gertvdijk's questions

For security reasons I would like to install a secondary completely independent operating system on a single-hard-drive computer using UEFI. Both installations should be able using different LUKS crypto partitions as root device (LVM on top of crypto) and need to be able to use their own kernels.

Having dual boot with two different operating systems is not a big deal. A Ubuntu and Fedora will not interfere and can share the EFI system partition, but two Ubuntu installations will fight with each other because they want both install their Grub instance on the same EFI location.

• Is there a way to change Ubuntu to install Grub on a different location than /boot/efi/EFI/ubuntu during installation? I need a persistent way, so that package updates and kernel/grub updates still install in that path. Just like how Fedora's Grub would install in /boot/efi/EFI/fedora. I can't find that configuration option in Grub. If I could, my Dell UEFI boot menu would allow me to choose which OS to load. Perhaps I'm missing some glue between efibootmgr and grub-efi here.

• Alternatively, would it be possible to use a single Grub installation from one of the two Ubuntu installations and have it automatically detect the right boot settings? I mean, I know os-prober will find the other /boot partition with kernels, but how can it know the right boot parameters for that OS (e.g. LVM volume inside locked crypto volume as root filesystem)? I can understand it can with unencrypted single-partition installations, but it can't match a /boot with its corresponding root filesystem, right? And it's completely impossible if encrypted. So that would require a lot of manual tweaking of the grub configuration files. I also dislike this option of a single shared bootloader configuration, making the two operating systems less isolated to some extent.

I'm fine with server text-based expert installation, by the way. Also some custom configuration as long as I don't have to repeat the steps for every kernel/Grub update.

I've just upgraded Ubuntu 14.04 to Ubuntu 15.04 (well, actually using KDE/Kubuntu, but it appears to be irrelevant) and I noticed colouring in the terminal is a bit messed up when using byobu. The most annoying one is symbolic links being colored with blinking. Looks like a website from the nineties.

So,

gert@e7240:~⟫ ln -s Desktop test
gert@e7240:~⟫ ls -l test
lrwxrwxrwx 1 gert gert 7 Apr 29 23:52 test -> Desktop

Then test from the ls -l output is blinking.

Why? And how can I disable this and restore the behaviour of Ubuntu 14.04 here?

It happens unrelated to the used terminal. In Konsole, xterm, etc.

Also, the highlighting of searches in less appears italic rather than highlighted before (negative).

LS_COLORS environment variable is different from inside and outside a byobu session:

inside:

env | grep -F LS_COLORS
LS_COLORS=rs=0:di=38;5;5:ln=4;5;37:mh=00:pi=40;33:so=01;35:do=01;35:bd=40;33;01:cd=40;33;01:or=40;31;01:su=37;41:sg=30;43:ca=30;41:tw=3;28:ow=34;42:st=37;44:ex=38;5;202:*.tar=01;31:*.tgz=01;31:*.arj=01;31:*.taz=01;31:*.lzh=01;31:*.lzma=01;31:*.tlz=01;31:*.txz=01;31:*.zip=01;31:*.z=01;31:*.Z=01;31:*.dz=01;31:*.gz=01;31:*.lz=01;31:*.xz=01;31:*.bz2=01;31:*.bz=01;31:*.tbz=01;31:*.tbz2=01;31:*.tz=01;31:*.deb=01;31:*.rpm=01;31:*.jar=01;31:*.war=01;31:*.ear=01;31:*.sar=01;31:*.rar=01;31:*.ace=01;31:*.zoo=01;31:*.cpio=01;31:*.7z=01;31:*.rz=01;31:*.jpg=01;35:*.jpeg=01;35:*.gif=01;35:*.bmp=01;35:*.pbm=01;35:*.pgm=01;35:*.ppm=01;35:*.tga=01;35:*.xbm=01;35:*.xpm=01;35:*.tif=01;35:*.tiff=01;35:*.png=01;35:*.svg=01;35:*.svgz=01;35:*.mng=01;35:*.pcx=01;35:*.mov=01;35:*.mpg=01;35:*.mpeg=01;35:*.m2v=01;35:*.mkv=01;35:*.webm=01;35:*.ogm=01;35:*.mp4=01;35:*.m4v=01;35:*.mp4v=01;35:*.vob=01;35:*.qt=01;35:*.nuv=01;35:*.wmv=01;35:*.asf=01;35:*.rm=01;35:*.rmvb=01;35:*.flc=01;35:*.avi=01;35:*.fli=01;35:*.flv=01;35:*.gl=01;35:*.dl=01;35:*.xcf=01;35:*.xwd=01;35:*.yuv=01;35:*.cgm=01;35:*.emf=01;35:*.axv=01;35:*.anx=01;35:*.ogv=01;35:*.ogx=01;35:*.aac=00;36:*.au=00;36:*.flac=00;36:*.mid=00;36:*.midi=00;36:*.mka=00;36:*.mp3=00;36:*.mpc=00;36:*.ogg=00;36:*.ra=00;36:*.wav=00;36:*.axa=00;36:*.oga=00;36:*.spx=00;36:*.xspf=00;36:

