=== alecu_ is now known as alecu [05:20] Hi [06:33] hey folks, I just tried the "alleged" final ISO and tried running the desktop under Microsoft HyperV and it installs fine, but on login after you enter your credentials the background just stays there and nothing happens ... but my CPU usage on HyperV Manager says 10% [06:34] do you know if you guys tested this scenario? It used to work in beta2. [09:47] good morning desktopers [09:52] pitti, isn't bug #1572863 a feature? what language did you pick? I though didrocks changed seeded complete support for the main locales with the goal to get ride of the "incomplete" dialog on login [09:52] bug 1572863 in language-selector (Ubuntu) "no "incomplete language support" note in offline xenial install" [Undecided,New] https://launchpad.net/bugs/1572863 [09:53] changes* === kathrin__ is now known as Jemand [09:58] seb128: no, check-language-support still gives me two metric tons of dictionaries and stuff, I only have the langpacks installed [09:58] morning all [09:58] weird [09:58] hey willcooke [09:58] seb128: well, maybe that is by design now, but that's not what the ISO test case says [09:59] and I'd like to have German spell checking :) [09:59] pitti, hunspell-de-* are on the iso [09:59] as is hyphen-de [10:02] desktoppers - ISO test please: http://iso.qa.ubuntu.com/qatracker/milestones/359/builds [10:04] desktoppers - for desktop a quick smoke test. Also look at flavours which don't have as many test runs [10:05] willcooke: I done broked it already :( [10:05] http://people.canonical.com/~davmor2/desktop-screenshots/wifi-broken.png === Drac0 is now known as Guest31263 [10:06] noooo [10:06] driver issue? [10:06] or more wide than that? [10:06] davmor2, move out of the wood back to civilization? [10:07] joke aside can you apport report a bug against network-manager? [10:07] so we get logs&co === hikiko is now known as hikiko|ln [10:10] davmor2, is that reproducible? when did that start? [10:11] seb128: yeap I'm just looking into causes as it works in the oem installer part [10:11] davmor2, is that an issue in the live session or installed system? does your wifi card requires a firmware? [10:13] seb128: intel wifi no driver, it works in live session, it works in the oem installer part so you can download updates, but it doesn't work in the installed end user session on first boot, if you reboot it then does [10:14] seb128: hence diggind into it as there are so many different bits to enable and not enable it in and it is finding out if it is every install or what [10:14] it doesn't help that the connect to network page has suddenly disappeared from ubiquity either though [10:17] No need to go on about that bug every 2 minutes, it's known [10:30] darkxst, rather asking here [10:30] is the GNOME iso supposed to boot to a "get me started" with no installer reference? === hikiko|ln is now known as hikiko [10:31] seb128, it should boot to the ubiquity greeter screen [10:31] I typed ubiquity but it has no result [10:31] the live session? [10:31] first item [10:31] I usually boot to live and run ubiquity from there [10:31] it feels nicer to have a session to play with while installing [10:31] oh in the live session, ubuiqity is in the panel on left in overview [10:32] which is hidden by default [10:32] that's quite un-intuitive [10:32] but ok [10:32] imho you should not display the "how to get started" on the iso [10:33] it's more likely than people want to play with things or install [10:33] rather than watch introduction video and read help [10:34] searching for ubiquity (or install) should also work? [10:34] not in the tutorial app which is maximized [10:34] like my default reflexe was to try to use what pop in front of my screen [10:34] oh I see, right [10:37] interesting the GCI kids all loved the tutorial app! it probably doesnt need to be on the live session though [10:37] or maybe would be better if it wasnt maximised? [10:39] I wonder if we could do a simple popup/notification "Press key to get started" [10:41] darkxst, unsure, but closing/not maximizing doesn't help either since your desktop is empty [10:41] like there is nowhere obvious on screen that you can hit the activity corner to get the dash [10:42] so if you are not a GNOME