[03:18] Good morning [03:37] pitti: okay... this is just getting ridiculous now [03:37] desrt: hey, how are you? [03:37] time to go to bed? :-) [03:37] good :) [03:37] no. it's not! [03:37] you wake up too early :p [03:38] "get woken up" is more accurate [03:38] did you spawn? [03:38] but it's light outside, and I've slept 7 hours, so I can just as well get up [03:39] ah. didn't think so :) [06:01] hey [06:02] * didrocks wonders if he's the only one to not be able to start the unity session and can start the gnome-shell, but freeze after few seconds [06:04] bonjour didrocks [06:04] didrocks: hmm, perhaps I shouldn't try to restart my desktop session after this morning's dist-upgrade? :-) [06:05] pitti: it was yesterday's upgrade TBH :) [06:05] didrocks: no, I can start a session for another user just fine [06:05] hem, that's really weird… [06:06] didrocks, gnome-shell is broken [06:06] darkxst: ok, but I'm still kicked out unity when logging in :) [06:06] some combination of gtk and/or clutter updates [06:08] interesting, lightdm things that authentification failed (even if I could enter the gnome-shell session) [06:13] downgrading lightdm didn't fix it [06:13] * didrocks continues digging [06:14] robert_ancell: hey, if you are still around, anyway to have more pam info when the authentification failed through lightdm? [06:14] * didrocks still wonders why he can log in a gnome-shell session [06:17] nevermind, the failed authentification should be something it's trying beforehand [06:20] it doesn't even seem to get to the point to start gnome-session… [06:25] even weirder: cp -a /usr/share/xsessions/ubuntu.desktop /usr/share/xsessions/ubuntu2.desktop -> this one works… [06:28] * didrocks is on his ubuntu2 session, changing in /etc/X11/Xsession.d/ to have ubuntu2 detected instead of ubuntu for now [06:56] good morning [06:57] salut jibel! [07:01] salut didrocks ! [07:04] fun! [07:05] * didrocks clearly has no idea what's going on… [07:05] didrocks: but will unity still be loaded in that case? considering unity support script checks for DESKTOP_SESSION=ubuntu [07:05] mlankhorst: I changed it in /etc/X11/Xsession.d/65compiz_profile-on-session [07:05] to load the right profile [07:06] but everything with DESKTOP_SESSION in path containing "ubuntu" either doesn't exist or have nothing relevant [07:06] ah [07:06] and not having gnome-session logs seems to show that gnome-session even doesnt' get any output [07:06] * didrocks should maybe right it with --debug === alan_g is now known as alan_g|afk [07:51] hey desktopers [07:52] Hello! [07:53] bonjour [07:57] oh, just for the record we debugged didrocks' login issue [07:57] https://bugs.launchpad.net/ubuntu/+source/upstart/+bug/1184878 [07:57] Ubuntu bug 1184878 in upstart (Ubuntu) "when upstart init the session jobs, it should be robust against environment changes (fallback?)" [Undecided,New] [07:57] mlankhorst: pitti ^ [07:57] it was due to a broken PATH in /etc/environment (not listing /sbin which makes "init" not be found) [07:57] it's quite brutal though :p [07:58] hey sil2100! [07:58] didrocks, lut ;-) [07:58] psh [07:58] hey seb128 ;) [07:59] shrug [08:00] gnome-desktop3 transition [08:00] we need to plan those transitions better [08:00] <- not happy about this one [08:00] darkxst, ^ [08:01] seb128, its kind of urgent now, that gnome-shell is completely broken [08:02] hello [08:02] * Laney phases back in [08:02] darkxst, how come in all those weeks/months gtk 3.8 was being prepared nobody tested gnome-shell? and you guys were the ones pushing to get the new gtk in to unblock stuff [08:02] that's ... no comment [08:02] seb128, we were testing it against 3.8 [08:03] shrug [08:03] and actually it maybe clutter and/or a combo with gtk that broke things [08:03] nobody has been testing gtk 3.8 on the archive version of gnome-shell before pushing for it to be uploaded? [08:03] well, in any case you guys not doing proper testing is not a reason to force everyone into a rushed transition [08:04] [08:04] " * Dropped changes: [08:04] - 02_refuse_to_break_GL_compositors.patch: [08:04] + Doesn't appear to be needed any more" [08:04] [08:04] I also doubt that's true [08:04] why is it not needed any more? [08:04] [08:04] Laney, hey, welcome back [08:04] I brought the rain back with me [08:05] keep it then (we are having sun since yesterday and would like to keep it this way :p) === alan_g|afk is now known as alan_g [08:06] seb128, not sure about that one, jbicha did that [08:07] it's not actually dropped [08:07] dpkg-source: info: unpacking gnome-desktop3_3.8.2-0ubuntu1.debian.tar.gz [08:07] dpkg-source: info: applying 02_refuse_to_break_GL_compositors.patch [08:07] Laney, is it in the serie? [08:07] yeah [08:09] hum [08:10] darkxst, I'm pondering rejecting that new gnome-desktop3 from the NEW queue (well the binaries, the source is in) [08:10] I'm really annoyed at how that was unprepared [08:10] darkxst, also please file proper merge requests for things you work on, I'm not going to go pull things from a ppa, unpack locally to figure the diff and do the work to manually copy the diff in the vcs [08:11] seb128, I did it more for testing, but then jbicha grabbed it and uploaded [08:12] darkxst, sorry you take the comments but he's not online... [08:12] has the transition been checked? [08:12] with gtksourceview I ended up needing to do some patching [08:13] Laney, it's not properly organized and not ready, which is what I'm complaining about [08:13] not organised I can deal with after complaining a bit but not ready is more worrying [08:14] Laney, I have tested locally, gnome-desktop3, g-s-d, g-c-c and nautilus [08:14] shrug [08:14] it's like 1/4 of the transition [08:15] I mean rejecting creates its own problems so it's not as if that's a very pleasant way out [08:15] seb128, I built all the other packages, but didnt try installing them just yet [08:15] yeah, I would go with a 3.8.is.really.3.6 if I do that [08:16] Laney, the transition is limited enough in number of packages that I think we should go through if they got g-s-d/g-c-c/nautilus sorted out [08:16] but I'm still unhappy that it was thrown that way in without discussion on the channel before and without being properly prepared [08:16] * Laney nods [08:19] * Laney wanted to look at evo/eds this week btw [08:19] Laney, uoa support \o/ [08:21] shrug, libgweather changed soname as well and nobody is taking care of the rebuilds either there [08:22] ok, I guess it's going to be "deal with GNOME transitions day" [08:25] seb128, didrocks: oh, interesting; I guess it's time to symlink init to /bin then :) [08:25] oh, just remembered that I wanted to tell didrocks that I'm coming on holiday to his city! [08:26] pitti: yeah, that seems a good plan as well :) [08:26] Laney: oh? when are you coming? :) [08:26] pitti, salut, ça va bien ? [08:27] seb128: oui, merci ! et toi ? [08:27] pitti, ca va bien ;-) [08:27] nous avons le soleil à nouveau \o/ [08:28] ici aussi (mais la pluie revient demain) [08:28] didrocks: August ... 