outside:

env | grep -F LS_COLORS
LS_COLORS=rs=0:di=01;34:ln=01;36:mh=00:pi=40;33:so=01;35:do=01;35:bd=40;33;01:cd=40;33;01:or=40;31;01:su=37;41:sg=30;43:ca=30;41:tw=30;42:ow=34;42:st=37;44:ex=01;32:*.tar=01;31:*.tgz=01;31:*.arc=01;31:*.arj=01;31:*.taz=01;31:*.lha=01;31:*.lz4=01;31:*.lzh=01;31:*.lzma=01;31:*.tlz=01;31:*.txz=01;31:*.tzo=01;31:*.t7z=01;31:*.zip=01;31:*.z=01;31:*.Z=01;31:*.dz=01;31:*.gz=01;31:*.lrz=01;31:*.lz=01;31:*.lzo=01;31:*.xz=01;31:*.bz2=01;31:*.bz=01;31:*.tbz=01;31:*.tbz2=01;31:*.tz=01;31:*.deb=01;31:*.rpm=01;31:*.jar=01;31:*.war=01;31:*.ear=01;31:*.sar=01;31:*.rar=01;31:*.alz=01;31:*.ace=01;31:*.zoo=01;31:*.cpio=01;31:*.7z=01;31:*.rz=01;31:*.cab=01;31:*.jpg=01;35:*.jpeg=01;35:*.gif=01;35:*.bmp=01;35:*.pbm=01;35:*.pgm=01;35:*.ppm=01;35:*.tga=01;35:*.xbm=01;35:*.xpm=01;35:*.tif=01;35:*.tiff=01;35:*.png=01;35:*.svg=01;35:*.svgz=01;35:*.mng=01;35:*.pcx=01;35:*.mov=01;35:*.mpg=01;35:*.mpeg=01;35:*.m2v=01;35:*.mkv=01;35:*.webm=01;35:*.ogm=01;35:*.mp4=01;35:*.m4v=01;35:*.mp4v=01;35:*.vob=01;35:*.qt=01;35:*.nuv=01;35:*.wmv=01;35:*.asf=01;35:*.rm=01;35:*.rmvb=01;35:*.flc=01;35:*.avi=01;35:*.fli=01;35:*.flv=01;35:*.gl=01;35:*.dl=01;35:*.xcf=01;35:*.xwd=01;35:*.yuv=01;35:*.cgm=01;35:*.emf=01;35:*.axv=01;35:*.anx=01;35:*.ogv=01;35:*.ogx=01;35:*.aac=00;36:*.au=00;36:*.flac=00;36:*.m4a=00;36:*.mid=00;36:*.midi=00;36:*.mka=00;36:*.mp3=00;36:*.mpc=00;36:*.ogg=00;36:*.ra=00;36:*.wav=00;36:*.axa=00;36:*.oga=00;36:*.spx=00;36:*.xspf=00;36:

I like the basic highlighting without using Byobu and now I want to stop it taking over that environment variable values. How do I do that?

It looks like this byobu change introduced the behaviour (as intended mostly). But please, it's ugly and it should honour the LS_COLORS from the environment in my opinion.

On my Kubuntu 14.04 system I'm trying to manage keys with my SSH agent, but somehow it appears to ignore my ssh-add commands. Look at this below and you'll see what I mean.

1. List the current keys

   ⟫ ssh-add -l
   2048 60:6f:58:ef:7c:b0:ec:94:fb:fa:59:21:86:3d:fc:4c gert@e6230 (RSA)
   

   This key is loaded at boot time, but I expected some ECDSA key, not RSA. I don't know this key...

