[01:00] desrt, still around? [04:21] Laney: I tried yesterday's desktop-next image which seb128 promised me to work, and it still doesn't -- still the same "no autologin" bug [06:31] good morning! [06:38] hello larsu ^5s === qengho is now known as CardinalFang === CardinalFang is now known as qengho [06:59] hey larsu ;) [07:28] didrocks: bonjour! [07:28] * larsu is sitting in the sun \o/ [07:28] larsu, nice! [07:30] hi robert_ancell! [07:31] * didrocks relocates in the balcony, like yesterday :) [07:40] Trevinho, is unity8 session broken in utopic? It's crashing on login for me (left with blank screen) [07:41] (and are you and Andrea maintaining this in this timezone?) [08:05] hey [08:05] pitti: grump, well I'll look today [08:06] morning Laney [08:07] hey Laney, good morning [08:07] seb128: so, false promises! [08:08] * pitti hugs seb128 [08:12] promise was that we won't hear about football ever ever again? :) [08:12] didrocks: RESOLVED/WONTFIX [08:13] didrocks: actually, that yesterday's ubuntu-next image would actually boot :_ [08:13] :) [08:13] pitti: yeah, some people will almost set that to INVALID anyway :) [08:13] ahah [08:13] or rather "OPINION" :) [08:13] didrocks: or in LP terms, "Opinion" :) [08:13] :p [08:13] * pitti te donne une accolade [08:14] didrocks: tu dois attendre avec juli 17 et pas parler avec des autres gens :) [08:14] * didrocks donne une accolade en retour [08:14] pitti: exactement, on va s'enfermer dans une cave ! [08:15] ah ah [08:15] ^ note that I used the French chuckling :) [08:15] \o/ [08:17] didrocks: est-ce que julie aime le football ? [08:17] pitti: non plus, donc elle est plutôt ravie que je n'aime pas :) [08:18] * pitti invokes google translate for that [08:25] pitti, right, saw that, I was just reading his summary email [08:26] can't switch to a vt [08:26] BAH! [08:26] Laney, I think the user metrics thing has a bug since the revert [08:26] time isn't updating [08:26] switch to vt wfm [08:26] k [08:26] I think it's probably crashed [08:26] so maybe it's blocked [08:26] pitti, what's your issue with the iso? [08:26] seb128: still the same -- it boots to lightdm, then I don't know which user/password to log in as [08:26] pitti, is that on real hardware? [08:26] oh, a robert_ancell [08:27] seb128: yes, on my X230 [08:27] pitti, :-( [08:27] seb128: same machine I tried it on on Malta [08:27] pitti, can you share /var/log/lightdm? [08:27] pitti, what issue did you download? yesterday's one? [08:27] seb128: hang on (sitting in a train) [08:27] seb128: yes, yesterday's [08:28] pitti, ubuntu-desktop-next/ubuntu is login/password [08:28] seb128: I can put the image on an USB stick (didn't bring the one from yesterday) and boot it [08:28] gives me a unity8 lock screen [08:28] ah, thanks [08:28] I do not know why this would only work for some people [08:28] pitti, but if autologin fails it's likely Mir being unhappy so it's not going to help you [08:28] although the syslinux screen thing doesn't show up [08:28] right, that's an issue since syslinux 6 landed [08:28] but if I boot it in uefi mode I get to choose install/boot anyway [08:28] Colin said he would have a look [08:28] seb128: ah, you figure it has tried to start the session, but failed? [08:28] if you hit enter on the empty screen it boots [08:29] pitti, yes [08:29] seb128: on Malta this apparently was a real issue, that Laney fixed, wasn't it? [08:29] pitti, that's what I was having under virtualbox first, before I tried on real hardware [08:29] autologin wouldn't work before [08:29] pitti, right, it was not autologin in until yesterday [08:29] but that is fixed [08:29] but I think it does the same thing if the actual login fails [08:29] i.e. takes you back to lightdm [08:29] the thing is that failed session start gives you a similar experience [08:29] it bounces you back to the greeter [08:29] Laney, ;-) [08:30] could probably show a message in this case though [08:30] failed autologin [08:30] Laney, is that on your nvidia box? [08:30] the crash? [08:30] yes [08:30] k, maybe Mir doesn't like nvidia or something [08:30] Is