3rd I think [08:28] la pluie revient cet après-midi ici :) [08:28] I looked at a map of where Debconf is (the week after) and picked somewhere not too far away for the previous week [08:28] Laney: ah, if I'm not at Guadec, let's organize something :) [08:28] for sure [08:29] where is debconf? [08:29] seb128: en Suisse apparemment: http://debconf.org/ [08:29] interesting [08:31] about 1h by train from Geneva [08:33] meh, and again right in the summer vacation :/ [08:48] seb128, good morning, would you have two minutes to give an okay to bug 213539? [08:48] Launchpad bug 213539 in ubuntu-meta (Ubuntu) "cheese is not a dependency of ubuntu-desktop" [Wishlist,Confirmed] https://launchpad.net/bugs/213539 [08:48] (or a reason for a not-okay) [08:49] mpt, hey, I need to have a look to its current state of upstream maintenance, bugs, etc before commenting ... I didn't try for a while but it was rather buggy last time I tried [08:49] ^ does anyone else have recent experience with cheese? [08:49] understood [08:50] seems fine here [08:50] I use it when I want to take a snap with the webcam [08:50] is it something we should have on the image? [08:51] we could also build g-c-c with it if we install it by default (it allows you to take a picture for your profile then) [08:51] Laney, I've no strong opinion, we could I guess, it's useful and we have clutter, etc on the CD nowadays [08:51] well, 'fine' - the framerate is a bit low but it's ok for photos [08:55] it has a dep on plugins-bad but that seems wrong [09:12] didrocks, sil2100: can I get https://code.launchpad.net/~seb128/gnome-control-center-unity/background-async-calls/+merge/160850 approved/landing? we need a g-c-c-unity no change upload in any case to pick the new libgnome-desktop soname [09:34] *grmble* i hate it when days start like this ... [09:35] PPA build failed, notes about a serious bug in our released version ... [09:38] seb128, g-c-c has 1 extra patch actually [09:39] darkxst, that was g-c-c-unity (the appareance capplet) I was talking about [09:39] seb128, ah ok, that has no changes [09:41] * popey wonders why he keeps getting signon_ui popups and browser tabs opening strange facebook URLs... [09:42] popey, https://developers.facebook.com/bugs/449221825171392 [09:42] sometimes I want to club our toolchain with a big trout ... repeatedly. First launchpadlibrarian rejects a patch with "would create existing file", which seemed to be true (but then should be reported by dpkg already), and when removing that patch the build breaks with this file missing. wth? [09:44] Sweetshark, what's the librarian error? [09:44] launchpadlibrarian vs. dpkg !?! [09:45] seb128: "The next patch would create the file liborcus/liborcus_0.1.0-boost-include.patch.1 which already exists!" [09:46] seb128: landing will wait for jenkins to be back, however, sil2100: do you mind reviewing it? [09:46] Sweetshark, can you point to the error log/email? that seems weird [09:46] sil2100: you seem to be really away lately or not looking at IRC :p [09:46] I havent quite figured out what goes wrong there, still on it. [09:47] ah, that sounds like quilt or dpkg, nothing to do with the librarian [09:47] the url will be a librarian one, but that's just where it's stored [09:47] unclean tree, or patches need to be updated to a newer version? [09:47] didrocks: ;p [09:47] didrocks, ok, thanks [09:47] didrocks: sorry about that ;p [09:48] sil2100: what are you working on? needing help? [09:48] I'm in packaging and paperwork frenzy [09:48] sil2100: packaging python-evdev, is it done? :) [09:48] didrocks: no no, just filling out all the paperwork for sponsoring [09:48] sil2100: hum? [09:48] didrocks: yes ;) Looking for mentors who could sponsor it in debian ;) [09:48] seb128: are you handling the rebuilds for gnome-desktop3? [09:48] ooh thanks seb128 [09:48] sil2100: but I think we shouldn't wait on that and still upload to saucy [09:49] didrocks: the upstream author wanted it to land in debian also, so I was hoping to push it to ubuntu through the sync - but I can do both if needed [09:49] Laney, so of those, but the binaries are not actually published yet so I just did a few uploads for nothing :/ [09:50] didrocks: was just following the guidelines, as we're not really upstream for python-evdev [09:50] heh [09:50] didrocks: I also filled in all the paperwork for the 100-scopes [09:50] sil2100: as this is on the path to get touch to saucy, I think we should land it, and then sync once available :) [09:50] sil2100: oh, the MIR is done? :) [09:51] didrocks: for now needs-packaging, I talked with cyphermox and he recommended to fill in MIR once MOTU approve the needs-packaging bug and ACK it into universe [09:51] we have quite a few Ubuntu people in the debian python teams - sponsoring shouldn't be hard to come by [09:51] https://bugs.launchpad.net/ubuntu/+bug/1184656 [09:51] Ubuntu bug 1184656 in Ubuntu "[needs-packaging] Unity 100-scopes" [Wishlist,In progress] [09:51] sil2100: hum? [09:51] sil2100: we already have one bug with all the components [09:51] for the 100 scopes [09:52] sil2100: and you don't really need that as we all synced on how to do it [09:52] didrocks: ok, so that's something Mathieu did not mention to me - so I should simply do a MIR straight away? [09:52] sil2100: bug #1154229 [09:52] Launchpad bug 1154229 in unity-scope-gdrive (Ubuntu Saucy) "New Unity Dash" [Undecided,New] https://launchpad.net/bugs/1154229 [09:52] sil2100: yeah [09:52] sil2100: the MIR is better [09:53] didrocks: since the previous bug is not a formal needs-packaging bug [09:53] the other is to track what we are adding and removing [09:53] sil2100: we don't really needs-packaging in ubuntu for what canonical package, as you know archive admins :) [09:53] didrocks: and has outdated branch-data and packages [09:53] sil2100: so we already have the context [09:53] Ah [09:53] Ok ;) [09:53] Then I'll just quickly do a similar MIR bug then [09:53] sil2100: rather update bug #1154229 with the new branches [09:53] Launchpad bug 1154229 in unity-scope-gdrive (Ubuntu Saucy) "New Unity Dash" [Undecided,New] https://launchpad.net/bugs/1154229 [09:54] and yeah, do a MIR [09:54] sil2100: meanwhile, I can review python-evdev [09:54] do you have a branch? [09:55] didrocks: not a branch, since I have been working on making it debian-compatible ;) Mirv cross-checked it already, but let me paste you the mentors link [09:55] Since it's basically that [09:55] sil2100: source package is good :) [09:55] (but the branch is orthogonal, you can still maintain in a branch, even for debian :p) [09:56] didrocks: yes, I will do that too ;p [09:56] didrocks: http://mentors.debian.net/package/evdev [09:56] didrocks: so, http://mentors.debian.net/debian/pool/main/e/evdev/ [09:57] sil2100: thanks, I'll change the version to have -0ubuntu1 and will get back to you :) [09:57] sil2100: if good, I'll upload a NEW it [09:57] ;) [10:01] sil2100: priotiy should be optional not extra [10:02] sil2100: any particular intend to not use debhelper 9? [10:02] (debian/compat and build-deps) [10:02] didrocks: not really, let's use that then [10:05] sil2100: you can reupload to mentors as many times as you want, so just dput again after changes [10:06] Mirv: did you check it already? [10:06] sil2100: no COPYING in the upstream tarball [10:07] sil2100: in fact, no copyright at all in any source nor a copyright file [10:07] Mirv: thanks \o/ [10:08] didrocks: there is no copyright in any source file, yes, but the git branch has a LICENSE file which states that all the sources are New BSD [10:08] sil2100: this should be shipped in the tarball then [10:08] didrocks: but the upstream developer has an invalid manifest file and those don't get uploaded into the tarball [10:08] sil2100: because the license isn't trivial when you get the source [10:09] didrocks: not very well, I trust too much lintian and helped addressing those warnings [10:09] sil2100: so, before creating the source tarball, better to patch it upstream and get a new one [10:09] didrocks: I have a branch I'll recommend to him [10:09] Mirv: I thought DD will make more check than that :p [10:09] didrocks: at first I used a proper tarball I created myself, with a modified manifest file with all the docs and LICENSE installed [10:10] didrocks: but as Mirv properly mentioned, to get it to debian we have to use an actual tarball that has been released alreay ;p [10:10] *already [10:10] sil2100: you can distro-patch with that, telling you are taking that commit from upstream [10:10] didrocks: ACK [10:10] better than a source that can't be accepted :p [10:10] nor in ubuntu or debian [10:10] ;) [10:11] didrocks: it wass a prgmatic "upstream tarball matches, packaging works", I don't have a good systematic routine of checking [10:11] should have though [10:11] Mirv: yeah, take care of licensing in particular, you tend to forget that one :) [10:11] ok, not dangerous code otherwise [10:12] Mirv: sil2100: once you get the tarball with those fixes, I'm happy to upload to saucy and NEW it [10:12] didrocks: will do! [10:12] Thanks :) [10:12] yw :) [10:12] * sil2100 fixes all the things mentioned and other stuff [10:13] pitti, seb128: those where the patches I got from xnox. there was a foo.patch and a foo-part2.patch -- with the second creating the same file as the first did -- so it was rejected for PPA upload. [10:14] pitti, seb128: so I removed this second patch, and now the build fails because the file is not created at all. [10:14] weird [10:15] darkxst, thanks for the merge requests, I commented on those, you are targetting the wrong Vcs ... also I think you make changes that means you can't build with gnome-desktop 3.6 so you need to bump the build-depends to >= 3.8 [10:16] Sweetshark: does the first patch actually create said file? [10:18] Sweetshark: all of my patches are cherrypicks.... so if you're making beta2 they shouldn't be necessory. it's something to do with how the package has debian patches - and patches in the tree applied by the build system. I first only applied foo.patch, then watched it fail to build & added whatever dpkg complained to me about into a second patch. [10:18] * xnox was confused about it as well. [10:21] xnox: this is still the 4.0.2 ftbfs fixup, but indeed I might just go directly for the beta2. [10:23] pitti: https://launchpadlibrarian.net/140879339/buildlog_ubuntu-saucy-amd64.libreoffice_1%3A4.0.2-0ubuntu2_FAILEDTOBUILD.txt.gz says the build is missing the file, but the 4-faf72664d11deaa570a0f8c7e702ff07bf5914c8.patch in the https://launchpad.net/~bjoern-michaelsen/+archive/libreoffice-3-4/+files/libreoffice_4.0.2-0ubuntu2.debian.tar.gz seems to have it and the log says its applied .... [10:29] seb128, I don't believe any of those changes require 3.8 [10:29] darkxst, ok, good, I though some of the .h got renamed in the new version [10:30] one header has been removed [10:33] does anyone know what the filtering of -Bsymbolic-functions in eds was for? cyphermox? [10:33] cyphermox, hey, could you push your eog changes from the other day (not sure if you didn't use the vcs or just forgot to push) [10:33] Laney, it was having runtime issues otherwise, things with plugins tend to not like -Bsymbolic-functions much [10:34] ah yes, I forgot Debian doesn't have that in their LDFLAGS [10:34] keeping that one then [10:35] Laney, see e.g https://bugzilla.gnome.org/show_bug.cgi?id=594473 [10:35] Gnome bug 594473 in Miscellaneous "-Bsymbolic-functions linker flag causes problems at runtime" [Normal,Unconfirmed] === vrruiz_ is now known as rvr [10:53] pitti, seb128: quilt/dpkg is to stupid to read git patches. [10:53] didrocks: hey. will you have time to check/upload the precise SRU packages this week, or next week (when your next patch pilot turn is)? [10:53] Mirv: I synced with dbarth, seems it can wait my patch pilot turn [10:54] pitti, seb128: (heh, but its still complaining that it will double create a file that it didnt create at all) [10:54] didrocks: alright, thanks [10:54] yw :) [10:55] pitti, seb128: quilt doesnt understand the "old filename is /dev/null"-format of git ... [10:55] aah [10:55] IMHO it should [10:55] I just got an one-line patch after two hours of debugging; I know your feeling [10:56] and during that, it seems I accidentally removed my GPG secring, argh! [10:56] "old filename is /dev/null for newly created files" that is ... [10:56] xnox: ^^ that was the cause of the mystery. [10:58] * Sweetshark gives pitti a hug! [10:58] * pitti hugs back Sweetshark [10:58] there, I can't upload anything any more [10:58] Sweetshark: interesting. well newer gnu patch suppose to support git-style patches better. not sure if we have it, and weather quilt/dpkg uses gnu patch directly to benefit from that feature. [10:58] * Sweetshark steals some more gpg keys while hugging. [10:58] it seems over the years I lost both of my physical backups, and of course it's not in any of my automatic online backups [10:59] your secret keyring? [10:59] ouch! [11:01] got a revcert at least? [11:01] no, that was on said backup [11:01] * pitti scratches head, thinking where he could still have a backup [11:01] I wonder if my printout still exists ... [11:02] I think I have a CD at my parent's house [11:02] "just" 600 km away now [11:02] * xnox was pondering to use high-dencity QR codes for rev-certs & key backups, because it's machine readable and has redundancy in case of ink fading etc.... [11:02] xnox: great idea [11:02] pitti: I'm sure they will be happy to see you =) [11:02] heh [11:04] xnox: Great idea with the QR code. And then someone comes along and offers a "simplified howto": just use a webbased QR-code generator for your l33t secret keys ;> [11:05] Sweetshark: there are web-based secret ssh & gpg key generators, which offer you to "download" your key. =)))))))) [11:06] surely not :( [11:06] xnox: to make it buzzword compliant, you should say it "backs up your data in the cloud!1!" [11:06] (with cloud being pastebin in this case) [11:08] Laney: some are better than others. e.g. amazon ec2 offers to generated ssh keys, but well ec2 can own you anyway by proxying all traffic to the VM, so this is not new. But I've seen "generic" private key generators on the web before. === MacSlow is now known as MacSlow|lunch [11:22] the new series nickname results in me having a jenkins job named "saucy-source". Maybe my Jenkins should start a boyband called "Saucy Source and the Funky Bunch" ... [11:36] way to ruin a nickname for me [11:47] hum [11:47] my gtk builds in the ppa failed yesterday with [11:48] "D-Bus library appears to be incorrectly set up; failed to read machine uuid: Failed to open "/etc/machine-id": No such file or directory" [11:48] or no, it's not the reason it failed ... but are those warnings normal? [12:05] they seem to have been around for some time [12:06] but I can't see where that path comes from - in dbus it's defined to /var/lib/dbus/machine-id === alan_g is now known as alan_g|lunch === MacSlow|lunch is now known as MacSlow [12:50] seb128: ok I think I've rebuilt everything needed for gnome-desktop3 except for gnome-control-center-unity [12:51] I've never worked with the magical autolanding branches [12:51] jbicha, hey, thanks, there was some grumpyness