/srv/irclogs.ubuntu.com/2013/10/28/#ubuntu-devel.txt

=== freeflying is now known as freeflying_away
=== freeflying_away is now known as freeflying
=== freeflying is now known as freeflying_away
NoskcajIs there any point trying to fix some of quickly's bugs since it's pretty much EOL05:40
pittiGood morning06:09
jibelpitti, some packages failed to install in adt VMs during latest provisioning. I reprovisioned them.07:19
jibelpitti, I'll restart the tests that failed once it is done.07:19
pittijibel: thanks; I've not gone through all of them yet (see #u-quality), but started07:19
pittijibel: there's a couple of new packages there which geuinnely fail07:20
pittijibel: oh, was visp one of those? that doesn't have an obvious failure07:27
jibelpitti, visp, quagga and gtkada07:28
pittigood, thanks; deleting those mails then07:28
jibelpitti, that's also a case where status in britney stays 'running' forever. Because the test fails early (like VM or test session fails to start), no results is generated.07:37
jibel-s07:37
=== freeflying_away is now known as freeflying
pittijibel: ah, git uninstallability is also on http://people.canonical.com/~ubuntu-archive/testing/trusty-proposed_probs.html08:31
pittiproblem is that https://launchpad.net/ubuntu/+source/git/1:1.8.4.1-1 built only on amd6408:31
pittitests fail on the othhers08:32
pittidebian bug 72722608:32
ubottuDebian bug 727226 in src:git "git tests fail on several architectures" [Serious,Open] http://bugs.debian.org/72722608:32
pittiseems we should perhaps just remove that version from trusty-proposed, and wait for a fix from Debian?08:33
pitticjwatson: ^ would you agree that this is a sensible approach? (unless someone on the Ubuntu side fancies fixing this)08:33
cjwatsonpitti: Could you check whether anything has grown versioned dependencies that would be broken by removing it?08:34
pittiyes, doing08:34
pitticjwatson: checked all components, none08:37
cjwatsonpitti: and AIUI this is causing some autopkgtests to fail?  (for the removal message)08:38
pitticjwatson: yes, at least guilt08:38
cjwatsonpitti: ok, removed08:39
pittithanks08:39
pitticjwatson: I didn't actually mean that you should do it for me, I was more interested in your general opinion whether we should just remove broken stuff from -proposed if it's unlikely that we'll fix it on our side08:40
pittibut I take this as a "yes" :)08:40
cjwatsonI'm not sure I hold that opinion in general, but in the case where it's having fallout I'd say probably08:40
cjwatsonas a general policy it worries me that it's hard to track08:41
pitticjwatson: heh, ironically it broke mercurial's tests, too (as they install git to verify git import)08:46
Fudgecjwatson  I sit here, I read but still have a hard time catching on :D08:51
debfxjodh: is it possible to run upstart services as a user with all its supplementary groups (without using wrappers like start-stop-daemon)?08:58
xnoxdebfx: http://upstart.ubuntu.com/cookbook/#setuid09:00
xnoxdebfx: http://upstart.ubuntu.com/cookbook/#setgid09:00
xnox?09:00
debfxyeah I know about setuid but that just sets the primary group09:02
xnoxi see.09:04
xnoxdebfx: i'm not aware of any other standard facility, apart from using sudo / start-stop-daemon / etc.09:05
xnoxdoko: with gcc-4.8, how come it's libstdc++-4.8-dev, yet libstdc++6-4.8-dbg09:05
pittimost packages don't put the soname into the -dev package for simplicity09:06
pitti(-dev package name, I mean)09:07
jodhdebfx: you mean+need setgid I think.09:10
xnoxpitti: sure, but it used to be libstdc++6 across the board (including all -dev packages) so not sure if the rename is incomplete.09:21
debfxjodh: setgid sets the primary group and leaves the supplementary group list empty, no?09:22
debfxbasically I want upstart to call initgroups() before execing or let me manually specify the supplementary groups.09:22
jodhdebfx: upstart does call initgroups.09:22
jodhdebfx: maybe you're using an older version that doesn't?09:23
debfxpossibly, tested it with precise09:23
