=== duflu_ is now known as duflu [05:53] Good morning [06:50] g'day mate! === doko_ is now known as doko [08:03] pitti: Guten Morgen! [08:03] bonjour didrocks, ça va ? [08:04] pitti: ça va bien, et toi? [08:05] didrocks: je vais bien aussi, merci [08:05] pitti: I wanted to ask you about xpathselect [08:05] pitti: IIRC, autopilot is using it [08:05] didrocks: right, just discussed with robru in #u-touch [08:06] didrocks: it was the second time I had to upload a no-change rebuild into trusty [08:06] didrocks: I added a landing ask, robru is on it [08:06] we need to release dependencies together with packages [08:06] didrocks, oh yeah, I just published that because pitti requested it [08:06] ok, and it's only a cross compilation commit, so fine [08:06] trunk just has some cross-building fixes and our tests run against trunk/daily PPA, so it's fine [08:07] yeah, I didn't look yet, but autopilot and xpathselect had the reputation to have landing asks with regressions :) [08:07] hence my fear [08:07] robru: please go ahead! :) [08:07] thanks pitti, robru ;) [08:07] didrocks, done ;-) [08:07] robru: so, you release xpathselect + autopilot-gtk again I guess? [08:07] or rather you "released" [08:08] ;) [08:08] didrocks, nope, just xpathselect. pitti said it would be fine [08:08] hum, really? that will pass -proposed? [08:08] ah, next one will pick the cross compile change I guess [08:09] robru: you need to merge back pitti's changelog for autopilot-gtk though (to not have a yellow prepare for autopilot-gtk) [08:09] didrocks, i dunno, but i trust pitti ;-) [08:09] didrocks, yeah, I figured that, haven't got to it yet [08:09] robru: thanks for tracking it to the end :) [08:10] didrocks, also just got ubuntu-html5-theme done. heading to bed soon ;-) [08:10] ah, you have done ubuntu-html5-theme at the same time [08:10] (just read it)! [08:10] great, thanks a lot [08:10] have a good night robru :) [08:10] didrocks, no worries, wanted to finish things properly tonight so I can start fresh work tomorrow instead of spending tomorrow cleaning today's mess ;-) [08:13] heh, indeed :) [08:13] didrocks, ok, changelog synced in lp:autopilot-gtk. goodnight! [08:14] robru: thanks [08:14] robru: sleep well! [08:14] pitti, thanks pitti ;-) [08:15] robru: have a good night! [08:16] didrocks, whoever, http://people.canonical.com/~ubuntu-archive/component-mismatches-proposed.svg needs your attention ... indicator-datetime lightdm, maybe fonts-wqy-microhe [08:17] doko: will have a look, thanks [08:17] I don't think we want gnome-panel back :p [08:17] oh, and alacarte! [08:17] neeeeeed ;) [08:21] didrocks, and nlpsolver is Sweetshark stuff, or Badshark's ;p [08:21] yeah, that would be for him :) (I'm sure he has an hilight on Badshark :p) [08:47] good morning desktopers! [08:48] Hello, world! [09:05] morning [09:06] Laney, good morning! how are you? [09:07] seb128: I'm pretty good thanks, had a nice climbing session last night ;-) [09:07] you? [09:07] good morning seb128, ça va? [09:11] lut didrocks [09:11] Laney, didrocks: I'm good thanks ;-) [09:22] ok, stupid git question [09:22] I've a checkout, I commited a patch, sent it for review [09:22] it got approved upstream, but other commits went it meanwhile [09:22] how do I put it on top of the commit stack? [09:22] when I "git pull" it asks me to put a merge msg [09:22] git pull --rebase [09:23] but I don't want to merge, I just want to put it on top of trunk [09:23] "Cannot pull with rebase: Your index contains uncommitted changes." [09:24] Laney, ^^ [09:24] are they the changes from the merge? [09:24] I ctrl-C-ed the merge [09:24] git diff [09:24] since I didn't want to enter a commit msg [09:24] but maybe it something [09:24] $ git diff [09:25] $ [09:25] I guess I'm going to rm -rf and checkout again [09:25] git reset --hard origin [09:25] git is such a pain to use :/ [09:25] (if you want to completely throw away your changes) [09:25] that'll kill his local patch [09:26] so, "git stash save" first [09:26] then reset, then "git stash pop" [09:26] (no I don't, I just want to push that patch that got approved) [09:26] thanks guys [09:26] I've rm-ed the dir and I'm doing a new checkout [09:26] it's over me how people can find git usable ;-) [09:28] you could probably have