Re: How to mount a LUKS partiotion with Nautilus with option discard?

[Jochen Spieker]

Joerg Desch:
>
> I'm using a separate LUKS encrypted partition on my SSD, which I only
> mount after login. Since there is now way to to this with crypttab/fstab,
> I have to use Nautilus to do this.

Hm? I don't understand why you say mounting manually is not possible
with crypttab/fstab. Just make the entries as usual and add the option
"noauto" in fstab. That way you are asked for the passphrase during
boot, though. The crypttab file supports the noauto option as well. Then
you have to run two commands for mounting the filesystem:

cryptdisks_start $mapped_device
mount $mount_point

> After clicking on the (unmounted) LUKS partition, the system asks for the 
> LUKS passphrase and than for the admin password. Now the partition is 
> mounted as expected. The only problem is the missing 'discard' option, so 
> I can't use fstrim with this partition.

You don't need the discard option for your filesystem to run fstrim. You only
need it in /etc/crypttab. How do you tell that you "can't use fstrim"?
Do you receive an error message when running fstrim?

> Does anyone knows some tricks how to add discard support?
> 
> This is the current output of mount:
> 
> /dev/mapper/udisks-luks-uuid-... on /media/PRIVAT_C type ext4 
> (rw,nosuid,nodev,relatime,user_xattr,barrier=1,data=ordered,uhelper=udisks)

As pointed out above, this only shows the filesystem option. The status
of your LUKS device can be queried like this:

# cryptsetup status home-decrypted
/dev/mapper/home-decrypted is active and is in use.
  type:LUKS1
  cipher:  aes-xts-plain64
  keysize: 256 bits
  device:  /dev/mapper/kida-home--crypt--lv
  offset:  4096 sectors
  size:88436736 sectors
  mode:read/write
  flags:   discards

J.
-- 
People only fall in love because they cannot bear the though of being
alone.
[Agree]   [Disagree]
 


signature.asc
Description: Digital signature

Patch wheezy's glibc to run on ancient kernel (2.6.16)?

[Clemens Eisserer]

Hi,

I would like to run debian wheezy on my nokia-770 (Linux-2.6.16.27) in a
chroot environment, unfourtunately chroot telling me the kernel is too old.
The latest version that worked this way is Debian Lenny, which is
unsupported since mid 2012.

Because updating the kernel almost impossible (binary wlan driver blob,
texas instrument`s patched OMAP source tree, ancient toolchain), is there
any way to patch a debootstrapped debian installation with a self-compiled
glibc that is compatible with older kernel versions?
I reason I ask is because the minimum kernel version also seems to be
stored in all the executeables too, e.g. I get the following output on a
MIPS box:

root@OpenWrt:/usr/bin# file xz
xz: ELF 32-bit MSB executable, MIPS, MIPS-II version 1 (SYSV), dynamically
linked (uses shared libs), for GNU/Linux 2.6.26, .

... however as the binary does not talk directly to the kernel (or, does
it?), I don't understand how the binary itself has a requirement on the
linux kernel.

Thank you in advance, Clemens

Re: How to mount a LUKS partiotion with Nautilus with option discard?

[Joerg Desch]

Am Mon, 01 Sep 2014 09:48:30 +0200 schrieb Jochen Spieker:

> Joerg Desch:
>>
>> I'm using a separate LUKS encrypted partition on my SSD, which I only
>> mount after login. Since there is now way to to this with
>> crypttab/fstab, I have to use Nautilus to do this.
> 
> Hm? I don't understand why you say mounting manually is not possible
> with crypttab/fstab. Just make the entries as usual and add the option
> "noauto" in fstab.

OK, here is my /etc/crypttab

private_luks UUID=... none luks,noauto,discard


And this is my /etc/fstab

/dev/mapper/private_luks /media/privates ext4 user,nofail,noauto,noatime


The LUKS partition is installed on a PC where several people have access!
So I don't want the system to ask for the passphrase at boot time! It 
must aks for the passphrase after the login!

If I use a LUKS encrypted USB stick, all this is done. But with a 
partition on the SSD, Wheezys GNOME don't do it. I don't know why.




> The crypttab file supports the noauto option as well. Then
> you have to run two commands for mounting the filesystem:
> 
> cryptdisks_start $mapped_device mount $mount_point

Is it possible to get this working with the GUI? 


> You don't need the discard option for your filesystem to run fstrim. You
> only need it in /etc/crypttab. How do you tell that you "can't use
> fstrim"?
> Do you receive an error message when running fstrim?

Yes! ioctl failes while TRIm is not supported.

fstrim: /media/PRIVAT_C/: FITRIM ioctl failed: Die Operation wird nicht 
unterstützt



> As pointed out above, this only shows the filesystem option. The status
> of your LUKS device can be queried like this:
> 
> # cryptsetup status home-decrypted /dev/mapper/home-decrypted is active
> and is in use.
>   type:LUKS1 cipher:  aes-xts-plain64 keysize: 256 bits device: 
>   /dev/mapper/kida-home--crypt--lv offset:  4096 sectors size:   
>   88436736 sectors mode:read/write flags:   discards
> 

/dev/mapper/udisks-luks-uuid- is active and is in use.
  type:LUKS1
  cipher:  aes-cbc-essiv:sha256
  keysize: 256 bits
  device:  /dev/sdb7
  offset:  4096 sectors
  size:408795136 sectors
  mode:read/write



-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/lu1fjn$abc$1...@ger.gmane.org

Re: How to mount a LUKS partiotion with Nautilus with option discard?

