07:53] hey willcooke [07:57] morning willcooke! [08:02] hey hey [08:02] it's friday! [08:02] I hope [08:02] * Laney checks [08:02] phew [08:04] hi Laney! [08:04] hey Laney [08:05] happy friday! [08:05] Laney: don't have the weekday shown in i-datetime? [08:05] * seb128 is having another of those week if he's unsure he's happy it's already friday [08:07] it means you get another chance next week :P [08:08] lol, yeah [08:08] Laney, do you plans for the w.e? [08:09] how was the murder mustery w.e btw? [08:12] no not really, home alone this weekend so i'll probably end up doing not much [08:12] maybe go for a ride or something [08:12] murder mystery was good [08:12] the plot was seriously complicated [08:13] we didn't come very close to getting to the truth [08:14] how are those handled? [08:14] who was organizing? [08:15] * seb128 read about the principle, unsure how it works in practice [08:16] some company writes the story and then hires the place to stage it [08:16] there were like 7 actors there all weekend who were in character [08:16] and you had to talk to them to try to find out what happened [08:16] there's a police evidence room with cryptic clues too [08:17] but it turned out to be massively more complex than we were expecting [08:17] we thought it would be like "oh, this person killed that person because they want their money" [08:17] but it was actually a big conspiracy [08:17] #fail [08:18] haha [08:18] sounds fun still :-) [08:18] that's really cool [08:20] yeah it was [08:22] she said the next one would be on a james bond theme ;-) [08:22] colocate with a sprint? [08:22] :) [08:22] yes! [08:23] haha [08:23] speaking of which - are we planning one alreadt/ [08:23] *already? [08:23] it was a hotel not unlike ones we normally go to [08:23] so how does it work in practice? you spend your w.e at the location? do you have other activities/lunch/diner organized? [08:23] meals are included and some games / party [08:24] that's when the main events happen [08:24] cool [09:05] pitti: do you know how far back britney/adt-britney looks to consider regression vs always failed? [09:05] https://jenkins.qa.ubuntu.com/view/Wily/view/AutoPkgTest/job/wily-adt-kde4libs/ is considered a regression [09:06] Laney: if it ever succeeded in the current release it's considered a regression [09:07] Laney: jenkins apparently dropped runs 1-20 already [09:07] Laney: it's in results.history on snakefruit [09:11] Laney: but given how long it has failed and how uninterested the Kubuntu guys seem to be to actually land their stuff, a force-badtest is probably in order? [09:13] that might let in the thing that actually broke kde4libs, maybe a skiptest though? [09:14] yeah, for non-K* packages blocked by it? [09:14] although actually it only holds back pcre3 [09:14] weird [09:15] Laney: I already overrode results.history for a few unrealted packages which were held back by it [09:16] I wonder if there's a way to do this differently [09:17] Like only run tests using the packages we are trying to migrate [09:19] yeah, we'd need smaller temporary release "bubbles" with sets of pacakges instaed of just one giant -proposed pool [09:19] but then you need to mark them somehow which you want to release as a group [09:20] we talked about this briefly in #u-devel yesterday [09:20] Such as doing it after the _output stage [09:20] Then it feeds back into the excuses stage for the next run [09:20] or something [09:20] we could try to run the test twice, once against "testing" (wily") and once against "unstable" (-proposed) in britney, [09:20] right, infinity mentioned that too [09:22] So if I upload a glib2.0 which regresses libsoup then next time I could still upload a new libsoup which fixes an independent bug and still have it go in [09:22] because the second run would not take the new glib when it runs the new libsoup's tests [09:23] but I could be fixing a genuine bug in libsoup that was exposed by the new glib so you need a way to figure that out and still let glib in === vrruiz_ is now known as rvr [09:24] maybe try (new libsoup+old glib) and (new libsoup+new glib) in this second run [09:24] because we stored that glib broke libsoup [09:25] * Laney flails :) [09:26] pitti: anyway, can I hack results.history here? just s/FAIL/PASS/? [09:27] Laney: proposed-migration/autopkgtest/data/adt/wily-proposed/amd64/work/results.history [09:27] Laney: but only for the "right" triggering package [09:27] yep [09:27] ^kde4libs.