user you get to sit down there watching tutorial or an empty desktop [10:42] oh well, I guess most of users are GNOME users and now where to go ;-) [10:45] well there is a slight learning curve compared to a traditional DE [10:48] and yet even windows users just go and click on "Activities" not try and press which even opens the menu on windows (maybe they don't know that either though!) [11:20] darkxst, good work, Ubuntu GNOME looks good ;-) [11:20] just did an i386 installation [11:20] worked great [11:21] seb128, thanks, should be a very solid release [11:22] mark them as ready! [11:22] Laney, ok, I just finished smoketesting the last images, look good here also [11:22] :3 [12:10] mdeslaur, lol, I didn't see that this trusty samba update was 4.1 to 4.3 serie, fun one! [12:10] seb128: yeah, the fix was over 400 commits, no hope to backport it into 4.1 [12:10] urg [12:10] everyone else updates to either 4.2 or 4.3 too [12:10] s/updates/updated/ [12:10] that's quite annoying what happens [12:11] its means security update disrupt your working infra [12:11] but I guess no choice there... [12:11] yeah, we didn't really have a choice [12:12] this is the perfect example of why we usually backport fixes instead of upgrading versions [12:13] right [12:14] especially on something complex/interacting with other systems like samba === francisco is now known as Guest15300 [13:07] whats with unity mouse acceleration? i dont know if i should open up a bug report or not [13:08] but disabling mouse acceleration in the mouse setting should disable the mouse acceleration [13:08] i shouldnt have to use xinput [13:08] xinput --set-prop 8 "Device Accel Velocity Scaling" 1 [13:13] and why does launchpad no longer allow you to report a bug ;( [13:56] Trevinho smells [13:56] mh [13:57] Trevinho, sorry, I made the mistake to let my laptop unlocked [13:58] somebody used the opportunity [13:58] seb128: yess, yesss...... [13:58] :) [13:59] :D [14:01] cyphermox, did you look at the missing "connect to wifi" page in ubiquity? [14:01] started looking yeah [14:01] you still have nm-applet, and it's too late for fix now. [14:01] still looking? did you figure you somebody [14:02] Laney is starting as well and just said he was sure he would duplicate work with somebody [14:02] seems like he's right :p [14:02] yeah [14:02] just don't want to dup too much work [14:02] how much did you look? [14:02] should we stop? [14:02] no, please keep at it [14:02] k [14:03] I only just got up minutes ago, and jumped right in a meeting [14:03] k [14:04] all I did was yesterday try to add printfs and such to find out if it was being started at all, but I didn't see any hints [14:04] I don't know why it's unhappy [14:21] cyphermox: I found the problem [14:21] stand by [14:43] happyaron, e.u.c daily xenial view has some youker-assistant reports about import error due to missing lxml [14:43] is anyone looking at that? [14:50] pitti: do you know things about dbus-python? [14:50] Laney: that's a trick question :) [14:50] The wifi bug is because we ask for a property called 'state' [14:50] but it's actually called 'State' [14:50] apparently dbus-python handled uppercasing in trusty but not in xenial [14:50] ??????????????????? [14:50] oh, did that change between 1.0 and 1.2? [14:50] oh, wow [14:50] I'm asking if it is known and/or deliberate :) [14:51] not to me, at least; so far whenever I used properties I just spelled them right [14:51] haha [14:51] so, single-character fix? [14:52] yep [14:52] well [14:52] I'm worried that it might be a problem elsewhere [14:52] * seb128 saw a wifi screen on Laney's screen [14:52] well done -- can't possibly get smaller for an SRU :) [14:52] Laney: definitely sounds like it could be a problem elsewhere [14:52] Laney: nice catch [14:52] this instance is obviously trivial to fix [14:52] but if other things expected the old behaviour... [14:53] I don't see any in ubiquity though [14:53] do we know, was this still working in wily? [14:54] ah, nevermind [14:54] dbus-python | 1.2.0-2build2 | trusty | source [14:54] dbus-python | 1.2.0-2build2 | vivid | source [14:54] dbus-python | 1.2.0-2build3 | wily | source [14:54] dbus-python | 1.2.0-3 | xenial | source [14:54] haha [14:54] it's not like this actually