[Joerg Desch]

Am Mon, 01 Sep 2014 08:18:09 +0200 schrieb Hans:

> cryptsetup luksOpen /dev/sda4 home2
> mount /dev/mapper/home2 /mnt
> 

Thanks. The manual way on the shell is working, but I has some hopes that 
the GUI would be working too! An mentioned in my other followup, LUKS 
encrypted USB sticks are handled as expected.



-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/lu1fo3$abc$2...@ger.gmane.org

Re: How to mount a LUKS partiotion with Nautilus with option discard?

[Hans]

Am Montag, 1. September 2014, 09:54:11 schrieb Joerg Desch:
> Am Mon, 01 Sep 2014 08:18:09 +0200 schrieb Hans:
> > cryptsetup luksOpen /dev/sda4 home2
> > mount /dev/mapper/home2 /mnt
> 
> Thanks. The manual way on the shell is working, but I has some hopes that
> the GUI would be working too! An mentioned in my other followup, LUKS
> encrypted USB sticks are handled as expected.
Hi Joerg!

As far as I remember (and someone may correct me, if I am wrong) if there is 
an entry in fstab, the devices are not mounted by the GUI (I guess you mean 
dolphin). 

Fstab is used for mounting at boot. Try to comment out the entry in fstab, but 
leave the entry in crypttab. 

Hint: You can use normal entries like /dev/sda6 or whatever in these files. For 
testing it ois not needed to use the UUID=.. entries.

If everything is working, you can easily do an 

dpkg-reconfigure  linux-base

which will automatically correct the entries to the UUID-stuff.

Feel free to ask for more.

Best

Hans 


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/2092094.8JDS8ZdR6x@protheus7

backup drive has died.

[Sharon Kimble]

Last night my backup drive died and is now totally unresponsive, but it
did give this error message when I tried to access it -

--8<---cut here---start->8---
Error mounting /dev/sde1 at /media/boudiccas/back1: Command-line
`mount -t "ext4" -o "uhelper=udisks2,nodev,nosuid" "/dev/sde1"
"/media/boudiccas/back1"' exited with non-zero exit status 32: mount: wrong fs
type, bad option, bad superblock on /dev/sde1, missing codepage or helper
program, or other error In some cases useful info is found in syslog - try
dmesg | tail or so
--8<---cut here---end--->8---

Is there any way in which I can access the contents of the drive please?

If not, is it safe to format it and re-use it please? If so, what should
I format it as ? ext3? ext4? Or something else?

It has for some months been showing this message in logwatch -

--8<---cut here---start->8---
 WARNING:  Kernel Errors Present
EXT4-fs (sde1): error count since last ...:  1 Time(s)
EXT4-fs (sde1): initial error at time 13973814 ...:  1 Time(s)
EXT4-fs (sde1): last error at time 13973814 ...:  1 Time(s)
--8<---cut here---end--->8---

and I did query what it meant and what I should do about it, but I was
reassured that it was OK and nothing needed to be done. In the light of
what's just happened, wrong!

Thanks
Sharon.
-- 
A taste of linux = http://www.sharons.org.uk
my git repo = https://bitbucket.org/boudiccas/dots
TGmeds = http://www.tgmeds.org.uk
Debian testing, fluxbox 1.3.5, emacs 24.3.93.1


signature.asc
Description: PGP signature

Re: backup drive has died.

[Gary Dale]

On 01/09/14 06:19 AM, Sharon Kimble wrote:

Last night my backup drive died and is now totally unresponsive, but it
did give this error message when I tried to access it -

--8<---cut here---start->8---
Error mounting /dev/sde1 at /media/boudiccas/back1: Command-line
`mount -t "ext4" -o "uhelper=udisks2,nodev,nosuid" "/dev/sde1"
"/media/boudiccas/back1"' exited with non-zero exit status 32: mount: wrong fs
type, bad option, bad superblock on /dev/sde1, missing codepage or helper
program, or other error In some cases useful info is found in syslog - try
dmesg | tail or so
--8<---cut here---end--->8---

Is there any way in which I can access the contents of the drive please?

If not, is it safe to format it and re-use it please? If so, what should
I format it as ? ext3? ext4? Or something else?

It has for some months been showing this message in logwatch -

--8<---cut here---start->8---
  WARNING:  Kernel Errors Present
 EXT4-fs (sde1): error count since last ...:  1 Time(s)
 EXT4-fs (sde1): initial error at time 13973814 ...:  1 Time(s)
 EXT4-fs (sde1): last error at time 13973814 ...:  1 Time(s)
--8<---cut here---end--->8---

and I did query what it meant and what I should do about it, but I was
reassured that it was OK and nothing needed to be done. In the light of
what's just happened, wrong!

Thanks
Sharon.
Install smartmontools on your system and check the backup drive's SMART 
status. Something like

  smartctl -H /dev/sde

If it passes, then reusing is OK. Otherwise, try to recover what you can 
by making a copy of the disk using dd-

rescue.