done git reset --hard ; git pull --rebase [09:31] the reset would have wiped out my commit no? [09:31] seb128: yes, I thought that's what you wanted to do [09:31] if you give it a commit id it resets to that one [09:31] seb128 | I've rm-ed the dir and I'm doing a new checkout [09:32] seb128: you're telling people that edit text in vi that git is hard to use?! [09:32] now that sounds like me and bzr [09:33] Laney: not having an index already kills bzr's (slightly) friendlier command line for me [09:37] hum [09:37] IRC split? [09:37] I was saying [09:37] pitti, I wanted to do the equivalent of "bzr uncommit; bzr pull; bzr commit; bzr push" [09:37] e.g just update my checking to current trunk [09:37] and then put back my diff on top to commit/push [09:37] [09:37] well, anyway, I did a fresh checkout and I'm just redoing my commit on top now [09:37] seb128: git pull --rebase [09:37] haha [09:37] that's what I said at the start :-) [09:37] pitti, that was giving me the fing^W "Cannot pull with rebase: Your index contains uncommitted changes." [09:37] seb128: so, stash them away first [09:38] git stash save [09:38] git pull --rebase [09:38] git stash pop [09:38] gives you current trunk plus your local commits on top, plus the uncommitted changes [09:38] noted for next time [09:38] I had no uncommited changes [09:38] seb128: err, but apparently you did? [09:38] out of whatever git might have done when I tried "git pull" (without the --rebase) [09:38] seb128: ah, so probably a half-done merge [09:38] which I ctrl-Ced when it asked me for a merge msg [09:38] since I didn't want to merge [09:39] ah, "git reset --hard" would probably have done it, but seems you solved it now [09:39] that would have thrown away my commit as well no? [09:39] I've done the "format-patch, store them away, rm -r, re-clone, and git am" exercise a lot of times, too [09:39] seb128: no, just the uncommitted changes (the noise from teh cancelled pull) [09:39] yeah, that's what I always end up doing :/ [09:40] seb128: "git reset --hard" == "bzr revert" [09:40] seb128: "git reset --hard origin" ~= "re-clone" [09:40] ok [09:40] that origin thing in git doesn't make any sense to me [09:41] (and yes, git doesn't make sense) [09:41] I should just write all those down [09:41] it's kind of like :parent [09:41] at least not for people who don't want to think much about RC [09:41] and not try to find a logic to it but just copy recipes [09:41] i don't know how to do the equivalent in bzr [09:41] Laney, I never use that either :p [09:41] bzr push --remember :parent is good [09:42] all these indexes, heads, origins, branches concepts are pretty hard to grok, and absolutely unnecessary for 95% of the projects out there [09:42] right, that's my main issue with git [09:42] I really like how bzr says "there are branches only, period" [09:42] I don't care about all those details, don't need them and don't want to have to think about them [09:42] I forgot "repositories" in above list [09:42] yet it's impossible to do anything without running into them [09:43] seb128: that's pretty much what I did, too -- find out how to do the most common workflow, curse about it, memorize it, and stop thinking about it [09:45] git rebase -i ♥ [09:45] Laney: indeed, that and bisect are the only two things I really like about git [09:46] Laney, what does rebase -i do? [09:46] Gives you a nice way to reorder, rename and change commits [09:46] * seb128 never really rebase anyway [09:47] I just uncommit;pull;commit [09:47] or uncommit/commit if I want to change a commit msg [09:47] git commit --amend [09:48] * larsu loves that and hates that he has to jump through hoops for the same thing in bzr [09:48] larsu: ok, good point, too [09:48] pitti: Indeed, those are great. I'd add having all the branches in the same place to my list too [09:48] I hate that, it doesn't work with my brain [09:49] Laney: that's actually a negative for me; git's concept of "checkout" (switch between branches) is utterly broken [09:49] I never know in what branch I am [09:49] as it only changes the tracked files, not the built files [09:49] so it's useless for anything that gets compiled [09:49] I keep screwing up and commiting to the wrong one/checking out the wrong one [09:49] my $PS1 