Re: test

[Christophe Musseau]

Il me semble. ça doit être le syndrome "achats de Noël" qui débute

Le 28 novembre 2016 à 08:40, steve  a écrit :

> pas de message depuis 3 jours. La liste est-elle encore en vie ?
>
>


-- 
Christophe MUSSEAU

(sous Debian GNU/Linux - https://www.debian.org)

test

[steve]

pas de message depuis 3 jours. La liste est-elle encore en vie ?

Re: Debian *not very good

[Latincom]

On Sat, 26 Nov 2016 18:38:58 +0100, Rob van der Putten wrote:

> Hi there
> 
> 
> On 25/11/16 22:26, Latincom wrote:
> 
>> Is there a step by step guide or How to on line?
>> I have 1 Wheezy without Systemd, and i would like to upgrade it.
>> Thanks.
> 
> You can do both an upgrade and an install from scratch without systemd;
> http://without-systemd.org/
> http://without-systemd.org/wiki/index.php/
How_to_remove_systemd_from_the_Netinst_CD
> http://without-systemd.org/wiki/index.php/
How_to_remove_systemd_from_a_Debian_jessie/sid_installation
> http://people.skolelinux.org/pere/blog/
How_to_stay_with_sysvinit_in_Debian_Jessie.html
> I had to make some other changes to get things to work though. Reading
> the release notes and keeping a close eye on the install process helps.
> 
> I run XFCE on my desktop. I had to add myself to sudo to make things
> work properly.
> 
> 

Thanks so much for all answers.


> Regards,
> Rob

Re: Stretch/Sid not resuming from suspend

[Lucio Crusca]

On April 20, 2016, I wrote:


Hello *,

I'm using Stretch/Sid amd64 and I update it every few days. Since 
about a month or so it fails to resume from suspend to RAM. It doesn't 
happen regularly, it fails about 50% of the times. When it doesn't 
resume, I get only a black screen in text mode with the blinking caret 
in the top left corner of the screen. No keyboard input is working.
I'm using Xfce and I suspend always by hand using the menu. My kernel 
is 4.4.0-1-amd64 with nvidia legacy driver (340). I'm not sure but I 
seem to recall the problem started when my driver moved from nvidia to 
nvidia-legacy. Or, maybe, it started with some kernel update, but I 
don't know which one.


I'd try with the nouveau driver but it proved to be even less reliable 
than the nvidia, causing freezes also during normal use of the system. 
The hardware is a Intel Core I5 notebook (Dell Precision M4500).


What can I do to pin down the issue?



After a few months, updates, and new kernels, the problem is still 
there, but I've managed to identify the most common conditions when it 
happens and I can reproduce it. When I suspend the notebook while 
plugged into the mains and then I resume it unplugged (e.g. I unplug it 
while suspended) it's *almost* a sure bet it happens.


Does that shed any light?

Re: Weirdness with two packages.........

[Charlie]

On Mon, 28 Nov 2016 00:19:38 -0500 The Wanderer sent:



>  I have - different from your own:
> > 
> > $ apt-cache policy libavutil55
> > libavutil55:
> >   Installed: 10:3.1.2-dmo2
> >   Candidate: 10:3.1.2-dmo2
> >   Version table:
> >  *** 10:3.1.2-dmo2 100
> > 100 /var/lib/dpkg/status
> >  7:3.2-2 500
> > 500 http://ftp.us.debian.org/debian stretch/main amd64
> > Packages  
> 
> This is almost certainly your problem. The version of libavutil which
> you have is from a different source, with a higher epoch version but a
> lower upstream version, and is no longer available from your selected
> repositories. In particular, it is not the same version as your other
> libav* libraries (as the snipped version information for libavformat
> indicates), and that mismatch is probably the source of the problem.
> 
> Do you have any idea how this mismatched package version may have come
> about?
> 
> I recommend that you explicitly install the "lower" version listed
> here (7:3.2-2), and see if your problem goes away.
> 
> I don't know what tools you normally use for package installation and
> upgrade, but I would do that with the following command (in a root
> terminal):
> 
> apt-get install libavutil55=7:3.2-2


After contemplation, my reply is:

I'm on a different machine now, at a different place when sending this.