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Archive: https://lists.debian.org/54044f0c.2060...@torfree.net

Re: backup drive has died.

[Hans]

Am Montag, 1. September 2014, 11:19:01 schrieb Sharon Kimble:
> Last night my backup drive died and is now totally unresponsive, but it
> did give this error message when I tried to access it -
> 
> --8<---cut here---start->8---
> Error mounting /dev/sde1 at /media/boudiccas/back1: Command-line
> `mount -t "ext4" -o "uhelper=udisks2,nodev,nosuid" "/dev/sde1"
> "/media/boudiccas/back1"' exited with non-zero exit status 32: mount: wrong
> fs type, bad option, bad superblock on /dev/sde1, missing codepage or
> helper program, or other error In some cases useful info is found in syslog
> - try dmesg | tail or so
> --8<---cut here---end--->8---
> 
> Is there any way in which I can access the contents of the drive please?
> 
> If not, is it safe to format it and re-use it please? If so, what should
> I format it as ? ext3? ext4? Or something else?
> 
> It has for some months been showing this message in logwatch -
> 
> --8<---cut here---start->8---
>  WARNING:  Kernel Errors Present
> EXT4-fs (sde1): error count since last ...:  1 Time(s)
> EXT4-fs (sde1): initial error at time 13973814 ...:  1 Time(s)
> EXT4-fs (sde1): last error at time 13973814 ...:  1 Time(s)
> --8<---cut here---end--->8---
> 
> and I did query what it meant and what I should do about it, but I was
> reassured that it was OK and nothing needed to be done. In the light of
> what's just happened, wrong!
> 
> Thanks
> Sharon.


Hi Sharon,

I had very good results with DEFT 7.2, which is for forensic purposes.
It is a livefile system. 

I suggest, to make an image of the partitions. On DEFT, there is a very nice 
tool called hbformost, which is a GUI for foremost and scalpel.

I am using nice GUYMANAGER from the kali-linux distro, then check the imager 
with hbformost. So you can easily restore most pictures, documents, music files 
whatever.

You might also try testdisk, which I am using from the Trinity-Rescue-Kit, 
another livefile system.

To your question for reuse: I would say yes, and if you can reformat it, I 
would use ext4. It is stable and fast.

For the future you might want to install smartd, which checks your harddrive 
at ruinning time, and warns you, if there are defects. It is posible, to lower 
the line, so that smartd is warning much earlier than normal (normal is, when 
I remember correctly, abou 10 percent defections)

You can of course check the complete harddrive before further use. There are 
some testing tools available, which are very hardware near. For those I am 
using some windows based tools (which are on another livefile cd, called 
UBCD4Win).

However, as you see, I am working with special livefile CDs which arer aimed 
for very special purposes, and I know, it costs time, to gather them all. This 
is just my way, other people might have better ways.

Hope this helps.

Good luck!

Hans


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/14390302.PpTsMezez3@protheus7

Re: backup drive has died.

[tv.deb...@googlemail.com]

On 01/09/14 06:19 AM, Sharon Kimble wrote:
> Last night my backup drive died and is now totally unresponsive, but it
> did give this error message when I tried to access it -
>
> --8<---cut here---start->8---
> Error mounting /dev/sde1 at /media/boudiccas/back1: Command-line
> `mount -t "ext4" -o "uhelper=udisks2,nodev,nosuid" "/dev/sde1"
> "/media/boudiccas/back1"' exited with non-zero exit status 32: mount:
> wrong fs
> type, bad option, bad superblock on /dev/sde1, missing codepage or helper
> program, or other error In some cases useful info is found in syslog -
> try
> dmesg | tail or so
> --8<---cut here---end--->8---
>
> Is there any way in which I can access the contents of the drive please?
>
> If not, is it safe to format it and re-use it please? If so, what should
> I format it as ? ext3? ext4? Or something else?
>
> It has for some months been showing this message in logwatch -
> --8<---cut here---start->8---
>   WARNING:  Kernel Errors Present
>  EXT4-fs (sde1): error count since last ...:  1 Time(s)
>  EXT4-fs (sde1): initial error at time 13973814 ...:  1 Time(s)
>  EXT4-fs (sde1): last error at time 13973814 ...:  1 Time(s)
> --8<---cut here---end--->8---
>
> and I did query what it meant and what I should do about it, but I was
> reassured that it was OK and nothing needed to be done. In the light of
> what's just happened, wrong!
>
> Thanks
> Sharon.

Hello, did you ever try to run fsck on this filesystem ? What is the
result ?


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/54045afa.9000...@googlemail.com

hwmonitor device name not persistant

[Bob Brewer]

I am running up-to-date debian sid amd64 and have found that fancontol 
sometimes fails to start after a reboot reporting that:

Device path of hwmon0 has changed
Device name of hwmon0 has changed
Configuration appears to be outdated, please run pwmconfig again

On inspecting /etc/fancontrol it would appear that the device name has 
changed between reboots.

