=== hggdh_ is now known as hggdh [03:29] Good morning [03:48] Laney: "Jenkins Fixed - saucy-adt-glib2.0 19" -- nice job! === VD is now known as Guest40853 [05:09] hi pitti [05:09] pitti: I have a question for you [05:09] I want to have my executable be elevated to root permissions after doing some initial messing around [05:10] so don't want to run the executable with sudo [05:10] but instead have it ask for sudo password part way through [05:10] is this possible? [05:10] thumper: no, not really [05:10] bugger [05:10] hmm... [05:11] thumper: you could split your program into a privileged part and an unprivileged one, and call the privileged one through sudo or pkexec [05:11] pitti: the program is juju :) [05:11] thumper: otherwise I suggest you start it as root, drop privileges at the beginning, and re-claim them only if you need to [05:11] pitti: hmm.. how do you do that? [05:12] a problem in particular is that some files which are created and should be owned by the user, are being owned by root [05:12] thumper: create a system user, find out its UID with getpwname(), then drop privs with setresuid(uid, uid, uid) [05:12] but I do need to write to /etc/init and /var [05:12] and when you need root, setresuid(0, 0, 0) temporarily [05:12] hmm.. can I do that in go? [05:13] that's a pretty basic posix call; I sure hope so [05:13] * thumper goes digging [05:13] pitti: thanks [05:13] thumper: eww, you start this as a user and write into /etc/ ? [05:13] thumper: maybe that should be re-designed first :) [05:13] pitti: it is to create a local provider [05:13] like, writing into ~/.config/juju instead, or creating a per-user dir in /var/lib/juju/ [05:13] so it creates two services for the mongo db and agent [05:14] these need to be started by upstart [05:14] and it also needs to create lxc containers in precise [05:14] so no user space [05:14] thumper: can't you ship a generic upstart job which tests for the existance of a particular file and only starts then? [05:15] hmm... [05:15] the thing is there could be multiple [05:16] if you had multiple local environments [05:16] I'm open to ideas === thumper is now known as thumper-afk === jhernand1z is now known as jhernandez [06:23] didrocks: to your queue, lp:~kubuntu-packagers/kubuntu-packaging/qtsystems-opensource-src + https://launchpad.net/~canonical-qt5-edgers/+archive/qt5-beta2/+files/qtsystems-opensource-src_5.0%7Egit20130614.orig.tar.gz would be ready (again, with the fixes) [06:23] qml plugins separated and license files added [06:24] Mirv: ah, thanks! :) [06:24] Mirv: will have a look later today [06:28] merci [06:46] good morning === thumper-afk is now known as thumper [07:58] hey [07:58] pitti: yeah - cool isn't it? [07:59] Laney: it sure is [07:59] Laney: so that's now with the proper runner? [07:59] yep [08:07] good morning desktopers [08:07] hey Laney, pitti [08:07] g'day mate [08:07] hey seb128 [08:07] hey mlankhorst [08:09] hey [08:10] more merge window fun, hopefully this nvd7 will work in 3.11 [08:10] :D [08:16] nothing like starting the day on a non working session [08:16] mlankhorst, I though for a bit that the new xorg was broken :p [08:16] gnome-session was complaining about errors loading the intel driver or something [08:17] but it turned out that it was a stalled .Xauthority in my userdir breaking things [08:21] * Laney remembers that he probably ought to fix his microphone [08:21] seb128: everything's an xorg bug ;) [08:21] but hey lets do it [08:27] Laney, if that's for the settings meeting you have another week, but yeah, would be nice ;-) [08:30] seb128: oh really? I thought it was today :P [08:30] Laney, we said every second week, though it would maybe be good to do it weekly now that things are quite active [08:31] Laney, but mardy is not there this week so... [08:31] yeah I knew that but thought it was two weeks already [08:31] could be [08:31] let me check :p [08:31] Laney, shrug, you are right ;-) [08:32] fixed anyhow [08:32] don't mind if you want to defer it since mardy isn't here [08:33] Laney, I made it bi-weekly in google calendar [08:33] Laney, well, we can have a quick round with those who are around, would be good to keep the cadence ;-) [08:33] k, cool [08:33] and we can discuss making it weekly [08:33] so we can have one with mardy next week as well :p [08:35] Laney, speaking of which, how much fun do you have with qml? ;-) want to pick another panel to work on? [08:36] yeah I was going to do some other bit of UI, maybe the security one? [08:36] Laney, well, in addition I mean ... or for when you are done with yours, or if you feel like doing a bit more "standard UI" [08:36] Laney, I think didrocks wanted to do security, what about bluetooth? [08:36] https://wiki.ubuntu.com/Bluetooth#phone [08:36] it seems like mostly standard UI [08:37] at least less crazy that the background one ;-) [08:37] and cyphermox seems to be overbusy [08:38] larsu, or check with didrocks if he wants help on the privacy UI === ara_ is now known as ara [08:41] I could do the basic bits for bluetooth [08:41] the specialist / advanced stuff would be best left for cyphermox though [08:43] oh interesting, this is one which is enabled dynamically [08:44] do we have a way of doing that? [08:45] using the entry component I guess [08:46] seb128: but at what time do you want to start uploads [08:47] mlankhorst, now? ;-) [08:47] do itt [08:48] mlankhorst, wait [08:48] Laney, can we put a britney lock on X*? [08:48] haha, not like that [08:48] if you give me a list of packages sure [08:48] right :p [08:48] mlankhorst, ^ [08:48] Laney, I think britney will not stop it because it Breaks: unity (<< new) [08:48] don't think it handles breaks like that [08:48] so we better put a manual lock there and wait to have things in place [08:48] yeah [08:49] just in case [08:50] seems like didrocks is busy with the Mir guys and I want to hear from him&Mirv before uploading [08:50] didrocks, Mirv: please ping back about new xorg when you are done being busy on other things [08:50] hmm, maybe it does though [08:50] infinity: does britney consider Breaks when determining installability? [08:51] seb128: just coordinate with Mirv, he's dealing with the transition AFAIK [08:51] mlankhorst: ^ [08:51] didrocks, ok [08:51] Mirv, hey [08:51] there was something wrong still with https://code.launchpad.net/~brandontschaefer/unity/move-pointer-barrier-to-xi-1.6.99.1/+merge/150175 I think? [08:51] Laney: everything from https://launchpad.net/~canonical-x/+archive/x-staging?field.series_filter=saucy except the unity and -omap-revert packages [08:51] Mirv: perhaps, unity needs to hard build-depend and depend on the version from that ppa, that should fix it. [08:52] can you give me a list? :( [08:52] the format is block package1 package2 ... [08:52] Mirv, not really, he forgot the commit message ... and it requires the new xorg which is in a ppa and going to be uploaded to saucy-proposed soon [08:52] mlankhorst: from seb128's comment I'd guess something like libxfixes-dev (>= 1.6.99.1) [08:52] Mirv, do you guys build on top of saucy or saucy-proposed? [08:53] Mirv: the epoch too.. [08:53] Laney: libxfixes libxi mesa-demos nvidia-graphics-drivers-tegra3 nvidia-tegra-codecs-cardhu unity x11proto-fixes x11proto-input xf86-input-mtrack xf86-input-multitouch xf86-input-tslib xf86-input-wacom xf86-video-omap xf86-video-omap-revert xorg xorg-server xorg-server-omap-revert xserver-xorg-input-acecad xserver-xorg-input-aiptek xserver-xorg-input-elographics xserver-xorg-input-evdev xserver-xorg-input-evdev-omap-revert xserver-xorg-input [08:53] ty [08:53] let me know where it truncates [08:53] xserver-xorg-input-evdev-omap-revert xserver-xorg-input [08:53] mlankhorst, pastebin? [08:54] splitlong.pl! [08:54] no truncate on pastebin, and no flooding of this channel :p [08:54] ok ill pastebin [08:54] :-) [08:54] seb128: if I haven't