=== asac_ is now known as asac [00:18] how do you flush your dns cache in ubuntu? [00:19] ccheney: what dns cache [00:20] i thought there was some sort of dns cache at the system level, maybe my problem is actually arp cache instead [00:20] restarting networking ? [00:21] my system can't connect to gtalk claiming no route to host i think due to it trying to connect before i authenticated with AP at the hotel [00:21] there is an arp cache [00:21] no route to host - you can see your route table by [00:21] ip route [00:21] you can get no route errors two ways [00:21] no route locally [00:21] yea just pidgin claims it not eg firefox [00:22] or ssh, etc [00:22] a router somewhere else sending back an ICMP error [00:22] ccheney: restart pidgin ? [00:22] routers can do that when they have no route. *or* when they don't want to let the traffic through [00:23] BUGabundo: it seems to not help restarting pidgin, i'll try rebooting my system to see if it changes [00:23] ccheney: that's a bit extreme [00:23] even for 10 sec boot systems [00:23] eheh [00:24] try restarting networking [00:25] hmm even total reboot didn't help :( [00:25] * ccheney kicks the hotel === pace_t_zulu_ is now known as pace_t_zulu [00:28] ccheney: mtr it ? [00:29] ccheney: you should be using empathy anyway ;) [00:29] pochu: i'm using jaunty [00:29] i can see talk.google.com [00:29] not an excusse eheeh [00:29] with mtr [00:30] there is a PPA for empaty trunk [00:30] hey pochu [00:30] ccheney: ahhh pidgin updates? [00:30] check for the record on the server tab [00:32] hmm no pidgin updates available for me, i'm mostly up to date (as of sat night) [00:32] it was working at the office too, so i think they are just blocking it somehow at the hotel [00:32] it was working last night though so its a bit weird [00:33] * ccheney heads off to dinner, bbl [00:33] ccheney: try https://imo.im [00:34] BUGabundo: thanks [00:35] np [00:36] * ccheney heads out now to dinner [02:08] hi - I hope I've got the right channel for my question, please redirect me if not - [02:10] I'm in the process of customising a desktop gnome environment based on jaunty - the customisations need to packaged, and I'd prefer to try and get this more correct than not [02:11] currently, I have replaced the ubuntu-artwork and ubuntu-wallpapers packages, providing modified gconf keys [02:11] so that takes care of the various themes and the desktop background [02:11] I need to make some fairly drastic changes to the panel though [02:12] This is not the channel you are looking for. [02:12] currently, the panel layout seems to be controlled from the gnome-panel-data package, which provides two files [02:13] /etc/gconf/schemas/panel-default-setup.entries and /usr/share/gconf/defaults/05_panel-default-setup.entries [02:14] my panel changes basically mean I remove the top panel and move relevant applets to the bottom panel [02:15] so I need to override those default files somehow (without modifying them directly, preferably) [02:16] I know I can provide my own 90_foo.entries file in /usr/share/gconf/defaults/ [02:17] however, because I don't plan to use the top panel gconf entry at all, I'm wondering if the 05_...entries file will still create it [02:17] what is the safest way for me to override these setting? [02:17] ScottK: sorry - completely missed your bit [02:17] ScottK: got a suggestion where I could go? [02:18] the problem is, this stuff seems to be distro specific [02:19] Not particularly, but that's definitlely not development of Ubuntu. [02:19] I think there is an #ubuntu-installer channel. [02:20] is there a desktop development focussed channel? [02:20] +gnome even [02:23] nvm, I'll have a look at #ubuntu-installer, and see if there's any relevant debian channels [02:23] thanks [02:38] cjwatson: Thanks for the pointers. I am trying to test my merged syslinux package. The Ubuntu (non-Suse: 12-, 13- and 14- ones) gfxboot patches can't apply to the new comboot16 module and upstream doesn't know if they are still relevant. Testing this whole thing is kind of tricky actually :) For reference: http://syslinux.zytor.com/archives/2009-July/012866.html [06:01] I