I only ever use apt-get for updates, upgrades, installs and purges.

But before I left that machine I attempted your excellent suggestion
and received this error message:

The following packages have unmet dependencies:
 libswresample2 : Depends: libavutil55 (>= 10:3.1.2) but 7:3.2-2 is to
be installed
E: Error, pkgProblemResolver::Resolve generated breaks, this may be
caused by held packages.

I assume that I might have missed a bug report on a update/upgrade? So
I will just have to purge libswresample2, and libavutil55 (>= 10:3.1.2)

Thank you for your help.
Charlie

-- 
Registered Linux User:- 329524
***

If you argue for your limitations, you get to keep
them. .Richard Bach

***

Debian GNU/Linux - Magic indeed.

-

Re: Weirdness with two packages.........

[The Wanderer]

On 2016-11-27 at 22:14, Charlie wrote:

> The sent: Re: Weirdness with two packages.
> 
> Wrote this:
>   >On 2016-11-27 at 18:57, Charles Schroeder wrote:
>>
>>> Hello Everyone,
>>> 
>>> In  3.16.0-4-amd64 #1 SMP Debian 3.16.7-ckt11-1+deb8u3
>>> (2015-08-04) x86_64 GNU/Linux
>>> 
>>> I get this message when I attempt to use Chromium:
>>> 
>>> $ chromium
>>> /usr/lib/chromium/chromium: relocation
>>> error: /usr/lib/x86_64-linux-gnu/libavformat.so.57: symbol
>>> avpriv_dict_set_timestamp, version LIBAVUTIL_55 not defined in file
>>> libavutil.so.55 with link time reference
>>> 
>>> Much the same when I attempt to use goldendict:
>>> 
>>> $ goldendict
>>> goldendict: relocation
>>> error: /usr/lib/x86_64-linux-gnu/libavformat.so.57: symbol
>>> avpriv_dict_set_timestamp, version LIBAVUTIL_55 not defined in file
>>> libavutil.so.55 with link time reference
>>> 
>>> I have no idea what is happening here or how I work round it?
>>> 
>>> I can't find those files in my: /usr/lib/x86_64-linux-gnu/
>>> directory?  
>>
>>They should be there, not as files, but as symlinks.
>>
>>> Trying to install both those packages tells me they are "already
>>> the newest version" so nothing to install?
>>> 
>>> Update, upgrade and reboot doesn't fix it.  
>>
>>What versions are those packages at, according to e.g. 'apt-cache
>>policy [packagename]'?
>>
>>What about the libavutil55 and libavformat57 packages?
>>
>>I have:

>>$ apt-cache policy libavutil55
>>libavutil55:
>>  Installed: 7:3.2-2
>>  Candidate: 7:3.2-2
>>  Version table:
>> *** 7:3.2-2 500
>>500 http://ftp.us.debian.org/debian testing/main amd64 Packages
>>100 /var/lib/dpkg/status
> 
> I have - different from your own:
> 
> $ apt-cache policy libavutil55
> libavutil55:
>   Installed: 10:3.1.2-dmo2
>   Candidate: 10:3.1.2-dmo2
>   Version table:
>  *** 10:3.1.2-dmo2 100
> 100 /var/lib/dpkg/status
>  7:3.2-2 500
> 500 http://ftp.us.debian.org/debian stretch/main amd64 Packages

This is almost certainly your problem. The version of libavutil which
you have is from a different source, with a higher epoch version but a
lower upstream version, and is no longer available from your selected
repositories. In particular, it is not the same version as your other
libav* libraries (as the snipped version information for libavformat
indicates), and that mismatch is probably the source of the problem.