debfxok, found it: https://code.launchpad.net/~stgraber/upstart/upstart-initgroups/+merge/13679409:26
darkxstclutter-gst is completely broken in trusty right now it seems09:50
darkxsthttps://launchpadlibrarian.net/155208131/buildlog_ubuntu-trusty-amd64.clutter-gst-2.0_2.0.8-1~trusty1_FAILEDTOBUILD.txt.gz09:50
darkxstother builds just hang09:51
darkxstalthough can't reproduce locally with pbuilder09:52
pittislangasek, stgraber, jibel: FYI, I renamed the BP to core-*, as per cjohnston's mail: https://blueprints.launchpad.net/ubuntu/+spec/core-1311-lts-upgrade-testing10:30
brainwashpitti: http://lpaste.net/94917 http://lpaste.net/94918 http://lpaste.net/9491910:33
brainwashregarding the suspend/resume issue10:33
brainwashsome random log files, did my best to gather some information when the failure occurs10:34
pittibrainwash: thanks! would you mind adding these to the bug, so that the information is at one place?10:36
pittibrainwash: as for the first, I take it you didn't do https://wiki.ubuntu.com/DebuggingDBus to see the method calls as well? The log only has signals10:37
brainwashoh, right.. forgot about this10:38
brainwashthere is no trace of this signal in the monitor log "SIGNAL EMISSION org.freedesktop.systemd1.Manager.JobRemoved()"10:39
brainwashsend by systemd-shim10:39
pittistgraber, slangasek: as discussed, I filed https://blueprints.launchpad.net/ubuntu/+spec/core-1311-ssd-trimming10:39
cjohnstonthanks pitti!10:49
=== _salem is now known as salem_
=== yofel_ is now known as yofel
tseliotpitti: hi. You seem to be part of the lightdm. Does this mean you get to review merge requests?11:24
=== steveire_ is now known as steveire
pittitseliot: I can review simple fixes, but I don't know the overall code; robert mainly added me to fix the apparmor profile11:25
tseliotpitti: oh, ok. I have a two lines patch but it's a feature11:25
=== MacSlow is now known as MacSlow|lunch
tseliotstgraber: are you around?11:53
mbaitHi there! I'd like to backport some package for Precise. I am using 'backportpackage' in a couple with PPA. The problem is the build is failed due to outdated dependencies. Is it possible to backport dependencies as well?12:08
=== greyback is now known as greyback|break
=== MacSlow|lunch is now known as MacSlow
=== greyback|break is now known as greyback
melodiehi13:32
melodieI am seeking an information related to the txt.cfg in the isolinux directory, in a Desktop Live. This is related to the name for a remix. If someone has knownledge about it? (I browsed lots of docs on the web and I missed the one detail I need, so far)13:33
sladenmelodie: it's a while I had to poke one;  there are a series of label/menu/title lines;13:57
sladenmelodie: it should be possible to spot the text that you'd like to change, and alter it13:57
LLckfan2I have an android tablet that I have put youtube app on. When I try to sign into youtubr it says an error occurred. Is there anything I can do to fix this? I did not get the app from Google Play as Google Play does not support my brand of tablet.13:57
=== racarr|sick is now known as racarr|well
melodiehi sladen13:58
sladenLLckfan2: Android?  This is #ubuntu-devel.  Ubuntu is a different operating system (though both Linux and Android use the Linux kernel for hardware drivers at the very base layer)13:58
melodieI have changed just the name "Ubuntu" for the name of my remix, and the next three isos had a little issue: after changing the language to my language, the two first boot stanzas where still in English.13:59
melodiethis is what I try to find about. Where should I look ?13:59
sladenmelodie: is there an 'i18n.cfg', or similiarly named file somewhere around?14:02
melodiesladen probably, I look14:02
sladenmelodie: also  https://help.ubuntu.com/community/InstallCDCustomization#Modify_installer_behaviour_using_a_Preseed_file14:03
melodieno there isn't14:03
melodieshould there be one?14:04
brainwashpitti: I've uploaded the log files now, including dbus-monitor with 99-eavesdrop.conf enabled14:05