just got done with a computer sci degree and am looking to take up an open source project, any ideas on where to start [06:44] Good morning [06:51] slangasek: you rock! [06:52] slangasek: want me to accept gvfs for hardy-proposed now, or stall everything until after .3 was released? [07:04] <\sh> moins [07:07] liw: Hello. Does apt-sync/zsync apply in any way to Packages files, or is it just about .deb files ? [07:21] I've multiple repositories in my sources.list.. how do I print the packages in one particular repository (say a launchpad ppa repo)? [07:23] good morning [07:24] * slangasek wavse [07:24] anilg: Hi, that's not really a development question (as the topic says: "Development of Ubuntu (not support, not app development on Ubuntu)"). You'd be better placed asking that in #ubuntu. (Though ftr I don't think there is any command for this, and I'd probably resort to applying grep-ctrl to apt's raw lists files) [07:25] hmm, why dont we have a proper commandline interface to tasksel ... [07:25] * ogra cant belive nobody ever wrote something like that [07:25] * dholbach wavse to slangasek too [07:25] maxb: thanks.. I did try #ubuntu [07:25] * ogra joins the wavsing to slangasek [07:28] ogra: We do, apt-get [07:29] StevenK, apt-get can make me a list of available tasks ? [07:29] * ogra checks the docs ... thats news to me [07:32] ogra: grep '^Task: ' /var/lib/apt/lists/silver.archive.ubuntu.com_ubuntu_dists_jaunty_*_binary-*_Packages | cut -d\ -f2- | tr ',' '\n' | sed -e 's/^ //' | sort -u [07:32] lol [07:33] well, then i rather parse /usr/share/tasksel/ubuntu-tasks.desc :) [07:33] and depend on tasksel data [07:34] ogra: For what? [07:34] rootstock [07:34] Ah [07:35] i want to show a gtk list with available tasks plus a textfield where you can add sinle packages as csv list before you fire up the build [07:36] so what i was actually lookinf for was something like "tasksel --list-tasks" and "tasksel --list-taks-descriptions" to just feed into the list :) [07:36] ogra: /usr/share/tasksel/ubuntu-tasks.desc will show you the *hosts* tasksel list, not the targets [07:37] the target doesnt exist at that point [07:37] ogra: For the example of running rootstock on a Jaunty machine, but building a Karmic root [07:38] hmm, yeah, i probably shuld pull the packages list from the archive and use your approach ... [07:38] prob is that this changes my UI approach :/ [07:39] since i need to know the target release before doing that ... [07:40] ogra: You can pre-populate it, given data on the running system first. [07:40] indeed, thats what i do anyway [07:40] ogra: But the task list might change if you're building for Karmic. [07:40] Actually, it *will* change, I can think of two tasks that Karmic has and Jaunty doesn't. [07:41] by default it builds ubuntu-minimal and pulls all other data (locale, kbd settings etc) from the host [07:43] * ogra ponders how to make that switchable on the fly without ending up with something like a wizard UI [07:45] ogra: mtp loves wizards! [07:46] mpt [07:46] hmm, i should probably make the script stop generating *all* en_* locales inside qemu-system-arm ... might speed it up a bit [07:46] dholbach, oh, since when ? [07:46] i always thought he didnt :) [07:46] didn't you see the wizard costume he was wearing in Boston? [07:46] heh [07:51] hrm, why did my trap code in the script stop working [07:57] I have a problem with menu.lst [07:57] Every time hardy updates kernel it mess up the grup option for testing ubuntu instalaltion [07:57] and vice versa [07:57] dholbach: I propose that all Fall UDSes that are in the US overlap halloween [07:58] YokoZar: I'm not sure I'm the right person to talk to :) [07:58] Failing that, they should end with a costume party [07:58] I am editing menu.lst for Every kernel update on testing or stable since 6.10 and I am finally sick of it [07:59] sorry I maybe should post this ono ubuntu or ubuntu+1, sorry. [08:00] btw dholbach, are there plans to let the 5-a-day stats page apply to individuals again? It's been "groups only" for a long time now [08:00] YokoZar: bdmurray is working on that [08:01] YokoZar: it needs an RT ticket