(Yesterdays /etc/fancontrol after running pwnconfig)

# Configuration file generated by pwmconfig, changes will be lost
INTERVAL=10
DEVPATH=hwmon0=devices/platform/nct6775.656
DEVNAME=hwmon0=nct6776
FCTEMPS=hwmon0/pwm2=hwmon0/temp2_input
FCFANS= hwmon0/pwm2=hwmon0/fan2_input
MINTEMP=hwmon0/pwm2=20
MAXTEMP=hwmon0/pwm2=60
MINSTART=hwmon0/pwm2=150
MINSTOP=hwmon0/pwm2=0


(Todays /etc/fancontrol after running pwnconfig)

# Configuration file generated by pwmconfig, changes will be lost
INTERVAL=10
DEVPATH=hwmon2=devices/platform/nct6775.656
DEVNAME=hwmon2=nct6776
FCTEMPS=hwmon2/pwm2=hwmon2/temp2_input
FCFANS= hwmon2/pwm2=hwmon2/fan2_input
MINTEMP=hwmon2/pwm2=20
MAXTEMP=hwmon2/pwm2=60
MINSTART=hwmon2/pwm2=150


So the device seems to have changed from hwmon0 to hwmon2 between successive 
boots.


An idea what I should do to make the device name persistent?

Thanks
Rob





-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/lu1lif$th0$1...@tuscan2.grantura.co.uk

Re: backup drive has died.

[Sharon Kimble]

Gary said:
Install smartmontools on your system and check the backup drive's
SMART status. Something like
  smartctl -H /dev/sde

If it passes, then reusing is OK. Otherwise, try to recover what you
can by making a copy of the disk using dd-
rescue.

Hans said:
To your question for reuse: I would say yes, and if you can reformat it, I 
would use ext4. It is stable and fast.

For the future you might want to install smartd, which checks your harddrive 
at ruinning time, and warns you, if there are defects. It is posible, to lower 
the line, so that smartd is warning much earlier than normal (normal is, when 
I remember correctly, abou 10 percent defections)

Thanks both, unfortunately its a 3tb USB external hard drive, with
smartd installed and working, except, it doesn't work on external USB
drives! And due to its size, it was 93% full with backups, doing a
direct copy is not feasible. The backups are all done with obnam, so it
just seems that I'll have to reformat, and start a new series of backups
on it. And if I get that "Kernel errors report" again, start making
alternative plans for replacing the drive, before it dies on me. It was
formatted ext4 before, which was why I queried if there was anything
better, but I'll have a play with it after dinner.

Thanks
Sharon.
-- 
A taste of linux = http://www.sharons.org.uk
my git repo = https://bitbucket.org/boudiccas/dots
TGmeds = http://www.tgmeds.org.uk
Debian testing, fluxbox 1.3.5, emacs 24.3.93.1


signature.asc
Description: PGP signature

Re: Sigil

[Johann Spies]

I had some success converting latex files to epub using the following:

#!/bin/sh

latexml --dest=$1.xml $1.tex
latexmlpost -dest=$1.html $1.xml
ebook-convert $1.html $1.epub --language en --no-default-epub-cover %

I do nearly all my authoring in LaTeX.


Also: abiword can save  in e-pub-format.

Regards
Johann
-- 
Because experiencing your loyal love is better than life itself,
my lips will praise you.  (Psalm 63:3)

Re: backup drive has died.

[Bzzzz]

On Mon, 01 Sep 2014 11:19:01 +0100
Sharon Kimble  wrote:

> Error mounting /dev/sde1 at /media/boudiccas/back1: Command-line
> `mount -t "ext4" -o "uhelper=udisks2,nodev,nosuid" "/dev/sde1"
> "/media/boudiccas/back1"' exited with non-zero exit status 32: mount:
> wrong fs type, bad option, bad superblock on /dev/sde1, missing
> codepage or helper program, or other error In some cases useful info
> is found in syslog - try dmesg | tail or so

What does say dmesg?


signature.asc
Description: PGP signature

Re: backup drive has died.

[Doug]

On 09/01/2014 07:39 AM, tv.deb...@googlemail.com wrote:

On 01/09/14 06:19 AM, Sharon Kimble wrote:

Last night my backup drive died and is now totally unresponsive, but it
did give this error message when I tried to access it -

--8<---cut here---start->8---
Error mounting /dev/sde1 at /media/boudiccas/back1: Command-line
`mount -t "ext4" -o "uhelper=udisks2,nodev,nosuid" "/dev/sde1"
"/media/boudiccas/back1"' exited with non-zero exit status 32: mount:
wrong fs
type, bad option, bad superblock on /dev/sde1, missing codepage or helper
program, or other error In some cases useful info is found in syslog -
try
dmesg | tail or so
--8<---cut here---end--->8---

Is there any way in which I can access the contents of the drive please?

If not, is it safe to format it and re-use it please? If so, what should
I format it as ? ext3? ext4? Or something else?


/snip/

Drives are cheap nowadays. Assuming you can get the data off the drive,
I can't see any good reason to trust it with your data again, even if you
can reformat it and partition it. were it me, I wouldn't!
--doug


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Archive: https://lists.debian.org/54048a55@optonline.net

Re: backup drive has died.

[Bzzzz]

On Mon, 01 Sep 2014 11:01:41 -0400
Doug  wrote:

> Drives are cheap nowadays. Assuming you can get the data off the
> drive, I can't see any good reason to trust it with your data
> again, even if you can reformat it and partition it. were it me, I
> wouldn't! --doug

Glitches happen, Doug (especially nowadays…), and much often if
machine+HD aren't plugged in a UPS.
So before jettison the HD, a complete re-format with the -cc
option would be interesting.

