[01:29] <smoser> anyone know how i can disable (via preseed) deb-src lines from getting into /etc/apt/sources.list
[01:32] <cjwatson> You can't in general right now, short of late_command.
[01:33] <smoser> cjwatson, welll...
[01:33] <smoser> will anything run an apt-get update bAESD ON THAT?
[01:33] <smoser> (sorry for caps)
[01:33] <cjwatson> probably not but you can run your own
[01:33] <smoser> i dont  (and can't get) src
[01:34] <smoser> can't get the mirror of src
[01:34] <cjwatson> well, it'll already automatically comment out sources it can't reach
[01:34] <cjwatson> oh, except not
[01:34] <cjwatson>         # Ubuntu change: network sources are always valid; apt will cope
[01:34] <cjwatson>         # gracefully later, even though the network may not be available
[01:34] <cjwatson>         # now.
[01:34] <cjwatson> *shrug* comment them out in late_command, see if it works :)
[01:34] <cjwatson> quicker than analysing
[01:35] <cjwatson> specially since I have a complicated bug to fix before I can go to bed
[04:45] <pitti> Good morning
[04:45] <rickspencer3> good morning pitti
[04:46] <rickspencer3> 1 week until release, how're things looking today?
[04:47] <pitti> slangasek: the stack trace does not tell anything about which python code this came from; I'll check the duplicates if anyone has a recipe for reproducing
[04:47] <pitti> hey rickspencer3
[06:09] <micahg> @pilot in
[07:07] <micahg> If someone adds a large section to a man page, should they be added to authors?
[07:14] <dholbach> good morning
[07:20] <micahg> jamespage: did you consider an SRU for Bug #598385?
[08:03] <micahg> stgraber: 1.299 and 1.301 are between 1.314 and 1.315 along with a lot of other commits
[08:04] <StevenK> RAOF: How many rubber chickens do I have to kill to get Do working on precise?
[08:04] <RAOF> StevenK: I'll get back to you in a moment :)
[08:06] <stgraber> micahg: weird indeed, the tags look correct though (looking at the diffs)
[08:09] <micahg> stgraber: hmm, looks like it's the importers fault ISTR cjwatson lamenting about this earlier in the week
[08:09] <micahg> or it might have about something else the importer did wrong
[08:14] <micahg> stgraber: seems that r1011 reverted everything back to the last import and then committed the fixes one by one
[08:14]  * micahg files a UDD bug
[08:18] <micahg> stgraber: bug 985465 if you're interested
[08:19] <stgraber> micahg: thanks
[08:19] <micahg> stgraber: also, as you TIL, do you want to have a look at https://code.launchpad.net/~roadmr/ubuntu/precise/casper/973794/+merge/100893 (as this affects the same script you touched)
[08:20] <stgraber> micahg: that's the bug I fixed ;) rejecting the branch
[08:21] <micahg> stgraber: great, thanks
[08:22] <RAOF> StevenK: What's not working with Do on precise?
[08:22] <RAOF> StevenK: Apart from unity eating anything that looks like <super> again.
[08:27] <Riddell> pitti: are you ok with this change? http://paste.kde.org/459788/
[08:27] <Riddell> pitti: and am I right in thinking that pkgbinarymangler has no bzr branch?
[08:27] <RAOF> StevenK: (Also known as: Do Works For Me™ on precise)
[08:27] <pitti> Riddell: it uses the standard ubuntu:pkgbinarymangler UDD branch
[08:28] <pitti> Riddell: I recommend to set NO_PKG_MANGLE=1 in calligra-l10n
[08:28] <pitti> Riddell: this blacklist is somewhat hard to handle, as IS needs to manually update the conffile in all the buildd chroots
[08:29] <Riddell> pitti: ah ok
[08:37] <StevenK> RAOF: Yeah, that is my problem. Can I move Unity off Super?
[08:38] <RAOF> StevenK: I'm not sure; you'd need to go ferreting around in ccsm.  I've caved, and set <ctrl><alt><space> as my summon keybinding.
[08:40] <StevenK> RAOF: That was easy.
[08:44] <Sweetshark> doko_: ping?
[08:45] <Sweetshark> doko_: there is currently flameage building up on #libreoffice-dev wrt to ubuntus choice to default to --as-needed when linking, could you join and flame back responsibly.
[08:45] <Sweetshark> ?
[08:48] <Mirv> slangasek: should bug #540816 perhaps be marked as High, similar to the duplicate 967229 you marked as high importance?
[08:53] <micahg> Sweetshark: Debian's planning on doing the same at some point: http://wiki.debian.org/ToolChain/DSOLinking
[08:56] <Sweetshark> micahg: "debian is doing it" doesnt mean its right (which is what upstream cares about) per se, I need someone to argue the "why".