resolved because we need to process HUGE amounts of data [08:01] ahh ok [08:02] like all ubuntu-bugs@ archives :) [08:02] the code is open [08:02] so if you want to help out, that's cool [08:03] hey mdz [08:06] c [08:08] dholbach, hi [08:09] * maxb boots intrepid and experiences nostalgia for how pretty it was :-/ [08:10] nostalgia ? [08:11] usplash and notifications were both prettier in intrepid than jaunty/karmic, IMO [08:11] well ... but its not even a year old [08:12] boot a warty or hoary liveCD ... thats nostalgic :) [08:12] I was running jaunty from alpha 3, intrepid seems a long time ago :-) [08:15] about names, just curious: so what will happen when ubuntu runs out of latin alphabet in release names? too far away to think about this? [08:15] That will not be until 2017.... [08:16] the 2K17 bug. very serious stuff indeed. [08:16] need a committee to solve the issue! [08:16] we will have taken over the world by then ... we'll just anhance the latin alphabet [08:16] *enhance [08:17] that's a good plan [08:17] hile: I guess it will start over with unpronouncable letters. and then you'll call it things like "the distribution formerly known as hoary" [08:18] I dunno, I quite like the sound of "Ancestral Alligator" [08:46] slangasek, \o/ for fixing the gvfs samba issue [08:47] seb128: thanks for your help setting me on the right track! [08:47] ;-) [08:53] ¡¡¡ please everybody help to clean up http://people.ubuntu.com/~dholbach/sponsoring/ !!! :-) [09:02] maxb, why ask in multiple forums, making sure that people miss the answer? I will answer via e-mail [09:02] howdy! could someone with upload rights please merge the two attached maintainer scripts at LP bug #399482 ? [09:02] Launchpad bug 399482 in bluez "bluez: upgrade from 4.45-0ubuntu1 to 4.45-0ubuntu2 fails" [Undecided,Fix released] https://launchpad.net/bugs/399482 [09:04] Q-FUNK: Doesn't that say "fix released"? [09:10] RAOF: it was entirely the wrong fix. === mdz_ is now known as mdz [09:37] Keybuk, erm, did the behavior of reboot change with upstart 0.6 ? [09:38] Keybuk, reboot -fp doesnt see to do what it did before in my qemu-system-arm VM [09:38] *seem [09:55] ogra: what do you mean? [09:57] hi all, i would appreciate an answer if someone know something about return codes when launching some proccesses... When i run some proccess and it finish its task, sometimes the return code in the terminal is like [1] Done | [3]- Done or [9]+ Exit 251 [09:57] Keybuk, i'm running a script that bootstraps a base system into a qemu image, fires up the image with a shellscript to proceed the second stage bootstrap and set up some stuff ... in the end the sript calls reboot -f which doesnt seem to behave like it did before [09:57] ok, that's a major bug, gksu no longer works. [09:57] it uses init= and indeed doesnt invoke upstart [09:57] i dont know the differences among [1] or a - / + after the number, or the exit code sometimes with + and other times with - .. [09:58] wincide, it's your shell outputting those, so a good first step would be its documentation ("man bash" if you're using that), even if it is intimidating [09:59] ogra: "doesn't seem to behave like it did before" ? [09:59] this is what I'm trying confused about [09:59] Keybuk, the vm doesnt shut down [09:59] did you try strace reboot -f ? [10:00] tricky to do in that setup [10:01] but i just had an insane idea, i can kill the vm from the outside after debootstrap is done with its second stage, do a normal boot and make the configuration script an upstart job :) [10:02] hi, is currently someone around who knows a bit more on the userspace side of v4l1->v4l2 transition and applications only capable of v4l1. There is a regression bug on the kernel side (bug 352765) which I believe is rather tied to the applications and the compat lib. [10:02] Launchpad bug 352765 in linux "Logitech webcam doesn't work (v4l2 support needed)" [Medium,Confirmed] https://launchpad.net/bugs/352765 [10:04] ogra: reboot -f/poweroff -f really don't do much other than call the reboot() syscall [10:04] Keybuk, ok, i thought they might try to route that through upstart or so now [10:04] thx