melodiewell in the latest lubuntu saucy isolinux there isn't one either14:05
brainwashpitti: it really looks like some dbus connection issue14:06
sladenmelodie: is  http://cdimage.ubuntu.com/lubuntu/releases/13.10/beta-2/lubuntu-13.10-beta2-desktop-i386.iso  what you are starting with?14:06
melodiesladen I have already read the link you are pointing me to14:06
melodieI started with a Ubuntu Mini Remix 12.04 months ago and I often look into Lubuntu and Xubuntu isos to search about information, but I don't use them directly14:07
melodieso after reading the link you pointed me to, it seems you think I want to customize for one language? This is not the case.14:07
melodieand I can't build seed files, because I don't use a regular method, just Ubuntu Builder and the command line for things it doesn't do14:08
melodieOnly having "Ubuntu" in the txt.cfg file changed to "My_remix_name" would be interesting right now, and later I will start to work on getting into other details, such as a meta package14:10
melodieI just don't understand why when I change the name there, 2 over the 5 boot stanzas stay in English.14:11
melodiethis is what I have:14:12
melodiehttp://pastebin.fr/2925014:12
melodiethese are the txt files in the isolinux directory14:12
=== cmagina_ is now known as cmagina
melodiesladen I wonder if the i18n.cfg was not what is now po4a.cfg ?14:15
melodieit contains the iso codes for the languages?14:15
sladenmelodie: there is normally base string in a configuration string; and then various translation files which contain the base string; and the translated replacement for each language14:18
sladenmelodie: this are normally .po files, yes; but I'm not sure how it works for syslinux/isolinux14:18
sladenmelodie: pitti or cjwatson may be able to remember off the top of their heads exactly how it is setup14:19
pittibrainwash: hang on, might it be that most of the people who are reporting this are running not Unity or GNOME, but XFCE or KDE?14:20
sladenmelodie: the translations being in the  /isolinux/NN.cfg14:20
sladenmelodie: the translations being in the  /isolinux/NN.tr14:20
pittibrainwash: I'm asking because gnome-settings-daemon inhibits logind's own handling of the lid14:21
pittibrainwash: does this also happen if you run suspend from the menu instead of closing the lid?14:21
brainwashpitti: I'm running xfce, but I don't use the lid to trigger suspend14:21
brainwashonly menu mostly14:22
brainwashsometimes via direct dbus call14:22
=== kentb-out is now known as kentb
melodiesladen ok, but still how come 2 first strings stay untranslated and the three after do... this is bizarre14:23
brainwashthe strange thing is that it only happens occasionally, like 20-25% of all suspend/resume cycles14:23
pittibrainwash: ok, then it's not that; thanks14:24
sladenmelodie: if you change a string;  but not the base string in a particular locale translation, then it will not match.  Without a match, the translation for that language can't be replaced14:25
brainwashpitti: and it appears to me that different issues got mixed in the bug report, but maybe they are all related14:25
melodiesladen ok I see. So until I know how to start from the sources the best I can do is leave it as is14:26
melodiethe NN.tr files are binaries so...14:27
melodieneeds sources and a compiling step :)14:27
sladenmelodie: po4a is "PO for All"14:27
melodieoh ok14:28
brainwashpitti: however, if systemd's actions don't get inhibited and the system suspends twice within a small time frame, things might fail also I guess14:29
melodiesladen do you know if this is what I would need to start from? https://launchpad.net/ubuntu/+source/syslinux-themes-ubuntu14:31
melodieor rather https://launchpad.net/ubuntu/+source/syslinux ?14:32
cjwatsonmelodie: The translations are in gfxboot-theme-ubuntu14:33
cjwatsonIt's tedious that this has to be done per-flavour, but necessary in general due to the structure of natural language14:34
cjwatsonNot a real example but compare e.g. "an Ubuntu mirror" with "a Lubuntu mirror" in English14:35