this unity 8 session? [08:31] Laney, well, autologin doesn't fail, technically it starts the session, session fails to start and exit ... so from a lightdm perspective it's not different from a log-in and log-out cycle [08:31] robert_ancell_, yes [08:31] robert_ancell_, desktop-next iso [08:31] It's crashing for me in utopic, but the upstart never returns so lightdm thinks it's running fine [08:31] robert_ancell_, hey btw, how are you? ;-) [08:31] I haven't tried that iso yet [08:32] seb128: doesn't it exit with a bad code? [08:32] robert_ancell_, the iso from yesterday (should still be current) is working for me on the inspiron 11, I even did a successful install [08:32] Laney, I don't think lightdm watches for exit codes and display an error when the session exit with one [08:32] It's bugged me for a long time that lightdm doesn't really know anything about the session. I wonder if there is a better interface that just monitoring the session process for exit [08:32] robert_ancell_, what is crashing/do you have a log/bt? [08:32] I'm suggesting maybe it could ;-) [08:32] seb128, good, how are you [08:33] Laney, right [08:33] robert_ancell_, I'm good thanks ;-) [08:33] robert_ancell_: up late? [08:33] ok, trying to boot it again, collecting logs [08:43] Laney, seb128: back, I now have the entire /var/log [08:43] Laney, seb128: so, I see text-mode plymouth (usually I have graphical dots), then immediately lightdm [08:43] but the logs do show a failed session [08:43] - other ideas [08:43] to be sure, I changed the password of u-d-n to "a" and tried to log in manually, then I got a blue error "session startup failed" in lightdm greeter [08:44] 3 would be best, but I doubt the unity team is going to work on that [08:44] lightdm.log: http://paste.ubuntu.com/7657351/ [08:44] x-0-greeter.log: http://paste.ubuntu.com/7657352/ [08:45] pitti, is that the correct log? It looks like a fresh boot log [08:45] x-0.log: http://paste.ubuntu.com/7657353/ [08:45] pitti, do you have unity-system-compositor.log ? [08:45] robert_ancell: yes, it's fresh from booting the ubuntu-desktop-next CD [08:45] seb128: not in /var/log/ [08:45] weird [08:45] I do have X.org logs apparently, though :) is that for lightdm itself? [08:46] yes [08:46] anyway, x-0-greeter.log shows a scary number of errors [08:46] ok, weird [08:46] the missing backgroud is probably ok [08:46] (nm-applet:2031): libnotify-WARNING **: Failed to connect to proxy [08:46] that probably also survived [08:46] for me in the VM case I had a unity-system-compositor.log with egl errors [08:47] the missing background is because the image is not seeded on the iso [08:47] seb128: welll, this is not VM, it's real errors [08:47] err, real iron [08:47] right [08:47] grr, network disconnects [08:47] robert_ancellhttp://paste.ubuntu.com/7657360/ [08:47] ups [08:47] seb128, Laney: so does yesterday's ISO actually work for you? [08:47] robert_ancell_, ^ [08:47] seb128, ta [08:47] pitti, yes, for some definition of "work" [08:47] I wonder if any of these "no such service" etc. are breaking it, or if it's a graphics card/Mir problem [08:48] on my inspiron 11, it boots fine in unity8, showing the tablet lock screen, I just need to click on the "phablet" button to unlock [08:48] installs works fine as well [08:48] in VM I've something similar to what you describe [08:48] oh, so x-0-greeter.log is for the greeter's own session, not from starting the user session? [08:48] Laney has it booting to the phablet lock but hanging there [08:48] pitti, correct [08:49] pitti, user session would be ~/.cache/upstart [08:49] so where does it show the attempted start of the u-d-n session? [08:49] in the ubuntu-desktop-next userdir [08:49] pitti, do you have x-0.log? [08:49] robert_ancell_, x-0.log: http://paste.ubuntu.com/7657353/ [08:49] sorry, I see the link now [08:49] robert_ancell_: yes, http://paste.ubuntu.com/7657353/ [08:49] default-session=ubuntu [08:50] shouldn't that be unity8-mir? [08:50] or is this something else? [08:50] robert_ancell_: but I thought this would also just be for the greeter