understood something wrongly, proposed (for example https://launchpadlibrarian.net/144130366/buildlog_ubuntu-saucy-amd64.unity_7.0.1%2B13.10.20130704-0ubuntu1_UPLOADING.txt.gz) [08:54] * seb128 shouldn't tell people about flooding about the "incident" where he flooder the channel recently :p [08:54] seb128: which would mean Xorg to -proposed, accept brandon's branch, rebuild stack after merged in [08:55] http://pastebin.com/JNHgkG9Y [08:55] if you build-depend on the right version it'll wait automatically [08:55] Mirv, sounds like a plan, we will get Xorg updated and then ping you for the accept brandon's patch/rebuild stack [08:56] seb128: ok, as mlankhorst pointed out, actually the brandon's branch is missing epoch also for libxi-dev, so I'll branch from there and add the 2: [08:56] Mirv, thanks [08:58] mlankhorst: so should the libxfixes-dev dependency be >= 1:5.0.1-1 ? [08:58] seb128: thanks [08:58] Mirv: yeah, and perhaps an explicit depends on libxfixes3 and libxi6 too [08:59] Laney: If it doesn't, it should. But ask Colin, he'd know for sure. [08:59] so symbols won't be messed up [08:59] mlankhorst: ok. that shouldn't be necessary, the -dev packages pick up the same version dependency automatically. [08:59] infinity: Yeah. I see mention of it in source so guess that it does. I did ask in #-release already. Thanks. [09:00] Mirv: not in this case, the xi2 symbols were introduced before [09:00] but we patched those out [09:00] Laney: Yeah, right after I wrote that, I read an email from Colin saying that he's out sick. So maybe don't ask him. :P [09:00] Heh [09:00] FINE. I'll ask #debian-release. [09:02] mlankhorst: interesting, ok. [09:03] sounds like we're safe [09:03] Laney, infinity: from IRC logs and discussion with Colin in the past, Britney checks for installability of ubuntu-desktop [09:04] but not including recommends [09:04] we had CD builds failing because of that [09:04] but that's ok, xorg and unity are both strong depends ;-) [09:05] seb128: yeah but for example libxi and libxfixes need to be held back manually [09:05] and the proto stuff too [09:05] now at https://code.launchpad.net/~timo-jyrinki/unity/move-pointer-barries-to-xi-1.6.99.1_try2/+merge/172975 [09:05] Laney, please put the big lock in place manually, better safe than sorry [09:06] Laney, mlankhorst: then we can start on uploads ... are those uploads a ppa copy, or manual upload or...? [09:06] ppa copy [09:07] mlankhorst, do you know the magic incantation that should be run? [09:07] is that basically copying the list of packages you just pastebined? [09:07] it's done [09:07] Laney, thanks ;-) [09:07] seb128: no idea what the magic invocation is [09:08] the answer is that it's enforced if the packages are required to be coinstallable [09:08] i.e. if there's a dependency relationship there [09:08] k [09:09] so if you can install unity without whatever-package-gets-the-breaks then it won't work [09:10] that makes sense, so it would block the server in this case [09:10] but not the libs [09:10] right [09:10] but updating the libs without the server should work [09:10] or you forgot some depends/breaks :p [09:10] copied from debian [09:10] anyway, let's get going [09:11] indeed [09:11] Laney, do you know if we can pocket copy including binaries from x-staging's ppa? [09:12] it has armhf builds [09:12] so I guess we can? [09:12] no [09:12] they're virtual arm builds aren't they? [09:13] copying has been used before [09:13] for 1.12 and 1.13 [09:13] infinity, ^ do you know? [09:13] he's asleep [09:13] RAOF: ^ [09:13] probably EOD [09:13] tjaalton, he commented 10 minutes ago, did he just feel asleep? [09:13] oh :) [09:14] in that case.. [09:15] Laney: but why wouldn't virtual arm work? [09:15] Oh, looks like it is non-virt but doesn't build for ppc [09:15] oh, ppc... [09:15] hm [09:15] it's known to be shaky [09:15] shrug [09:15] we need ppc binaries [09:15] so I guess we need rebuilds [09:15] the