[08:56] <micahg> Sweetshark: that wiki tells you why it's "right"
[08:57] <micahg> quite simply, it violates encapsulation
[09:00] <Sweetshark> which is why the is a --as-needed switch. however, there are scenarios where you do not want that. Using --as-needed where possible is good, mindlessly fumbling with defaults is not,
[09:02] <Sweetshark> micahg: feel free to join #libreoffice-dev and discuss this with sberg
[09:02] <micahg> Sweetshark: if I didn't need to go to sleep, I'd consider it :)
[09:02] <micahg> Sweetshark: at the very least, you can show the wiki page as the "why"
[09:03] <Sweetshark> micahg: http://nabble.documentfoundation.org/make-check-problem-in-libtest-smoketest-building-master-td3902404.html#a3905049 i did already long ago, we are way past that.
[09:08] <micahg> Sweetshark: looks like fun :), doko can answer the questions much better than I in any event
[09:16] <micahg> @pilot out
[09:37] <davmor2> hey guys is it me or has scroll stopped working for people?
[09:50] <pitti> WFM
[09:53] <RAOF> davmor2: Possibly - can you try running ‘sudo modprobe -r psmouse && sudo modprobe psmouse’ and then try scrolling again?
[09:54]  * RAOF will be happy to get home, where the internet is stable enough for evolution to properly filter through the 3000-odd email backlog.
[09:56] <davmor2> RAOF: that seems to of done it I think most windows now seem to scroll thanks
[09:57] <RAOF> davmor2: Right.  You're running into the problem where the scroll valuator eventually gets stuck on -2,000,000ish.
[09:59] <davmor2> RAOF: ah could be
[10:36] <cjwatson> Is anyone in a position to test bug 969343?
[10:36] <cjwatson> (last comment)
[10:41] <doko> Sweetshark, joined -- however no flaming. disappointed ;-/
[10:42] <azeem> maybe _rene_ is on vacation
[10:42] <jjardon> dobey: Hey, Is the info I put here enough ;)? https://code.launchpad.net/~jjardon/intltool/autoreconf/+merge/99802
[10:46] <ogra_> nice ... http://www.heise.de/imgs/18/8/2/1/2/6/2/USB-Power-c433ad30a541bc3f.png ... use your laptop and a USB Y-Cable as a welder :)
[10:50] <doko> mvo, ping
[10:50] <kklimonda> on the http://wiki.debian.org/Multiarch/Implementation it's written that if you use /usr/lib subdirectories as a target for the .links file, this file has to be autogenerated in debian/rules - is there some example I could check to see how to do it?
[10:57] <ogra_> jodh, your serial.conf looks good, will test it soon, what i was wondering is if we couldnt bind it to the device rather than to runlevel [23] ... (not for precise, but i think it might make sense to have somethng like a serial-device-up event or so)
[10:58] <jodh> ogra_: agreed!
[10:58] <jeinor> Hello. I have problems getting Ubuntu Core up and running. I try to run the 11.10 release of Ubuntu Core for armel on a Cubox (http://www.solid-run.com/), which runs a patched kernel of version 2.6.32.9. I copied the SD card included with the Cubox using dd if=/dev/sdb of=/data/somefile.img. I then restored this .img to a new SD card, mounted the second partition (there are two, one boot and one rootfs) rootfs and removed all f
[10:58] <ogra_> so we dont tie to multiuser modes ;)
[10:58] <jeinor> I then untared the .tar.gz to the mounted partition, and also copied the /lib/modules/* from the original rootfs to the new ubuntu core rootfs.
[10:59] <jeinor> Could that possibly work? Or am I doing something terribly wrong?
[10:59] <jeinor> When I boot, the kernel boots, but when I switch to the console terminal using Ctrl+Alt+F1, all I get is a black screen.
[11:00]  * ogra_ invites jeinor to #ubuntu-arm :)
[11:00] <jeinor> woops
[11:01] <jeinor> I entered the channel mentioned at the bottom of https://wiki.ubuntu.com/Core.
[11:01] <jeinor> Wrong? Should I repost my question in the ubuntu-arm channel?
[11:01] <ogra_> sure, we can talk here as well ...
[11:01] <ogra_> your first sentence was cut off
[11:01] <ogra_> .... d one rootfs) rootfs and removed all f
[11:01] <ogra_> thats how it ended
[11:02] <jeinor> ok, it should have been "and removed all files."
[11:02] <ogra_> note though that ubuntu core is totally unconfigured, you need to do all bits an installer would usually do manually
[11:03] <jeinor> I understand. I actually got a Ubuntu Core up and running using qemu and a kernel from Ubuntu 11.04 before.
[11:03] <ogra_> (its main focus is to be used as a base for building custom images (i.e. modifying it) and chroots, it is not intended to be used out of teh box as a rootfs)
[11:03] <jeinor> I see. I am looking to customize it later on, but first I just want it up and running.