Do you have any idea how this mismatched package version may have come
about?

I recommend that you explicitly install the "lower" version listed here
(7:3.2-2), and see if your problem goes away.

I don't know what tools you normally use for package installation and
upgrade, but I would do that with the following command (in a root
terminal):

apt-get install libavutil55=7:3.2-2

-- 
   The Wanderer

The reasonable man adapts himself to the world; the unreasonable one
persists in trying to adapt the world to himself. Therefore all
progress depends on the unreasonable man. -- George Bernard Shaw



signature.asc
Description: OpenPGP digital signature

Re: Dual Boot

[Jan Bakuwel]

Hi Pascal,

On 28/11/16 12:48, Pascal Hambourg wrote:
> Le 27/11/2016 à 21:02, Jan Bakuwel a écrit :
>>
>> On 26/11/16 11:01, Pascal Hambourg wrote:
>>>
>>> When embedding is not possible, the core image is stored as a regular
>>> file in /boot/grub. Then /boot/grub must be on the same drive as the
>>> boot image. However blocklists are not reliable with files, because
>>> the filesystem may move blocks containing a file around.
>>
>> I find it extremely useful (to the point where it once literally saved
>> me from disaster which is a story beyond the context of this thread) to
>> have multiple OS-es installed on any machine. I've been doing that for
>> years and even with grub it has - so far - always worked fine. But I
>> hear you that that's not a guarantee, and thus I started using extlinux.
>
> I think that I experienced that problem once (but not sure), and
> needed to reinstall GRUB. To be honest, the chances that the blocs
> containing the core image may be moved around are very thin if you
> have a separate ext2 /boot partition, which is rarely modified.
>
>> Which brings me to the following question: what is the recommended way
>> to boot multiple OSes with grub, for example with a partition layout as
>> below. Or is there simply no sane way to do this with grub?
>
> I would not pretend to know "the recommended way".
> However I would recommend a few layouts for some configurations.
>
> If EFI boot is an option, then install all systems for EFI boot (for
> Windows this implies a GPT partition table on the boot disk). This way
> all bootloaders can be installed in separate directories in the EFI
> system partition without overwriting one another. You can define one
> of them as the primary boot loader with higher boot priority in the
> EFI boot list.

One day I'll have to bite that bullet but that day has not come yet :-)

>
> For BIOS boot, without Windows I would recommend to install a primary
> GRUB in the MBR with embedding. Install secondary GRUBs in partitions
> boot records, without embedding. Then you have several options :
>
> - add all entries for Windows and other Linux system kernels in the
> main GRUB menu, which can be done automatically with os-prober called
> by update-grub. The drawback is that the main boot menu must be
> updated everytime a secondary boot loader is updated.
>
> - add entries to chainload the secondary boot loaders in the main GRUB
> menu, so that the main boot menu does not have to be updated when a
> secondary boot loader is updated. The "multiboot" command can be used
> to load a secondary GRUB core image file, so the embedding requirement
> does not apply here.
>
> - include or load secondary GRUB config files in the main boot menu.
> This does not work well with different versions of GRUB, so I do not
> recommend it.
>
> One must also decide whether to use the GRUB belonging to one of the
> Linux systems or an independent GRUB as the main boot loader. Of
> course there is less risk to damage an independent boot loader.
>
> So my advice for a safe setup would be to :
> - install the main GRUB in the MBR and its own partition for /boot/grub
> - install each secondary GRUB in its OS partition
> - chainload secondary GRUBs in the main GRUB's boot menu

This is how I've done it (before using extlinux). I still have a few
servers on grub using this method.

>
> Unfortunately with Windows the MBR is not a safe place because Windows
> overwrites the MBR at installation and sometimes when updating, so be
> prepared to restore a backup or reinstall GRUB. 

Yep. Haven't seen Windows mucking up the MBR with an update though - yet.

