UbuntuIRC / 2014 /10 /21 /#ubuntu-devel.txt
niansa
Initial commit
4aa5fce
=== kirkland is now known as Guest45646
=== timrc is now known as timrc-afk
=== wendar_ is now known as wendar
=== ubott2 is now known as ubottu
[08:26] <tseliot> doko: any news on the two MIRs? (I saw you unsubscribed the MIR team)
[08:59] <LocutusOfBorg1> hi, nobody wants to sponsor ettercap?
[08:59] <LocutusOfBorg1> bug 1382848 :(
[08:59] <ubottu> bug 1382848 in ettercap (Ubuntu) "[FFe] Sync ettercap 1:0.8.1-1 (universe) from Debian unstable (main)" [Wishlist,New] https://launchpad.net/bugs/1382848
[09:40] <Madkiss> hi folks
[09:41] <Madkiss> I'm building a pacakge for a PPA that needs dependencies from another PPA (the cloud team ppa for openstack icehouse on precise). How can I make the PPA auto builders use additional sources.list entries?
[09:43] <LocutusOfBorg1> Madkiss, "Edit PPA dependencies" on your ppa page
[09:44] <Madkiss> oih, hadn't seen that
[09:44] <Madkiss> thanks!
[09:44] <LocutusOfBorg1> you are welcome ;)
[09:53] <flexiondotorg> I think I am seeing a regression in LightDM. Anyone here who I can quickly discuss that with? Maybe due to the recent fixed to address a double free in 1.12.
[10:03] <Madkiss> LocutusOfBorg1: Okay. And how do I add external dependencies to other repos? the ubuntu cloud archive stuff isn't a PPA obviously
[10:04] <LocutusOfBorg1> I don't get the question, but I think the answer is "you can't"
[10:05] <Madkiss> I need to do " sudo add-apt-repository cloud-archive:icehouse" for the PPA.
[10:06] <LocutusOfBorg1> add ppa dependency
[10:06] <LocutusOfBorg1> you select "choose"
[10:07] <LocutusOfBorg1> oh, indeed, then I don't think you can do it
[10:07] <Madkiss> oh boy.
[10:08] <LocutusOfBorg1> maybe you can with ubuntu-cloud-archive/icehouse-staging
[10:08] <LocutusOfBorg1> you can with it, but I don't know if it is the same
[10:09] <LocutusOfBorg1> maybe try to ask on #launchpad
[10:09] <Madkiss> yeah, that could work
=== timrc-afk is now known as timrc
[13:39] <arges> doko: hi. i'm testing ubuntu-toolchain-r. are there cross-compiler builds as well somewhere?
[13:39] <doko> arges, sorry, not yet
[13:40] <arges> doko: ok
[13:40] <doko> arges, can you point me to a package which I could upload to a ppa? or maybe do it yourself?
[13:41] <arges> doko: what needs to be uploaded? i was really just trying to test the ubuntu-toolchain-r with kernel compiles, and i wanted to test multiple arch builds. I could just try to find real hardware or emulate
[13:47] <doko> arges, I assume a linux source package, of the most current kernel in trusty-updates/-security
[13:52] <arges> doko: oh oh, you mean use a PPA and add your toolchain as a dep
[13:53] <doko> arges, yes, but if you give me the package, I can upload to the ubuntu-toolchain-r ppa too
[13:54] <tseliot> doko: what's the status of the two MIRs?
[13:54] <doko> tseliot, looking
[13:54] <tseliot> thanks
[13:57] <arges> doko: i should be able to handle it. i don't want to mess up your ppa with a kernel build
=== freeflying__ is now known as freeflying
=== Trevinho_ is now known as Trevinho
[15:48] <Hurri877> Hi I was wondering were I could get some help with git buildpackage?
[15:53] <jnxd> didrocks, I recently installed UDTC and then through it android studio. Just wanted to know in case I wanted to remove the latter, what steps are necessary?