2. Remove the key from the agent.

   ⟫ ssh-add -D
   All identities removed.
   

   yey! But... is it?

   ⟫ ssh-add -l
   2048 60:6f:58:ef:7c:b0:ec:94:fb:fa:59:21:86:3d:fc:4c gert@e6230 (RSA)
   

   What the hell? It just lies to me.

3. What's going on here?

   ⟫ env | grep -i ssh
   SSH_AUTH_SOCK=/run/user/1000/keyring-eDJggO/ssh
   

   Let's see what process is running that socket.

   ⟫ sudo fuser -u /run/user/1000/keyring-eDJggO/ssh
   [sudo] password for gert: 
   /run/user/1000/keyring-eDJggO/ssh:  9434(gert)
   ⟫ ps -p 9434 u
   USER       PID %CPU %MEM    VSZ   RSS TTY      STAT START   TIME COMMAND
   gert      9434  0.0  0.0 292528  7192 ?        Sl   00:05   0:00 gnome-keyring-daemon [...]
   

   What the hell is GNOME keyring doing on my KDE system? Shouldn't KDE wallet be my SSH agent here?

This leads to more questions than answers and I'm left with a non-functional ssh-agent.

On another system I don't observe this behaviour and I fail to find a configuration difference. Both have only KDE installed and the packages installed are nearly identical (managed by Puppet).

After the BEAST attack and Heartbleed bug, now I've heard about a new vulnerability in SSL/TLS called POODLE. How do I protect myself against being exploited?

• Are only servers or also clients affected?
• Is this OpenSSL/GnuTLS specific?
• What kind of services are affected? Only HTTPS or also IMAPS, SMTPS, OpenVPN, etc.?

Please show me examples on how to avoid this vulnerability.

When using kinit to acquire a Kerberos ticket I have configured it to use a default realm, e.g. GERT.LAN by editing /etc/krb5.conf:

[libdefaults]
        default_realm = GERT.LAN

That's great since I don't have to supply that all the time on the command line.

⟫ kinit
[email protected]'s Password:

However, my local username gert does not match the remote username gertvdijk. Now I have to supply the full principal name as an argument still. If this is just kinit I could create a bash alias, but more Kerberos tools appear to try my local username. For example Kredentials does not allow me to use another than default principal.

So, basically, what I want is to create a mapping between the local user gert and the remote principal [email protected].

Ironically, when using a more complicated setup with PAM I am able to achieve this. In krb5.conf:

[appdefaults]
        pam = {
                mappings = gert [email protected]
        }

But I don't want to use the Kerberos PAM module anymore since I've locked out myself so many times by thinking the Kerberos server isn't reachable and I'm trying to enter the local password...

So, long story short, is there a way to configure a default principal or a mapping from local usernames?

For some reason I need to disable a system service on boot. In my case this is the Avahi Daemon. I just don't want it to start on boot.

In previous Ubuntu releases (e.g. 12.04) I could do

sudo update-rc.d avahi-daemon disable

However on 14.04 it shows

update-rc.d: warning:  start runlevel arguments (none) do not match avahi-daemon Default-Start values (2 3 4 5)
update-rc.d: warning:  stop runlevel arguments (none) do not match avahi-daemon Default-Stop values (0 1 6)
 System start/stop links for /etc/init.d/avahi-daemon do not exist.

And Avahi is started on boot. How can I disable this? I'm looking for a command line way of doing this.

And please care to explain what happened here? Some change in a recent Ubuntu release broke the use of update-rc.d? I suspect I'm missing something here.

When using full disk encryption with Ubuntu (opposed to homedir encryption), dm-crypt with LUKS is used for encrypting the volume. In the installer (at least on 12.04 alternate) you can choose between setting it up using a passphrase or a keyfile. I'd like to use a combination of the two; not either, but require both.

Why? Because this enhances security (two-factor); you'll need to have something and you need to know something to unlock it. Then I want to put the keyfile on a small removable storage device (USB flash drive) and only plug it in during boot time. The result should be that it is required to put in the right flash drive and provide the right passphrase to unlock the root partition.

So, put in other words, I want to be asked during boot for the passphrase to which the keyfile on an external drive is encrypted.

I see a /usr/share/initramfs-tools/hooks/cryptgnupg helper script that may help to accomplish it, but I have no clue how to use it.