One more thing: HDz, except in large secured, redundant and
error corrected bays are NOT a reliable meaning of backup;
tape is (LTO).

-- 
 but I'm a princess awaiting her prince charmant! 
 Wait! What do you think? A guy who has to taken alone a fortress,
 snuffed out a frigging dragon, exhausted a mean witch and
 everything just to lay, a nice guy with a minty breath 
 and shining tooth with a TINT! and hair flying in the wind? 
 …
 Believe me girl, Shrek is the uppermost version of a realistic
 fairy tale… It is close to a documentary… 
 …
 Now if you want, I can also quickly explain you Santa Claus,
 God and the Tooth Fairy.


signature.asc
Description: PGP signature

embrace, extend, extinguish

[Dan Ritter]

Reading:
http://0pointer.net/blog/revisiting-how-we-put-together-linux-systems.html

systemd's upstream is explicitly interested in taking over all
Linux distros, not in the minor sense of being supported on
every system but in the major sense of making package management 
conform to their own views.

-dsr- 


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/20140901172609.ga20...@randomstring.org

Re: embrace, extend, extinguish

[Brian]

On Mon 01 Sep 2014 at 13:26:09 -0400, Dan Ritter wrote:

> 
> Reading:
> http://0pointer.net/blog/revisiting-how-we-put-together-linux-systems.html
> 
> systemd's upstream is explicitly interested in taking over all
> Linux distros, not in the minor sense of being supported on
> every system but in the major sense of making package management 
> conform to their own views.

Hello Dan,

We have had four threads on broadly the same topic within less than
three months. The most recent one starts at

  https://lists.debian.org/20140825184423.56668...@mydesq2.domain.cxm

We are sure you have a valuable contribution to make to that debate, so
feel free to add to it.

If you have a technical question relating to init systems and how Debian
deals with them a new post is probably advisable so it doesn't get lost
amongst the discussion there.


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/20140901182950.gg4...@copernicus.demon.co.uk

Re: embrace, extend, extinguish

[Erwan David]

Le 01/09/2014 20:29, Brian a écrit :
> On Mon 01 Sep 2014 at 13:26:09 -0400, Dan Ritter wrote:
>
>> Reading:
>> http://0pointer.net/blog/revisiting-how-we-put-together-linux-systems.html
>>
>> systemd's upstream is explicitly interested in taking over all
>> Linux distros, not in the minor sense of being supported on
>> every system but in the major sense of making package management 
>> conform to their own views.
> Hello Dan,
>
> We have had four threads on broadly the same topic within less than
> three months. The most recent one starts at
>
>   https://lists.debian.org/20140825184423.56668...@mydesq2.domain.cxm
>
> We are sure you have a valuable contribution to make to that debate, so
> feel free to add to it.
>
> If you have a technical question relating to init systems and how Debian
> deals with them a new post is probably advisable so it doesn't get lost
> amongst the discussion there.
>
>

Do you think that the views expressed here should not be discussed by
debian users ? After it presents a vision for the future of the
distribution on which we, as users should be able to give our opinon, no ?

It is a topic for users of any distribution. And yes, I say *users* not
only developpers.



-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/5404bc53.5060...@rail.eu.org

Re: Patch wheezy's glibc to run on ancient kernel (2.6.16)?

[Sven Joachim]

On 2014-09-01 10:28 +0200, Clemens Eisserer wrote:

> I would like to run debian wheezy on my nokia-770 (Linux-2.6.16.27) in a
> chroot environment, unfourtunately chroot telling me the kernel is too old.
> The latest version that worked this way is Debian Lenny, which is
> unsupported since mid 2012.

This is not too worrying since the 2.6.16 kernel has been unsupported
since 2008, and 2.6.16.27 is even two years older.

> Because updating the kernel almost impossible (binary wlan driver blob,
> texas instrument`s patched OMAP source tree, ancient toolchain), is there
> any way to patch a debootstrapped debian installation with a self-compiled
> glibc that is compatible with older kernel versions?

Probably.  To patch eglibc for supporting older kernels, change the 
MIN_KERNEL_SUPPORTED variable in debian/sysdeps/linux.mk and update the
check in debian/debhelper.in/libc.preinst for the minimum kernel version
accordingly.

> I reason I ask is because the minimum kernel version also seems to be
> stored in all the executeables too, e.g. I get the following output on a
> MIPS box:
>
> root@OpenWrt:/usr/bin# file xz
> xz: ELF 32-bit MSB executable, MIPS, MIPS-II version 1 (SYSV), dynamically
> linked (uses shared libs), for GNU/Linux 2.6.26, .

The requirement is stored in a section in the binary labeled
NT_GNU_ABI_TAG and is usually the same as the one for the glibc version
the binary was linked with.  AFAIK the dynamic linker does not use this
information for executable programs, but it does for libraries (refuses
to load them if the kernel is too old).