[11:04] <jeinor> Or am I doing it wrong?
[11:04] <ogra_> though i would suspect your kernel is missing parts we need in ubuntu
[11:04] <jeinor> The only configuration I did was to change the password for the root user so I should be able to login.
[11:04] <ogra_> like devtmpfs for example
[11:04] <jeinor> aha
[11:05] <jeinor> Well, I guess I can compile a kernel and try to boot using that. What version would I need for the ubuntu-core 11.10 release, do you know that?
[11:05] <ogra_> (if you dont use a initrd and dont have devtmpfs i'm not sure you have /dev/console at all for example)
[11:05] <jeinor> That would explain it.
[11:05]  * ogra_ would have to check the -core tarball if we ship a hardcoded node for it, but i dont think so
[11:06] <jeinor> However, the kernel shipped with Cubox is used (by default) to boot Ubuntu 10 or something. Perhaps that doesn't matter?
[11:06] <ogra_> best to take one of our arm kernels and compare the generic config options
[11:06] <ogra_> also is there any particualr reason you use armel and not armhf ?
[11:07] <ogra_> *particular
[11:07] <jeinor> I only found armel here: http://cdimage.ubuntu.com/ubuntu-core/releases/11.10/release/
[11:07] <jeinor> What's the difference?
[11:07] <ogra_> oh, yeah, 11.10
[11:07] <ogra_> speed :)
[11:07] <jeinor> Ah :)
[11:07] <ogra_> but armhf only exists in 12.04
[11:08] <jeinor> I see. Any system supporting armel should support armhl as well?
[11:08] <jeinor> armhf*
[11:08] <ogra_> well, all ARMv7 CPUs definitely support hf
[11:08] <ogra_> cubox is an armada, it should be able to
[11:09] <jeinor> "It is based  on the advanced Marvell 88AP510 SoC featuring  an ARMv7 core running up to 800MHz"
[11:09] <jeinor> found in specs
[11:09] <ogra_> right
[11:09] <jeinor> so I guess it should work
[11:09] <ogra_> well, check your kernel config first i'd say
[11:10] <jeinor> hmm.. I might need more information on how to do that, do you have any good resource I could read?
[11:10] <ogra_> things like having SIGNALFD ennabled are essential for example
[11:11] <jeinor> that's the one read by "make cubox_defconfig" for example, right?
[11:11] <jeinor> when compiling the kernel
[11:12] <ogra_> that would be the default config to make the board HW work ... not sure how it sets the generic options though
[11:13] <ogra_> http://paste.ubuntu.com/936755/ has a config of a well working ubuntu arm kernel
[11:13] <ogra_> (ignore the arch specific stuff and compare the generic options)
[11:14] <jeinor> ok
[11:14] <jeinor> I'll start with that
[11:14] <ogra_> and indeed, also make sure your cmdline is right (console= should have the right args to even see anything on the serial console etc)
[11:14] <jeinor> where do I specify the cmdline? (you can see I'm pretty new at this... :) )
[11:15] <ogra_> in the bootloader configuartion
[11:15] <ogra_> no idea what cubox uses or how thats configured though
[11:15] <jeinor> ok
[11:15] <ogra_> (likely u-boot since its marvell...)
[11:15] <jeinor> yes, it's u-boot
[11:16] <jeinor> in general, it should be possible to get Ubuntu Core running on that kernel? Or might it be too old?
[11:16] <jeinor> 2.6.32.9
[11:16] <ogra_> well, i used a .32 kernel in oneiric with one of my arm devices ... back then it worked fine
[11:17] <jeinor> ok
[11:17] <jeinor> thanks a lot for your help orgra_, i'll start with comparing those configurations
[11:17] <ogra_> great, good luck
[11:17] <jeinor> will probably be back later :)
[11:17] <ogra_> feel free :)
[11:18] <ogra_> btw, if the cubox comes with an ubuntu installed, you could just use that and upgrade ;)
[11:18] <jeinor> well, it's the desktop version installed now. I just want Core :)
[11:21] <jeinor> I'll check back later, bye for now
[11:27] <mvo> doko: pong
[12:36] <jibel> mvo, ping
[12:37] <mvo> jibel: pong
[14:12] <DreadKnight> can someone take a look at this usability bug report I've made? https://bugs.launchpad.net/unity/+bug/985675 thanks
[14:12] <seb128> DreadKnight, try #ubuntu-unity to discuss unity design
[14:13] <DreadKnight> seb128, thanks!
[14:13] <seb128> yw
[15:32] <slangasek> Mirv: there are two different bugs: one is the lack of smooth transition (which is low-priority and intractable), the other is that there's text when X exits.  The latter is high-priority and solvable, though it requires constant vigilance against stray console messages :/
[15:34] <Mirv> slangasek: ah, good point. indeed so.