*yourpackage [09:27] kde4libs 4:4.14.6-4~ubuntu4 FAIL pcre3 2:8.35-7ubuntu1 [09:27] I want a hint type for this [09:27] e. g. that [09:28] yeah, we need a "ignore failed test of foo for package bar [09:28] the results.history hacking will soon stop working when we switch to the cloudy stuff [09:30] bbl, running [09:30] zombie horde? [09:31] * larsu looks around [09:31] can they climb stairs? [10:59] is archive.ubuntu.com slow for anyone else or just me? [11:04] 3.209 kB/s [11:04] can't complain [11:04] oh, a new LibO [11:05] nice apt-get download benchmark :) → 2.482 kB/s [11:08] it seems ok again (I'm at 1,390 kB/s) [11:08] woah. look at that locale! [11:09] ah, that's a '000 separator, not a decimal point of course [11:09] color me happy [11:11] it was slow for me yesterday, but ok now [11:13] larsu: ah yes, German here :) [11:13] ya, I figured :) === MacSlow is now known as MacSlow|lunch [12:22] ah, GNOME [12:22] let's force-enable the touchpad when it's the only pointing device present, becasue otherwise people might get confused [12:22] result: people get confused because their enable/disable touchpad key doesn't work anymore [12:23] * larsu thinks we should just kick that logic out [12:23] +1 [12:23] I was pondering suggesting that on the bug as well [12:23] when someone disables their touchpad, it shouldn't magically be reenabled just because they plugged/unplugged a mouse [12:23] just ship a mouse with the image and force-disable the touchpad ;) [12:24] haha [12:27] seb128: maybe we should get some input (haha, "input") from mpt? [12:27] but this is fairly clear-cut I think [12:28] larsu, yeah, he's off in vac for the next week though [12:28] maybe suggest on the upstream bug to disable the logic meanwhile? [12:28] and wait a bit to see if they comment [12:28] this was an oem issue you said? [12:29] I'm in favor of just throwing it out in u-s-d [12:29] the code upstream is different now anyways (I'll still comment, though) [12:31] it's an oem issue yes [12:32] larsu, did you see https://bugs.launchpad.net/ubuntu/+source/unity-settings-daemon/+bug/1454950/comments/9 [12:32] Launchpad bug 1454950 in unity-settings-daemon (Ubuntu) "It needs to press touchpad hotkey twice to re-enable touchpad" [High,Confirmed] [12:32] larsu, so yeah, code change but it seems they still have the issue [12:33] seb128: I did, yes. The logic is still the same - just saying the same patch wouldn't apply [12:33] right [12:38] i disagree, btw [12:38] touchpad being reenabled makes sense [12:38] for the same reason as it makes sense to reenable the internal display if you turned it off when you connected an external monitor [12:39] for keyboard-centric folks like us it may seem silly, but to a newb, a computer with no pointing device is utterly unusable [12:39] and maybe they would want to have used the trackpad to navigate to the UI to... reenable the trackpad [12:39] which presents an annoying catch-22 [12:40] so what do you do with people that disable the trackpad from their keyboard? [12:40] however... it should probably do the same thing as the monitors panel in another way too: it should remember that the last time i plugged in a certain monitor i disabled the internal device [12:40] because this is what's causing the issue right now [12:40] and it should do the same for the trackpad [12:41] since that would be really useful: my computer learns that when i connect my mouse it ought to disable the trackpad just as it knows to automatically turn off the internal screen when i connect my monitor [12:41] that's the real bug, i think [12:41] it's annyoing if unplugging/replugging the mouse results in a state change (enabling of the trackpad) which i guess it now does? [12:41] yes [12:41] so file the real bug [12:42] it is filed... [12:42] wait, which one do you mean? [12:42] about re-disabling the trackpad via memory? [12:42] because that's the real bug [12:42] ah. That's what you think. I disagree [12:42] enabling it when it's the lack pointing device is pretty logical [12:42] *last [12:42] the bug we have right now is that it's impossible to disable the trackpad [12:42] I should be able to disable the trackpad if I want [12:42] if it's the only thing left [12:43] which people tend to want to do... [12:43] what's the logic there? [12:43] it's annoying you when you're doing data entry? [12:43] I have a laptop with a trackpad and I accidentally touch it while typing?! [12:43] turn on the typing detection? [12:43] GNOME removes the switch in g-c-c if it's the last thing [12:43] i mean... i