You can set the LD_ASSUME_KERNEL environment variable to tell the
dynamic linker about that you're actually a different kernel.  Here I'm
telling it that my kernel is actually too old:

,
| $ LD_ASSUME_KERNEL=2.6.25 /bin/true
| /bin/true: error while loading shared libraries: libc.so.6: cannot open 
shared object file: No such file or directory
`

> ... however as the binary does not talk directly to the kernel (or, does
> it?), I don't understand how the binary itself has a requirement on the
> linux kernel.

The binary might make syscalls which are not implemented in older
kernels, but most programs use glibc wrappers instead.  So I think a
wheezy chroot should mostly work once you have rebuilt eglibc.

Cheers,
   Sven


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/87ppff18ds@turtle.gmx.de

[OT] - Re: embrace, extend, extinguish

[Bzzzz]

On Mon, 1 Sep 2014 13:26:09 -0400
Dan Ritter  wrote:

> http://0pointer.net/blog/revisiting-how-we-put-together-linux-systems.html

An article written by… one of the systemd devs………

We happen to learn that it'll also be _dependent_ on BTRFS
and (may be?, when?) support EXT4 & XFS (bad luck, for its
features, capacities and stability, I prefer to use ZFS;
unsupported?).

It also carefully avoids the embedded machines questions.

It finishes with: "The future is going to be awesome!", for
its devs, this is an absolutely sure thing ; from what it
describes, I'm seeing it more like: "The future is going to
be cumbersome" (but this is only my own impression & opinion).

This confirms the prima vista tentacular image.

We'll see if that can resist to power users, especially those
that administers thousands of servers…

As P. Venezia stated (more or less:): a good idea but a bad realization.


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/20140901205325.7ab40aee@msi.defcon1

Re: embrace, extend, extinguish

[Brian]

On Mon 01 Sep 2014 at 20:34:59 +0200, Erwan David wrote:

> Le 01/09/2014 20:29, Brian a écrit :
> > On Mon 01 Sep 2014 at 13:26:09 -0400, Dan Ritter wrote:
> >
> >> Reading:
> >> http://0pointer.net/blog/revisiting-how-we-put-together-linux-systems.html
> >>
> >> systemd's upstream is explicitly interested in taking over all
> >> Linux distros, not in the minor sense of being supported on
> >> every system but in the major sense of making package management 
> >> conform to their own views.
> > Hello Dan,
> >
> > We have had four threads on broadly the same topic within less than
> > three months. The most recent one starts at
> >
> >   https://lists.debian.org/20140825184423.56668...@mydesq2.domain.cxm
> >
> > We are sure you have a valuable contribution to make to that debate, so
> > feel free to add to it.
> >
> > If you have a technical question relating to init systems and how Debian
> > deals with them a new post is probably advisable so it doesn't get lost
> > amongst the discussion there.
> 
> Do you think that the views expressed here should not be discussed by
> debian users ? After it presents a vision for the future of the
> distribution on which we, as users should be able to give our opinon, no ?

Of course not. You will note I gave the OP a very suitable thread to
post to. After catching up with the archive contents he may choose to
use one of the other quite recent discussion threads.


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: 
https://lists.debian.org/01092014194930.0e54a42ff...@desktop.copernicus.demon.co.uk

Re: portability of jigdo

[songbird]

Rusi Mody wrote:
> On Monday, September 1, 2014 8:40:01 AM UTC+5:30, songbird wrote:
>> Rusi Mody wrote:
>> > Context:
>> > On the tex user group, someone was asking/complaining about 
>> > the difficulties of downloading texlive.
>> > - One has to download one (few?) large (in GBs) dvd image
>> > - The user was on a slow/flaky line
>
>>   texlive should not be that large?
>
> https://www.tug.org/texlive/acquire-iso.html 
> says 2+ Gb

  ah, looking at my local system i see i only
have the base system installed and not much
extra, so it runs about 80M to download.


>> > In response the texlive folks admitted that there was a problem without 
>> > a clear solution.
>> ...
>
>>   on a slow connection debdelta is useful at times.
>>   seems they don't know it exists.
>
> debdelta can work for windows?

  not too likely.  thought this was a debian linux
related question...


  songbird


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/chredb-v02@id-306963.user.uni-berlin.de

Re: embrace, extend, extinguish

[Steve Litt]

On Mon, 1 Sep 2014 13:26:09 -0400
Dan Ritter  wrote:

> 
> Reading:
> http://0pointer.net/blog/revisiting-how-we-put-together-linux-systems.html
> 
> systemd's upstream is explicitly interested in taking over all
> Linux distros, not in the minor sense of being supported on
> every system but in the major sense of making package management 
> conform to their own views.
> 
> -dsr- 

Let's rename the Linux Kernel to the Systemd Kernel. Then let's make
the package manager part of the kernel.

SteveT

Steve Litt*  http://www.troubleshooters.com/
Troubleshooting Training  *  Human Performance


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/2014090115.73942...@mydesq2.domain.cxm

Re: backup drive has died.

[David Christensen]

On 09/01/2014 03:19 AM, Sharon Kimble wrote:

Last night my backup drive died and is now totally unresponsive,


Some hard drive manufacturers offer a bootable CD ISO image containing 
diagnostic tools that you can use to check their products.  For example, 
SeaTools for DOS:


http://www.seagate.com/support/downloads/seatools/


What is your drive make and model?


What did the diagnostic(s) report?