[15:34] <Mirv> the text is what's worrying, a plain cursor blink would not too bad
[15:35] <Mirv> marking as separate again
[15:36] <slangasek> thanks
[15:39] <Shinobi> Where can I get the src for the ubuntu supported Banshee release?
[15:43] <infinity> jodh: Around?
[15:43] <jodh> infinity: yo!
[15:43] <infinity> jodh: You pinged me about bug #876626
[15:44] <infinity> jodh: I haven't had much time to look at it at all, so if you're looking at it, that's fine by me. :P
[15:44] <jodh> infinity: gr8, thanks. I'm currently looking at it.
[16:25] <slangasek> pitti: hey, would it make sense for the apport retracer to paste backtraces to each of the bugs it decides are duplicates, instead of duping them and leaving no info about the individual crashes?
[16:26] <pitti> slangasek: we could, but then we would need to keep the duplicates private
[16:26] <pitti> I'm not entirely sure which is worse
[16:26] <slangasek> well, what good does it do to have public bugs that are duplicates of a private bug? :-)
[16:26] <pitti> the master bug can be made public, but usually developers don't bother making the dupes public
[16:27] <pitti> and it's a good metric for how urgent a bug is
[16:27] <pitti> we have whoopsie now, of course, to tell us that
[16:28] <pitti> slangasek: so, it's technically easy to do
[16:38] <slangasek> pitti: right, so maybe whoopsie is the better route after all?
[16:42] <seb128> slangasek, better route for what?
[16:43] <seb128> slangasek, will we have full debug stacktrace with whoopsie or mozilla's like stacktrace, i.e just function names?
[16:43] <slangasek> for making sure we can retain a view of the individual crashes
[16:43] <slangasek> well, I hope it's meant to be full debug :)
[16:43] <slangasek> ev: ^^ ?
[16:45] <pitti> slangasek: is the whoopsie data accessible somehow? I thought it was meant to be kept very restricted
[16:54] <slangasek> pitti: well, I'm not sure
[16:54] <slangasek> I thought ubuntu devs were meant to be able to get access to it to poke around... else how would we be able to develop the kinds of hooks that would let us extract more info we need for debugging?
[16:57] <pitti> slangasek: ev would know better, but I thought initially this was meant to only give statistics, not individual bugs
[16:58] <slangasek> it's meant to allow us to aggregate crash info
[16:58] <slangasek> but I didn't think that meant we wouldn't have access to drill down
[16:58] <slangasek> but I'm speculating here :)
[17:00] <jodh> slangasek: fyi - bug 985755.
[17:01] <cjwatson> the mockups ev's shown me permit a degree of drilling down
[17:02] <slangasek> jodh: thanks
[17:25] <bryceh> @pilot in
[17:28] <kklimonda> slangasek: hey, do you have an example of multiarch package that uses /usr/lib as a target for the .links file? according to the http://wiki.debian.org/Multiarch/Implementation this file should be generated in the install target and I'm wondering how to best do that
[17:29] <slangasek> kklimonda: I don't think I have any good examples - this has been somewhat overtaken by developments in debhelper
[17:29] <slangasek> kklimonda: nowadays, I would do this with an executable .links file that uses dh-exec
[17:30] <slangasek> but I haven't actually done this *yet*, because all the packages I've converted were done before that was an option :)
[17:30] <cjwatson> kklimonda: it's not for multiarch specifically, but groff generates .dirs and .install files, and you could use an identical approach
[17:31] <kklimonda> thanks, I'll check both and see which one is less invasive :)
[18:18] <FAalbers> I've been searching a doc on how to setup scons in ubunto to build QT apps. But can't find any help. Is there someone her that could help ?
[18:26] <jeinor> Hello! And hello again ogra_ if you're there :)
[18:26] <jeinor> I tried joining #ubuntu-arm, but it said "Cannot send to channel" when I tried to.
[18:27] <jeinor> Don't know if I have to have permission to write there or something?
[18:27] <jeinor> Anyway, I am still stuck on not getting Ubuntu Core to boot on the Cubox (http://www.solid-run.com/)
[18:28] <jeinor> The Cubox is shipped with Ubuntu, so I figured the kernel I have from there should work for Ubuntu Core..?
[18:28] <jeinor> I just copied the SD card that came with the Cubox (using dd, full disk image using dd if=/dev/sdb of=/data/cubox.img)
[18:29] <jeinor> I then copied that back onto another SD card using dd if=/data/cubox.img of=/dev/sdb
[18:29] <jeinor> I then mounted the second partition (the Cubox default SD card has two partitions, one boot and one for the rootfs)
[18:30] <jeinor> So I mounted the rootfs, ran sudo rm -rf * on it and then copied Ubuntu Core 11.10 onto there (untared it directly to the root)
[18:30] <jeinor> However, when I boot, all I get is a command prompt with only a flashing underscore on ctrl+alt+f7 and a black screen at ctrl+alt+f1
[18:32] <jeinor> Anyone have any ideas? ogra_ told me to compare the kernel configuration with a known working kernel configuration, but I don't know how to that since the configuration file is 2800 lines with a lot of options
[18:34] <jeinor> And I am curious though, if Ubuntu 10.04 runs on the Cubox, shouldn't Ubuntu Core run as well then?