missing build records should be created when you copy [09:15] ok [09:15] Laney, should we try one on source to see how it goes? [09:16] go on then ... [09:16] try xorg metapackage [09:17] mlankhorst, is that arch: any? [09:17] not xserver-xorg [09:17] which is built from xorg [09:17] mlankhorst, I was going to try libxfixes [09:18] seb128: yeah that one will fail because it needs the new protos first [09:18] Then again a lot of these packages were built back in May which might be on an older toolchain [09:18] so you might want rebuilds anyway [09:18] seb128: Please never, ever, ever copy from a virt PPA. Ever. [09:18] Laney: shrug just upload the proto first in that case [09:19] rest should be ok [09:19] infinity, is https://launchpad.net/~canonical-x/+archive/x-staging virtual? [09:19] mlankhorst: don't you have upload to these anyway? :P [09:19] infinity: can you enable ppc on that ppa? [09:20] seb128: Nope, that's non-virt. You can tell from which builders it uses. But (and this is a big but), non-virt PPAs still aren't, by default, built the same was as the distro. [09:20] seb128: As far as translation stripping, etc. [09:20] s/was/way/ [09:20] https://launchpad.net/~canonical-x/+archive/x-staging/+build/4766545 [09:20] oh indeed [09:20] let's just rebuild those [09:21] mlankhorst: We could do an X-staging PPA that meant for copying, but no harm in just doing source copies this time. [09:21] it should be ok to rebuild, in that case x11proto first, then wait until it's done and the libxi/xfixes [09:21] then* [09:22] can you not dump all at once, libx* should have sufficient b-dep checks to prevent them from building before proto? [09:22] Interestingly enough, in most ways, that PPA seems to be correct already, except for the missing PPC. You should probably get that enabled. [09:22] (for next time) [09:23] sanity check [09:23] Laney, infinity: does that looks correct? [09:23] $ ./copy-package -s saucy --ppa-name canonical-x/x-staging --to-primary --to-suite=saucy-proposed x11proto-input [09:24] I think it's --ppa=canonical-x --ppa-name=x-staging [09:24] yeah [09:24] seb128: Does owner/name work? I always use --ppa= ... What Laney said. [09:24] ok, let me try that [09:24] You'll get a confirmation anyway [09:24] I'm new to those ppa to archive copies :p [09:24] Though I like yours. Someone should make the tool take that. :P [09:25] infinity: how do you specify what release to use for copy_package? [09:25] mlankhorst: Hrm? [09:25] (to take the saucy packages in there) [09:25] mlankhorst: -s saucy [09:25] in there -> from the ppa [09:25] mlankhorst: And if extra paranoid '-e $version' too. [09:25] hmz [09:26] no need for that I hope, ubuntu0.3 will fail to build on saucy due to toolchain changes [09:26] and the raring versions are mostly older, so they won't work anyway [09:27] -s saucy will get the right thing [09:27] ok [09:27] It will, yes. [09:27] The "extra paranoid" bit was if you were doing some scripting or something and trying to avoid races with other people. [09:27] In this case, I assume you know no one will upload to the PPA while you're doing this. :P [09:33] tjaalton, mlankhorst: I guess it's normal some of those have raring in their changelog target? e.g https://launchpad.net/ubuntu/+source/x11proto-input/2.3-0ubuntu1 ? [09:33] yeah, doesn't matter [09:34] yeah the x1.14 stack was originally created for raring, can just be carried over :) [09:34] or could that be pulled from debian now? [09:34] synced [09:34] hm actually.. it could [09:34] can I ppa copy for somebody else? that upload has my mine on it... [09:34] tjaalton, mlankhorst: which ones can be synced from debian? [09:35] I guess the one without ubuntu revision can be syncpackaged? [09:35] why don't I just extend the xorg packageset (mlankhorst requested this a month ago) and mlankhorst can self service? [09:36] Laney, that would be good [09:36] that would work, also the lts-* [09:36] ok [09:36] I