[15:58] <Riddell> mvo: release-upgrade doesn't work when ran from a kubuntu package manager, https://bugs.launchpad.net/ubuntu/+source/ubuntu-release-upgrader/+bug/1383786
[15:58] <ubottu> Ubuntu bug 1383786 in ubuntu-release-upgrader (Ubuntu) "Kubuntu 14.04 → 14.10 "Could not calculate the upgrade"" [Undecided,New]
[15:58] <Riddell> mvo: but it's ok when ran from do-release-upgrade
[15:58] <Riddell> mvo: wibble, help
[16:00] <LocutusOfBorg1> interesting, backportpackage seems to be not working anymore
[16:00] <LocutusOfBorg1> bdrung, any idea?
[16:00] <LocutusOfBorg1> distro_info.DistroDataOutdated: Distribution data outdated. Please check for an update for distro-info-data. See /usr/share/doc/distro-info-data/README.Debian for details.
[16:02] <bdrung_work> LocutusOfBorg1, you need to update distro-info-data (and probably distro-info, too)
[16:02] <cjwatson> yes they're on their way
[16:02] <cjwatson> waiting to be able to sync distro-info
[16:02] <cjwatson> oh I think somebody uploaded that
[16:03] <cjwatson> yeah I just need to unblock those, doing no
[16:03] <cjwatson> w
[16:03] <LocutusOfBorg1> yes, thanks :)
[16:04] <bdrung_work> cjwatson, we have to apply the fix for distro-info to trusty (precise is not affected)
[16:05] <cjwatson> bdrung_work: ok, if somebody uploads an SRU I'll review it
[16:09] <Riddell> mvo: oh you're at a conference?
[16:09] <mvo> Riddell: yes
[16:09] <Riddell> whyever is there a conference on release week
[16:09] <mvo> Riddell: do you need anything from me?
[16:09] <mvo> Riddell: I don't know
[16:09] <Riddell> mvo: bug 1383786
[16:09] <ubottu> bug 1383786 in ubuntu-release-upgrader (Ubuntu) "Kubuntu 14.04 → 14.10 "Could not calculate the upgrade"" [Undecided,New] https://launchpad.net/bugs/1383786
[16:09] <stgraber> Riddell: the release sprint is at a company sprint and yeah that's pretty weird
[16:09] <Riddell> mvo: "do-release-upgrade -m desktop -f DistUpgradeViewKDE" fails but works with the "-d"
[16:10] <Riddell> mvo: is that expected?
[16:11] <mvo> Riddell: hmmmm, it seems like its breaking on "startpar" which is now required
[16:11] <Riddell> what's that?
[16:11] <mvo> Riddell: I don't know
[16:11] <Riddell> :(
[16:11] <cjwatson> conference on release week> not our idea
[16:13] <mvo> Riddell: what does apt list startpar give you?
[16:14] <Riddell> Listing... Done
[16:14] <Riddell> (on trusty)
[16:14] <Riddell> mvo: ↑
[16:15] <mvo> Riddell: hmmmm
[16:16] <Riddell> mvo: I don't understand how it can be different due to the -d , I've set the dns to point to a server with utopic included in meta-release so it should be the same thing
[16:18] <rbasak> doko: python-eventlet 0.13.0-1ubuntu3 fixed bug 1371291, right?
[16:18] <ubottu> bug 1371291 in python-eventlet (Ubuntu Utopic) "FTBFS caused by PEP 466" [High,Triaged] https://launchpad.net/bugs/1371291
[16:19] <mvo> Riddell: yeah, I don't see startpar at all anywhere, I wonder where it cmes from.
[16:20] <mvo> Riddell: also " MarkInstall startpar [ amd64 ] < none -> 0.59-3ubuntu2 > ( universe/admin ) FU=1" <- its in universe and required? very confusing
[16:20] <arges> doko: so i am trying to test your ubuntu-toolchain-r, i setup cross-chroots with your PPA, but they seem to be failing on python installation issues (when installing deps for the kernel) this affects ppc64el and arm64 so far
[16:21] <arges> doko: http://pastebin.ubuntu.com/8616765/
[16:21] <Riddell> mvo: sorry I need to go out for a couple of hours, any idea you have appreciated
[16:22] <mvo> Riddell: I look into it
[16:25] <jibel> Riddell, does upgrade work with -proposed disabled?