changed much [14:54] * Laney didn't look there [14:54] did we switch py2->3? [14:54] and http://launchpadlibrarian.net/234773210/dbus-python_1.2.0-2build4_1.2.0-3.diff.gz has no code changes at all [14:55] hmm, nope, I was reproducing on python3 [14:55] umm, not reproducing [14:55] dbus change? [14:56] more probable, yes [14:56] nothing obvious in the commits log or bugzilla* [14:56] but that means it would affect dbus-glib, gdbus etc. too [14:56] Laney: did that affect wily already? [14:56] I suppose not [14:56] I think not [14:56] yes [14:56] then we at least don't look at the 1.6 -> 1.10 changes, only at 1.10.0 to 1.10.6 [14:56] it does? [14:56] I used d-feet and it works [14:57] yes to my previous question [14:57] $ busctl introspect org.freedesktop.NetworkManager /org/freedesktop/NetworkManager | grep State [14:58] .State property u 70 emits-change [14:58] that says it's "State", not "state" ?? [14:58] State is the right name [14:58] oh, what you said, sorry [14:58] but on old dbus you could do Get(iface, 'state') [14:58] and it worked [14:58] * Laney goes to ask smcv [14:58] Laney: that should be simple enough to test with dbus-send on a trusty VM? [14:59] that's what I just did [14:59] I'll just fix ubiquity, but want to confirm that this was intentional [15:06] cyphermox: can haz bug number? [15:07] not sure, too many tabs [15:07] Bug 1572793 [15:07] bug 1572793 in ubiquity (Ubuntu) "Ubiquity no longer prompts to join available WiFi networks" [Undecided,Confirmed] https://launchpad.net/bugs/1572793 [15:08] ta [15:08] if we all think very hard, can we flip that one bit on our released image? [15:09] I can go and sed them [15:09] pitti: you don't want to respin? :) [15:09] don't worry [15:09] Laney: live binary sed at download-time, everywhere [15:09] oh, the NSA could help there. [15:10] cyphermox: no, just direct some strategic solar radiation on the world's mirrors :) [15:10] you mean bit-flipping using a magnifying glass? [15:10] whatever helps :) [15:11] I thought that's what "stategic solar radiation" meant, at least if you apply the same logic to ants. [15:11] strategically aimed/placed at that bit, I meant [15:12] anyway, 16.04.1 it is [15:15] yep [15:18] charles, tedg, did any of you had a chance to look at the geoclue issue? [15:18] * tedg has not [15:19] :-( [15:22] cyphermox: for your reviewing pleasure [15:22] SHIP IT! [15:22] :) [15:22] might be a good idea for someone other than me to test [15:22] yep I will in a second [15:23] i. e. this is boot life session, change that s to an S (or install -proposed package), run ubiquity, right? [15:23] sure [15:23] not that there is any package yet [15:24] right, but that change is simple to apply inline on a running system [15:24] yep, just saying so that you don't go looking [15:27] davmor2: ok, that other bug is as I said, in the oem-config, after we get back to a "real" session, wpasupplicant has been killed. [15:27] no supplicant = no wifi [15:29] cyphermox: \o/ [15:31] cyphermox: so that is a fixable thing right? [15:31] davmor2: yeah, given enough eyes, all bugs are shallow [15:31] given enough hammers, anyone can fix systemd units. [15:32] cyphermox: I have hammers it how I break things :) [15:32] (needs really big mallets, though) [15:32] but I think maybe I'd just make things reboot rather than trying to get to a session straight out of the oem wizards [15:33] perhaps all we need is for wpasupplicant to ship a unit, so that things don't get killed off if it's running already [15:33] oh, except it does [15:34] and the busname looks wrong. [15:35] I have fi.epitest.hostap.WPASupplicant and fi.w1.wpa_supplicant1 [15:35] [Service] [15:35] Type=dbus [15:35] BusName=fi.epitest.hostap.WPASupplicant [15:35] ExecStart=/sbin/wpa_supplicant -u -s -O /run/wpa_supplicant [15:36] it should be fi.w1 ^ [15:36] where do you have both pitti? [15:36] /usr/share/dbus-1/system-services/fi.epitest.hostap.WPASupplicant.service and /usr/share/dbus-1/system-services/fi.w1.wpa_supplicant1.service [15:36] right [15:36] and tab-completing "busctl introspect fi." [15:37] I'm looking at wpa_supplicant.service under /lib/systemd/system [15:37] right [15:37] that name does exist, but there's another one, not sure which is the "right" one [15:38] https://blog.mozilla.org/futurereleases/2016/04/21/firefox-default-browser-for-linux-users-ubuntu-new-snap-format-coming-soon/ [15:38] fi.w1... seems to me like the right one. that's where we talk to wpa from NM, in any case [15:39] cyphermox: nice catch; then it should (re)start on demand again [15:39] so is /usr/share/dbus-1/system-services/fi.epitest.hostap.WPASupplicant.service entirely obsolete then? [15:40] maybe, lemme see if it's not really used only for hostap [15:40] oh, or maybe it actually does listen to both names [15:40] the old name for backwards compat reasons [15:41] yep [15:42] for some reason upstream only ships the old name as a systemd unit?! [15:42] seb128: So apparently we want to SRU unity8-policy-kit, which is a new package and on CI-Train. I don't know how all those work together. [15:43] seb128: I don't think I can make an SRU bug until the package exists, no? [15:45] tedg, you can open a bug against "ubuntu" and reference it to the changelog, the important bit is to have a page for the testing/tracking, no to have the right component/autoclosing on upload [15:46] seb128: Okay, will do [15:46] Thanks! [15:46] tedg, yw! [15:47] tedg, what was the review url you pinged me about the other day? [15:47] I can have a look now [15:47] seb128: Cool, it is the packaging for this package: https://code.launchpad.net/~ted/unity8-policy-kit/initial-work/+merge/291371 [15:48] tedg, charles, any chance I can one of you to look at this geoclue issue in exchange? :p [15:49] seb128: Heh, probably not today as we're both doing time critical things today, but I should be free tomorrow. [15:49] sounds like a good friday afternoon thing ;-) [15:49] thanks! [15:50] Oh, now that we have a VPN I wonder if I can test being in the UK.... [15:50] pitti: this didn't look to have worked :/ [15:50] * tedg starts substituting all his 'z' and 's's [15:50] seb128, as tedg said, today doesn't look good [15:50] charles, I think you are in Prague next week? I can nag you in person there :p [15:51] seb128: Heh, don't let him change his timezone without a fix! [15:54] tedg, right! [15:57] tedg, " This package provides a service for the Pay Service" ?! [15:57] tedg, did you copy/paste the description from the pay-service and didn't change it? [15:57] Oops, copy paste :-) [15:58] tedg, also policykit / policykit-gnome ... can you do "unity8-policykit" for consistency? (without the "-") === ahayzen_ is now known as ahayzen [15:59] (I guess it's a bit more annoying because it would mean renaming the launchpad project) [16:00] Yeah, though we could make the deb and the LP project be different. [16:00] would be confusing as well [16:00] like users would try to go from the package to the launchpad page using the name and get a not found [16:01] Heh, I'd like to see a user figure out there's an issue with the policykit agent :-) [16:01] tedg, copyright should be updated to -2016 (nitpick) [16:01] More worried about Ubuntu Devs [16:02] right, those are your users as well :p [16:06] Heh [16:06] seb128: Copyright and description updated. What do you want to do with the package name? [16:06] We're gonna have to rebuild, so we could change the LP project as well. [16:09] willcooke, pitti, cyphermox: do we know what is happening with this now? [16:09] tedg, if you rename can you make it policykit-unity8? consistent with -gnome/-kde/... [16:09] tedg, but yeah renaming is better if that's not too much work [16:09] davmor2, "this"? [16:09] davmor2: what's "this" ? [16:09] willcooke, cyphermox: the wifi issues [16:10] davmor2, yes! [16:10] in part, yeah [16:10] my idea for wpasupplicant doesn't seem to work though [16:10] seb128: K, I'll just make a new LP project and kill the old one. [16:10] no don't say that [16:10] tedg, thanks [16:11] tedg, the packaging/license/etc looks fine otherwise [16:11] seb128: Cool, thanks! [16:11] yw! [16:22] congratulations desktoppers [16:23] Yay! [16:24] * Trevinho takes drinks out from the fridge [16:25] \o/ [16:25] thanks! [16:25] well done team! [16:25] and other teams as well ;-) [16:28] \o/ [16:28] * pitti goes and grabs a xerus ice cream [16:29] ... for purely medical reasons, of course! [16:32] congrats to the successful release! :) [16:32] and I must say, GNOME