> The boot record of the main GRUB's partition is a safer place, but it
> prevents embedding, so the risk of moving blocks comes back. However
> this partition should not be mounted and modified often, so the risk
> is low.

Only keeping Windows on my workstation for video editing with Windows
running on the bare metal.
Never on the servers though.

>
>> /dev/sda1: boot/rescue system
>>
>> /dev/sda2: Windows 7
>> /dev/sda3: Windows 10
>>
>> /dev/sda5: /boot - Linux system 1
>> /dev/sda6: / - Linux system 1
>> /dev/sda7: /var - Linux system 1
>> /dev/sda8: /var/log - Linux system 1
>>
>> /dev/sda9: /boot - Linux system 2
>> /dev/sda10: / - Linux system 2
>> /dev/sda11: /var - Linux system 2
>> /dev/sda12: /var/log - Linux system 2
>>
>> /dev/sda13: swap
>> /dev/sda14: LVM
>
> From the missing /dev/sda4, I guess it is an extended partition and
> /dev/sda{5..14} are logical partitions in a DOS/MBR scheme. 10 logical
> partition is quite a lot. The linked list extended partition structure
> is quite fragile and rigid. A GPT partition table would be more
> robust, but it requires to reinstall reinstall Windows in EFI mode,
> and of course that the motherboard firmware is a UEFI. Otherwise,
> using LVM for most filesystems and swap is an option to reduce the
> number of partitions.

For new servers I surely will be using more LVM and less 

Re: Weirdness with two packages.........

[Charlie]

From my keyboard

The sent: Re: Weirdness with two packages.

Wrote this:
  >On 2016-11-27 at 18:57, Charles Schroeder wrote:
>
>> Hello Everyone,
>> 
>>  In  3.16.0-4-amd64 #1 SMP Debian 3.16.7-ckt11-1+deb8u3
>>  (2015-08-04) x86_64 GNU/Linux
>> 
>> I get this message when I attempt to use Chromium:
>> 
>> $ chromium
>> /usr/lib/chromium/chromium: relocation
>> error: /usr/lib/x86_64-linux-gnu/libavformat.so.57: symbol
>> avpriv_dict_set_timestamp, version LIBAVUTIL_55 not defined in file
>> libavutil.so.55 with link time reference
>> 
>> Much the same when I attempt to use goldendict:
>> 
>> $ goldendict
>> goldendict: relocation
>> error: /usr/lib/x86_64-linux-gnu/libavformat.so.57: symbol
>> avpriv_dict_set_timestamp, version LIBAVUTIL_55 not defined in file
>> libavutil.so.55 with link time reference
>> 
>> I have no idea what is happening here or how I work round it?
>> 
>> I can't find those files in my: /usr/lib/x86_64-linux-gnu/
>> directory?  
>
>They should be there, not as files, but as symlinks.
>
>> Trying to install both those packages tells me they are "already
>> the newest version" so nothing to install?
>> 
>> Update, upgrade and reboot doesn't fix it.  
>
>What versions are those packages at, according to e.g. 'apt-cache
>policy [packagename]'?
>
>What about the libavutil55 and libavformat57 packages?
>
>I have:
>
>$ apt-cache policy chromium
>chromium:
>  Installed: 53.0.2785.143-1
>  Candidate: 53.0.2785.143-1
>  Version table:
> *** 53.0.2785.143-1 500
>500 http://ftp.us.debian.org/debian testing/main amd64 Packages
>100 /var/lib/dpkg/status
> 53.0.2785.143-1~deb8u1 500
>500 http://security.debian.org stable/updates/main amd64
> Packages 53.0.2785.89-1~deb8u1 500
>500 http://ftp.us.debian.org/debian stable/main amd64 Packages

I have:

$ apt-cache policy chromium
chromium:
  Installed: 53.0.2785.143-1
  Candidate: 53.0.2785.143-1
  Version table:
 *** 53.0.2785.143-1 500
500 http://ftp.us.debian.org/debian stretch/main amd64 Packages
100 /var/lib/dpkg/status