[16:25] <jibel> Riddell, startpar is only in utopic-proposed
[16:25] <mvo> I think i know what is going on
[16:26] <doko> arges, strange ... this ppa survived a whole archive test rebuild
[16:29] <mvo> Riddell: so startpar is only available in proposed and priority is required, I uploaded a new version that lowers the priority, this should fix the upgrade
[16:31] <cjwatson> mvo: eh that's ineffective
[16:31] <cjwatson> mvo: correct fix is to get an AA to change-override :)
[16:32] <cjwatson> mvo: I'll do that after your upload builds everywhere
[16:32] <mvo> cjwatson: oh, sorry. you are right of course :) please reject and fix it there
[16:32] <cjwatson> mvo: too late
[16:32] <cjwatson> should be optional too :)
[16:32] <mvo> cjwatson: want me to do another one?
[16:32] <cjwatson> no
[16:32] <mvo> ok, thanks
[16:33] <cjwatson> actually this is weird
[16:33] <cjwatson> priority-mismatches should be whining about this
[16:33] <cjwatson> oh, we probably don't run that against -proposed
[16:33] <cjwatson> ok, I'll fix after lunch
=== roadmr is now known as roadmr_afk
=== roadmr_afk is now known as roadmr
[18:49] <xnox> pitti: do i remember correct, there was some init script or upstart job, to change the power/cpu scheduler after X seconds post-boot to improve battery life or some such.
[18:49] <xnox> pitti: do you recall something like that? and if yes, where was it?
[18:52] <pitti> xnox: it rings a bell, but I don't remember either where that happened :/
[18:52] <xnox> pitti: i vague recall you were the author of that trick.
[18:52] <xnox> unless i have it complete wrong.
[18:53] * xnox ponders what was that package which whole bunch of default sysctl commands et.al.
[18:53] <sarnold_> procps?
[18:55] <xnox> pitti: /etc/init.d/ondemand
[18:56] <j1729> Hello guys. How are you doing ?
[18:57] <j1729> are there bugs to fix or someway I can get started with contributing as a developer ?
[18:58] <xnox> sarnold_: yes, that's the other package i thought it might be in.
[19:04] <pitti> xnox: hah, indeed; with a lovely sleep 60
=== _salem is now known as salem_
=== salem_ is now known as _salem
[19:11] <infinity> xnox: You're thinking of /etc/init.d/ondemand from initscripts
[19:11] <infinity> xnox: Oh, pitti got there.
[19:13] <xnox> infinity: funny how it sleeps first, and only then determines if there are desired governors and then bails.
[19:13] <xnox> infinity: would be nice to bail, before going for a sleep =)
[19:13] <infinity> xnox: Meh, not that it matters much, really.
[19:17] <ogra_> xnox, iirc we needed the sleep on nexus7 back then
[19:18] <ogra_> (i dont remember why, but i guess it could go)
[19:44] <bdrung> Laney: re #1381995, why not fix distro-info in trusty?
[19:44] <Laney> I think it will be
[19:45] <Laney> But don't want to block on that
[19:45] <bdrung> it's a one line fix.
[19:45] <Laney> Does that affect the SRU aging period?
[19:46] <bdrung> no
[19:46] <Laney> Also I am (maybe) going to -security and I don't know if that fix would.
[19:46] <Laney> Soooo
[19:46] <Laney> Yeah
[19:47] <bdrung> okay, then let's keep the workaround, get distro-info fixed in a SRU and then let's see which date will be the final date
[19:56] <cjwatson> pitti: language-pack-af upload seems to have gone missing: http://people.canonical.com/~ubuntu-archive/proposed-migration/update_excuses.html#language-pack-af-base
[19:56] <xnox> apw++
[19:58] <alexbligh1> I'm trying to debug an upstart problem where 14.04 does not start an upstart script that runs on udev-finish in a container in lxc. When not in a container, it works. pstree doesn't show anything unexpected running. Any ideas where to start? The confusing thing is that this used to work ...