shows which git branch I'm on [09:49] yep [09:49] I clean a lot I guess [09:49] 0 martin@donald:~ [09:49] $ cd upstream/upower/ [09:50] 0 (master) martin@donald:~/upstream/upower [09:50] it makes comparing/cherry-picking/creating/merging quite easy for me [09:51] Laney, seems like you are an heavy vcs user ;-) [09:51] * seb128 mostly does packaging and u-s-s [09:51] I find that a lot easier in bzr with separate dirs per branch, but that's a matter of taste I guess [09:51] neither of those have branches or stuff to cherry pick or such [09:51] (and I use built sources a lot) [09:51] yeah I guess it is [09:52] I did most of my learning with Debian packaging branches which I build out of tree anyway [10:25] seb128, ok, switching [10:25] diwic, larsu: hey [10:26] diwic, we didn't fork g-s-d to u-s-d yet but that's likely to happen this cycle [10:26] having the dialog in there is fine [10:26] but we don't have g-s-d in unity8 [10:26] so if you do that here it means it's not futur proof and that you are going to need to redo it somewhere else in the next cycles [10:27] seb128, yeah, it's not that much code so I don't think that's a problem [10:29] seb128, so at this point, should I prepare my code to be a patch/branch against g-s-d and you will merge it? [10:29] seb128, or want me to wait for the fork? [10:30] diwic, g-s-d is fine, the fork is basically going to be a copy of g-s-d with renames when we do it [10:30] well, at first [10:37] seb128, ok, then we have a plan :-) [10:40] diwic, great [10:40] seb128, is it lp:ubuntu/gnome-settings-daemon ? [10:41] diwic, no, lp:~ubuntu-desktop/gnome-settings-daemon/ubuntu [10:42] seb128, ok, thanks === vrruiz_ is now known as rvr [11:18] xnox: upstart halp [11:19] Laney: ok =) trading that for locales help =) [11:19] I want to write a user session job for session-migration but I'm not sure what it's start on should be [11:19] it wants to be 'started dbus' but also ideally to block session starting [11:19] but not to assume gnome-session [11:19] Laney: let me see, i think we can do that. [11:20] I found 'wait jobs' in the cookbook ... === Ursinha is now known as Ursinha-afk [11:22] Laney: start on (started dbus and starting xsession-init) [11:23] Laney: this will result in dbus + all bridges available + pulseaudio, and will block xsession event which is before any desktopy stuff starts. [11:23] I think. [11:23] Laney: initctl2dot is your friend to visuallise things. [11:23] even though dbus is also 'starting xsession-init'? [11:24] Laney: yeah, at the moment dbus blocks xsession-init, and you want to "continue blocking xsession-init" until you are done but with dbus available. [11:25] oh okay I was worried they would deadlock [11:25] are there themes other than light-themes using unico? [11:25] Laney: in practice, it's equivalent to changing "xsession-init" to: start on started session-migration [11:25] * larsu is about to hack something really badly [11:26] Laney: assuming that session-migration will be type "task" [11:26] indeed [11:26] let me try that [11:26] hm, looks like mythbuntu-default-settings and lubuntu-artwork depend on it [11:26] Laney: =) [11:26] hopefully it'll work on the device after this ... [11:26] Laney: my problem however is that /etc/default/locale is set, /etc/environment has two lang variables, yet my booted phablet user does not have locale set =( [11:27] got anything in ~/.pam_environment? [11:29] Laney: *sigh* so it sets locale when I do serial tty login.... [11:29] Laney: but not when doing adb shell, nor when doing adb shell sudo -u phablet -i [11:30] indeed, it's the same on the device [11:30] doesn't seem to read /etc/environment [11:31] Laney: i think i want to change android-tools-adbd job to source /etc/environment I think... [11:31] or someother way make adb act more like login / pam aware thing [11:32] I think a lack of PAM-ness is the root problem indeed [11:32] can you use ssh instead? [11:34] not really no. [11:35] emulator networking stack is crap & unreliable. And we don't want ssh open by default, so we are tied to adb. [11:36] okay :/ [11:36] I guess you could get far enough by doing what you need on an ad-hoc basis [11:36] xnox, ssh is installed, just add something that calls "start ssh" in the emulator bootup [11:37] I guess that's "we don't want ssh open by default" [11:37] i would vote for having ssh enabled by default in the emulator [11:37] I don't know if it'd be possible to have adb shell run the pam stack somehow but that feels best to me [11:38] no, wont [11:38] at least not without a lot hackery [11:38] well, I didn't imagine it would come for free [11:39] especially since i was told we need to keep adb for android compatibility we cant really change it to much [11:39] mmm, since it's not forked I suppose [11:39] right [11:39] it needs to stay command compatible at least [11:39] I don't know enough of the architecture to answer what happens on the phone side [11:39] that's where I guess it would happen [11:40] Laney: exec sudo -i adb => should result in all the pam stack environment to be present ;-) [11:40] jibel: I committed the pci_add → {device,drive}_add change, I also tested it under precise [11:40] * xnox updates g+ mood to Evil [11:40] heh [11:41] O_O [11:41] xnox, dont the emulator scripts already modify the rootfs ? i think it would make sense to rm /etc/init/ssh.override [11:41] anyway, that starting condition seems to work [11:41] thanks! [11:41] Laney: your welcome. [11:43] ogra_: i disagree that we want ssh, in either dev device / final device / emulator. And it's a problem that $ adb shell, does not have correct locale set, across the board - regardless ssh having it's pam/LOCALE ducks in a row [11:43] xnox, well, we need ssh so it will always be installed [11:44] (just not enabled due to the default .override file) [11:44] (the sdk enables and uses it) [11:45] ogra_: with emulator we don't have "normal networking" nor "usb networking" only via qemud_pipes, which is multiple times worse than $ adb shell, which is qemud_pipes "native". [11:46] ah, k [11:46] ogra_: or driving serial tty, which beats them all =) [11:47] "... with Ctrl+C killing emulator, nice feature" [11:47] heh === Sweetsha1k is now known as Sweetshark === alan_g is now known as alan_g|afk === alan_g|afk is now known as alan_g|lunch [12:55] Hm. Just updated my trusty desktop and it's broken, dropping to VESA video mode ☹ [12:57] tseliot: so I'm not getting the nvidia crash on start up now woohoo [12:58] oh, not just me? [13:00] popey: reboot into 3.12 kernel, i don't think all binary drivers were upload for 3.13 yet. [13:01] ah [13:02] this shouldn't happen. ☹ [13:02] nope. 3.12 is broken too [13:03] http://paste.ubuntu.com/6714683/ [13:04] popey: man I wish I read your line before I updated :( [13:04] is there a bug for this? [13:05] [ 16.920] (EE) Failed to initialize GLX extension (Compatible NVIDIA X driver not found) [13:05] popey: go into the terminal and disable nvidia [13:06] disable as in remove the package? [13:07] popey: see if the neuvau driver works at least it will give you a desktop again [13:07] popey, wrong channel, try #ubuntu-devel or #ubuntu-kernel [13:07] well "wrong channel" [13:07] you probably want to notify the kernel guys [13:08] well. I was trying to see if any other desktop users had the same issue, hence being here. Happy to file it with whoever you think. [13:09] I think most of us use intel [13:09] which obviously doesn't have those binary driver issues... [13:09] davmor2: great :) [13:09] popey: changing to kernel 3.12 worked for me [13:10] popey: where did you get 3.13? [13:10] well, I'm on 3.12 now and it's still failing [13:10] tjaalton: apt-get update [13:10] upgrade even [13:10] popey: in 14.04? Can you get me some logs please? [13:10] on -proposed? [13:11] no [13:11] trusty main [13:11] not here [13:11] tjaalton, 3.13 was published in trusty earlier today [13:11] huh [13:11] ok [13:11] tjaalton, you use a mirror? [13:11] http://paste.ubuntu.com/6714725/ [13:11] my own and it seems fi. [13:11] used to be the main mirror [13:12] tseliot: http://paste.ubuntu.com/6714683/ [13:12] tseliot: what logs do you need? [13:12] ah no, just my own :P [13:12] tseliot: what should I file a bug against? kernel? [13:12] popey: you have linux-generic installed, and the headers? [13:13] one mo, just rebooting after removing nvidia* [13:13] popey: yes, it's probably a matter of installing the headers, for 3.12, at least. I'll upload a fix for 3.13 later today [13:14] linux-image-generic: Installed: 3.13.0.1.3 [13:14] linux-generic? [13:14] the meta package that pulls headers [13:14] linux-generic: Installed: (none) [13:14] there you go [13:14] why dont I have that? [13:14] don't remove it ;) [13:14] dont assume I did [13:14] check dpkg.log === Ursinha-afk is now known as Ursinha [13:15] alan@wopr:~$ dpkg -l linux-generic [13:15] dpkg-query: no packages found matching linux-generic [13:15] no refernece to linux-generic in dpkg.log [13:16] what was the system installed with? [13:16] release [13:16] 12.04 probably [13:16] how can I tell? [13:16] and upgraded from it? [13:16] ya [13:16] /var/log/installer [13:16] maybe older [13:17] did you upgrade straight to trusty? [13:17] jan 2 2012 [13:17] no [13:17] i upgrade like a good boy [13:17] then I can't see how nvidia worked before if you didn't have the headers meta pkg installed.. [13:18] dpkg.log.* probably doesn't go far enough [13:18] it worked this morning [13:18] and has worked for months [13:18] on what? [13:18] 14.04 [13:18] uh [13:18] i just did my usual daily update today and boom [13:19] and I _do_ have headers installed [13:19] so I think this is a red herring anyway [13:19] ii linux-headers-3.13. 3.13.0-1.16 amd64 Linux kernel headers for version 3.13.0 on 6 [13:19] headers for what? [13:19] linux-headers-generic? [13:19] linux-headers-3.13.0-1-generic: Installed: 3.13.0-1.16 [13:19] linux-headers-3.13.0-1: Installed: 3.13.0-1.16 [13:20] installing ubuntu-desktop task doesn't want to add anything [13:20] so it's not like I have anything missing [13:20] it's not what should pull it [13:21] hardwired to the installer [13:21] but no linux-headers-generic? [13:22] i have linux-headers-generic, as I said, linux-generic I don't have [13:22] 2014-01-08 12:50:00 status installed linux-headers-generic:amd64 3.13.0.1.3 [13:22] ok [13:22] so that's enough [13:23] check /var/lib/dkms if the build logs are still there [13:23] i only see bbswitch and vboxhost [13:23] (and dkms_dbversion) [13:23] so purging nvidia got rid of those, nice [13:23] :) [13:23] hah [13:23] try reinstalling [13:23] k [13:23] nvidia-current? [13:23] yep [13:24] hmm no [13:24] it's old [13:24] bah [13:24] nvidia-331 [13:24] ok [13:25] tseliot: is the nvidia-current transitional pkg still needed? [13:25] DKMS: install completed. [13:26] * popey reboots [13:27] tjaalton: http://paste.ubuntu.com/6714792/ dkms log [13:27] that's for 3.13, did it build for 3.12 ok? [13:28] i see no directory for 3.12 [13:28] http://paste.ubuntu.com/6714800/ [13:29] tseliot: shouldn't it build the module for all the kernels? [13:30] tjaalton: nope [13:30] i thought it always used to [13:30] it's always the kernel in use and the latest kernel [13:31] if the latest kernel is the one in use then it builds only for one [13:31] imagine having 10 kernels installed and having to build the module 10 times every time... [13:32] imagine having a broken desktop ☻ [13:32] right, rebooted and no dice, still broken. [13:32] popey: so boot the previous kernel.. does it build then? [13:32] ok [13:32] popey: BTW don't use it with 3.13 [13:32] ok [13:33] kinda weird that it was pushed through without checking the dkms builds first.. [13:34] tseliot: no really might it not work with 3.13 grrrrrrr [13:35] * popey has built, reboots to 3.112 [13:35] davmor2: I have written patches for it and I'm about to test them. Unfortunately this is isn't the only thing I'm working on [13:35] * tseliot is very very busy [13:35] 3.12 works [13:36] tseliot: is fglrx busted too? [13:36] tseliot: where do you want this filed? [13:36] tjaalton: I think so [13:36] popey: "this" being what? [13:36] this is a bug. [13:36] i can't be the only nvidia/ati owner on 14.04? [13:37] popey: are talking about 3.13 or the fact that linux-generic wasn't installed? [13:37] either/both [13:37] the latter is probably not a bug [13:38] the former is just that I didn't have enough time to fix the drivers and I don't think we need a bug report about that. I'm already on it [13:38] replied to k-t about this.. [13:39] tseliot: drop everything else ;) [13:39] these are a priority now [13:39] tjaalton: even the hwe stuff? [13:40] what's that? [13:40] not in a public room ;) [13:40] i'd say so.. how long does it take to fix these? [13:40]