liw, buy i've reading all the bash man, and the closest term related to what i trying to find out, is just the number of proccess launched, but ,anything about this kind of return codes with + - or exit with + or - .... [10:04] ogra: no, only if you don't call with -f [10:04] the thing is that the kernel and VM are the same as before, only the image content is karmic [10:05] without -f they call shutdown [10:05] so its some software thats blocking here [10:05] Keybuk: Is this a good time to start thinking about switiching from init.d scripts to upstart? [10:05] but i think i'll go with abusing upstart as installer, thats a funny idea [10:06] and will actually give me all initscripts out of the box instead of me having to invoke them from the script one by one [10:06] Daviey: to a limited degree, yes [10:07] Keybuk, is there a limitation on the amount of runlevels or could i just add a runlevel 7 or 8 ? [10:07] ogra: there are no runlevels 7 or 8 [10:07] Keybuk: thanks. [10:07] use 3, 4 or 5 - they're spare [10:07] Keybuk, no, i want them to stay untouched [10:08] you could use a different event than runlevel [10:08] being able to add runlevel 7 and just dump my job scripts in there would be cool [10:09] and delete everything related to 7 after i'm done [10:09] pam: I'm a bit confused. Why not just start with the working assumption that the upstream gfxboot module is sufficient, and test that? You could test for screen-clearing after 'localboot' (i.e. "Boot from first hard disk") easily enough, and I'd be surprised if force-prompt weren't handled upstream since otherwise it doesn't work very well with the standard menu system [10:09] hmm [10:10] pam: surely we shouldn't be getting a reputation of asking upstream to support Ubuntu's changes; that's bad form [10:11] wincide: it's described in the "JOB CONTROL" section of the bash manual page, specifically towards the end of the paragraph starting with "There are a number of ways to refer to jobs in the shell" [10:12] im taking a look again, lets check it... thx [10:14] Keybuk, oh, i suspect its even easier, i could just temporary replace /etc/init/rcS.conf [10:15] and call my config the same way sulogin is started [10:16] wow, thats so insane that i actually start to like it :) [10:21] you could do that [10:21] though that would mean forcing start into single-user mode [10:25] Keybuk, thats fine what bothers me since a while is that my script needs to invoke all the rcS stuff, thats error prone, using upstart works around that :) [10:26] and i can properly shut down [10:26] ogra: the rcS stuff? your script needs the recovery menu? [10:27] no, but mounted / and proc etc etc [10:27] ah, I see your confusing [10:27] rcS.conf is for what to do *while* in single-user mode [10:27] not the sysinit scripts [10:27] oh, k [10:27] (which Debian always confusingly put in rcS.d) [10:27] rc-sysinit.conf is the one you want to look at [10:27] ok, will do [10:28] it runs rcS.d [10:28] then runs telinit to switch to the next runlevel [10:28] yep i see [10:28] oh and telinit dies with my test image :/ [10:29] "dies" ? [10:30] yeah [10:30] but the image seems corrupt as well [10:30] grmbl [10:30] do you have a core dump? [10:31] telinit "did not recieve a reply etc etc" [10:32] i guess its the kernel i use for the VM [10:32] * ogra tries with a jaunty versatile one [10:35] bye, thx for all [10:35] weird, same issue === cjwatson_ is now known as cjwatson [10:44] Keybuk, was a corrupted fs ... [10:44] works with a different image [10:45] good to know [10:47] grmbl [10:47] whats up with my network today === ogra_ is now known as ogra [11:19] Ummm... why was a package seemingly just accepted into the jaunty release pocket? [11:20] oh, uploads to partner get sent to jaunty-changes too [11:20] cody-somerville, partner archive ? [11:20] yeah [11:27] udevd[9589]: udev: missing sysfs features; please update the kernel or disable the kernel's CONFIG_SYSFS_DEPRECATED option; udev may fail to work correctly [11:27] mumble === ember_ is now known as ember [11:33] quadrispro: around? [11:37] james_w: hi james [11:37] hi [11:38] I'm looking at ncurses [11:38] what does "Install wide-character patches into /{,usr/}lib32." mean? [11:39] james_w: only a moment, bug? [11:39] found [11:39] mvo: I saw your patch, I've been overly busy these days, I'll take care of it now [11:39] bug 399625 [11:39] Launchpad bug 399625 in ncurses "Merge ncurses 5.7+20090613-1 (main) from Debian unstable (main)." [Wishlist,Confirmed] https://launchpad.net/bugs/399625 [11:43] james_w: nothing, maybe I forgot to do some change... i'm checking it (and unsubscribing sponsors team for now) [11:44] quadrispro: thanks [11:44] it looks as though that changelog entry might have just been carried forwards for lots of uploads when it wasn't needed [11:50] enrico: thanks, no problem about the delay [11:55] StevenK: what do you think about the reply from fbreader upstream regarding their contributor guidelines, is this going to be a problem? -> http://www.fbreader.org/docs/contributions.php [12:01] mvo: uploaded, thanks a lot for the patch" [12:01] s/"/!/ [12:04] enrico: cool, thanks! [12:33] oh [12:34] did the initscripts massively change in karmic ? [12:34] especially the stop tasks in them ? === MacSlow is now known as MacSlow|lunch === akgraner_ is now known as akgraner === beuno_ is now known as beuno === MacSlow|lunch is now known as MacSlow === dwatson is now known as davewatson [13:35] seb128: could somebody please add a versioned Replaces on gnome-session to gnome-session-bin? I don't really want to accept the package in NEW since the missing Replaces will cause problems [13:35] cjwatson, will do, thanks for noticing [13:35] ta [13:39] cjwatson, in fact isn't the Conflicts enough? [13:40] seb128: I'm not convinced [13:40] cjwatson, I think Josselin argued before that using Conflicts only is enough and will assure things are upgraded in the right order [13:40] the safe rule is *always* to add a Replaces and it does no harm [13:41] right, I'm doing that, I just failed to convince Josselin apparently during previous discussions [13:41] I will ping him again about that [13:41] is Conflicts actually necessary? it makes the upgrade painful [13:41] using Conflicts means the old gnome-session package has to be entirely removed before installing the new gnome-session-bin [13:42] well I would say only Replaces is required [13:42] if possible, it would be better to use Replaces plus maybe Breaks [13:42] Breaks is not required, it's really a Replaces situation I think [13:43] right, in that case it's probably just the old-fashioned confusion between Conflicts and Replaces [13:43] yes [13:43] it's possible that he's technically right that Conflicts is enough, but it's definitely a suboptimal way to do it [13:46] cjwatson, Replaces version uploaded, you can accept the binaries now the Conflicts should work too or wait for the new build [13:46] ok, thanks [13:46] np, thanks for checking the binaries and letting me know about the issue [13:49] could someone finally fix lvm2 bug 246324 ; it's trivial and makes life easier... [13:49] Launchpad bug 246324 in lvm2 "initramfs-tools lvm2-hookscript won't include lvm.conf and claim devices it shouldn't " [High,Confirmed] https://launchpad.net/bugs/246324 [13:52] pitti: I've pushed a seed change to desktop-common to pull apmd on armel; apmd was demoted from main to universe in karmic, so I don't think I need a MIR to promote it back, do I? [13:53] ivoks: done [13:53] cjwatson: awsome; thanks [13:59] lool: right, we can just re-promote it; I just thought it was horribly outdated and ditched it from main [13:59] pitti: The Marvell SoC we target for karmic will rely on APM, yeah! [14:00] pitti: (I discovered that 10 minutes ago BTW) [14:00] modern! [14:01] * ogra wonders if devicekit-power can do APM at all === agateau_ is now known as agateau === dholbach_ is now known as dholbach [14:27] enrico: did you push your upload to git yet? [14:36] is fast-user-switch-applet being dropped in karmic? [14:37] mvo: ah, right, I forgot [14:38] mvo: done. [14:38] thanks! [14:42] james_w: thanks for the package branch announcement [14:42] c_korn, it's in gdm now [14:43] james_w: one thing that's not quite clear to me, should the official