hate trackpads as much as the next person, and for all of the same reasons [12:44] and it breaks XF86TouchpadToggle [12:44] (nt completely though, becasue that's buggy as well) [12:44] (have to hit it twice to disable/enable or something) [12:44] but trackpad-shittiness-mitigation is working fairly well thesse days [12:44] not all people have the newest laptops [12:44] so you think that button on their keyboard should just do ... nothing? [12:44] syndaemon is pretty universal [12:45] delays are different between users [12:45] my 'change monitor config' button also does nothing if my internal display is the only one [12:45] i guess you would expect that this should disable my internal screen? [12:45] maybe i'm just listening to music... [12:45] the screen key is "cycle between config" [12:45] not "turn the screen off" [12:45] so it does what it's meant to do [12:45] cycle between the valid config [12:45] there is only 1 in that case [12:45] the touchpad key is "toggle touchpad state" [12:46] and it doesn't do that [12:46] we do disallow turning the screen off when it's the last one... [12:46] i think we're splitting hairs :) [12:46] so desrt has a point there [12:46] there is a different [12:46] I still don't think it's a good one... [12:46] you can't use a computer without a screen [12:46] you can use without a pointing device [12:46] seb128: listening to music... [12:46] and again: for most humans, that's not true [12:46] a pointer is really essential [12:46] and there is a point that both of you are failing to gasp: accidental settings happen [12:46] and they have a key [12:47] on their KEYBOARD [12:47] also: intentional settings that are later forgotten [12:47] to reenable it [12:47] this is a real thing and it's a serious problem [12:47] desrt, you can't really start the music without a screen to see what to select [12:47] that's generally why we try to avoid "[x] break my computer" settings unless the breakage is always immediately obvious and reversable [12:47] desrt: "You've turned off the last pointing device. Hit ESC to turn it back on" [12:47] because someone will be poking around some panel, and they'll set it, then forget that they ever did it [12:47] and later their computer will not be working [12:47] [ ] don't show this message again [12:48] well, a setting is different from a keyboard key [12:48] and in particular, they're very likely not to notice it if they did it when they had a mouse in [12:48] since they're not using the trackpad [12:48] maybe they pressed the key by mistake... [12:48] let's disable the "disconnect from wifi" key as well [12:48] you get a notification telling you what you did [12:48] larsu: hardware wifi switches are a serious usability issue as well [12:48] also the mute key [12:48] larsu: which is why there is UI dedicated to 'disabled in hardware' in GNOME now [12:49] what does that do? [12:49] it tells you if your hardware killswitch is on [12:49] oh, just tell you that it's disabled from the keyboard? [12:49] also: people are used to selecting wifi networks [12:49] so it gives a natural place to put that in some UI [12:50] in any case, I see your point, but I'm not convinced. Turning off the trackpad is a legitimate feature that people use [12:50] but one personal point: my mother accidentally kill-switched her laptop and she had to call me to tell her how to fix it [12:50] and it even took me about 10 minutes before i figured it out (over the phone) [12:50] hopefully she won't have to do that when we;ve all switched to voip [12:51] (as a user she is not used to selecting wifi networks because she only uses the computer at home and my sister set it up for her and she never touched that part of the UI again) [12:51] I guess this is the real solution: https://support.apple.com/en-us/HT201933 [12:51] ya. that was my argument [12:52] remember the user's preference for disabling the trackpad when an external device is connected [12:52] it's a different setting though [12:52] brb [12:53] anyway... trackpad-hate is semi-legit, so i guess i understand the gripe [12:53] but syndaemon is working nicely these days... [12:53] and that works on more or less any laptop made in the last 15 years... [12:57] desrt, that's assuming that the issue is click while typing [12:57] I'm more often annoying by the "clicks while paused/thinking" [12:58] annoyed* [12:58] * desrt never has this issue [12:58] you probably don't stop with the hands on the keyboard/in position [12:58] * desrt has only 'difficult to