>
>$ apt-cache policy libavutil55
>libavutil55:
>  Installed: 7:3.2-2
>  Candidate: 7:3.2-2
>  Version table:
> *** 7:3.2-2 500
>500 http://ftp.us.debian.org/debian testing/main amd64 Packages
>100 /var/lib/dpkg/status

I have - different from your own:

$ apt-cache policy libavutil55
libavutil55:
  Installed: 10:3.1.2-dmo2
  Candidate: 10:3.1.2-dmo2
  Version table:
 *** 10:3.1.2-dmo2 100
100 /var/lib/dpkg/status
 7:3.2-2 500
500 http://ftp.us.debian.org/debian stretch/main amd64 Packages

>
>$ apt-cache policy libavformat57
>libavformat57:
>  Installed: 7:3.2-2
>  Candidate: 7:3.2-2
>  Version table:
> *** 7:3.2-2 500
>500 http://ftp.us.debian.org/debian testing/main amd64 Packages
>100 /var/lib/dpkg/status

I have:

$ apt-cache policy libavformat57
libavformat57:
  Installed: 7:3.2-2
  Candidate: 7:3.2-2
  Version table:
 *** 7:3.2-2 500
500 http://ftp.us.debian.org/debian stretch/main amd64 Packages
100 /var/lib/dpkg/status

>
>and I don't have goldendict installed.

~$ apt-cache policy goldendict
goldendict:
  Installed: 1.5.0~git20160508.g92b5485-1.1
  Candidate: 1.5.0~git20160508.g92b5485-1.1
  Version table:
 *** 1.5.0~git20160508.g92b5485-1.1 500
500 http://ftp.us.debian.org/debian stretch/main amd64 Packages
100 /var/lib/dpkg/status
>
>When I launch chromium from a terminal, I don't get the error you
>report. (I do get a different message, but it looks like "normal"
>Chromium console spam; I'm not sure I've ever had a nontrivial Chromium
>session with no console messages.)
>
>-- 
>   The Wanderer
>
>The reasonable man adapts himself to the world; the unreasonable one
>persists in trying to adapt the world to himself. Therefore all
>progress depends on the unreasonable man. -- George Bernard
>Shaw
>
 
Thank you for your reply.

Charlie





http://www.egwildlife.com.au/

-- 
Registered Linux User:- 329524
***

Let each one turn his gaze inward and regard himself with awe
and wonder, with mystery and reverence; let each one work his
own influence, his own havoc, his own miracles. ---Henry
Miller

***

Debian GNU/Linux - just the best way to create magic

-


Am sending this through webmail.

Re: Weirdness with two packages.........

[The Wanderer]

On 2016-11-27 at 18:57, Charles Schroeder wrote:

> Hello Everyone,
> 
>   In  3.16.0-4-amd64 #1 SMP Debian 3.16.7-ckt11-1+deb8u3
>   (2015-08-04) x86_64 GNU/Linux
> 
> I get this message when I attempt to use Chromium:
> 
> $ chromium
> /usr/lib/chromium/chromium: relocation
> error: /usr/lib/x86_64-linux-gnu/libavformat.so.57: symbol
> avpriv_dict_set_timestamp, version LIBAVUTIL_55 not defined in file
> libavutil.so.55 with link time reference
> 
> Much the same when I attempt to use goldendict:
> 
> $ goldendict
> goldendict: relocation
> error: /usr/lib/x86_64-linux-gnu/libavformat.so.57: symbol
> avpriv_dict_set_timestamp, version LIBAVUTIL_55 not defined in file
> libavutil.so.55 with link time reference
> 
> I have no idea what is happening here or how I work round it?
> 
> I can't find those files in my: /usr/lib/x86_64-linux-gnu/ directory?

They should be there, not as files, but as symlinks.

> Trying to install both those packages tells me they are "already
> the newest version" so nothing to install?
> 
> Update, upgrade and reboot doesn't fix it.