brnaches be committed/pushed to, or are they read-only and only an auto-import? [14:43] they should be pushed to [14:43] you just can't yet [14:43] see the part about permissions [14:43] james_w: ah, so right now, they are read-only [14:43] yeah [14:43] james_w: other question, should they only be full-source, or packaging-only branches as well? [14:44] so any commit you do are just "squashed" in to one [14:44] full-source is the one true way! [14:44] there was discussion at UDS of how to incorporate history from older packaging-only branches [14:45] pitti: I think we will have to update our desktop team workflow progressively to full-source branches :) [14:45] james_w: right, but both debian and ubuntu have a huge set of packaging-only branches [14:45] seb128: http://www.ubuntu-pics.de/bild/18739/screenshot_003_qv7BQY.png you mean that? [14:45] and at this point they are still better than using the auto-imports [14:46] didrocks: yes, probably [14:46] why are they better? [14:46] c_korn, yes, cf gdm changelog for details [14:46] c_korn, the ubuntu changes need to be ported to the gdm version [14:46] (not being argumentative, I'd like to fix the issues) [14:46] c_korn, what you have now is the upstream variant, tedg is working on porting the ubuntu changes [14:47] didrocks, no; no change in the ubuntu desktop workflow [14:47] ah that explains the depends on gdm [14:47] pitti: (FWIW there is no harm in continuing to use the old packaging-only branches outside the lp:ubuntu/karmic/ namespace until such time as the full-source branches are good enough) [14:47] seb128: not now, just when the branches will be ready to use (not read only as of now), don't you think? [14:48] didrocks, we have something working I think we should not be the first to run in new things, we have enough to do without fighting tools and learning new way every week [14:48] seb128: I can't agree more :) [14:52] james_w, one thing better with packaging only in vcs is the time it takes to fetch the source [14:53] james_w, it takes over 15 minutes to get the nautilus source there where it takes 1 minutes to get the tarball and the debian dir from a vcs [14:53] yes [14:53] StevenK: what's happening with https://launchpad.net/about-window ? [14:53] I don't think that should be the only deciding factor though, and it is something that is being worked on and has improved a lot. [14:54] james_w, well it's a matter of pro and con, and that's an annoying con without any real pro for full source in what we do now [14:54] well, let me try and convince you of the pros [14:56] james_w, ok ;-) [14:56] seb128: nautilus isn't imported yet, and I'd like to do an experiment [14:56] any other package of a similar size you work with a lot? [14:57] james_w, gnome-control-center? [14:57] nope :-) [14:58] seems a lot of gnome isn't done yet :-) [14:58] gtk+2.0 is a bit an extreme example [14:58] gedit? [14:58] or gvfs [14:58] pitti: scott ritchie reports that pcspkr sounds have been replaced with normal alert sounds in karmic https://bugs.edge.launchpad.net/ubuntu/+bug/77010 [14:58] Ubuntu bug 77010 in hundredpapercuts "Overuse of system beep without volume control" [Undecided,Confirmed] [14:58] gvfs is recent though [14:58] djsiegel_: indeed, I saw his comment [14:58] didn't check on that bug yet, though [14:58] james_w, rhythmbox? just pick anything desktopish... ;-) [14:58] pitti: glory glory hallelujah [14:59] cjwatson: This is what I am trying to do (test if the patches are still relevant). About the mail: since Stefen is rewritting the patch again in C, just wanted to ping him about these fixes and incorporate them in the new version if necessary. [14:59] pitti: ok, I will check and mark fixed when I get a chance, should have a machine to test tomorrow [14:59] pam: why not just plug it into an existing image and see if it works? :) [15:00] I don't have access to the C version yet, not released AFAIK [15:00] cjwatson: for the com16, sure [15:01] kirkland: why am I getting the motd every time I start a terminal window? that's incredibly annoying [15:02] seb128: empathy 38s vs 9s, so still some work to do [15:03] seb128: but when