melodiehi cjwatson14:35
melodiewhat is  gfxboot-theme-ubuntu ? I am looking into a syslinux-theme-ubuntu archive from https://launchpad.net/ubuntu/precise/+source/syslinux-themes-ubuntu : this is not what you refer to, is that right?14:37
melodiea package maybe?14:37
cjwatsonIt's a source package14:37
melodieI seek for it, thank you14:37
sladenmelodie: apt-get source gfxboot-theme-ubuntu && cd /gfxboot-theme-ubuntu-*/po14:38
sladenmelodie: in eg. 'fr.po' there is  msgid "^Try Lubuntu without installing" <-> msgstr "^Essayer Lubuntu sans l'installer"14:39
melodiesladen cjwatson I have the source package for the version I am working on now, I just wonder if there is a howto compile that once the strings changed? :)14:40
melodieis is just ./configure && make or such?14:41
sladenmelodie: sudo apt-get build-dep gfxboot-theme-ubuntu && debuild  should work14:43
melodienice! I'll try that14:44
melodiein the README it says how to add text, and how to remove text, is that also valid for replacing a string?14:46
slangasekpitti: ack, thanks14:47
cjwatsonmelodie: I usually do those replacements with some sed rune I make up on the spot14:48
melodieI see :)14:49
melodiehow does the syslinux bootloader know which one of the versions will be booted, as in a NN.po file each flavor is provided?14:54
melodieshould a new block file be added in the po files?14:55
melodiea new block text*14:55
melodiethat seems to lead to a larger job :)14:57
cjwatsonI generally add one per flavour, yes14:58
cjwatsonThe PO files just include all known flavours14:58
cjwatsonAnd syslinux knows about it because its other configuration files are set up appropriately14:58
melodiecjwatson ok15:00
melodieI'll have to learn lots more about all this15:00
melodieduring the coming winter :)15:00
melodiethank you very much15:02
melodieand sladen thank you very much also15:02
melodieI wonder what is still good from that page nowadays and what is not? (ie syslinux.cfg used to contain what txt.cfg does now): https://help.ubuntu.com/community/LiveCDCustomizationFromScratch15:08
sladenmelodie: a really useful thing would be to keep track of your successes, and failures, and write-up some (more up to date) documentation as you go15:21
melodiesladen unfortunately I don't have (at least so far) the knowledge for that. I didn't start from Scratch because when I started I just tried head into the doc and when arrived at the interesting page in the doc (something about diverting sysctl) I just got lost, not knowing how to understand what was after: was it to be used literally or adapted, I had no idea. Since then, I use ready to go scripts, the one I find working best for now being Ub15:28
melodieuntu Builder (a gambas gui) with Ubuntu Mini Remix as a start.15:28
melodieelse, I do take notes and put the files used and information on a web space15:28
sladenmelodie: you are acquiring the knowledge.  Write notes as you go, and it will be useful for you, and anyone working with you; but *also* useful for everything afterewards15:29
melodieI might copy them to bitbucket or github later, for now it's here: http://meets.free.fr/Downloads/BentoVillageProject/15:29
melodienot too messy, but will need to refine again later :)15:30
tseliotslangasek: are you available to review/approve an upload (with a small diff) which should fix 2 SRUs?15:30
melodiesladen I might learn more when I will dive into seeds for the preseed directory, I found a doc and some good pointers on the Edubuntu wiki yesterday. I would have questions about preseeds too. Are they used as well for building the isos as for updating at install time when the option is ticked in the installer?15:40
=== brainwash_ is now known as brainwash
melodiethanks for all, bye15:58
=== racarr|well is now known as racarr
tseliotstgraber: re-ping16:11
stgrabertseliot: hi16:12
=== Ursinha is now known as Ursinha-afk
brainwashpitti: the report is getting flooded with comments once again, just to clarify my comment, the entry is missing, because systemd-shim didn't print it https://bugs.launchpad.net/ubuntu/+source/network-manager/+bug/1184262/comments/5816:25