_{Just to avoid confusion: I am not asking for a way to add an additional key to the volume to unlock it.}

I'm trying to run an application in another language as my current environment. To start out simple, I wanted to change the language of ls.

Here's what I read all over the place (see below for related questions): set LANG to one of the supported locales, e.g.:

LANG=nl_NL.UTF-8 ls /nonexistent

to have it show Dutch (NL) output.

However, I still get English output:

$ LANG=nl_NL.UTF-8 ls -al /nonexistent
ls: cannot access /nonexistent: No such file or directory

My current environment:

$ locale
LANG=en_GB.UTF-8
LANGUAGE=en_GB:en
LC_CTYPE="en_GB.UTF-8"
LC_NUMERIC="en_GB.UTF-8"
LC_TIME="en_GB.UTF-8"
LC_COLLATE="en_GB.UTF-8"
LC_MONETARY="en_GB.UTF-8"
LC_MESSAGES="en_GB.UTF-8"
LC_PAPER="en_GB.UTF-8"
LC_NAME="en_GB.UTF-8"
LC_ADDRESS="en_GB.UTF-8"
LC_TELEPHONE="en_GB.UTF-8"
LC_MEASUREMENT="en_GB.UTF-8"
LC_IDENTIFICATION="en_GB.UTF-8"
LC_ALL=

To be sure, I've checked:

• The locale is supported:

  $ grep nl_NL /usr/share/i18n/SUPPORTED 
  nl_NL.UTF-8 UTF-8
  nl_NL ISO-8859-1
  nl_NL@euro ISO-8859-15
  

• The locale is generated:

  sudo locale-gen nl_NL.UTF-8
  

  and sudo dpkg-reconfigure locales also shows it's generated.

• Installed the Dutch language pack (already installed):

  sudo apt-get install language-pack-nl
  

What else have I tried?

• Using export to set both LANG and LANGUAGE instead of prepending the command.
• Setting also LC_ALL.

Ironically, some (only some!) GUI applications are actually in Dutch, but I haven't configured that at all!

I'm on Kubuntu 12.04(.2), as far as that matters.

Similar questions (to no avail):

• Is it possible to change language for user interface temporarily?
• How to override the system locale on a single command?
• Temporarily change language for terminal messages/warnings/errors (demonstrated to be working, but not working for me)
• How do I add locale to ubuntu server?

What's wrong on my system? Where to debug this further?

In the top right corner of a standard KDE/Kubuntu installation you'll see this:

But seriously, I seem to be unable to understand what those "activities" are and what it would enable me to profit from. It just pops up like an "add widget" screen with just four options:

• New Activity - default plasma widgets desktop - is this a bug that it's called like that?
• Search and Launch - sort of over-spacious netbook interface?
• Photos Activity - I don't have any pictures local on my machine, useless for me.
• Desktop Icons - traditional style desktop-icons-only

Choosing one of those options seems to create a completely new desktop "plasma". Then why is this called Activity? Am I missing something?

Is this is nifty feature I've been missing on for a few years now or is this not that exciting? How do I use this as it's supposed to work?

Googling for what this is supposed to be give me all kind of vague descriptions like this one:

No feature defines the KDE 4 release series more than Activities. At the same time, no feature is so little understood -- Fedora even has a package for removing the desktop toolkit, which provides mouse access to Activities.

But, when you take the time to learn about Activities, you'll find them a natural extension of the desktop metaphor that just might help you to work more efficiently.

Activities are a super-set of Virtual Desktops. They don't replace Virtual Desktops -- in fact, each Activity can have its own set of Virtual Desktops if you choose. Instead, Activities are alternative desktops, each of which can have its own wallpaper, icons, and widgets.

Sorry, but after reading such a vague and impractical story I'm out. Can someone give me an overview of what I can do with it? I guess this is supposed to provide a way to separate private and business work, but then, how?

Tasksel or plain Apt?

In the past I've always used tasksel to install so-called "tasks". It seems, however, that this has been integrated in APT:

apt-cache dumpavail | grep ^Task
[...] snip
Task: lamp-server
[...]

and that I can install such "tasks" by appending a caret (^) to it, like this:

sudo apt-get install lamp-server^

Is this equivalent to the following?

sudo tasksel install lamp-server

And which is preferred?