there's a new one to grab it should be quicker in bzr [15:03] james_w, that's already quite good my tests there where rather 10 times slower [15:03] that's why we are using the new format :-) [15:03] well there is a limit anyway which is the quantity of information you need to get [15:04] even git which is reputed to be fast is some 6-7 times slower than getting just the source in tarball [15:04] all the commits history is quite some datas [15:16] cjwatson: that will go away as soon as slangasek accepts the pam_motd patch [15:17] cjwatson: rm /etc/profile.d/update-motd.sh if you can't wait [15:17] cjwatson: update-motd package going away, see ubuntu-devel@ mailing list [15:18] * kirkland -> returns to his holiday [15:28] djsiegel_: You know, that's been stalled for about 5 releases? [15:28] StevenK: yes I know [15:29] StevenK: do you have any interest or bandwidth to unstall it? [15:34] kirkland: ok, thanks === kiko-afk is now known as kiko [15:46] * manjo moving to ubuntu-kernel as per rtg's command === mterry_ is now known as mterry [16:27] pitti, how long does it take to have a -dbgsym for a package? (say libdbus-1-3-dbgsym) [16:29] fta: until it's on ddebs.u.c.? Should be within 6 hours [16:30] 8, actually [16:30] 10 3,11,19 * * * [16:30] pitti, hm, ok. thanks. i guess i'll wait [16:30] 19:10, is that UTC? [16:31] Probably London. [16:31] BST [16:31] so, 1810 UTC [16:31] less than 3h.. ok, thanks [16:43] Keybuk: I see that upstart 0.6 provides /lib/init/upstart-job, but there's no Provides: upstart-job; oversight? want a bug report? [16:44] slangasek: ah, yes please [16:45] Should also have a Conflicts/Replaces I guess [16:46] yeah should be a super-switch thing [16:46] mbiebl: btw. did you get any further with the "better" upstart-job than my hacky shell script? [16:47] It's not much better, but it has support for being called by /etc/rc?.d/???servicename [16:47] and it also tries to much all action that are supposed to be available (like force-reload) [16:47] to corresponding upstart actions [16:49] mbiebl: should it? we shouldn't ever have more than one upstart-job implementation on a given arch? [16:50] mbiebl: can you file a bug and attach it? [16:50] slangasek: upstart vs. sysvinit both providing it [16:50] since when is sysvinit going to provide it? I thought the alternative implementation was going to be built from upstart source [16:50] yeah, the idea was to provide a upstart-job implementation for sysvinit [16:50] since it needs to understand upstart job formats [16:51] for people that don't want (or can make) the switch [16:52] can't make, meaning upstart is not yet available on their arch [16:52] ah, I forgot that we were giving people a choice in some cases. :) [16:53] right, but there are two alternate implementations in Debian [16:53] or at least will be [16:53] this alternate implementation is giving me some headaches though. [16:54] not sure yet, how the stop action can be implemented [16:58] mbiebl: killall5 should be fine [16:58] slangasek: I switched to #pkg-sysvinit === Sarvatt_ is now known as Sarvatt [17:31] pitti: do you have https://blueprints.launchpad.net/ubuntu/+spec/security-karmic-apport-abort on your schedule? I'd like to start using that feature now that it's in glibc. [17:33] Keybuk, "init: rc-sysinit main process (380) killed by TERM signal" ... i assume i'm missing a kernel feature (fiddling with qemu arm kernels) would anything strike you as something i could miss from the top of your head ? [17:33] apart from that message the VM works perfectly fine [17:34] ogra: when did you receive that message? [17:35] Keybuk, when the login prompt shows up [17:35] "the login prompt" ? [17:35] * Setting up console font and keymap... [ OK ] [17:35] * Starting system log daemon... init: rc-sysinit main process (380) killed by TERM signal [17:35] [ OK ] [17:35] * Starting kernel log daemon... [ OK ] [17:35] Ubuntu karmic (development branch) dove ttyAMA0 [17:35] dove login: test [17:36] thats the boot (through a serial console) [17:37] ah right [17:37] it's just what was in the syslog buffer then [17:37] ah [17:37] so it actually died before [17:37]