[18:41] <\sh> moins
[18:42] <\sh> kklimonda, ping freeipa :)
[18:42] <kklimonda> \sh: pong
[18:43] <\sh> kklimonda, I just got some attention to a project you are eventually involved...bringing freeipa to debian/ubuntu ;) it happens to be that I was encountered by a core dev of freeipa here at COL
[18:44] <\sh> kklimonda, just wanted to ask if you are still involved and if you need help with this
[18:45] <stgraber> \sh: you may also want to read https://blueprints.launchpad.net/ubuntu/+spec/servercloud-p-freeipa-tech-preview and talk to tjaalton
[18:45] <tjaalton> hmm I should mark that one finished, and postpone the last task
[18:46] <\sh> I just encountered one issue while compiling the client side:
[18:46] <\sh> /usr/bin/ld: warning: libkrb5.so.26, needed by /usr/lib/x86_64-linux-gnu/libgssapi.so.3, may conflict with libkrb5.so.3
[18:46] <stgraber> \sh: you realize you can do "apt-get install freeipa-client" on precise, right? :)
[18:46] <tjaalton> right
[18:47] <tjaalton> there are a couple of issues with the ipa-client-install script that I still intend to fix for precise..
[18:47] <kklimonda> \sh: we always need help :)
[18:47]  * kklimonda wonders what COL is
[18:47] <\sh> stgraber, yes...so this ld warning is interesting to me and eventually for you =? :)
[18:47] <\sh> kklimonda, citrix online
[18:47] <tjaalton> \sh: building against heimdal? don't do that
[18:48] <tjaalton> don't think upstream supports it
[18:49] <\sh> tjaalton, so I forget about that...now I need to backport it to 10.04 ,-)
[18:49] <kklimonda> ouch
[18:49] <tjaalton> uh oh
[18:49] <\sh> yes yes
[18:49] <tjaalton> the client?
[18:50] <tjaalton> nevermind
[18:50] <tjaalton> \sh: who did you meet?
[18:51] <\sh> tjaalton, jr aquino
[18:53] <tjaalton> ah, at citrix
[18:53] <\sh> tjaalton, where else ,-)
[18:53] <tjaalton> redhat? :)
[18:53] <\sh> tjaalton, never...was there 2001 ,-)
[18:55] <tjaalton> referred to the fact that most of the upstream is working for red hat, but anyway
[19:05] <\sh> tjaalton, did you fix the uintptr_t issue in ipa-join.c?
[19:06] <tjaalton> \sh: hm what?
[19:06] <tjaalton> bug# ref please
[19:10] <tjaalton> \sh: oh, got it..
[19:10] <tjaalton> yes, it's patched
[19:11] <tjaalton> included in 2.2 and later, precise has 2.1.4
[19:19] <kees> hallyn: cap question for you...
[19:19] <kees> ns_capable(child_cred->user->user_ns, CAP_SYS_PTRACE)    this means "does current have CAP_SYS_PTRACE in the child's namespace?" right?
[19:20] <\sh> tjaalton, cool...:)
[19:26] <bodhi_zazen> When making a live Ubuntu CD, how do you get the rootfs.squasfs so small ?
[19:26] <bodhi_zazen> What compression options are used ?
[19:26] <sladen> bodhi_zazen: lzma
[19:29] <bodhi_zazen> sladen: any idea of the exact command ?
[19:29] <bodhi_zazen> =)
[19:30] <sladen> bodhi_zazen: apt-get install live-build
[19:30] <bodhi_zazen> OK, I am familiar with live-build
[19:30] <sladen> bodhi_zazen: http://lists.debian.org/debian-live/2011/06/msg00152.html
[19:30] <bodhi_zazen> will look at those scripts
[19:34] <arges_> micahg, hello
[19:36] <micahg> hi arges
[19:37] <arges_> micahg, looking at pad.lv/943502 whois backport to oneiric
[19:37] <bdmurray> @pilot in
[19:37] <arges_> micahg, i see you moved some stuff around in the bug, and was wondering how I should help? I've started to build the PPA and was going to test it
[19:38] <micahg> arges_: yes, please continue, I just closed the SRU task since you decided to go the backport route and added a backport task
[19:41] <arges_> micahg, ok cool. Just making sure
[19:42] <arges_> micahg, also: pad.lv/968612   I was trying to backport openldap... I was able to build and install most of the packages, but wasn't sure how to test them each. Are there guidelines for this?