[20:02] <zbenjamin> slangasek: thanks for helping with the gnueabi vs gnueabihf. I have a working qmake now :)
[20:03] <xnox> alexbligh1: increase upstart verbosity, and see if the desired events are emitted.
[20:03] <xnox> alexbligh1: also test if the job is valid.
[20:03] <alexbligh1> xnox, if I start the job manually, all is well
[20:03] <xnox> alexbligh1: sudo lxc-start -n container -- /sbin/init --verbose
[20:04] <xnox> alexbligh1: in that case start-on conditions probably never fire. Oh, there is no udev nor udev-finish in containers =)
[20:04] <slangasek> zbenjamin: huzzah
[20:04] <alexbligh1> xnox, what /stops/ there being a udev in containers? I seem to remember coming across that before and disabling whatever stopped it, because I want udev in this instance
[20:04] <cjwatson> pitti: accepted, thanks
[20:04] <xnox> alexbligh1: you might want your job to start on (...) or not-container
[20:04] <pitti> cjwatson: ah, you beat me to it; sorry about that
[20:05] <pitti> (with accepting, I mean)
[20:05] <xnox> alexbligh1: sorry or container that is.
[20:05] <xnox> alexbligh1: chat with people on #lxcontainers about udev and device access....
[20:05] <cjwatson> pitti: np
[20:05] <cjwatson> I was in the queue anyway
[20:05] <xnox> alexbligh1: in general udev runs on the host's kernel and doesn't affect what happens inside the container and one cannot per-se connect/disconnect devices from a container.
[20:06] <xnox> one can choose to bind-mount / export them into the container, but that's about it.
[20:06] <alexbligh1> xnox, I'm guessing something subtle has changed between 14.04.00 and 14.04.01
[20:07] <xnox> alexbligh1: lxc gets a lot of bugfixes and new stable releases are pushed to trusty. chat on #lxcontainer they'll be able to help. And are very familiar with upstart inside the lxc containers.
[20:07] <alexbligh1> xnox, thanks. upstart + containers = confusing.
[20:09] <zyga> I have a small patch for python3.4 that memsets one structure that gets passed to a syscall, currently that struct has some uninitialized data (which is safe but makes some of my testing unreliable)
[20:11] <mapreri> pitti: :) hi! might I hope for a last-day sync of scribus? it fixes the nasty #1383407 but it's a small patch (<http://www.mapreri.org/debian/debdiffs/scribus/1.4.4+dfsg1-1_1.4.4+dfsg1-2.debdiff>) quite harmless.
[20:12] <pitti> mapreri: can I redirect you to #ubuntu-release, please? in meeting/busy/no powers any more to break the release freeze without asking
[20:12] <mapreri> pitti: fine, thanks
[20:17] <Laney> pitti: mapreri: bugfix, unseeded> still can sync
[20:18] <Laney> oh, but scribus isn't unseeded :(
[20:19] <mapreri> Laney: IIRC it's in edubuntu....
[20:20] <mapreri> Laney: I'm submitting a FFE bug... I just wanted to get rid of https://bugs.launchpad.net/ubuntu/+source/scribus/+bug/1383407 before release...
[20:20] <ubottu> Ubuntu bug 1383407 in scribus (Ubuntu) "Scribus spell check and hyphenation doesnt work" [Undecided,New]
[20:20] <alexbligh1> xnox, thanks - you led me to the solution there.
[20:26] <Laney> mapreri: No need for ffe
[20:27] <mapreri> indeed it's not feature freeze, but it's the final freeze
[20:27] <Laney> Yes, don't worry about a bug
[20:28] <mapreri> Laney: I'm new to last-day uploads, what kind of permission I need? Might I use your message above as an ack and send a green light to my sponsor?