accurately position' and 'accidentally rightclick when trying to leftclick', but that has more to do with lack of proper physical buttons [12:59] well, in any case discussing personal experiences is not really relevant there [12:59] I think it doesn't matter whether syndaemon is nice enough. If there's a button on the keyboard, it should work [12:59] right [12:59] maybe [13:00] it should do something at least [13:00] ya [13:00] but imho, disconnecting an external device should always force the setting back to 'enabled' [13:00] maybe open a dialog that you need to click to confirm [13:00] if you want to re-disable it, you can [13:00] NO DIALOG [13:00] don't make me hurt you [13:00] you are the one asking for that" [13:00] no [13:00] "don't let the key do the action it's made to do" [13:00] i am asking that disconnecting the mouse reenables the trackpad [13:00] possibly with the usual visual notification that accompanies that [13:01] * larsu wonders why the current code doesn't even get the "feature" right [13:01] desrt, yeah, that's an orthogonal behaviour/request [13:01] ideally it would also remember to re-disable the trackpad next time the mouse goes in (again, with the correct notification) [13:01] we're so overthinking this [13:01] but ya... if the mouse is out and you explicitly press the 'disable' key, it should disable i guess [13:02] in any case it shouldn't get in a state where the gsettings key value doesn't match the actual valye [13:02] disagree :D [13:03] lol [13:03] although that raises an interesting other angle that i failed to think about before :) [13:04] i now have to consider if adding/removing hardware counts as 'user interaction' as per my customary rule :) [13:05] meh [13:05] there are a lot of fine points here, but there is really only one thing i consider to be absolutely mandatory in any solution: disabling the trackpad while a mouse is connected and later disconnecting that mouse should result in the trackpad being reenabled [13:05] the other details are less important [13:06] right [13:06] the existing solution meets that criteria (and i guess that this point above is the reason for the solution being how it is) [13:06] but clearly there is potential room for improvement [13:07] just don't throw the baby out with the bathwater :) [13:07] well, not only improvements, there is a bug to fix there [13:08] having a key that doesn't do anything on first press and act when you do it again in buggy [13:08] you can look at it whatever way you want, there is no rational way to describe that as a wanted behaviour [13:10] well [13:10] the bug doesn't impact me [13:10] no trackpad-disable key here :D [13:10] :-) [13:22] desrt: we shouldn't call the ui "disable" then [13:22] but "disable when mouse is present" [13:24] i'm not even sure that explicit UI like that is required [13:24] but ya... maybe it could help [13:25] i really like the implicit approach taken by the displays panel [13:25] hm? it works the same way, no? [13:25] no [13:25] it grays out the switch when only one monitor is present for me [13:25] it remembers the serial numbers from displays that you connect [13:25] and writes them into monitors.xml [13:25] DEL [13:25] DELL U3014 [13:25] P1V6N33K783L [13:25] oh, I thought you meant ui-wise [13:26] and then it saves configurations according to the precise combination of devices that were connected when you changed the settings [13:26] too magic [13:26] perfectly magic [13:26] i connect my stuff and it remembers what i did last time i had that device connected [13:26] for you, because you know how it works [13:26] and it does it again [13:26] it's perfect [13:27] i only know how it works because i wondered how it was possible that it always did what i wanted [13:27] so i investigatedd [13:27] it's confusing if you don't know about that detaul [13:27] *detail [13:27] i dunno [13:27] i think it's the epitome of 'just works' [13:29] and not really transferable to the input stuff [13:29] because there you can have more combinations I think [13:29] in practice, not really [13:29] your laptop probably only has one external connector [13:29] and really, this one case where you unplug your mouse after disabling your touchpad is really, really weird [13:30] desrt: bluetooth mice?! [13:30] although it is good enough to keep separate the "i connected my home monitor" from the "i connected a projector at a conference" cases [13:30] vendor/product ids? [13:30] I'm not saying it's impossible, I'm saying it's not