[19:42] <micahg> arges: just install and run or build, install, and run for the reverse dependencies
[19:43] <micahg> *reverse build dependencies
[19:44] <micahg> whatever "run" is for the package in question (start it, use an app that exercises the library)
[19:45] <arges_> micahg, ok. that one will take some time
[19:46] <micahg> arges_: right, which is why requests for packages like those mostly go unanswered
[19:47] <arges_> micahg, yea. i started writing some python script to take that request output and turn it into 'apt-get install' statements, so I could test
[19:47] <arges_> because it became very tedious
[19:50] <micahg> arges_: indeed
[19:50] <micahg> slangasek: Bug #985842 (I thought we shouldn't have these issues anymore :()
[19:52] <slangasek> micahg: it's not a gzip issue, it's a differing-contents issue
[19:52] <micahg> ah, ok, that's a relief of sorts :)
[19:52] <slangasek> micahg: we never fixed that for gcc-4.4; why is someone trying to do this with gcc-4.4?
[19:53] <micahg> no idea, just saw the bug come in and thought it was going to be a problem
[19:54] <micahg> why should the README.Debian have different contents per arch though?
[19:54] <slangasek> look and see for yourself ;)
[19:54] <slangasek> (it documents the per-arch patches applied to the source)
[19:55]  * micahg would think someone would want to possibly cross-compile with gcc-4.4 since it's multiarch enabled, they might have deceptively thought that they could install both :)
[19:56] <slangasek> heh, ok
[19:57] <micahg> I read some of the thread on debian-devel about this, so I know it's not so easy to fix
[20:03] <jeinor> If I have a Ubuntu system which boots to a flashing prompt. If I go ctrl+alt+f1, the screen turns completely black. What could cause this? :)
[20:08] <hallyn> kees: vacation today, afk, will be in tomorrow
[20:09] <sladen> jeinor: sounds like a bug -> bug tracker please.  We need traces to get it debuggered further
[20:10] <jeinor> sladen, well, it most probably is not a bug... probably me not knowing enough on kernels boot-loaders and the alike. Not exactly a standard installation I'm trying to get up and running.
[20:11] <sladen> jeinor: how is it non-standard?
[20:11] <jeinor> it's an Cubox arm system based on a pretty old 2.6 kernel which I am trying to get Ubuntu Core 11.10 running on
[20:11] <kees> hallyn: okay, have fun!
[20:12] <jeinor> I THINK I have done everything right, but apparently not
[20:13] <sladen> jeinor: I don't know that I can help.  The starting place would be to find somebody else with the same hardware and compare the two setups
[20:14] <jeinor> sladen: Thanks for the tip, though I have a working installation running Ubuntu 10.04 on the same hardware. I just copied that kernel and replaced the rootfs, but didn't work out.
[20:14] <jeinor> sladen: the working installation was shipped with the cubox (http://www.solid-run.com)
[20:16] <sladen> jeinor: I would try again, replacing the minimum at a time and bisect the point at which it explodes
[20:17] <sladen> jeinor: then at that point (with minimum delta between $known_working and $known_broken) start poking to see if it's just the graphics driving dying, or the whole kernel going up in flames
[20:18] <jeinor> sladen: yea, something like that. However, how do I gradually replace Ubuntu Core with Ubuntu Desktop 10.04? I guess the breaking point will be when I remove X. I think the problem is that it is trying to boot into X using tty7, but I can't get the text console on any tty
[20:18] <jeinor> *the other way around, "replace Ubuntu Desktop 10.04 with Ubuntu Core"
[20:19] <slangasek> jeinor: the gettys are configured with upstart using /etc/init/tty*.conf.  If you have no getty on any VT, you probably are failing to reach runlevel 2
[20:19] <slangasek> jeinor: what's the kernel commandline being passed?
[20:21] <jibel> during an upgrade from lucid to precise from the command line, some packages show this error/warning: dpkg-deb: file `/var/cache/apt/archives/ubuntu-docs_12.04.4_all.deb' contains ununderstood data member data.tar.xz     , giving up
[20:21] <jibel> what does it mean ?
[20:21] <jeinor> slangasek: 'console=ttyS0,115200n8 vmalloc=384M root=/dev/mmcblk0p2 video=dovefb:lcd0:1920x1080-32@60-edid clcd.lcd0_enable=1 clcd.lcd1_enable=0'
[20:22] <jeinor> what does "have no getty on any VT" mean? VT=virtual terminal, but getty?
[20:22] <slangasek> getty is the program that gives you a login prompt
[20:23] <jeinor> I see
[20:23] <slangasek> jibel: it means approximately this: http://lintian.debian.org/tags/data.tar.xz-member-without-dpkg-pre-depends.html
[20:23] <slangasek> jibel: however, that's supposed to be an archive-side reject rule in launchpad, so there shouldn't have been any such packages (!)