Metapackages

Additionally, how do metapackages come into play here? Some tasks seem not to be a metapackage while others are:

apt-cache show lamp-server
N: Unable to locate package lamp-server

apt-cache show kubuntu-desktop
Package: kubuntu-desktop
[...]
Task: kubuntu-desktop

So, using the last example, what is the difference between the following three possible ways to install?

• sudo apt-get install kubuntu-desktop
• sudo apt-get install kubuntu-desktop^
• sudo tasksel install kubuntu-desktop

Suppose I know the numerical product/vendor IDs of hardware I don't have here and I want to know if there's support for me in the kernel, I can do this (USB bluetooth receiver example):

modinfo btusb
filename:       /lib/modules/3.7.5-030705-generic/kernel/drivers/bluetooth/btusb.ko
[...]
alias:          usb:v0A5Cp*d*dc*dsc*dp*icFFisc01ip01in*
alias:          usb:v0489p*d*dc*dsc*dp*icFFisc01ip01in*
alias:          usb:v413Cp8197d*dc*dsc*dp*ic*isc*ip*in*
[...]

From this I can see which devices are to be claimed by btusb and it requires me to know the module name on beforehand. In some cases however, I don't have a clue about the module(s) name(s) supporting such a device.

Networking hardware is one of those categories which seems very hard to find out about. Also, the lspci and lsusb names are sometimes inconclusive or simply wrong. I'm currently grepping through the whole kernel source tree, which is not that elegant.

• Is there a way to list all modaliases of all modules (including built-in to the kernel) in one go so I can find a match?
• Or, alternatively, is there a way to query the kernel to provide me the kernel modules claiming such a device I don't actually own?

Use cases:

• Users asking questions here which I can answer/dupe with by actually verifying that "support is in if you upgrade to..." or "Try rmmod module1; modprobe module2 as module2 seems to support your device as well."
• Checking support before buying hardware with prior knowledge of the IDs. Additional to searching with the IDs, I can then look for bug reports on the kernel module itself.

Lately I saw some of my friends using a mouse to copy & paste text around in the virtual terminal (Ctrl+Alt+F1). I was amazed that this was possible and would like to use this too.

He wasn't sure how this was possible on his machine, but it doesn't work on my machine. How do I get this to work?

^{Actually I just made up this story about a friend. I do know how this works and I answered my own question.}

Why older kernel?

For whatever reason out there, you might be forced to run another kernel than the ones provided by Ubuntu. It may even take you back a few years for a kernel that is compatible with specific precompiled kernel modules, your Xen/container-based VPS provider may force you to use his kernel, etc.

I have this question for a long time, but this sparked it again today.

In such a case it would be very useful to be able to say whether you can blame the kernel for issues or whether you should even bother to try to set up a more recent version of Ubuntu in your situation.

Policy, documentation?

I'm particularly interested in what the objectives of the developers/QA are in this regarding LTS releases and newer stable running the LTS-kernel. Some closely related questions:

• What is the policy on compatibility with earlier kernel versions? E.g. no bug reports accepted, must work with all kernel versions back to and including previous LTS, etc.
• Example case, practically: How likely will I be in trouble when running Lucid's kernel on Precise?
• To what extent is software relatively close to the kernel (udev, gvfs, mdadm etc.) being tested on other than the version provided with the release?
• How does Desktop/Server edition differ in this?

The most obvious place to look at this would be the Release notes. However, besides updates/changes to the Ubuntu flavoured kernel, this doesn't mention anything about compatibility with other kernels, while kernel-related features are being mentioned in other parts of the notes, e.g.

Software RAID now supports bad block management (MD).

Is Ubuntu simply not bothering about these cases or am I missing a resource on this? Besides the release notes, I have used Google to quite some extent using keywords: Ubuntu 12.04 minimal kernel version required and several variants to it. Yet, no statement about it appears to be made on those results. I now find this question popping up as only relevant resource. I did find this answer, tough, and it looks very promising, yet it's about a specific issue/environment and not really about server/desktop usage.

Userspace vs kernel

I know most pieces of userland software should not bother about kernel versions, but it's getting more complicated for VPN software or applications interacting with hardware, like the one above, but also for example V4L2, Network Manager, Alsa, etc.

Debian vs Ubuntu

Debian is really clear on this. Already for Wheezy, we know that if you rely on udev, it will require 2.6.26 to run properly from it's release notes (in the works):

The udev version in wheezy requires a kernel of version 2.6.26 or newer with [...]

