[03:32] <gloved_hands> sudo apt-get build-dep linux-image-$(uname -r) is not working
[03:33] <gloved_hands> kernel-wedge cannot be found
[03:55] <gloved_hands> sudo apt-get build-dep linux-image-$(uname -r) is not working
[03:55] <gloved_hands> kernel-wedge cannot be found
[04:07] <gloved_hands> hey
[04:09] <gloved_hands> obake
[04:10] <gloved_hands> apw where is obake
[04:44] <gloved_hands> apw: key up?
[05:36] <fingertips> apw
[06:27] <fingertips> apw more secure?
[07:26] <fingertips> apw: have you been on the www lately?
[08:14] <apw> gloved_hands (N,BFTL), well kernel-wedge _is_ a package so that makes no sense, i would expect you to be using sudo apt-get build-dep linux, though.
[08:14] <apw> fingertips, i have literally no idea what that refers to
[08:25] <fingertips1> apw: the original 4
[08:25] <fingertips1> darpa
[08:26] <fingertips1> This is sme amazon cloud
[08:26] <fingertips1> It doesn't even connect to darpa
[08:27] <fingertips1> the internet was started with 4 machines
[08:28] <fingertips1> this is some virtual cloud routing system 
[08:30] <fingertips1> apw: why doesn't lucid have a kernel-wedge?
[08:30] <apw>  kernel-wedge | 2.29ubuntu3 | lucid            | source
[08:30] <apw> it does ...
[08:33] <fingertips1> apw: build-dep complains when using apt-get but not aptitude
[08:33] <apw> perhaps it is a bug in teh apt resolver in that release, dunno, seems supprising as that is how
[08:34] <apw> the deps get installed on a builder when we build the kernel in the archive ...
[08:37] <fingertips1> apw using this routine wiki.ubuntu.com/Kernel/BuildYourOwnKernel?
[08:38] <apw> the builders follow a subset of that, that in concept, unpack a chroot for the series, unpack <.dsc>, apt-get build-dep <srcpackge>, dpkg-buildpackage -b, upload results
[08:44] <fingertips1> apw: it doesn't even download the same source package as uname -r reports
[08:45] <fingertips1> What is the lucid amd64 version string?
[08:46] <fingertips1> apw: When you say 'we' what is that supposed to mean?
[08:47] <fingertips1> apw: Try to communicate.
[08:49] <fingertips1> apw: A bit faster.
[08:54] <fingertips1> apw: we need to get secured with myself
[08:56] <fingertips1> apw: key up?
[09:06] <apw> we == "the ubuntu distro" in that sentence
[09:12] <fingertips1> apw: apt-get source linux-image-$(uname -r) this is downloading some other version
[09:12] <apw> some other version?  which version and what is uname -r saying
[09:14] <fingertips1> Yes a version different from what uname -r shows.
[09:15] <apw> and which are they, else i have no way to try and reproduce that
[09:15] <apw> no way to understnad the problem
[09:16] <fingertips1> 2.6.32
[09:16] <fingertips1> and the string goes on after that wih a -
[09:17] <apw> uname -r won't be saying that
[09:17] <fingertips1> 2.6.32-
[09:17] <apw> right .. i need the whole of it
[09:17] <apw> else i can't reproduce your exacty command
[09:18] <fingertips1> 2.6.32-xx
[09:18] <fingertips1> the -xx part differs
[09:18] <apw> well i want the one you say is producing a download of the wrong version
[09:19] <fingertips1> apw: it looks like it is compiled against the debian sources with some scripts to make changes
[09:21] <fingertips1> apw: It doesn't matter what the exacty version strings are they differ.
[09:25] <apw> fingertips1, ok tested with my local version 3.13.0-39-generic which downloaded linux 3.13.0-39.66, which is correct, so it is specifi to you
[09:26] <fingertips1> and what commands build it?
[09:27] <fingertips1> did build deps run?