itself? [08:50] Laney, [+0.17s] DEBUG: Loaded session /usr/share/xsessions/unity8-mir.desktop (Unity8-Mir, ) [08:50] I see no trace at all for the attempted lightdm startup of the "actual" session [08:50] seb128: yeah that means it's installed [08:50] yes, but merely loading the .desktop file doesn't mean much [08:50] but I'd expect it to be the default too? [08:51] can you look at your /var/log/lightdm/? [08:51] pitti, this is autologin? [08:51] see if you have the same things [08:51] oh, right [08:51] robert_ancell_: yes; as I said, this is a clean live session boot of yesterday's utopic-desktop-next-amd64.iso [08:52] robert_ancell_: It could be that autologin is still broken, of course; but manual login said "failed to start session" [08:52] pitti, do you have a /etc/lightdm/lightdm.conf? [08:52] pitti, because the lightdm log indicates this is not an autologin, it has started a greeter and prompted you for something (probably the password) and not got a reponse [08:52] seb128: yes, it did have that [08:52] pitti, what is in there? [08:52] pitti, dpkg -l | grep unity8-desktop-session [08:52] seb128: I don't know any more, I didn't save it; but presumably the same you have? [08:53] seb128: I'm afraid I can't boot back and forth multiple times now, I need my battery (this is just an IC without power plugs, no ICE :/) [08:53] pitti, well, it looks like you don't have Laney's changes that landed this w.e [08:53] pitti, are you sure you got the correct iso? [08:53] * robert_ancell_ thinks I should make a lightdm --dump-config that combines all the config files into stdout for easy checking for configuration errors [08:53] seb128: that's why I'm asking you to check that default-session in your logs [08:53] should confirm that theory [08:53] Laney, it's unity8-mir for me [08:54] * Laney nod [08:54] pitti, what's the md5 of your iso? [08:54] seb128: oh, erk -- indeed [08:54] because this is just what those changes were trying to fix [08:54] I did dowlnoad the new image yesterday, but apparently it got sent to Nirvana :/ [08:54] thanks firefox [08:54] zsync time! [08:55] seb128, Laney: argh, sorry for the noise then; will re-download this afternoon when I'm back with bandwidth [08:56] pitti, no worry [08:56] robert_ancell_, in the "trying to boot that iso in a VM" case things are weird [08:56] someone said that vmware was supposed to work [08:57] robert_ancell_, like there is no actual session log, and you get back to the greeter, it seems like it's unity-system-compositor that fails but it's not really communicated properly anywhere in the logs [08:57] Laney, I tried virtualbox only [08:57] yeah, just saying [08:57] * seb128 tries to download vmware just to see [08:57] seb128, u-s-c is still running but the u-s-c log doesn't show anything? [08:57] I think it was some mir guys at the sprint maybe [08:58] I also believe vmware was working at some point [08:58] robert_ancell_, no, u-s-c is not running [08:58] Not sure if that's what's been released into ubuntu though, there tends to be a delay [08:58] seb128, that's bad if lightdm hasn't detected that [08:58] the lightdm logdir has a u-s-c.log with a mir exception [08:58] and lightdm sits on a login prompt [08:59] which means autologin failed to start the session, sent back to the greeter without telling anything [09:00] the u-s-c.log has "std::exception::what: Error opening DRM device 19, "No such device"" [09:00] (that's under virtualbox) [09:00] seb128, oh, it falls back to VT switched X by default [09:01] But yes, we should have some mechanism for a greeter to have an error message when it starts that is set by lightdm [09:01] So we have a better user experience under failure [09:03] yeah, and if I enter ubuntu-desktop-next a l/p then I get the greeter "frozen" [09:03] like the username/password entry vanish but it stays on the greeter [09:03] I can still open/close the a11y indicator [09:03] oh well, better handling of error cases is a wishlist [09:04] robert_ancell_, looking to your email ... meanwhile, did you have an opinion on the gnome-desktop