around this morning, please stop starting stacked transitions like that without discussing it in the channel first [12:52] I pinged didrocks/sil2100 about g-c-c-unity this morning, they will handle it once jenkins is back running [12:52] (the setup is having some issues) === alan_g|lunch is now known as alan_g [13:05] logging in from lightdm doesnt bring up a unity panel at all anymore (just background)... [13:05] when running unity --replace it complains that unity-panel-service is not found [13:05] any idea? [13:05] * asac runs fluxbox now [13:05] logging in a guest session is fine [13:06] so... if i knew where settings live that tweak how unity starts up, that could help [13:08] asac, can you run unity and copy the output to pastebin? [13:08] seb128: not easy... if i log in the unity session i can only run that from console [13:08] but there i cannot copy paste [13:08] or wait ... i can try [13:09] in fluxbox i cannot run unity [13:09] asac, can't you just run unity from fluxbox? [13:09] why not? [13:09] let me try [13:09] unity [13:09] compiz (core) - Info: Loading plugin: core [13:09] compiz (core) - Info: Starting plugin: core [13:09] unity-panel-service: no process found [13:09] compiz (core) - Error: Another window manager is already running on screen: 0 [13:09] compiz (core) - Info: Stopping plugin: core [13:09] compiz (core) - Info: Unloading plugin: core [13:09] thats all i get [13:09] unity --replace ? [13:09] the error is there because of fluxbox [13:09] same i think ... have to stop fluxbox to get more [13:09] yeah --replace is exactly the same [13:10] try logging into an unity session [13:10] unity-services: /usr/lib/unity/unity-panel-service [13:10] and copy .xsession-errors or ~/.cache/upstart/gnome-session.log somewhere [13:10] i am on raring btw [13:10] ok, so ~/.xsession-errors [13:10] ok let me do that [13:10] first clean that [13:10] then log in [13:10] then copy back [13:10] and come back [13:13] FYI regarding evening's meeting, I'll be at my niece's kindergarten spring party something. I'll take my laptop with me but I cannot guarantee I've access to it at that time (I don't know the schedule). spreadsheet updated. [13:13] ok back :) [13:13] cat .xsession-errors-no-unity | pastebinit [13:13] http://paste.ubuntu.com/5710120/ [13:13] seb128: ^^ [13:14] hmm... I/O warning [13:14] didrocks: ^ not sure if making to the meeting today [13:14] asac, the unity plugin is not loaded [13:14] because of the session IO issue? [13:15] Mirv: ok, as long as you have updated the spreadsheet with all the infos, and there are IRC logs later on :) [13:15] asac, dconf read /org/compiz/profiles/unity/plugins/core/active-plugins [13:15] I/O warning : failed to load external entity "/home/asac/.compiz/session/10d121c0a95afbd308136974671220768100000172470031" [13:15] asac, I don't think so, that warning is harmless afaik [13:15] dconf read /org/compiz/profiles/unity/plugins/core/active-plugins [13:15] ['core', 'composite', 'opengl', 'decor', 'snap', 'compiztoolbox', 'copytex', 'vpswitch', 'resize', 'grid', 'move', 'gnomecompat', 'mousepoll', 'place', 'imgpng', 'regex', 'session', 'animation', 'workarounds', 'wall', 'expo', 'fade', 'ezoom', 'scale'] [13:15] asac, try to dconf reset /org/compiz/profiles/unity/plugins/core/active-plugins [13:15] then get again, does it list unity? [13:16] after reset its all empty now [13:16] 'unityshell' [13:16] ok [13:16] returns nothing [13:16] try restarting your session [13:16] ok... [13:16] under unity [13:16] see if it works [13:17] seb128: i am back :) [13:17] thanks so much [13:17] i still feel this is flaky shit :) [13:17] but ... i accidentially installed the phablets ppa :) [13:18] thats where my adventure started 5 hours ago :) [13:18] we... 3-4h [13:18] haha [13:18] so i cannot hold it against anyone [13:18] i just hope i amnaged to get everything back now [13:18] asac, yeah, compiz is flaky :/ but we are replacing it in unity-next ;-) [13:18] at least i have a fresh ubuntu-desktop stack ... stripped everything down to ubuntu-minimal and reinstalled ubuntu-kdestop [13:19] desktop [13:19] guess got rid of 1k packages that i didnt want anyway [13:19] hmm. my sip client is gone :) ... one collateral victim i guess [13:20] ok let me doo a clean reboot and see if everything is good now [13:24] seb128: very thankful!! [13:25] hehe [13:25] while we are at it, i guess you have no idea why lightdm doesnt use my native 1080p resolution? [13:25] asac, glad it worked ;-) [13:25] asac, do you use a docked laptop or something? my guess would be that it's in mirror mode [13:25] 1920x1080 would be correct, but its running on something 1024ish [13:25] its docked thinkpad, ues [13:25] yes [13:25] yeah, it's mirror :/ [13:25] yeah you are right [13:25] that's annoying [13:25] it uses the resolution of my thinkpad :) [13:26] i remember i already guessed that as the reason at some poitn ... ok [13:26] guess nothing for me to do here then :) [13:34] * ogra_ would think about a reinstall [13:35] using the ubuntu touch PPAs on a desktop install might have broken more than you see at first sight [13:35] ogra_: well... you never know ... [13:35] i certainly dont have those packages anymore :) [13:35] but they might have hackist maintainer scripts :) [13:36] *hackish [13:36] *shrug* [13:36] my desktop is up and running [13:36] :) [13:36] ok [13:36] i will see what happens on suspend [13:36] :) [13:36] asac, try to "sudo cp /home/asac/.config/monitors.xml /etc/gnome-settings-daemon/xrandr/monitors.xml" [13:37] asac, that will make the unity greeter use the same screen config as your user session [13:40] seb128: not sure i wnat that ... if i undock etc. it will be bad i guess [13:41] hmm. maybe not? [13:42] Laney: seb128: correct, that's the reason for filtering out -Bsymbolic-functions [13:42] asac, that's the default, it will behave like your session does [13:44] cyphermox: righto, thanks [13:45] didrocks: I think I got all the things you have pointed out [13:45] didrocks: I also pushed the packaging as a bzr branch, and sent a pull request to upstream for the LICENSE inclusion [13:46] didrocks: the bzr branch is https://code.launchpad.net/~sil2100/python-evdev/debian [13:46] didrocks: but also the same thing is pushed to mentors ;) [13:49] sil2100: ok, I'll give it a look, please update the spreadsheet === m_conley_away is now known as m_conley [14:09] ok, I'm out for some exercice, back for the team meeting === davmor2_ is now known as help === help is now known as davmor2 [14:22] didrocks: hm, regarding the MIR thing... === alan_g is now known as alan_g|tea [14:22] didrocks: how am I to fill the MIR bug if I can't fill in a bug for the distro project of the scopes? [14:23] sil2100: just file it against unity, add the tasks I added for the existing one on the FFe bug [14:23] and write for the other components [14:23] that will help tracking [14:23] if that suits mterry ^ [14:23] seb128: thanks for the sponsor, should I just remove those Merge requests since I did the wrong target? [14:23] mterry: ping ^ [14:24] sure... you're saying that the scopes don't have ubuntu packages yet, so can't file bugs against them? [14:24] sil2100, ^ [14:25] mterry: exactly, I was preparing an MIR bug for all of those, but I can't since there are no targets for those yet [14:25] mterry: so I can simply submit that to unity (Ubuntu), yes? [14:26] sil2100, sure, for now. But they are also going through