What versions are those packages at, according to e.g. 'apt-cache policy
[packagename]'?

What about the libavutil55 and libavformat57 packages?

I have:

$ apt-cache policy chromium
chromium:
  Installed: 53.0.2785.143-1
  Candidate: 53.0.2785.143-1
  Version table:
 *** 53.0.2785.143-1 500
500 http://ftp.us.debian.org/debian testing/main amd64 Packages
100 /var/lib/dpkg/status
 53.0.2785.143-1~deb8u1 500
500 http://security.debian.org stable/updates/main amd64 Packages
 53.0.2785.89-1~deb8u1 500
500 http://ftp.us.debian.org/debian stable/main amd64 Packages

$ apt-cache policy libavutil55
libavutil55:
  Installed: 7:3.2-2
  Candidate: 7:3.2-2
  Version table:
 *** 7:3.2-2 500
500 http://ftp.us.debian.org/debian testing/main amd64 Packages
100 /var/lib/dpkg/status

$ apt-cache policy libavformat57
libavformat57:
  Installed: 7:3.2-2
  Candidate: 7:3.2-2
  Version table:
 *** 7:3.2-2 500
500 http://ftp.us.debian.org/debian testing/main amd64 Packages
100 /var/lib/dpkg/status

and I don't have goldendict installed.

When I launch chromium from a terminal, I don't get the error you
report. (I do get a different message, but it looks like "normal"
Chromium console spam; I'm not sure I've ever had a nontrivial Chromium
session with no console messages.)

-- 
   The Wanderer

The reasonable man adapts himself to the world; the unreasonable one
persists in trying to adapt the world to himself. Therefore all
progress depends on the unreasonable man. -- George Bernard Shaw



signature.asc
Description: OpenPGP digital signature

Re: before first use of a brand new flash drive

[Doug]

On 11/27/2016 05:11 PM, rhkra...@gmail.com wrote:

On Sunday, November 27, 2016 03:02:19 PM Brad Rogers wrote:

f3 is in the Debian repos.

For anybody that goes looking for it, it is apparently a "current" package for
Jessie and Sid, but it is in backports for Wheezy.

https://packages.debian.org/wheezy-backports/f3


There are also a number of rpm versions in rpmfind. I installed one, but 
I will have to make a PATH to it--the system can't find it.


--doug

Weirdness with two packages.........

[Charles Schroeder]

Hello Everyone,

In  3.16.0-4-amd64 #1 SMP Debian 3.16.7-ckt11-1+deb8u3
(2015-08-04) x86_64 GNU/Linux

I get this message when I attempt to use Chromium:

$ chromium
/usr/lib/chromium/chromium: relocation
error: /usr/lib/x86_64-linux-gnu/libavformat.so.57: symbol
avpriv_dict_set_timestamp, version LIBAVUTIL_55 not defined in file
libavutil.so.55 with link time reference

Much the same when I attempt to use goldendict:

$ goldendict
goldendict: relocation
error: /usr/lib/x86_64-linux-gnu/libavformat.so.57: symbol
avpriv_dict_set_timestamp, version LIBAVUTIL_55 not defined in file
libavutil.so.55 with link time reference

I have no idea what is happening here or how I work round it?

I can't find those files in my: /usr/lib/x86_64-linux-gnu/ directory?

Trying to install both those packages tells me they are "already
the newest version" so nothing to install?

Update, upgrade and reboot doesn't fix it.

Any clues please.