options? [09:04] seb128, I think I missed the question after the list of options [09:04] robert_ancell_, I didn't state a question I think :p [09:05] But agreed implementing in Unity is the best by far. The service seems an acceptable option, though it just feels risky for some reason [09:05] robert_ancell_, which one do you prefer/do you have a better idea, would be the question [09:05] No better options that I can think of [09:05] yeah, I don't really like the new service [09:05] it feels like it has potential to bring bugs [09:05] It shouldn't be particularly hard to implement in Unity 7 (for someone who is experienced in that codebase) === chrisccoulson_ is now known as chrisccoulson [09:06] yeah, but that's going to take some time even if we do it [09:06] Do we break any other desktops by doing this update (and do we care?) [09:06] the GNOME guys being blocked meanwhile [09:06] We should unblock them [09:06] the easiest option might be to upload a gnome-desktop3.8 then [09:06] no we don't [09:06] The other option I had was to just carry a big patch to provide the old functionality [09:06] I don't think that's that bad [09:06] in gnome-desktop? [09:06] yeah [09:07] we are going to need to rebase [09:07] the patch might also be non trivial [09:07] darkxst: you dropped the plugins.background enable gsettings key from g-s-d [09:07] It will probably be big, but not very interleaved with exisitng code [09:07] i.e. it will be mostly a separate module [09:08] I think I would prefer to either ship gnome-desktop3.8 as a separate lib/copy the functions we need in u-s-c/u-c-c [09:08] until we implement the dbus interface in Unity [09:08] I feel that would be more expensive [09:08] then we can drop that [09:08] k [09:08] Though I don't have a feel on how many apps use the libgnome-desktop api [09:08] the xrandr one? [09:08] yeah [09:08] only g-s-d and g-c-c [09:09] If that's the case I'd be happy to copy-paste the code into u-s-d/u-c-c and be done with it [09:09] right, that's what I was thinking [09:09] +1 from me then [09:09] I'm going to have a try to it [09:09] you might get a request to review the changes if that works out ;-) [09:09] no problem [09:10] If any trouble, give me a call. I feel fairly in-depth with xrandr now (I've been deep diving into X protocol recently) [09:10] ok, thanks [09:10] seb128: I just applied for vacation during the next two weeks, I only had 3 days off this year so far and those two weeks are the ~only ones where this doesnt hurt too bad (between beta2 and the final rcs of the first upstream final). [09:11] Sweet5hark, k, works for me [09:11] Sweet5hark, you've only had three days off this year!? Man, take some time off! [09:12] gtg, bye all [09:12] robert_ancell_, summer is just starting for us, most people didn't take that many days off yet ;-) [09:12] seb128, ah... [09:12] bye robert_ancell_ [09:12] robert_ancell_: sounds like a real manager :) [09:12] robert_ancell_, have a nice evening! [09:12] robert_ancell_: yeah, its not too tragic ;) [09:18] * Sweet5hark will have 5 hours of management meetings again (plus the desktop meeting with you guys, but thats enjoyable) this week, after that I need a vacation. ;) [09:18] Sweet5hark: oh, so *you* are the new desktop manager? [09:19] darkxst: I'm going to upload g-s-d, you missed the patch from series it seems [09:19] pitti: Im not drunk enough yet to fall for that trap ;) [09:19] TBH I doubt that would be possible. [09:19] pitti: libroffice management meetings :) [09:20] pitti, don't try to sneak out from that job, it's yours without discussion ;-) [09:20] Sweet5hark: ah, LibO management [09:20] seb128: yeah, yeah, true. We made that decision in Malta, didnt we? [09:21] Sweet5hark: you can synergize with having multiple management positions at the same time, and leverage your competencies in Ubuntu as well! [09:21] pitti: E_TOOMANYBUZZWORDs [09:22] yeah, I'm afraid it's about as many as my little engineer head can put into one sentence :/ [09:22] can we just hire 'managerRequest = const "yes"' as our new manager? [09:22] Laney: what kind of