NEW right now right? (to get into universe) [14:26] mterry: yes, they should be - didrocks and others had a bug for that already, but for instance yesterday I filled another one by mistake for the MOTU guys (;p) [14:27] mterry: you'll finish the MIR on it so that on the same day, we can upload that, get the scopes out of proposed and live in a pieceful world? :p [14:28] didrocks, there was already a MIR bug for this, right? [14:28] didrocks, that I partly filled out? [14:28] partly reviewed, that is [14:28] mterry: I think so [14:28] mterry: not really sure if it was the FFe bug or another one [14:31] Mirv: sil2100: cyphermox: fine to postpone the meeting to tomorrow? ken is still away, robru on holidays and Mirv isn't sure to be there today [14:31] sil2100: cyphermox: Mirv: so proposing tomorrow, same time? [14:32] didrocks: ack [14:32] seb128: FYI if you lurk at our meeting ^ [14:33] didrocks: hmm, ok, I'll do something and do my best to be on it! [14:33] ok, moving it, thanks :) [14:34] Ok, launchpad is getting on my nerves [14:34] Second time it timesout on submitting the MIR bug and I had to create it from scratch [14:34] sil2100: look if you don't find one already from what mterry told ^ [14:36] didrocks: I was browsing launchpad for a MIR bug related to 100 scopes but did not find any, maybe if it was not marked as MIR? [14:36] -if [14:36] didrocks: ok [14:37] mterry: I anyway filled one in, feel free to mark it 'duplicate' if you find an existing branch already === alan_g|tea is now known as alan_g [14:37] sil2100: me neither, just yours :) === m_conley is now known as m_conley_away [14:55] seb128: ohh I'm so tempted to make a bike ride now instead of waiting until after meeting, any objections if I miss the meeting? [14:57] status report: sent out patch v4 of locking patch to upstream, preparing xorg-server sru to raring with fixes for nvidia prime, x1.14 is going to be copied to archive together with the next unity, posted dm application to debian-newmaint [14:59] and tested upstream nouveau, some hardware that didn't work before now started working, and my own system now suspends properly \o/ === tkamppeter_ is now known as tkamppeter [15:21] does ubuntu remove /tmp files at shutdown or at boot? [15:21] czajkowski or stgraber: Would you be so kind as to rescore evolution-data-server in https://launchpad.net/~laney/+archive/evolution-3.8 for me please? I want to stage the transition in there. [15:22] the thing is, I have files I want to keep in /tmp but the laptop ran out of battery (stupid me), if I boot it, it will remove files? I should boot from liveusb? [15:22] didrocks, ^ [15:22] * xclaesse just hopes its not a tmpfs in memory :/ [15:23] xclaesse: at boot, see /etc/init/mounted-tmp.conf [15:23] ok, liveusb then :) [15:23] thanks [15:24] Laney: it only has to wait 16 mins [15:24] wtf [15:24] I looked a minute ago and it said 6 hours [15:24] * czajkowski peers at Laney [15:24] srsly [15:25] now it says 1 hour [15:25] Laney: done [15:25] it said 1h and 3h here [15:25] merci [15:25] maybe czajkowski saw the 16 after you rescored [15:25] although that should be "in a moment" [15:25] never mind [15:26] Laney: does folks also need rescoring? [15:26] nah [15:26] depwait should take care of empathy building against the right thing [15:29] re, just on time for the meeting [15:30] Sweetshark, qengho, Laney, mlankhorst, tkamppeter, attente, desrt, larsu: hey, it's meeting time [15:30] \o [15:30] https://wiki.ubuntu.com/DesktopTeam/Meeting/2013-05-28 [15:31] o/ [15:31] ok, first let's welcome larsu to the team [15:31] he's officially desktop team since yesterday [15:31] w00t [15:31] :) [15:31] thanks seb128 [15:31] Laney: raising the right arm like that looks ... icky. [15:32] oooh [15:32] welcome larsu! [15:32] Sweetshark: that's my left arm, you're just standing behind me [15:32] * Laney coughs [15:32] really? Welcome larsu, if you havent been a member you have been an excellent spy before ;) [15:32] ok, let's get started [15:32] Sweetshark, weekly status update? [15:33] saucy ftbfs fix, big-merge from Debian for 4.1 beta, some security issue handling, been at LinuxTag, meeting with LiMux personell and a member of the federal german parlament [15:33] EOF [15:33] how is saucy ftbfs going? some people would like a new poppler to be uploaded but libreoffice rebuild will be in the way of that transition... [15:34] Sweetshark, ^ [15:34] seb128: well, build is running locally, should be finish in a few minutes. [15:35] seb128: I would throw it at a PPA then for bdrung to sponsor from there then. [15:35] you could stage up teh rest of the transition in -proposed? [15:35] blocking stuff in proposed for an indeterminate length of time isn't very desirable [15:35] seb128: how is libreoffice in the way of poppler? [15:36] Sweetshark, britney enforce all the rdepends to be rebuilt to pick the new soname when a library soname change [15:36] Sweetshark, and libreoffice is linking to poppler [15:36] seb128: ah, because they need a rebuild and ... yeah [15:37] pitti, I don't want to have stuff blocked in proposed for 3 weeks if that's the time it take to defeat libreoffice [15:37] seb128: so from my POV the fix should be there RSN -- that is today. [15:37] seb128: *nod* [15:37] Sweetshark, great, thanks [15:37] qengho, hey [15:37] seb128: my jenkins says "dpkg-deb: building package `libreoffice-dbg' in `../libreoffice-dbg_4.0.2-0ubuntu2_amd64.deb'" so its only some 15 more minutes [15:38] \o/ [15:38] no qengho? [15:38] Laney, hey [15:38] hey [15:38] • Short week; public holiday [15:38] • Finished msva-perl Xsession.d replacement user job and uploaded - required some strange upstart contortions that might be educational if you want to check out the diff. Think the archive is clean wrt. Xsession.d and user session jobs. [15:38] • Uploaded gnome-session with a fix allowing for any gnome-session to be run under user sessions, instead of just the ubuntu (unity) one as it was before. [15:38] • Bit more patch piloting [15:38] • Most of the rest of the week was spent on doing GNOME updates, including helping out with the gtksourceview transition (some packages needed patching), testing gtk3.8 and prodding at gnome-terminal (prereqs done; waiting for some kind of resolution on the vte.sh issue). [15:38] • Currently staging an eds (+ folks + evolution) migration in https://launchpad.net/~laney/+archive/evolution-3.8 - look out for a call for testing this week. [15:38] • DMB: started a thread and pad about a proposal to drop the coupling between Ubuntu membership and upload rights: https://lists.ubuntu.com/archives/devel-permissions/2013-May/000489.html [15:38] eof [15:39] Laney: I'll be happy to help test evo [15:39] Laney, speaking about DMB, do you know if Sweetshark's application is going anywhere? [15:39] Laney: did you end up taking the new gnome-terminal with the --disable-asshole thing? [15:39] Laney, I'm that close > < to just add him to desktop set so he can upload [15:39] desrt: hahaha, not just yet [15:39] or was this just 3.8 catchup still? [15:39] that was vte wasn't it? [15:40] g-t, i thought. but maybe? [15:40] bonus points if it was vte -- affects more modules that way, i guess [15:40] hmm, anyway - no [15:40] if gnome-terminal: delayed a bit until doko comes up with a bright idea [15:41] if vte: patched out [15:41] bright idea? [15:41] https://bugzilla.gnome.org/show_bug.cgi?id=697475 [15:41] Gnome bug 