[20:23] <slangasek> cjwatson: ^^
[20:24] <slangasek> jibel: and indeed, ubuntu-docs has Pre-Depends: dpkg (>= 1.15.6)
[20:24] <slangasek> jibel: so what's the context of the error message?  That shouldn't have happened as part of the actual upgrade
[20:26] <jeinor> slangasek: there are files from tty1.conf to tty6.conf. If I press ctrl+alt+f1 at the flashing underscore, the flashing stops and the screen is just black. If I press back to ctrl+alt+f7, the flashing returns. Should I perhaps add a tty7.conf for that terminal?
[20:27] <jibel> slangasek, messages are displayed the early in the upgrade, right after the new packages are downloaded and before preconfiguration
[20:28] <jibel> there are 8 packages with this error in the test: cloop-utils_2.6.39.2-1ubuntu1_i386.deb libgl1-mesa-dri-dbg_8.0.2-0ubuntu3_i386.deb libgl1-mesa-glx-dbg_8.0.2-0ubuntu3_i386.deb libyaml-syck-perl_1.19-1_i386.deb python-m2crypto_0.21.1-2ubuntu2_i386.deb ttf-arphic-uming_0.2.20080216.2-4_all.deb ubuntu-docs_12.04.4_all.deb whois_5.0.15ubuntu2_i386.deb
[20:29] <slangasek> jeinor: no.  your problem is that you're *not reaching runlevel 2*.  try adding --verbose to the kernel commandline; that should spew a good amount of data at you
[20:29] <slangasek> jibel: aha, I thought it might be preconfiguration
[20:29] <slangasek> jeinor: some of the data may even tell you where it's blocked :)
[20:30] <slangasek> jibel: so the apt debconf hook tries to manually unpack the .deb files using the currently-available dpkg, so that it can extract and run the debconf .config scripts.  This will fail if dpkg doesn't yet understand the package format
[20:30] <slangasek> jibel: I'm assuming there's no impact other than the error message?
[20:31] <jeinor> slangasek: I see :) so I adjust my kernel command line to include --verbose. On which screen will it outut? Will I need to specify console=tty0 as well perhaps? (really grateful for your answers here, I've been stuck at this for a while now..)
[20:34] <jibel> slangasek, apparently not. what kind of impact could this error produce ? package is silently kept to the old version ?
[20:34] <slangasek> jibel: not at all; it just means that, if the package has any debconf questions that need to be shown, they won't be shown until much later in the install
[20:34] <bryceh> slangasek, there's a couple patches in the sponsor queue for multi-arch transitions.  Are multi-arch changes permissible as SRU?
[20:34] <jeinor> slangasek: I tried adding --verbose, but that didn't give me any output at all. I don't see any text at all until I see the flashing prompt.
[20:35] <slangasek> jeinor: do you have the serial console hooked up?
[20:35] <jibel> slangasek, ah ok. so no impact other than the error message
[20:35] <slangasek> jibel: correct - it just means package *pre*configuration will fail, which is an optimization only
[20:35] <slangasek> jeinor: (note that your first kernel commandline option tells everything to talk to the serial console, not the graphical console...)
[20:36] <slangasek> bryceh: I discussed that with pitti earlier this week; we're agreed to allow them, provided there's due diligence on making sure they're not regressing anything
[20:37] <bryceh> slangasek, thanks
[20:37] <jeinor> slangasek: I have the computer connected by HDMI to a screen, but I have not connected any serial console. I actually tried changing that first command line option to console=tty0 before, but that didn't give me any output either.
[20:37] <slangasek> bryceh: which one are you looking at, OOI?
[20:37] <bryceh> 977952 libbonoboui and 977940 gnome-vfs
[20:37] <slangasek> jeinor: Ubuntu tries to keep the console quiet by default, so that may be part of it; dropping the 'console' option entirely and adding --verbose may give you better results
[20:38] <slangasek> bryceh: ok
[20:38] <slangasek> bryceh: are you thinking to sponsor them?
[20:38] <bryceh> slangasek, also since we're in Final Freeze, should everything that isn't a targeted bug be going through -proposed and the SRU process now?
[20:38] <slangasek> yes
[20:38] <slangasek> and half of what is a targeted bug, too :)
[20:38] <bryceh> slangasek, well I was going to kick them out if a firm rule existed
[20:39] <slangasek> yeah, it doesn't
[20:39] <slangasek> if you don't sponsor them, I'll be looking at them later :)
[20:39] <bryceh> there's 3-4 I'm going to look at first, and my pilot session is long in the tooth so I likely will be leaving them for you :-)
[20:39] <slangasek> no problem
[20:39] <bryceh> er 3-4 other patches
[20:41] <jeinor> slangesek: I tried that now, still just a black screen :( not even the flashing prompt
[20:41] <jeinor> slangasek: perhaps I should try connecting to that terminal console
[20:44] <jeinor> slangasek: does the order of the kernel command line parameters matter? for example, the resolution and such are set after the console commands, perhaps if I put console=tty0 last it would apply that same resolution to tty0?