language is that? [09:23] Laney: nah, we need some logic to handle request from above slightly different :> [09:23] so, who here wants to have meetings and hangouts all day, regularly fly to quarterly business reviews, forget how to program, and get desparate about falling behind the plan? [09:24] pitti: its always the guy who is asking! [09:24] Prelude> let managerRequest = const "yes" [09:24] Prelude> managerRequest "can I have a pony?" [09:24] "yes" [09:25] * pitti should perhaps not work in recruitment marketing :) [09:25] pitti: haskell ;-) [09:25] pitti: that's what I almost tested some months ago… so "no" :) [09:25] (without having the title ;)) [09:25] Laney: aah, should've known that :) [09:26] Laney, ah ok, oops, I know i did rebase the patch to keep the schemas ;) [09:26] darkxst: you are a naughty boy! [09:26] do you still use the vcs? [09:26] looks like it [09:26] ok, going offline for some time to work on some code and change train in a bit [09:26] Laney, ubuntu-desktop? yes [09:26] thought you might have moved it to ubuntu-gnome-dev [09:27] but actually it is still in u-desktop [09:27] so this makes sense [09:27] pitti, ttyl [09:27] Laney, well we could, but that makes no sense until it moves into our packageset [09:27] okay [09:28] * didrocks is going for a run to think about how to test this dynamic module loader (category/frameworks) effectively [09:29] Laney, we need to build the schemas from u-s-d before doing that [09:29] didrocks, enjoy ;-) [09:29] seb128: thanks! [09:30] maybe [09:30] * didrocks fixes some pep8 before though [09:30] that's pretty annoying for them [09:35] right, probably more new keys get added, than ones get removed [09:36] we could keep the schemas updated in u-s-d [09:36] yes, that's the annoying part :P [09:36] like take the 3.12 version and put it in there [09:36] buh, why is my local mirror giving 404s? [09:44] larsu, could you have a look to https://errors.ubuntu.com/problem/692f277e8623fe5be647a3a9ffcb59f6756371cc ? [09:44] I wonder if that could be an issue with gtk 3.12, that started recently [09:45] gtk_application_set_app_menu() issue === vrruiz_ is now known as rvr [10:15] seb128: sure, I'll have a look [10:16] seb128: want to look at newing e-d-s if you have a minute? === MacSlow is now known as MacSlow|lunch [10:31] Laney, sure [10:31] thanks [10:31] larsu, thanks [10:31] also, looks like turning on recommends didn't add that much to next [10:33] * seb128 nudges Laney with a -v [10:33] I used copy-package from a ppa [10:33] maybe I forgot to -v there ... [10:33] https://launchpadlibrarian.net/177555901/evolution-data-server_3.12.2-1ubuntu1_source.changes [10:33] looks okay [10:35] limitation of the pocket copy I guess [10:37] might be okay in the end [10:37] ... [10:37] Laney, NEWed [10:38] great, merci buckets [10:38] de rien ;-) [10:43] yeah looks like it got the right one [10:43] phew! [10:44] not on -changes though :( [10:44] right what? changelog? [10:44] https://launchpad.net/ubuntu/utopic/+source/evolution-data-server/3.12.2-1ubuntu1 → view changes file [10:44] oh, right [10:45] must be a copier bug [10:46] yeah, I think it's a known issue, Colin mentioned it in the past iirc [10:47] ho hum [11:36] pitti, could you have a glance to https://code.launchpad.net/~gunnarhj/ubuntu-docs/hibernate-multiple-users/+merge/223300 to look if it's correct [11:36] pitti, looks fine to me, but you know logind better [11:41] seb128: replied there, with some nitpicking [11:41] pitti, danke [11:42] mvo, hey, I just wanted to check what's the status of "get the correct backend used to install clicks" [11:43] mvo, the desktop next iso is working now, I was trying to get clicks to install on it/test the new platform api to make sure it fixed the abort issue, but I'm unsure how to hack the aptcc backend exactly [11:43] chrisccoulson, hey, what's the status of your fix/workaround for webbrowser on desktop mir sessions? [11:44] seb128: where is the download url? you probably still need to change the backend to dummy, but I would love to test that myself as I still have