HTH,

David


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Archive: https://lists.debian.org/5404e4ba.6030...@holgerdanske.com

Re: backup drive has died.

[Gary Dale]

On 01/09/14 09:01 AM, Sharon Kimble wrote:

Gary said:
Install smartmontools on your system and check the backup drive's
SMART status. Something like
   smartctl -H /dev/sde

If it passes, then reusing is OK. Otherwise, try to recover what you
can by making a copy of the disk using dd-
rescue.

Hans said:
To your question for reuse: I would say yes, and if you can reformat it, I
would use ext4. It is stable and fast.

For the future you might want to install smartd, which checks your harddrive
at ruinning time, and warns you, if there are defects. It is posible, to lower
the line, so that smartd is warning much earlier than normal (normal is, when
I remember correctly, abou 10 percent defections)

Thanks both, unfortunately its a 3tb USB external hard drive, with
smartd installed and working, except, it doesn't work on external USB
drives! And due to its size, it was 93% full with backups, doing a
direct copy is not feasible. The backups are all done with obnam, so it
just seems that I'll have to reformat, and start a new series of backups
on it. And if I get that "Kernel errors report" again, start making
alternative plans for replacing the drive, before it dies on me. It was
formatted ext4 before, which was why I queried if there was anything
better, but I'll have a play with it after dinner.

Thanks
Sharon.
Of course smartctl works with external drives. That's the point of "Self 
Monitoring And ReporTing". You may not be able to report on a drive that 
isn't connected, but smartctl can certainly give you a report anytime 
that it is connected.


smartd.conf also allows you to configure it to monitor removable drives.


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Archive: https://lists.debian.org/5404ea4c.1030...@torfree.net

Re: embrace, extend, extinguish

[green]

Dan Ritter wrote at 2014-09-01 12:26 -0500:
> Reading:
> http://0pointer.net/blog/revisiting-how-we-put-together-linux-systems.html

Dan, thank you for posting this link.  It is especially interesting
considering it is by Lennart Poettering.  Also, I consider it relevant
to Debian users and firmly on-topic here.


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/20140901220410.GB4968@swansys

Re: embrace, extend, extinguish

[Haines Brown]

I hestitate to enter this thread, given its nasty tone, so please don't
attack me if I pose some non-partisan questions.

An issue came up whether debian Jessie can run without systemd. I'm
currently running Wheezy without systemd installed, but I get this:

  $ dpkg -l "*systemd*" | grep ii
  ii libsystemd-daemon0:i386 44-11+deb7u4 i386 systemd utility library
  ii libsystemd-login0:i386  44-11+deb7u4 i386 systemd login utility library

As best I can make out, these libraries are there so that daemons and
log-in services that are designed to run under systemd will continue to
function. Can I remove these two packages and still have my old line
applications, like emacs, continue to function?

Next spring I'll likely upgrade to Jessie, and it seems that Jessie will
use systemd. A question came up but was never answered whether Jessie can
install sysvinit and make use of it instead of systemd. What is the
answer?

My wife used to run Windows and has now fled to a Mac. When things went
wrong on her Windows machine, I found them very difficult to fix. More
often than not I could only reboot and cross my fingers. On the other
hand, if I spend some time, I usually can fix Debian problems. Will the
expansion of systemd to acquire a monothic control over my system mean
that a non-expert like myself will be less able to maintain their
system?

Haines Brown


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: 
https://lists.debian.org/20140901221729.gb8...@historicalmaterialism.info

Re: embrace, extend, extinguish

[Brian]

On Mon 01 Sep 2014 at 18:17:29 -0400, Haines Brown wrote:

> I hestitate to enter this thread, given its nasty tone, so please don't
> attack me if I pose some non-partisan questions.

This is not an attack. :)

Non-partisan questions are much better off being asked in a standalone
post separate from what are obviously invitations to partisan activity.
That way people might view their responses as serving a useful purpose
for the poster and the list and not being submerged in an advocacy
quagmire.

[Questions which might be resubmitted snipped]


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/20140901225842.gh4...@copernicus.demon.co.uk

Re: embrace, extend, extinguish

[Nate Bargmann]

* On 2014 01 Sep 12:43 -0500, Dan Ritter wrote:
> 
> Reading:
> http://0pointer.net/blog/revisiting-how-we-put-together-linux-systems.html

I stopped at his second bullet point.  He wants to lump RPM and .deb
systems together and it's clear he hasn't got a clue.  I'm also an
upstream project manager and have not seen what he's blabbering about.
Plus the project I'm involved with builds and runs on BSD, OS/X, and 
Windows in addition to the miriad Linux distributions.  Nice straw
man...

- Nate

-- 

"The optimist proclaims that we live in the best of all
possible worlds.  The pessimist fears this is true."

Ham radio, Linux, bikes, and more: http://www.n0nb.us


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/20140901233613.gd3...@n0nb.us

Re: backup drive has died.

[David]

On 1 September 2014 23:01, Sharon Kimble  wrote:
>
> Thanks both, unfortunately its a 3tb USB external hard drive, with
> smartd installed and working, except, it doesn't work on external USB
> drives!