[09:27] <apw> i am referring to apt-get source linux-image-$(uname -r) 
[09:28] <apw> if you want someone to figure out your isue with it, you need to tell me the actual version your uname -r returns, if ou won't do that, i can't help
[09:31] <fingertips1> apw: will you install it?
[09:32] <apw> fingertips1, there is almost no way i could know what the "it" in that sentence refers, so ... i have no idea
[09:36] <fingertips1> apw: connect to my jabber server
[09:43] <fingertips1> apw: this is downstream
[09:43] <fingertips1> apw: there is something changing the code
[09:44] <fingertips1> do you want to come oon this side of the wall?
[09:44] <apw> this is a simple question with a simple non-secret answer, there is no point in being coi about it
[09:44] <apw> i am perfectly happy out here in public thanks, that is the nature of open source
[09:45] <fingertips1> To help identify what is going on.
[09:48] <fingertips1> brb apw
[10:45] <liveuser> making progress
[11:18] <liveuser> apw does ubunu support skype?
[11:21] <liveuser> CAN THE MIND GO INTO an intel cpu?
[12:02] <liveuser> apw: Why would anybody wan't Ubuntu on a chromebook?
[12:03] <amitk> liveuser: to use it for development?
[12:13] <liveuser> netsplit google?
[12:14] <liveuser> the crypto keys installed oon chromeboks
[12:39] <liveuser> Why does lspci persist in reporting the wrong model for the wireless chip? This happened after loading a non working kernel module driver. Before loading it no model number was reported.Now, after unloading it the wrong model persists to bve reported.Is this a sign that wrong firmware was sent to the chip?
[12:40] <apw> it is a sign that the kernel module you loaded thought it identified the device and recorded that in the kernel
[12:41] <apw> likely it will remain the same until something else identifies it differently
[12:43] <ogra_> or reboot ... 
[14:24] <VadimT> Hi. I have some problem with kernel. Error: Read-error on swap-device. What it can be? P.S: Ubuntu 14.04.1 LTS. Kernel: 3.13.0-24-generic
[14:25] <VadimT> Thank you.
[14:28] <apw> VadimT, could be a lot of things including a bad disk under the swap partition, it would depend somewhat on the errors just before
[14:33] <VadimT> apw, Thank you for answer. 
[14:43] <fingertips> apw: wrong eeprom
[14:48] <apw> VadimT, np, you might want to pastebin the error, and we might be able to help, never know
[14:52] <VadimT> apw, I have only a photo with this error... http://1drv.ms/1ytucTU
[14:57] <VadimT> apw, I can find this error in /var/log/syslog and in other same places too.
[14:59] <VadimT> *can`t
[15:01] <smb> VadimT, from the picture your kernel crashed. It is relatively normal that in those cases there is no time to write to any logs. What it says is that your second hard drive seems to have gone away from the kernels point of view. One could not say why for sure from that information.
[15:02] <apw> VadimT, well that error says that the drive returned a failure with additional sense information, but that that information was nothing valid according to the spec.  therefore the IO was failed and that happened to be a page from init, so the world ended
[15:04] <VadimT> apw, Hmm... Some tips?
[15:07] <apw> VadimT, that really says the disk behind your swap space had a hickup, if it was mine
[15:07] <apw> VadimT, i would test the swap space and see if it recurrs, if not, i'd write it off as sunspots or similar
[15:13] <fingertips> apw wrong eeprom
[15:16] <apw> that is a matter of oppinion
[15:16] <VadimT> apw, thx for answer.
[15:17] <fingertips> Oh it doesn't work apw
[15:17] <fingertips> There is no wlan0 interface created.
[15:19] <VadimT> apw, Okay. I`ll try you advice. Good day you. One more time thank you. Bye 
[15:21] <didrocks> hey
[15:21] <fingertips> apw I need skynet back up
[15:22] <fingertips> apw I can run netsplits on a compute cloud
[15:22] <apw> didrocks, hi?