[21:36] <bdmurray> slangasek: bug 985735 isn't really about update-manager is it?
[21:37] <slangasek> bdmurray: nope, that prompt will come from eglibc
[21:37] <slangasek> because apparently we stopped restarting it from pam, but still do from eglibc?
[21:51] <slangasek> bdmurray: triaged
[21:53] <slangasek> Riddell: ^^ I'm not sure if kubuntu-dev gets bug mail; can I draw your attention to bug #985735, which needs input from Kubuntu ASAP?
[21:54] <slangasek> also, how did no one run into this for upgrades to maverick or natty and report it?  did it get lost in the noise of the pam restart?
[21:57] <Dr_Who> doko: ping
[22:10] <Riddell> slangasek: replied
[22:13] <bryceh> bdmurray, we're cutting a swath through the sponsor queue today :-)
[22:16] <ScottK> slangasek: FWIW, I agree with Riddell.
[22:21] <bdmurray> slangasek: maybe it has to do with bug 944876 and not knowing about updates?
[22:22] <slangasek> Riddell, ScottK: ack; will pull kdm from the service restart list
[22:22] <slangasek> bdmurray: srsly?  augh
[22:23] <bdmurray> I'm not positive but it could be
[22:23] <slangasek> well, I guess *that* should've been noticed if it was affecting the whole Kubuntu userbase
[22:24] <slangasek> Riddell, ScottK: any insight into why this has only just been reported now?
[22:25] <ScottK> You mean now as in a month and a half ago?
[22:26] <Riddell> I'm not sure why that software-properties bug is related to the kdm restart issue
[22:26] <slangasek> ScottK: I'm talking about bug #985735 - if this really breaks kdm on upgrade, why did nobody raise it before?
[22:27] <sladen> nasty
[22:27] <Riddell> it never showed up in beta 2 upgrade testing
[22:27] <Riddell> and ScottK tested an upgrade today and never reported it
[22:27] <ScottK> Oh.  That one.  I thought you meant the software-properties one.
[22:27] <ScottK> The upgrade I did was in a chroot.  KDM wasn't actually running.
[22:28] <slangasek> heh
[22:28] <ScottK> I have done real upgrades and not had the problem though.
[22:28] <Riddell> has it only appeared after beta 2 ?
[22:29] <ScottK> I upgraded my laptop about a week and a half ago and didn't see it, but these things also come down sometimes to package update sequence and that's not deterministic.
[22:29]  * ScottK has to go.
[22:30] <slangasek> ScottK: I don't see any way it could be dependent on update sequence
[22:30] <ScottK> OK.
[22:31] <slangasek> libc6 is always going to be upgraded and always want to restart
[22:31] <ScottK> True.
[22:36] <slangasek> Riddell, ScottK: so there were changes to the restart code in eglibc as of Apr 12, but nothing that should have changed *whether* kdm is restarted
[22:36] <slangasek> only when
[22:37] <Riddell> I did plenty of upgrade testing before beta 2 so a bit of a mystery that
[22:37] <Riddell> I'll probably do some more upgrade testing tomorrow so I'll check this bug to see if I should expect it to work or not
[22:40] <stgraber> that's one kind of issue the daily upgrade testing won't find as it just runs do-release-upgrade from an ssh session, though I guess we could configure the VM to auto-login and ensure that the session is still alive post-upgrade...
[22:45] <bryceh> anyone know offhand if apport has been turned off now?
[22:45] <bryceh> @pilot out
[22:46] <bdmurray> bryceh: yes
[22:46] <bdmurray> bryceh: er, yes I know and yes it has
[22:46] <bryceh> bdmurray, thanks
[22:46] <bdmurray>   * etc/apport/crashdb.conf: Disable Launchpad crash/kernel reports for the
[22:46] <bdmurray>     final release. Leave Apport running for whoopsie, though, so use the new
[22:46] <bdmurray>     mechanism for selective problem types.
[22:47] <slangasek> infinity, Riddell, ScottK: blast, it's my bug - sorry
[22:50] <infinity> slangasek: Oh, you found where it went wrong?
[22:51] <slangasek> yah
[22:51] <infinity> Good, cause I hadn't gotten there yet. :P
[23:27] <bdmurray> @pilot out
[23:33] <FAalbers> ly ! Got qt4 working on my Linux box with scons ! :)
[23:57] <\sh> oh damn...adding a devmapper device (via multipath) prevents 12.04 from booting.