What I'm not asking for

I am very well aware about the backports provided for newer kernels from newer releases to the current LTS version. This question is about the opposite.

Please avoid any discussion like "why would one want to run an older kernel?" - you just don't have a choice sometimes and it's not about what we want, but how one can deal with such a given situation.

I'd like to see to what extent my system is LTS-supported by means of what packages are supported for 5 years and which are not. I could disregard some non-5yr-supported packages, as some are rarely used or very unlikely to get into (security) issues.

I think this is useful as one can get a report and draw a conclusion, e.g. "my system is 100% LTS", "due to packages X,Y,Z, my system is just 99% LTS", "due to the use of KDE, my system is now 50% LTS".

As related to my answer in the question Does 12.04 LXDE have LTS?, I posted a way to see which packages of Ubuntu feature five years of support. E.g.:

$ apt-cache show unity | grep ^Supported
Supported: 5y

$ apt-cache show lxde-core | grep ^Supported
<no output>

I could write a script to get all information for all the packages, however, the apt-cache commands are horribly slow:

real    0m1.535s
user    0m1.484s
sys     0m0.036s

With 2700+ packages installed, this would take roughly 70 minutes (!).

How can I speed up things and get a report for all non-5yr-supported packages on my system?

I'd prefer a simple apt-* shell command for the use in a simple shell script. If it would require more advanced scripting like going into Python, this is fine too. Eventually, I would like to release a (small) script to create a report on a system easily and quickly.

^{Note: I'm not interested in the discussion about whether or not a specific flavour of Ubuntu provides LTS or not - this is really just packages. You can just mix LTS and non-LTS packages on a system.}

On this site I see hundreds if not a thousand questions being asked about issues with a Wireless LAN adapter not being detected/recognized by Ubuntu. Some mainstream adapters appear to be utterly broken in Linux support, while others just require some simple steps to get it to work.

The options to get it to work appear to be infinitely huge and I have no clue where to get started!

What is considered 'best practice'? What order of things to try is most efficient in terms of time, effort and imposes minimal risk to my Ubuntu installation? Please help me save time and guide me what to do, step by step.

I have an Ubuntu 12.04 server that just crashed because of a very obvious cause: 30+ of apt-check processes consuming all memory, the OOM killer kicking in, killing vital services. I'm not sure where the apt-check processes come from, but I guess my Nagios/Icinga plugins check_apt might use it, as well as the byobu status line may want to display its output. I guess something locked up and all of the processes were just waiting, yet holding memory.

How can I prevent to have so many instances of apt-check on the system? It doesn't make sense to me and it should just quit as soon as it can't get a read lock on the dpkg database.

It seems that I'm not the only one running into trouble here. All suggestions for apt-check are pretty negative:

^{(clean browser, not logged in, no personalised search)}

Trying to download Django source package using dget and the .dsc file from here.

Then dget fails to verify the source, because it is missing a public key.

$ dget https://launchpad.net/ubuntu...
[...]
dscverify: python-django_1.4.1-2.dsc failed signature check:
gpg: Signature made Tue 21 Aug 2012 09:12:04 CEST using RSA key ID F2AC729A
gpg: Can't check signature: public key not found
Validation FAILED!!

Fine, I added this key to my keyring, because I think I can trust this one:

$ gpg --keyserver keyserver.ubuntu.com --recv-key 0xF2AC729A
gpg: requesting key F2AC729A from hkp server keyserver.ubuntu.com
gpg: key F2AC729A: "Raphaël Hertzog <[email protected]>" not changed
gpg: Total number processed: 1
gpg:              unchanged: 1

^{(well, this output is run afterwards, when it was already in my keyring)}

And it is really there:

$ gpg --fingerprint F2AC729A
pub   4096R/F2AC729A 2009-05-07
      Key fingerprint = 3E4F B711 7877 F589 DBCF  06D6 E619 045D F2AC 729A
uid                  Raphaël Hertzog <[email protected]>
uid                  Raphaël Hertzog (Debian) <[email protected]>
uid                  Raphaël Hertzog (Freexian SARL) <[email protected]>
sub   2048R/71F23DEE 2009-05-07

But still fails for the same reason:

gpg: Signature made Tue 21 Aug 2012 09:12:04 CEST using RSA key ID F2AC729A
gpg: Can't check signature: public key not found

I'm running all these commands as my regular user. I also tried adding this key to APT's keyring and root's keyring. Both to no avail.