=== roadmr is now known as roadmr_afk
[20:29] <Laney> I'll sync it, I just asked in #ubuntu-release if there will be new images comig anyway
[20:30] <mapreri> Laney: ok, thanks. (I saw the message, but didn't understand it :))
[20:45] <hallyn> arges: hi, the SRUs for qemu and libvirt to trusty are verified adn have baked for awhile, coudl they be promoted? (security team has updates they want to push)
[20:46] <Laney> mapreri: it is synced
[20:50] <hallyn> infinity: https://launchpad.net/ubuntu/trusty/+queue?queue_state=0&queue_text=ipxe-precise what is the next step to getting that out of NEW?
[20:51] <infinity> hallyn: Someone needs to review it.
[20:51] <infinity> hallyn: And yeah, I can promote the others.
[20:52] <mapreri> Laney: thanks very much. will wait for acceptance now :)
[20:57] <hallyn> infinity: cool, thanks
=== roadmr_afk is now known as roadmr
[21:19] <arges> hallyn: i'll take a look
[21:20] <arges> hallyn: oh looks like infinity got to it already
[21:20] <arges> infinity: thanks
[21:25] <goodwill> ppetraki: ping
[21:25] <hallyn> arges: yup, thanks
[21:25] <goodwill> hallyn: maybe you can help ... what in debian installer triggers multipath-udeb install?
[21:26] <cjwatson> goodwill: disk-detect.sh in hw-detect
[21:27] <cjwatson> from a user point of view you enable it by preseeding disk-detect/multipath/enable=true
[21:27] <goodwill> cjwatson: right ... so I am trying to try out the new fixed udeb for multipath in proposed? should I install them in early_command then?
[21:29] <goodwill> cjwatson: I ran disk detect manually ... but there is not /sbin/multipath
[21:29] <ppetraki> goodwill, pong
[21:30] <goodwill> ppetraki: still fighting multipath ;) ^^^
[21:30] <cjwatson> goodwill: if it's in -proposed, usual advice is to just use apt-setup/proposed=true
[21:30] <cjwatson> goo don't run installer components manually
[21:30] <cjwatson> (sorry, bit of lag)
[21:31] <goodwill> cjwatson: I used the standard menu prompt and installed the DM-Multiparth support, which appears to install the module drivers
[21:31] <goodwill> cjwatson: then I ran detect and that installed kpartx udeb
[21:32] <goodwill> cjwatson: so far I failed to figure out how the multipath-udeb is installed though
[21:32] <cjwatson> I told you above
[21:32] <cjwatson> preseed disk-detect/multipath/enable=true
[21:33] <cjwatson> that is, boot the installer with the additional kernel arguments disk-detect/multipath/enable=true apt-setup/proposed=true
[21:33] <goodwill> I guess a better question is: it preseeded only and not available in the user prompts?
[21:33] <cjwatson> assuming this is netboot
[21:33] <cjwatson> as far as I can see it is preseeded only yes
[21:33] <goodwill> I see
[21:33] <goodwill> okie
[21:33] <cjwatson> (I don't know much about multipath specifically, just about the installer)
[21:33] <goodwill> that explains it
[21:33] <goodwill> cjwatson: was booting from mini iso for testing
[21:33] <cjwatson> netboot mini.iso?
[21:33] <goodwill> yeah
[21:34] <cjwatson> yes that's the same as netboot for this purpose
[21:34] <goodwill> right
[21:34] <cjwatson> apt-setup/proposed=true causes the installer to fetch components of itself from -proposed as needed
[21:34] <cjwatson> that's used for verifying SRUs
[21:34] <goodwill> cjwatson: I am sorry what does SRU stand for?
[21:34] <cjwatson> stable release updates
[21:56] <goodwill> cjwatson: is there a way to do db_set on a command line?
[22:36] <cjwatson> goodwill: debconf-set but it needs to be in the right context (don't have time to explain now); it's easier to just put that variable on the kernel command line, or use a proper preseed file if this is read late enough
[22:36] <goodwill> cjwatson: I see
=== IdleOne is now known as Guest88208