ubottuUbuntu bug 1184262 in network-manager (Ubuntu) "[logind] stuck in PrepareForSleep, causing network and other services to not resume" [High,Confirmed]16:26
brainwashpitti: besides that, is any other log/debug informantion needed to understand the problem and identify the culprit?16:28
tseliotslangasek: never mind16:45
=== Ursinha-afk is now known as Ursinha
=== MacSlow is now known as MacSlow|eveningb
=== Ursinha is now known as Ursinha-afk
=== MacSlow|eveningb is now known as MacSlow
=== gdfuego_ is now known as gdfuego
=== Ursinha-afk is now known as Ursinha
psusislangasek: I'm looking at a bug report by a user saying he has no /dev/.blkid.tab... and I don't see how it *ever* works.  persistent-storage.rules always runs blkid -p, which prevents it from writing out the cache, so how does it ever get created?  Shouldn't that not have a -p there?18:18
dobeysarnold, jdstrand: hi. do you guys know who would be best to chat with about creating an apparmor profile for tarmac? or more specifically, to prevent external processes spawned by tarmac from doing certain things18:35
sarnoldhi dobey :) I've never heard of 'tarmac' before. What is it? what does it do?18:37
dobeysarnold: it's a tool for managing branch landing on launchpad18:38
sarnolddobey: if tarmac uses the usual fork()+exec() method of starting processes, AppArmor's "child" profiles may be suitable: http://wiki.apparmor.net/index.php/QuickProfileLanguage#Child_profiles18:38
dobeyit uses the python subprocess module; so i presume that's standard enough18:39
sarnolddobey: you -may- need to add something like /bin/bash ix,  to the tarmac profile if the python subprocess winds up shelling out to system(3) to execute programs; if it builds arguments as arrays and uses execve(2) directly, you may not need the /bin/bash ix, and just use the /path/to/helper Cx, rules18:41
dobeysarnold: hrmm. does "profile * {}" work for that?18:43
=== racarr is now known as racarr|lunch
sarnolddobey: I don't think it would work as you'd hope -- if you want to allow tarmac to execute -anything- but just confine what those "anythings" are allowed to do, it might look something like this:18:44
sarnold/usr/bin/tarmac {18:44
sarnold  /usr/bin/* cx -> tarmac_child,18:45
sarnold  profile tarmac_child {18:45
sarnold    /path/to/launcpad/branches/** rw,18:45
sarnold  }18:45
sarnold}18:45
dobeysarnold: it needs to run pretty much anything, as what it's doing is building a source tree and running the tests for it; and that can vary infinitely wildly across projects18:45
dobeyhmm18:46
dobeysarnold: but it has to be a full path like /usr/bin? can it be "* cx -> tarmac_child" ?18:46
sarnolddobey: so you'd want to add something like /**/ r, /** rm, /** ix, /path/to/allowed/writes** w,18:47
sarnolddobey: it could be /** cx -> tarmac_child18:47
dobeyok, maybe i can use apparmor to do this then18:47
sarnoldwoo :)18:48
dobeywould rather use apparmor than a complex chroot setup18:49
dobeysarnold: and /**/tarmac will work to match the tarmac script itself, regardless of the install path?18:54
sarnolddobey: /**/ will only match directories; you'll also need /** for the files18:57
sarnold(it's a bit confusing the first few times..)18:57
sarnolddobey: I've got to run, I'll be back in ~two hours.. I hope this is a good enough start for you to get into some trouble :)18:58
dobeysarnold: yeah, and there are plenty of existing profiles to look at18:58
dobeysarnold: thanks. i'll probably ask you to review my profile :)18:58
sarnolddobey: thanks! :) have fun18:58
=== racarr|lunch is now known as racarr
=== Lutin is now known as Guest20807
=== iulian is now known as Guest86495
Noskcajroaksoax, kirkland: Can one of you release testdrive? I think we've got enough fixes to warrant it19:49
NoskcajAnd do either of you have time to help with gtk3 and/or python3 porting?19:49
=== sergiusens_ is now known as sergiusens