697475 in general "New tab is not opened in same directory as previous tab" [Normal,Unconfirmed] [15:41] oh. that. [15:41] thanks [15:41] seb128: bdrung / barry should have/will be contacting with next steps [15:42] and by "just adding" you mean using the proper process :-) [15:42] Laney, ok [15:42] yeah, asking for 3 people to +1 him [15:42] which I'm sure would be something I can find ;-) [15:42] Laney, thanks [15:43] mlankhorst, hey, there or did you go for your bike ride? [15:43] seb128: Sorry, didn't see it. I was away on holiday for the last week. I'm packaging chromium v27 this week. EOF [15:43] I assume he went, next one [15:43] tkamppeter, hey [15:44] - Report is about one week before, last week I was on vacation [15:44] - OpenPrinting Summit 2013: Participated via phone due to denial of ESTA [15:44] - UDS: Sessions about mobile printing, most canceled due to lack of interest of people [15:44] - Private hangout with mobile guys about printing: Printing will most probably go only into the second release of Ubuntu Touch, at OpenPrinting a design for a mobile printing dialog was worked out, made slides about that available to mobile people, prototype needed, I will also continue with mobile features in cups/cups-filters packaging and in cups-filters upstream development. [15:44] - e-mail discussion: Poppler vs. MuPDF looks more like Poppler: Well established and also used by the normal desktop, so better for convergence [15:44] - e-mail discussion: Auto discovery of IPP printers needs to be done in a way that printers do not get woken up, but jobs need to get spooled, problems with CUPS queues. [15:44] - 4 students working for OpenPrinting in GSoC 2013. [15:45] tkamppeter: I really love to hear the story about ESTA some day, read it on http://naruoga-en.blogspot.de/2013/05/openprinting-summit-pwg-face-to-face.html and did not believe it .. [15:45] tkamppeter, thanks [15:46] why's the discussion / slides private? [15:46] Sweetshark, some people told they got 10-year visa for the US, I hope to get one, too, 10 years no need of ESTA ... [15:47] * desrt smiles contently [15:48] tkamppeter, thanks [15:48] attente, hey [15:48] hi seb128 [15:48] worked on the custom widget for setting the keyboard shortcut in the region panel, still having trouble with a couple of keyboard/mouse grab bugs to squash [15:48] spent some time looking into the gvim with unity-gtk-module issue, it's just a missing gtk 2 setting [15:48] looked into larsu's MP for custom menu item factories [15:48] EOF [15:48] attente, unity-gtk-module finally got uploaded to Ubuntu [15:48] attente: where's the part about bzr bd do? :) [15:49] it needs NEW review next, I will do that after the meeting [15:49] seb128, oh cool [15:49] attente, can you write a MIR (https://wiki.ubuntu.com/MainInclusionProcess) for it? [15:49] seb128, sure [15:49] then we just need to get didrocks or mterry to review it and we should see the end of that one [15:49] attente, thanks [15:49] attente, I saw your gdk-pixbuf's mr update, that's also on my list of today (and I uploaded larsu's patch to the desktop ppa) [15:50] oh, ok, thanks seb128 [15:50] desrt, hey [15:50] attente: what gdk-pixbuf work? [15:50] desrt, backport of the 2.29 icon patche"s [15:50] ah [15:50] seb128: hi [15:51] spent a good deal of the week discussing design of action descriptions with larsu and chatting to the PS guys about that and how/if we will use them from QML [15:51] desrt: they made a new proposal, which is almost completely different [15:51] implemented the first part of the patch. public API is done(ish), dbus exporter is done, and proxy creation is working for GMenuItem, at least -- the gtk work (GtkToolButton, etc) can come later [15:52] larsu: are we in london or something? ;) [15:52] :-( [15:52] let's keep the argumentation for after the meeting [15:52] desrt: feels like it. I hung out with them this morning. There's another hangout at 12 with everyone [15:52] spent some time getting the new glib support for installed tests working [15:53] since we now have three cases for tests to worry about: normal, builddir!=srcdir and installed [15:53] Laney, ^ glib 2.37.1 tarball available if you feel like doing the update btw ;-) [15:53] will be landing a patch to deal with that in a hour or so after i'm done making sure it works in all the cases [15:53] oh ya. did a glib release too :) [15:53] general bugfixes, etc.... (eof) [15:53] ;-) [15:53] yeah, seen, tomorrow or when I get pissed off with eds [15:53] desrt, thanks [15:54] larsu, hey, welcome again ;-) [15:54] seb128: hi, thanks again :) [15:54] last week I put out MRs for gtk+, ido, and libdindicator to support the crazy menu items we have in the indicators; gtk widgets are now in ido (I put the first one in: the indicator-session user switcher item) [15:55] reviews are trickling in, some minor issues but I think we're good for landing them this week [15:55] so that all the other menu items can be moved over as well [15:55] \o/ [15:55] next one for me is indicator-sound [15:55] ted said network is coming along nicely [15:56] so we're almost all set for indicators in unity 8 [15:56] larsu, I got your gtk patch in the ppa, I had to do a small change (add the symbols to gtk/gtk.symbols for make check to pass) [15:56] excellent [15:56] ya, sorry about that :) [15:56] how long do you usually have it in that ppa before it gets into ubuntu? [15:57] I also reviewed the new action API proposal by Wellark and pete-woods [15:57] a day with some people testing it is enough for a patch [15:57] also, mardy fixed a bug in qt that was blocking dconf-qt [15:57] seb128: awesome [15:57] oh, you got that fixed? [15:57] \o/ [15:57] mardy did all the work, and it's not in a ppa yet [15:57] I was starting being concerned that it would block the sdk guys [15:57] Kaleo pinged several times about it [15:57] and I started looking at qmenumodel, but the indicator stuff was more important for now [15:58] seb128: yep, talked to him [15:58] \o/ [15:58] that's good news [15:58] many people pinged about it ... as predicted in oakland :) [15:58] okay, that's all from me [15:58] larsu, thanks [15:58] meeting over. arguments begin! [15:58] we are just in time for the second half of the meeting for didrocks' [15:59] oh. nvm ;) [15:59] desrt, no, #ps integration meeting :p [15:59] and 2 minutes to discuss the wiki [15:59] desrt: I'll be in that call now, will argue enough there. Want to join= [15:59] seb128: no meeting for us, it will be tomorrow (see my previous ping) :) [15:59] does anyone still see a point on having most important work of the week listed on the wiki? [15:59] https://wiki.ubuntu.com/DesktopTeam/Meeting/2013-05-28 [15:59] it seems most of the team members don't do it [16:00] and I'm not sure anyone goes to pull on wiki pages to read those summaries [16:00] larsu: not particularly :) [16:00] seb128: I get value out of compiling a list every week, I don't care whether it's on irc or the wiki [16:00] desrt: I figured [16:00] should we rather try to encourage groups to email/blog about specific topics [16:00] ? [16:00] larsu: i have real work to do, unfortunately [16:00] please let me know how it goes [16:00] yes, I already would email if I had something I wanted people to look at :-) [16:01] larsu, we will keep doing the IRC status update like we just did [16:01] desrt: it's not that important that you join, I got it. Will let you know [16:01] larsu, I'm