[15:22] <didrocks> apw: some question, we are striking on bug #1387090 with some systemd case. After asking on system-devel, there are multiple proposals that Lennart is suggesting at http://lists.freedesktop.org/archives/systemd-devel/2014-November/025370.html
[15:22] <fingertips> apw: make the eeprom
[15:23] <didrocks> apw: on his proposal a), which is having the / writable from the initrd, is there practical reason to not do that?
[15:23] <didrocks> I can think of fsck happening quite early, before plymouth as a drawback
[15:23] <didrocks> anything else I'm missing?
[15:24] <apw> didrocks, well fsck can't happen before plymouth as we use its output channels
[15:25] <didrocks> yeah, so, it would silence it and we will have no feedback on fsck status
[15:26] <didrocks> apw: this would be the only reason? (I'm trying to have some reasonable answer on why preferring e) to a) on the ML)
[15:27] <apw> didrocks, on a call, i'll read it in bit
[15:27] <didrocks> apw: no hurry, thanks :)
[15:29] <didrocks> pitti is suggesting we do a) (mounting / as rw in initrd, and so fsck) only if /etc/machine-id is empty
[15:37] <ogra_> didrocks, what if there are errors fsck wants input for ? 
[15:38] <didrocks> ogra_: exactly my point and why I'm arguing against it :)
[15:38] <didrocks> that and having 2 code paths/behavior/user feedback channels
[15:38] <didrocks> (on the condition /etc/machine-id being empty, like after a factory reset)
[15:39] <didrocks> but I want to ensure that my arguments are rights, hence the questions here :)
[15:39] <ogra_> drop the old code path and pull machine-id into the initrd 
[15:41] <ogra_> note that on the phone i plan to pull Mir into the initrd at some point, for exactly this reason
[15:41] <didrocks> ogra_: machine-id is per machine identification
[15:41] <didrocks> not something static
[15:42] <didrocks> it's like the dbus id if you prefer, which is stable for a machine once initialized
[15:42] <ogra_> yes, pull it fromm the rootfs ... at this point you already know where it lives ... mount it ro ... cp it over and then run your fsck
[15:43] <ogra_> i assume with systemd mountall will be dead in debian ... and i doubt we want to carry it along either ... 
[15:43] <didrocks> ogra_: hum, however, there are tricks, if /etc/machine-id is empty, but you have the dbus-id in /var/lib/dbus/machine-id, this one is reused
[15:43] <ogra_> so the old code path is a dead end 
[15:43] <didrocks> yeah, I'm not looking at the old code path for now
[16:43] <apw> didrocks, ok what i am not clear on is how we get into this situation of not having a system-id, is this in a virgin image booting for the first time ?
[16:44] <didrocks> apw: yeah, the foundation team is trying to have some clean instance that you can use and spawn
[16:45] <apw> didrocks, so what stops us say, having it empty n th
[16:45] <apw> didrocks, so what stops us say, having it empty in the virgin image, let systemd do its thing, putting a tmpfs on it and making one up
[16:46] <apw> and then postprocessing that at ro->rw time to copy it down into the image
[16:46] <didrocks> apw: that's proposal e) and I'm in favor of that one
[16:46] <ogra_> ++
[16:46] <apw> into the persistant part
[16:46] <didrocks> apw: just wanted to ensure we couldn't do sanely a) first
[16:47] <apw> didrocks, i'd say you are asking a lot to do that, making those images rather different
[16:48] <didrocks> apw: agreed, I'm happy thus to do e), it's an interesting exercise to make it race-free anyway :)
[16:48] <apw> i think e) is interesting in your namespace world, unmount the overmount in there and create a file underneath :)
[16:49] <didrocks> right ;)
[16:58] <didrocks> thanks for your feedback btw! :)
[17:53] <jodh> apw: I'm seeing an entry in the mount table that cannot be unmounted (mount claims: "not mounted"). It could be related to the fact we're bind mounting on top of a sym link. Ever seen this?
[18:30] <apw> jodh (N,BFTL), not seen that no, if you have a reproduce by, perhaps i can try and figure it out