Software / Ubuntu Software is working well here :) [16:35] it is quite nice [16:49] seb128: So we have the branch to add a libpam-cgfs dep for UAL so that it gets pulled in for unity8, but that then fails on vivid because there is no libpam-cgfs. Is there a way to write that dep so that it'll work on vivid and xenial? === ahayzen_ is now known as ahayzen [17:05] \o/ === vrruiz_ is now known as rvr [17:21] out [17:47] in ou tin === JoseeAntonioR is now known as jose === dpm is now known as dpm-afk === muktupavels is now known as muktupavels_ [19:36] attente: when you're about - https://bugs.launchpad.net/ubuntu/+source/gnome-software/+bug/1551707/comments/3 [19:36] Launchpad bug 1551707 in gnome-software (Ubuntu) "Installing local debs gives an error dialog" [High,Fix released] [19:36] if it is in fact different I'll report it - unless it gets reported in the meantime [19:44] flocculant: hey, i'm not sure what's going on there, i just tried it with some other debs, but maybe there's some deps missing that would cause the error message... [19:45] attente: could be - if I try with gdebi - it wants to get 2 dependencies [19:46] flocculant: ok, yeah. so i'm not sure what to do there. on the one hand it sounds like erroring out is the right thing to do if we can't install the deb due to missing deps, on the other, maybe we could be better about looking for those deps [19:47] I'd have to assume that the second would be the way to go :) [19:47] but no coder here ;) [19:49] attente: mmm - so I installed the deps that gdebi needed, but it still fails with software [19:50] flocculant: oh... really? [19:50] yea [19:51] just going to look in a clean install [19:51] flocculant: ok, i'll look into it [19:51] flocculant: not sure that will change much [19:52] well - this install has been kicked about all over the place for the last 2 or 3 months - so I prefer to check :) [19:53] flocculant: sure, be my guest :) i'm not expecting anything different though... [20:00] attente: well the issue is 'slightly' different on that - doesn't install, but don't get the error popup [20:01] flocculant: it actually doesn't install? or is it just the button that isn't updating properly? [20:01] that second one is a known issue [20:01] do you see the progress bar at all for installation on the button? [20:02] attente: it doesn't install - I saw the progress bar - which I would think finished much too quickly to have been the install [20:03] ok. did it pop up a dialog asking you to authenticate with user password? [20:03] no [20:03] flocculant: can you link to the deb? i want to try that specific one [20:04] https://www.google.com/chrome/browser/desktop/ 64 bit [20:05] locally I needed to install libappindicator1 and libindicator7 to allow gdebi to install it [20:06] attente: also I'm using xubuntu - not that it should make any difference [20:07] flocculant: yeah, i can reproduce with that deb [20:07] attente: ok cheers :) [20:07] attente: you need anything from me at all now? [20:08] flocculant: nope. i see the error message too :) [20:08] GDBus.Error:org.debian.apt.TransactionFailed: error-unreadable-package-file: [Invalid UTF-8] [20:08] okey doke - I shall wander off into the void again then :) [20:08] thanks for looking [20:08] flocculant: nooo not the void! :P [20:09] flocculant: thanks for letting me know :) === ralsina_ is now known as ralsina [20:09] welcome :) [20:10] nice to actually have some idea of who to tell to be honest - this is a rather useful channel for a flavour's qa to idle in :) [20:11] hehe, indeed! [22:27] Holy crap! Just discovered that systemctl can use policykit so you don't have to run it with sudo. Wow thats awesome! :) [22:28] Although you do have to autnehticate every time. [22:28] Still, useful to know. [22:53] * Laney can't remember if polkit has caching [22:54] but I thought that the gnome agent did [22:54] could be just making that up though [22:57] Seems not for systemctl at least. [22:57] The GNOME agent may, but this was under Unity. [22:58] that's the gnome one [22:58] Just checked, doesn't here [22:58] Ah ok. [22:59] still, cool of systemd to use that [23:00] Indeed. [23:01] I discovered it totally by accident, calling systemctl to work with a service without using sudo.