desirable [13:31] and really out of scope :) [13:31] man [13:31] kiss kiss is so great [13:32] seb128: what's your takeaway from this conversation? [13:32] seb128: I'm still in favor of just removing the feature [13:32] let's see how many bugs we get... [13:33] larsu, so currently if you have a mouse connected, disable the touchpad, unplug the mouse, the touchpad is re-enabled? [13:33] yes === MacSlow|lunch is now known as MacSlow [13:33] would your suggest change undo that? [13:34] suggested* [13:34] yes [13:34] I think it's suboptimal [13:34] I wonder if we can fix this bug in another way [13:34] but I think we'd always end up with reenabling the touchpad [13:34] at least after logging in again [13:34] that's fine [13:34] but that hw button should be permanent, no? [13:34] if it's a login [13:35] as long as the key works for the active session [13:35] not sure when this function is running, let me see [13:35] it's not a switch control with a position on/off [13:35] it's a key [13:35] ok it runs on startup and when a device is removed [13:35] I would be fine if it works only for the active session [13:35] hum [13:35] oh [13:36] it is a switch in u-c-c [13:36] using the key "remove a device" [13:36] right? [13:36] what kind of key do you mean? [13:36] the keyboard hotkey [13:36] I'm just wondering why the function triggers [13:36] ah, right. The key is XF86TouchpadToggle [13:36] if you sau it does at startup and on device changes [13:37] we do neither of those here [13:37] the function triggers because when hitting that key while the touchpad is on, the device is removed [13:37] right [13:37] can we detect that special case somehow? [13:38] not easily [13:38] there's only this one gsettings key which gets set from the media-keys plugin [13:38] and the mouse plugin reacts to changes in the devices [13:39] do we know what's the device change? [13:39] hm, it doesn't call that function when only the key is changed [13:39] weird... [13:39] maybe this does work... [13:39] seb128: ya [13:40] can we tell a touchpad apart from a mouse? [13:40] yes [13:40] in which case we might want to re-enable only if !touchpad [13:40] when do you unplug a touchpad? [13:41] that would let mouse unplug still re-enable the pad [13:41] but the pad being removed not leading to it being added again [13:42] yeah this is a good idea [15:09] * Laney stares at happyaron [15:22] Sweet5hark, did you see my comment yesterday about the dmb? [15:23] seb128: yeah, saw it. I will try to look into it again. [15:23] Sweet5hark, thanks [15:36] cyphermox, hey, did you say you would backport https://git.gnome.org/browse/network-manager-applet/commit/?id=98dc7a7657b2609fcac05134db99455a9de6610a ? it seems it could fix the segfault we discussed before vivid which is a top of euc [15:37] yeah, I said I would, I just didn't get to it [15:37] is there a bug number so I can get this fixed right now? :) [15:38] got it [15:43] willcooke: https://skyfromme.wordpress.com/2015/07/10/going-mobile/ [15:44] Sweet5hark, niice! Thanks a lot [15:45] willcooke: kudos go to sverzegnassi for the qml stuff. I only wrote 30 lines of c++ ;) [15:45] :) your help is gratefully received [15:51] seb128: building in sbuild now. as soon as it's good I'll upload to wily, and do the SRU to vivid [15:51] cyphermox, thanks [15:55] desrt: thanks; and what do you think about the proposal Alberts posted today? [16:08] Laney, cyphermox, unsure how/why but seems gnome-common (main) depends on autoconf-archive (universe) in wily [16:08] or am I looking at something wrongly? [16:08] no, I'm getting to the same conclusion... [16:09] what's the problem? [16:09] oh wait [16:09] https://bugs.launchpad.net/ubuntu/+source/autoconf-archive/+bug/1154130 [16:09] Launchpad bug 1154130 in autoconf-archive (Ubuntu) "[MIR] autoconf-archive" [Undecided,Fix committed] [16:09] Laney, how did we end up in that situation with britney? [16:09] # apt-get install gnome-common [16:09] what situation? [16:09] gnome-common : Depends: autoconf-archive (>= 20150224) but it is not installable [16:10] britney doesn't know about main/universe [16:10] k, that explains :p [16:11] so it just need to be promoted, right? [16:11] yeah [16:12] done [16:26] come on murray!!!! [16:27] ok, that's quite enough of being a tennis watcher [16:35] Laney, heh [16:37] stressful [16:37] indeed [16:38] it touched the line! [16:38] bah, wrong line :p [16:39] come on andy! [16:40] bah [16:41] what a shot! [16:41] Laney, you stopped watching? [16:41] haha no [16:41]