not had time to dig into the deatils [11:47] mvo, download url for what? the click? [11:47] or the iso? [11:47] http://cdimage.ubuntu.com/ubuntu-desktop-next/daily-live/current/ [11:48] what Laney said if you were asking about the iso [11:48] * Laney guesses so [11:49] yeah, me too ;-) [11:49] ta [11:50] downloading [11:51] pitti, oh, also, that udisks bug ... the timeout issue, you pointed to an util-linux bug, which is "fix released" in Debian by an Ubuntu patch from you they backported, does it mean it's not really fixed or...? [11:51] mvo, that's not going to be different from running unity8-desktop-session-mir on your install [11:51] seb128: yeah, I suppose it's something else [11:51] seb128: the ubuntu bug report for that is still open, haven't yet reproduced/debugged that one [11:52] k [11:53] mvo, where do you look for errors from the apps lens, I tried to install yes/no, I get it to display 100%, and then "installed", but no way to run it [11:54] click list doesn't list it [11:54] so I guess the "installed" is a lie [11:55] seb128: probably [11:55] seb128: .cache/upstart/logs iirc [11:56] mvo, ok, issue 1 was "needs to restart session after configuring online accounts" ... I'm going to open a bug about that (do you know if there is already one?) [11:59] hum, signond abort in fact, debugging === alan_g is now known as alan_g|lunch [12:01] seb128: I don't think there is one open already, if you have that, I will put it on the click-ftd blueprint [12:01] seb128: I was bitten by it too [12:01] mvo, I can do that, debugging the abrt first, then opening bugs/updating the blueprint [12:03] ok [12:10] just tried in vmware [12:10] didn't work [12:10] Failed to create GBM device [12:12] Laney: does it work with kvm/qemu? [12:12] nein [12:12] :( [12:13] how do I test it? [12:16] mvo, boot a media on real hardware, until Mir supports VMs (which is scheduled for this cycle) [12:16] Laney is trying to get 3d to work in vmware, that might be a solution as well, let's see [12:22] seb128: quick question, did you want me to redo the xdg-utils MR or can you apply my patch as is? [12:22] ochosi, hey, I can apply it, I just didn't get to that yet [12:23] sure, no problem, just wanted to make sure you weren't waiting on me to do something :) [12:24] I'm not, I've just been dealing with other things [12:27] good, no rush === davidcalle_ is now known as davidcalle [12:27] just working through my todo-list before going away for a few days === MacSlow|lunch is now known as MacSlow === alan_g|lunch is now known as alan_g [13:02] Laney, hum, "/usr/lib/gcc/arm-linux-gnueabihf/4.8/../../../arm-linux-gnueabihf/libwebkitgtk-3.0.so: undefined reference to `std::__once_call@GLIBCXX_3.4.11'" ... seems like evo got doked (that's the armhf build) [13:08] yey, we're not the only ones [13:09] lol [13:18] robert_ancell: mh no, I don't work on that... At least not yet. [13:19] seb128: did you coin a new term for a broken build? "it got dokoed"? [13:20] mvo, that's the vocabulary to use for buggy toolchains ;-) [13:20] hahaha [13:20] Trevinho, robert_ancell is off for some hours, what was he asking you? [13:21] seb128: is there a pattern already? :) [13:21] * didrocks got dokoed on Travis CI, but a rebuild did it [13:22] didrocks, between doko and buggy toolchains? sure is yes [13:22] seb128: if I maintain unity8 session [13:22] seems like libstdc++ is buggy [13:22] Trevinho, oh ;-) [13:22] libstdc++… add some C++11 on top of it and it rings some black time and hours to me :) [13:23] indeed [13:24] seb128: if that could be fixed would be nice, since that C++11 bits are quite nice to have [13:25] yeah [13:25] (like the ability to use automatically-disconnected-on-object-destruction-lambdas :P) [13:33] * Laney gets a "I've been dokoed" tattoo [13:33] it's a picture of his face [13:34] who would like that?! [13:34] badge of honour [13:34] it gets you into all the secret clubs [13:34] but are you supposed to talk about it then? ;-) [13:35] being dokoed is a very public thing [13:35] guess so, it just happened to a bunch of people today [13:35]