think we had some discussion about it and then forgot to do it [09:36] maybe i could finally do the lts raring sru uploads myself then :> [09:37] but I guess if we don't need to do a binary copy I can do it myself [09:38] tjaalton: please don't do uploads :) [09:38] give me a minute to remember how to drive the tool [09:39] oops [09:39] laney@iota> edit-acl -P xorg -S precise -s libxrandr-lts-raring add ~/temp [09:39] Added: [09:39] libxrandr-lts-raring [09:39] cool [09:40] mlankhorst: harhar [09:41] * mlankhorst wonders why it said your name [09:41] mlankhorst, I did 2 syncs and used tjaalton for -s ... sorry if that created confusion [09:41] I tried to match the ppa uploaders [09:42] ah :-) [09:42] mlankhorst, well anyway, I just put x11proto-input/fixes and libxi/xfixes in so far, I will let you do the rest [09:42] I'll have to wait for libxi/libxfixes to fail first [09:44] mlankhorst, libxfixes built fine... [09:44] okay then, try that [09:44] mlankhorst, https://launchpad.net/ubuntu/saucy/+source/libxfixes/1:5.0.1-1 [09:44] edit-acl -P xorg -S saucy query lets you see what's in it [09:45] libxi did not :-) [09:46] mlankhorst, ^ did you see Laney's line? are you set for those copies/uploads? ;-) [09:47] should be good [09:47] mlankhorst, I did the first copies with "./copy-package -s saucy --ppa=canonical-x --ppa-name=x-staging --to-primary --to-suite=saucy-proposed " [09:47] except the arm stuff, but I can ask tjaalton for that [09:47] if that's useful [09:47] yeah [09:52] unity can be uploaded at this point, btw [09:57] mlankhorst, better to wait a bit, the binaries are not published yet and I'm not sure their merger/jenkins does depwait [09:58] ah [10:15] now I just need to wait for xorg-server to show up in proposed then before I can do the rest, it's mostly a rebuild bump from here on [10:25] mlankhorst, if you have updated build-depends those will depwait ... if not, you better wait yes ;-) it's built on i386/amd64, depending if armhf and ppc finish before :30 you might need to wait 1 or 2 extra publisher cycles [10:25] e.g ~1h [10:26] if the past build times are any indicator there is still a good half an hour of build on armhf, so you can go for lunch easily ;-) [10:26] can I create a version number lower than 0~ ? [10:26] would 0~~ work? [10:27] yep [10:27] oh good [10:27] try dpkg --compare-versions 0~ gt 0~~; echo $? [10:27] laney@iota> dpkg --compare-versions 0~~ lt 0~ && echo yep ~/temp [10:27] yep [10:28] ah great, it should be possible to force an upgrade then :P [10:28] force an upgrade of what? [10:31] seb128: mind approving? https://code.launchpad.net/~didrocks/libusermetrics/simplify-packaging/+merge/172994 [10:31] (while I'm are it :p) [10:33] didrocks, done [10:33] thx [10:34] Mirv, bah, the CI doesn't use proposed... https://jenkins.qa.ubuntu.com/job/unity-saucy-amd64-autolanding/75/console [10:34] going to be fun :p [10:35] that's about https://code.launchpad.net/~timo-jyrinki/unity/move-pointer-barries-to-xi-1.6.99.1_try2/+merge/172975 [10:35] it's one of those cases here it would be easier to just upload the patch to saucy-proposed manually and merge back [10:35] but didrocks is not going to like that suggestion I'm sure :p [10:36] seb128: we can push it to trunk directly rather [10:36] and merge manually [10:36] then quick a daily [10:36] the tests are using proposed [10:37] kick* [10:37] didrocks, right [10:37] Mirv, ^ can you do that? [10:40] seb128: ok.. feels evil to push something to trunk but ok :) the cu2d will run all the tests anyhow that's true [10:41] Mirv, thanks [10:51] ok, it's there, launching a build of unity in cu2d === MacSlow is now known as MacSlow|lunch [10:57] built successfully, it's running autopilot now (1h) [10:59] I mean prepared successfully, building + autopilot next :) [11:10] Mirv, sorry, I was grabbing a bite [11:10] Mirv, great, let's see how that goes [11:14] amd64 built