On 2 September 2014 07:51, Gary Dale  wrote:
>
> Of course smartctl works with external drives. That's the point of "Self
> Monitoring And ReporTing". You may not be able to report on a drive that
> isn't connected, but smartctl can certainly give you a report anytime that
> it is connected.

smartmontools.org writes quite a lot on this topic:

http://www.smartmontools.org/wiki/FAQ#SmartmontoolsforFireWireUSBandSATAdiskssystems
http://www.smartmontools.org/wiki/USB
http://www.smartmontools.org/wiki/Supported_USB-Devices


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: 
https://lists.debian.org/CAMPXz=pwDN4rnAaqxNzX5HU=nv2w6bqgcpmmdf7mqstrbaz...@mail.gmail.com

Re: backup drive has died.

[Bzzzz]

On Tue, 2 Sep 2014 10:33:05 +1000
David  wrote:

> smartmontools.org writes quite a lot on this topic:
> 
> http://www.smartmontools.org/wiki/FAQ#SmartmontoolsforFireWireUSBandSATAdiskssystems
> http://www.smartmontools.org/wiki/USB
> http://www.smartmontools.org/wiki/Supported_USB-Devices

Unfortunately, some USB HD boxes chipsets don't support a HD SMART test
(or may be, the kernel chipset support isn't complete yet).

I just tested on mine, which gave me a shiver when I saw it wasn't
working:(


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/20140902024406.79a3b8b5@msi.defcon1

Re: backup drive has died.

[Bzzzz]

On Tue, 2 Sep 2014 10:33:05 +1000
David  wrote:

OOPS, my bad (and many thanks for your links); it is working
with the right switch.


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/20140902024613.7206267c@msi.defcon1

Re: embrace, extend, extinguish

[Zenaan Harkness]

> http://0pointer.net/blog/revisiting-how-we-put-together-linux-systems.html

That looks -awesome-! Great potential. Object-oriented distributions/
installations. It'll be quite a journey from here to there though :)

So many comedic opportunities shall undoubtedly present themselves.


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: 
https://lists.debian.org/caosgnssf4veo-yeognue14utyjcr--tn6sa7k4nnxjqadzd...@mail.gmail.com

Re: embrace, extend, extinguish

[John Hasler]

Steve Litt writes:
> Let's rename the Linux Kernel to the Systemd Kernel. Then let's make
> the package manager part of the kernel.

No, no.  Make the kernel part of Systemd.  And X as well.
-- 
John Hasler 
jhas...@newsguy.com
Elmwood, WI USA


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/878um2hh2b@thumper.dhh.gt.org

Re: embrace, extend, extinguish

[Bzzzz]

On Mon, 01 Sep 2014 21:50:04 -0500
John Hasler  wrote:

> No, no.  Make the kernel part of Systemd.  And X as well.

In this case, why not making only one package of the whole distro:
systemd-all-in-one.deb-rpm-gz ;)


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/20140902045805.6f26942c@msi.defcon1

Re: embrace, extend, extinguish

[Jimmy Johnson]

Dan Ritter wrote:

Reading:
http://0pointer.net/blog/revisiting-how-we-put-together-linux-systems.html

systemd's upstream is explicitly interested in taking over all
Linux distros, not in the minor sense of being supported on
every system but in the major sense of making package management 
conform to their own views.


-dsr- 



These people(Kay Sievers, Harald Hoyer, Daniel Mack,Tom Gundersen, David 
Herrmann, and Lennart Poettering) infuriate me..They should just port 
their own system, that is the Linux way and if it's good people will use 
it..I run twelve Debian Systems, eight Wheezy, two Jessie, one Sid and 
one Squeeze, I'm always running other distros too and I'm watching..

--
Jimmy Johnson

Debian-Live - Wheezy - KDE 4.8.4 - AMD64 - EXT4 at sda1
Registered Linux User #380263


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Archive: https://lists.debian.org/54053d38.3010...@gmail.com

Re: How to mount a LUKS partiotion with Nautilus with option discard?

[Joerg Desch]

Am Mon, 01 Sep 2014 12:19:39 +0200 schrieb Hans:

> As far as I remember (and someone may correct me, if I am wrong) if
> there is an entry in fstab, the devices are not mounted by the GUI (I
> guess you mean dolphin).

I'm using Nautilus in GNOME. Dolphin is KDE. IMO fstab entries without 
the option 'auto' are visible in the sidebar could be mounted by the GUI.

This seems to be not the case for LUKS entries.


> Fstab is used for mounting at boot. Try to comment out the entry in
> fstab, but leave the entry in crypttab.

Tested. It doesn't work. There is no visible change. Even a single entry 
in crypttab isn't shown in the Desktop GUI.


Any further tips?


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: https://lists.debian.org/lu3j9f$9m1$1...@ger.gmane.org

Re: [SOLVED] - Re: upgrade stuck (and machine too)

[Johann Spies]

On 23 August 2014 18:44, B  wrote:

> On Thu, 21 Aug 2014 22:01:13 +0200
> B  wrote:
>
> Welcome to the wonderful world of systemd :(
>
> I had one line in /etc/fstab about mounting an USB key
> that was set in automatic mount.
>
> This line never caused any problem w/ sysV, but systemd
> considered not being able to mount it automatically a
> failure and stucked the whole system :(
>

I had the same problem on more than one computer and it took me some time
to find the culprit.

Regards
Johann