What am I missing here?

Somehow after upgrading to 12.04, my virtual machines always boot with an attempt to boot from the network first. See this:

while I don't have any PXE configuration set:

I've tried:

• to disable SPICE, by changing the emulator to /usr/bin/kvm from /usr/bin/kvm-spice by editing the XML.
• Ctrl+B to configure the iPXE, but it doesn't let disable this as a boot option.
• setting another type of NIC - not an option, I need virtio for performance reasons. However, e1000e doesn't work either.
• removing the NIC: works. However, I need network.
• Googling around. Hard. Lots of result is about failing configured PXE boots.

Not a big issue, but it increases boot times by 50-100% here (booting from SSD), so it's relatively long and annoys me.

How can I disable this and boot from virtual hard disk directly?

I own a Logitech M555b mouse for use with my HP Elitebook 8570w laptop running Kubuntu 12.04. Works fine right after connecting using the KDE Bluetooth control module.

However, after some time (seemingly random), it starts to lag. Movements are being delayed for roughly 500ms for a short period of time. Usually it recovers after some time too, but it can take minutes. All actions are being delayed: movements, click, scrolls. Additionally, the movements can be choppy during these times.

A workaround that always works for the same short period of time is to disconnect an re-connect the mouse. This can be done using the same KDE Bluetooth control module.

What did I try already?

• Running this at boot time:

  echo on > `readlink -f /sys/class/bluetooth/hci0`/../../../power/level
  

  To disable any power saving features on the Bluetooth hci0 device.

• Check the mouse's batteries (it's just a week old, other new batteries: same result)

• Checking logs and kernel messages about Bluetooth-related entries: none aside the expected messages on connect time.
• I'm running kernel 3.5.0-13-generic as provided in the xorg-edgers PPA. Booting the regular 3.2 Precise kernel results in the same behaviour.

Some other information that may help:

• It happens when no other Bluetooth connections are active on the machine.
• Similar symptoms also occur on my Bluetooth stereo (A2DP) headset, but then it's audio lagging and skipping. Swapping Bluetooth profiles as described here then helps. Conclusion: it's not the mouse that's faulty.
• The headset always worked fine using my now dead Thinkpad T61p with built-in Bluetooth.

• The bluetooth module in my laptop is connected via USB and shows up as

  0a5c:21e1 Broadcom Corp. 
  

• Turning the built-in Bluetooth adapter off and using another one works fine, without lag.

  0a5c:2046 Broadcom Corp. Bluetooth Device
  

I'm mobile and several people around me are using Bluetooth at work (A2DP mostly). It also occurs at home, where my neighbours are probably using Bluetooth as well. It could just be radio interference, but I think Bluetooth connections should just hop to another channel. And, moreover, it just works properly instantly when re-connecting.

Therefore I think it's a software driver issue and I'd like to debug it. Is there any way to get more verbose logging on the Bluetooth(-hid) modules?

First of all, I'm aware a similar question for GNOME is asked here: "Switch off laptop backlight when locking screen".

Objective

I would like to turn off my screen on locking the session for power saving reasons.

Actual problem

Locking the screen on Kubuntu (KDE) inevitably triggers the screensaver as far as I can see. There's no screensaver option other than 'Blank screen' together with its background colour set to black that comes just close to my goal. It blanks the screen, but doesn't turn off the screen. Screen's backlight will still be on and not saving any power.

Current workaround

A workaround via a script + shortcut key is possible, however, it's just a workaround since it doesn't trigger on all ways to lock the session. Therefore, I think it should be possible to have it done more elegantly, for example by providing this option in KDE's configuration dialog of the screensaver.

The workaround I am now using is the following. A script that locks the screen and turns off the screen:

#!/bin/bash

qdbus org.freedesktop.ScreenSaver /ScreenSaver Lock
xset dpms force standby

and let it run with a shortcut key via a custom menu entry. It works.

Here's why I consider it to be a workaround rather than a solution. It doesn't work for other ways to trigger the locking of the session.

My actual question(s)

Do I need to touching/patching KDE's source?

• If not what are my options?
• If so,
  • could someone point me to where I can get started?
  • what do you think is the recommended place in the GUI for configuration?

I'm using Kubuntu 12.04 and willing to upgrade to KDE 4.9 or waiting for the 12.10 release.