https://launchpad.net/~ubuntu-unity/+archive/daily-build/+build/4769091 , dependencies seems correct. powerbc built as well. [11:23] Mirv, great [11:29] mlankhorst: what does 'Xlib: extension "NV-GLX" missing on display ":0".' mean and is it caused by new X? :-) [11:30] I noticed that I was getting weird corruption in QML apps: http://ubuntuone.com/7ibGVZ0LQ1yt1SqQohjkMb and that message appears on the console [11:35] Laney: your nvidia drivers are messedup, mostly [11:35] did you manually install them, by any chance? [11:35] dunno! [11:36] Laney, why do you hate Ubuntu, why not using the packages! [11:36] ;-) [11:36] use Mir ! [11:36] I appear to have nvidia-319 installed [11:37] ogra_, hey; I think I forgot to say it yesterday, congrats on getting the container flip by default landed ;-) [11:37] hey, thanks :) [11:41] mlankhorst: looks much better using nouveau :P [11:41] mlankhorst, xserver-xorg-core | 2:1.14.1-0ubuntu0.8 | saucy-proposed | amd64, armhf, i386, powerpc [11:41] mlankhorst, you can upload the drivers I think [11:41] lightdm does not, however - that has the wrong resolution and is mirrored [11:42] (that's "rmadison -S xorg-server -s saucy-proposed") [11:42] Laney, right, we default to mirror... [11:42] maybe it was always mirrored :P [11:42] but it had the right res with nvidia [11:42] oh well [11:42] nvidia has its own config no? [11:42] moar freedom for me [11:43] Laney, cp .config/monitors.xml /etc/gnome-settings-daemon/xrandr/monitors.xml [11:43] I don't know how to configure lightdm [11:43] that's what I did here [11:43] did I ever do that? [11:44] seb128: ah yes, very nice - that's how it looked before [11:44] ;-) [11:44] what voodoo was nvidia doing ... [11:44] is their config tool running as root? [11:44] did it write config in /etc? [11:44] or maybe they default to non-mirror [11:45] anyway, system-settings looks right again [11:45] great [11:45] do you ever see it segfault on start btw? [11:45] yes [11:45] like one every 10 start or something [11:46] just did it three times in a row [11:46] * Laney gets qt debug symbols [11:46] seems rather a qt bug than a system-settings one though [11:48] Laney, is the stacktrace for your also loading an icon? [11:48] the rebuilt unity would seem to segfault for me and autopilot test seem like concurring (at least it's respawning, no segfault message) http://10.97.0.1:8080/job/autopilot-saucy-daily_release/329/label=autopilot-intel/console [11:48] #0 QPMCache::createKey (this=this@entry=0x7fffd0024560) at image/qpixmapcache.cpp:396 [11:48] I update the minimal amount required to manually install the unity from daily-build PPA [11:49] d [11:49] Mirv, crap, do you have a stacktrace of the segfault? [11:49] seb128: http://pastebin.ubuntu.com/5843267/ [11:50] Mirv, can your ldd -r /usr/lib/compiz/libunityshell.so and pastbin it? [11:50] * seb128 has a deja vu here [11:52] seb128: right, http://pastebin.ubuntu.com/5843270/ [11:52] undefined symbol: XFixesSelectBarrierInput (/usr/lib/compiz/libunityshell.so) [11:52] Mirv, dpkg -l | grep xfixes? [11:53] seb128: http://pastebin.ubuntu.com/5843275/ [11:53] mlankhorst, ^ do you have any idea on that? [11:54] Mirv, are you sure you have the patch? [11:54] Mirv, https://code.launchpad.net/~timo-jyrinki/unity/move-pointer-barries-to-xi-1.6.99.1_try2/+merge/172975 has [11:54] - XFixesSelectBarrierInput(dpy, DefaultRootWindow(dpy), 0xdeadbeef); [11:54] + barrier_ = XFixesCreatePointerBarrier(dpy, [11:54] seb128: sounds like you were using old unity, still [11:56] Mirv, dpkg -l | grep unity? [11:56] Mirv, https://launchpad.net/~ubuntu-unity/+archive/daily-build/+packages?field.name_filter=unity&field.status_filter=published&field.series_filter=saucy has [11:56] unity - 7.0.1+13.10.20130704.1-0ubuntu1 [11:56] just a second [11:56] Mirv, it should be 7.0.2 [11:57] I thought the Breaks was supposed to prevent that happening [11:57] ah, so it should, argh (7.0.2) [11:57] Laney, he didn't update xserver [11:57]