roaksoaxNoskcaj: saucy just openned a few days ago19:58
roaksoaxthere's no immediate need to release until at least we have alpha one19:58
roaksoaxhowever, i'll be releaswing later on the week19:58
stgrabers/saucy/trusty/20:02
Noskcajroaksoax, makes sense, thanks for the info20:09
roaksoaxs/suacy/trusty20:11
=== tkamppeter_ is now known as tkamppeter
=== Guest86495 is now known as iulian
dobeycan i get someone to push rhythmbox-ubuntuone from precise-proposed to -updates? it's been in proposed for almost 5 months with no complaints on the bug report about it. and just noticed the verification tag was still -needed, but changed to -done now.20:23
ScottKdobey: Done20:28
dobeythanks ScottK20:29
ScottKyw.20:30
dobeyScottK: are you sru team btw? i have another favor to ask if so :)20:30
=== vila is now known as lt-columbo
=== lt-columbo is now known as vila
dobeythere's an ubuntu-sso-client in saucy unapproved queue, which is a critical bug fix, that we'd like to get through asap. i uploaded it on friday, but hasn't gotten reviewed yet20:32
ScottKI'l take a look and if I'm reasonably comfortable with the diff, I'll accept it.  No promises.20:33
dobeythanks. it's a small and obvious diff, so i'm confident :)20:34
=== salem_ is now known as _salem
ScottKdobey: Accepted.20:36
dobeyScottK: thanks again :)20:39
ScottKyw again.20:39
=== Guest20807 is now known as Lutin
cjwatsonjdstrand: your latest click-apparmor upload doesn't seem to be in lp:click-apparmor20:45
mdzcjwatson, is google calendar telling the truth, that tech board is supposed to meet now?21:02
cjwatsonmdz: Sounds plausible-ish but I hadn't planned to turn up to future TB meetings21:03
cjwatsonmdz: It's 2100 in the UK if that's any help21:03
mdzwho would know for sure? there's no chair on the wiki, etc.21:04
mdzoh hi pitti21:04
pittihey mdz, how are you?21:05
jdstrandcjwatson: fixed. you keep pointing this out to me. perhaps I will learn some day...21:13
cjwatsonheh, thanks :)21:13
cjwatsonOf course I only notice when I want to propose a change21:13
mdzpitti, good good21:18
mdzpitti, going to be in the US anytime soon?21:18
=== ChrisTownsend1 is now known as ChrisTownsend
pittimdz: not this year any more21:20
cjwatsonmdz: you going to be at the BAD meeting on Wed?21:20
mdzcjwatson, didn't know about it, and have plans Wednesday evening21:20
cjwatsonah, shame21:20
cjwatsonin Oakland this week21:20
mdzI saw jonathan and asheesh last night though21:21
mdzI didn't know they were having regular get togethers21:21
sergiusenscjwatson, hey, do you know of any good location to get python3-requests for precise?21:22
* pitti waves good night21:22
cjwatsonmdz: don't know if this is a regular thing, I only know about it from slangasek21:22
cjwatsonsergiusens: nope, I guess it might be possible to backport it21:23
cjwatsonsergiusens: but I'd say that if dependencies are unavailable on precise then that's a good reason to have bilingual code and run it under Python 2 on precise21:23
mdzcjwatson, slangasek isn't in the bay area now, is he?21:23
mdzhttp://bad.debian.net/ lists August 8th 2012 as an 'upcoming event' :-)21:24
mdzgood night pitti21:24
sergiusenscjwatson, sounds reasonable21:24
cjwatsonmdz: he's with us in Oakland21:24
mdzoh, just at the moment21:24
cjwatsonyeah21:24
mdznot relocated21:24
cjwatsoncan't imagine prying him loose from Portland21:24
mdzyeah21:24
mdzwho's there and how long? anyone interested in meeting up?21:25
cjwatsonUE managers and tech leads21:25
mdzthis week is pretty busy for me but we could probably work something out21:25
cjwatsona few old-timers, I can ask around21:25
hallynslangasek: see edk2 thread titled 'OVMF support for QEMU's PC System Flash' - i could be wrong, but i *think* that gives us what we wanted with being able to persist nvvars changes across qemu starts21:49
paulproteusmdz: Huh, BAD is on Wednesday?22:09
paulproteusIntriguing.22:09
=== Zic is now known as Guest53304
=== Guest53304 is now known as Zic

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!