just suggesting that having the wiki in addition doesn't work/isn't useful [16:01] seb128: ah, understood. /me is new here [16:01] if we had meetings in the meeting channel we'd get parcelled up logs for free, fwiw [16:01] ok [16:02] would that be useful to anyone compared to the irclogs.ubuntu.com logs we have atm? [16:02] bit more discoverable [16:02] but otherwise, no clue [16:02] ok [16:02] let's think a bit about that, I like having it in #ubuntu-desktop [16:03] but seems like nobody wants to fight for the wiki update, so let's drop that ;-) [16:03] thanks everyone [16:03] err [16:03] I'll ask if the bot can come in here [16:03] you didn't ask me ;) [16:03] cyphermox, ask about? [16:03] well, I don't only do indicators ;) [16:04] cyphermox, sorry ECONTEXT [16:04] in the round [16:04] pretty simple update really -- I was working on teach NM to speak to ofono [16:04] the wiki weekly updates? [16:04] nah, the status updates [16:04] cyphermox, oh, sorry, didrocks said he wants to do the second round of the meeting tomorrow for some reason [16:04] didrocks: the upstream developer of python-evdev recommended to also create a python3 binary package for it [16:05] yeah, but AIUI it's mostly for daily release stuff [16:05] cyphermox, thanks for the update ;-) [16:05] didrocks: I'll modify the bzr packaging branch and if you could take a look later it would be awesome ;) [16:05] but yeah, soon we'll have shiny new toys, basically [16:05] cyphermox, right, that's a good point, I will ask for generic status updates at the end of the first half of the meeting next week [16:05] great [16:05] cyphermox, do you need help testing? [16:05] I will, soonish, from those who have 3G modems [16:05] just ironing out the kinks first [16:06] cyphermox, thanks [16:06] anyone else having an update or questions/comments? [16:07] seems not [16:07] that's a wrap then [16:07] thanks everyone [16:07] sil2100: ah, will do then! :) [16:08] Laney, seb128: is it possible to multiarch one binary of a source package but not another? [16:08] yep [16:08] kind of common infact, with -dev packages [16:09] ah... but i mean wrt. libdir [16:09] like can we have /usr/lib/foopkg libdir for one part and /usr/lib/x86_64-whatever/foopkg for another? [16:09] no reason you couldn't, but why? [16:09] ah, well you can override the destination when the files are installed into binary packages [16:09] thinking about installed tests [16:10] assuming libdir isn't embedded into them, that would be the easiest way [16:10] so we get files in /usr/share that look like so [16:10] shouldn't you install them in libexecdir? [16:10] Exec=env G_TEST_DATA=/home/desrt/.cache/lcl/libexec/glib/installed-tests /home/desrt/.cache/lcl/libexec/glib/installed-tests/regex [16:10] and i know that we change libexecdir to libdir in debian... [16:11] so then instead of $prefix/libexec/... we have $libdir/glib/... [16:11] which is a problem if $libdir is different per-arch [16:11] because now we have arch-dependent data in a file in /usr/share [16:12] which means we can't put it in a -common package, as we ought to be able to [16:12] and it also means that even if we put it in an arch:any package, they won't be parallel installable anymore [16:13] what's the reason for not using libexecdir? [16:13] we _do_ use libexecdir [16:13] $libexecdir/glib/installed-tests/____ [16:13] oh, so it's easy [16:13] typically we multiarch libdir, not libexecdir [16:13] oh [16:14] yes. that would fix it, indeed [16:14] then you get conflicts on /usr/lib/glib/installed-tests/ [16:14] --libdir=/usr/lib/$(DEB_HOST_MULTIARCH) \ [16:14] --libexecdir=/usr/lib/dconf \ [16:14] on debian though libexecdir == ${packagename}/lib right? [16:14] for dconf [16:14] just as you would for files in /usr/bin [16:14] woh. it's walters. [16:15] walters, right, but that's the same between arches [16:15] it would be a bit weird since we're installing into $libexecdir/glib/ [16:15] do the data file in -common would point to the right location [16:15] so you'd get /usr/lib/glib/glib/installed-tests [16:15] well, only for glib [16:15] seb128: https://launchpad.net/~bjoern-michaelsen/+archive/libreoffice-oneirictest-20110718/+builds?build_state=pending [16:16] seb128: i guess almost everyone will end up following this same template [16:16] walters: why $libexecdir/pkgname/installed-tests and not the other way around? [16:16] Sweetshark, \o/ [16:17] desrt, mmm...then you'd need some package to "own" that directory [16:17] walters: meh. [16:17] who owns /usr/share/man? [16:18] or locale, or help [16:18] nobody or every package installing something in that directory [16:18] depends what you call "own" [16:18] the first package to be installed to have a file in there will create it [16:18] we have quite a lot of cases for 'well known subdirs of share/lib/etc' [16:18] the last one to be removed will deleted it [16:19] walters: seems that having a common dir in libexec will do a lot to help keep things neat [16:19] instead of having every single package installing its own subdir directly in libexec mixed in with all the binaries already there === greyback is now known as greyback|food [16:21] i'm uncertain what we're fixing here exactly [16:22] is it just cosmetic? [16:22] well [16:22] remember packages can be using $(pkglibexecdir) already for other things [16:22] it also makes it very easy to type rm -r /usr/share/installed-tests /usr/libexec/installed-tests [16:23] dpkg -r glib-2.0-tests ? [16:23] (note may not be -r, this bit is long swapped out of my brain) [16:23] it also makes for some nice symmetry between the two paths... [16:24] -P !! [16:24] ah yeah, purge [16:27] i guess i was thinking i just wanted to avoid creating new "toplevel" directories [16:29] but what you're really suggesting is /usr/lib/installed-tests right? Not /usr/libexec, because no such thing exists in Debian. [16:31] \o walters [16:32] heya jcastro [16:33] walters: i'm suggesting how we do this from a purely upstream point of view [16:34] debian packaging can do as it wishes, after the fact :) [16:34] anyway... i can understand if you don't want to change it because we already have the existing setup in a few package [16:34] but i think it's worth getting it right [16:35] also: do you find it weird that you're installing random data files in libexec? [16:36] like this seems odd: libexec/glib/installed-tests/cert-tests/cert1.pem [16:37] i actually would prefer if you kept it this way, honestly... because the alternative (putting these files in /usr/share somewhere) is much more difficult to deal with [16:38] pretty sure autotools would get angry at you if you ever wrote libexec_DATA=... though [16:40] jbicha: ping [16:42] larsu, hi [16:43] didrocks: hey [16:43] didrocks: you do all the test story for apps and mir etc.? [16:43] test/CI story [16:43] asac: I don't write the tests, but jibel and I are writing the tool for launching them [16:44] if this is the question :) [16:44] didrocks: yeah. i think that means you know everything that i want to knwoa bout :) [16:45] asac: that's a huge presumption, I'm under pressure now :p [16:46] didrocks: so ... where are all the tests that get run? === m_conley_away is now known as m_conley [16:46] can i see them on one big dashboard already? [16:46] didrocks: also, how long would they run if we run all in one shot? [16:46] e.g. take an image, run all the tests we have accumulated in your area and run them [16:47] asac: well, we need to wire up them first, this has been delayed as the QA lab in lexington got down [16:47] asac: I can only give you estimation [16:47] didrocks: your estimation is good enough [16:47] asac: for the dashboard, I need to get time to build it, it's a multimonth project (because I want daily releases reflected here) [16:47] 1h, 3h, 6h, 12h, 24h ? [16:48] asac: I would say 3h at most, as with the new system, we don't reprovision an image and start from a snapshot [16:48] good [16:48] for one release [16:48] didrocks: so why not tell QA team to run them as smoke tests? [16:48] so saucy [16:48] every day against our image? [16:48] didrocks: ping! I pushed the updated packaging of python-evdev to https://code.launchpad.net/~sil2100/python-evdev/debian [16:48] didrocks: any reason? [16:48] count ~ the same for another one [16:48] this would come on top of all your premerge testing etc. [16:49] asac: well, right now, we need to have the system reliable [16:49] asac: which usn't the case until we wire up the new system [16:49] then, we can talk about running them more often [16:49] i am not sure what you mean [16:49] before, running the tests (when the provisionning worked) took ~7h+ [16:50] you are talking strictly about phablet here, right? [16:50] asac: no, in general, all the testing we have for PS [16:50] every test gets a fresh provisioned system? [16:50] before, using UTAH, every test instance (so, let's say, we test unity), it was installing the iso [16:50] so yeah, fresh provisionning [16:50] so ... [16:50] and flacky as well [16:50] qa is still running stuff on utah [16:51] http://reports.qa.ubuntu.com/ [16:51] that stuff [16:51] asac: right, and we are going to unwire UTAH for the integration tests for PS [16:51] and use otto [16:51] which doesn't have that limiation [16:51] one sec [16:52] http://reports.qa.ubuntu.com/ [16:52] limitation [16:52] that one [16:52] boots our images [16:52] and it runs tests [16:52] that i call smoke tests [16:52] thats NOT pre-integration [16:52] asac: yeah, I'm not aware at all about that [16:52] thats daily [16:52] i just want for now to have all the tests that you run during development and during integration to be run there :) [16:52] it's something that the QA team is working on, not wired up in releasing our components [16:52] right.... i am not talking about release testting your components [16:53] i am talking about using your tests to continue testing the image that is out [16:53] e.g. after release of your components [16:53] asac: I think this is something to ask the QA team to do [16:53] makes sense? [16:53] right :) [16:53] hence i just wanted your time estimate :) [16:53] making sense, I just doubt they will get that right enough with UTAH [16:53] so i know if i can jsut ask the to add all [16:53] knowing the pain we experienced :) [16:53] or if i need to make a selection to get going [16:54] ok, so count ~30min of provisionning [16:54] i dont mind if they switch to something else in the back later [16:54] and then 3h for running [16:54] but what i know is that i want real device testing [16:54] if everything is getting installed [16:54] as well... [16:54] so we cannot just abandon that :) [16:54] its certainly not good for your component testing [16:54] asac: this is on a desktop hardware FYI (those estimates) [16:54] to run on real phone hardware, i agree [16:54] asac: nobody is having integration tests running on the device automatically yet [16:54] didrocks: yeah its fine. [16:55] didrocks: i have problems with terminology [16:55] when you talk about integration testing ... you refer to what? [16:55] the smoke tests we run on the archive produced images? [16:55] or the tests you run on images ou produced with your component pre-integrated? [16:55] (before uploading to archive ... which i call releasing fwiw) [16:56] the second one [16:56] because the tests are functionals one [16:56] let's call them "big tests" :) [16:58] asac: we can discuss that in more detailed tomorrow if you want :) [17:00] right [17:00] mterry: hi, do you have a moment? ;) [17:00] for now i declare all those tests as smoke tests [17:00] we can later remove stuff :) [17:00] i want a good dashboard that shows me if your work is still w orking fine [17:00] sil2100, hi === m_conley is now known as m_conley_away [17:01] asac: I should show you the dashboard plans as well. But things in order: utah-replacement first, then this dashboard for me :) [17:01] * didrocks tackles one issue after another one :) [17:01] sil2100: thanks! [17:03] didrocks: cant you just go and teach utah to run your lxc container? [17:03] i really hate inventing new systems [17:03] should be hidden behind a single api/facac [17:03] de3e [17:03] asac: what's the point of utah then? [17:04] asac: no test is in utah format [17:04] it's in autopilot format === alan_g is now known as alan_g|life [17:05] asac: knowing that the most prominent issue was the reliability of UTAH itself (due to their ssh and preseed hack) and the support we got as a consumer [17:06] I don't want to readd a point of failure when we have a system which is way less hackish and so, less tied to failing [17:06] (knowing that we have archive and restore know, which enables you to rerun the same container, in the same state just downloading the archive) locally [17:09] ok, enough for today, /me waves good evening :) [17:09] Bye ;) [17:09] didrocks, night [17:09] mterry: do you have a moment for a packaging review? [17:10] sil2100, in a few minutes, yeah. Which branch? Is this the MIR? [17:11] mterry, hey, speaking of MIR, any chance you could add https://bugs.launchpad.net/ubuntu/+source/unity-gtk-module/+bug/1185117 to your list? [17:11] Ubuntu bug 1185117 in unity-gtk-module (Ubuntu) "[MIR] unity-gtk-module" [Undecided,New] [17:12] mterry, it's code write by attente, should allow us to finally drop those gtk patches for appmenu and have a proper system there [17:13] mterry: ok, no, actually it's a packaging branch for python-evdev - Didier already reviewed it, but we added a double build for python3 [17:13] mterry: and since Didier is now EOD, I'd like someone with experience to take a look ;) [17:14] mterry: we're pushing it to Ubuntu and to Debian, so I want to double-check it before pushing it to the mentors [17:15] seb128, do i add the package to the seed first? or wait until approval by ubuntu-mir? [17:16] mterry: https://code.launchpad.net/~sil2100/python-evdev/debian [17:17] mterry: here is the bzr version of the packaging that's basically submitted to mentors in Debian ;) [17:17] mterry: I didn't push the latest one there [17:17] mterry: I mean, didn't push the latest one to mentors, bzr has the latest one [17:17] attente, you wait for approval ... I'm also not sure if the seed is the right place or if we should make unity recommends it [17:17] mterry: once you don't see any obvious mistakes, I would push it to mentors again and look for a sponsor ;) [17:19] mterry: thanks! === jjardon__ is now known as jjardon === greyback|food is now known as greyback [17:43] sil2100, you still around? [17:43] mterry: yep! [17:43] mterry: I just pushed a small update related to the .docs part === damjanko is now known as hmsck === greyback is now known as greyback|away [20:58] Laney: FYI, meetingology is here now [21:03] oh joy [21:38] ty === greyback|away is now known as greyback === m_conley_away is now known as m_conley