TIA,
Charlie

Re: Stretch Alpha 8 netinst fails

[Cindy-Sue Causey]

On 11/17/16, marathon.duran...@gmail.com  wrote:
> On Tue, 2016-11-15 at 19:19 -0500, Michael Owen wrote:
>> Installation fails installing the base system:
>>
>> /lib/partman/choose_partition/60partition_tree/do_option: line 88
>>
>> /lib/partman/active_choices/copy/choices: not found
>>
>> I've tried different hard drives, both installing on a blank drive and a
>> selected partition and always the exact same error message. I tried the
>> main Alpha 8 and a daily build...same results.
>>
>> Hardware is older core 2 duo on SATA drive, nothing esoteric. I've
>> installed many distros and never got anything like this. Alpha 7
>> installs fine.
>>
> Can't help, must I gave up on Alpha 8 as I couldn't get it to install
> either.
> Mucho disk errors on a new install and when I eventually got it installed,
> there
> was a problem with the distro list not being signed. So, after asking for
> help,
> not getting any here, decided to go with Ubuntu. Seems more suited for a
> desktop
> install IMHO. Unfortunately, as I've been a longtime Debian user.


I'm working through a few glitches of my own because I'm working in
Sid Unstable. Stretch Alpha 8. Hm. That sounds pretty unstable, too.

As soon as those glitches I'm experiencing are worked through, perhaps
we could start a new thread on using debootstrap to install
Debian.

Cindy :)

-- 
Cindy-Sue Causey
Talking Rock, Pickens County, Georgia, USA

* runs with plastic sporks *

Re: Dual Boot

[Pascal Hambourg]

Le 27/11/2016 à 21:02, Jan Bakuwel a écrit :


On 26/11/16 11:01, Pascal Hambourg wrote:


When embedding is not possible, the core image is stored as a regular
file in /boot/grub. Then /boot/grub must be on the same drive as the
boot image. However blocklists are not reliable with files, because
the filesystem may move blocks containing a file around.


I find it extremely useful (to the point where it once literally saved
me from disaster which is a story beyond the context of this thread) to
have multiple OS-es installed on any machine. I've been doing that for
years and even with grub it has - so far - always worked fine. But I
hear you that that's not a guarantee, and thus I started using extlinux.


I think that I experienced that problem once (but not sure), and needed 
to reinstall GRUB. To be honest, the chances that the blocs containing 
the core image may be moved around are very thin if you have a separate 
ext2 /boot partition, which is rarely modified.



Which brings me to the following question: what is the recommended way
to boot multiple OSes with grub, for example with a partition layout as
below. Or is there simply no sane way to do this with grub?


I would not pretend to know "the recommended way".
However I would recommend a few layouts for some configurations.

If EFI boot is an option, then install all systems for EFI boot (for 
Windows this implies a GPT partition table on the boot disk). This way 
all bootloaders can be installed in separate directories in the EFI 
system partition without overwriting one another. You can define one of 
them as the primary boot loader with higher boot priority in the EFI 
boot list.


For BIOS boot, without Windows I would recommend to install a primary 
GRUB in the MBR with embedding. Install secondary GRUBs in partitions 
boot records, without embedding. Then you have several options :


- add all entries for Windows and other Linux system kernels in the main 
GRUB menu, which can be done automatically with os-prober called by 
update-grub. The drawback is that the main boot menu must be updated 
everytime a secondary boot loader is updated.


- add entries to chainload the secondary boot loaders in the main GRUB 
menu, so that the main boot menu does not have to be updated when a 
secondary boot loader is updated. The "multiboot" command can be used to 
load a secondary GRUB core image file, so the embedding requirement does 
not apply here.


- include or load secondary GRUB config files in the main boot menu. 
This does not work well with different versions of GRUB, so I do not 
recommend it.


One must also decide whether to use the GRUB belonging to one of the 
Linux systems or an independent GRUB as the main boot loader. Of course 
there is less risk to damage an independent boot loader.


So my advice for a safe setup would be to :
- install the main GRUB in the MBR and its own partition for /boot/grub
- install each secondary GRUB in its OS partition
- chainload secondary GRUBs in the main GRUB's boot menu

Unfortunately with Windows the MBR is not a safe place because Windows 
overwrites the MBR at installation and sometimes when updating, so be 
prepared to restore a backup or reinstall GRUB. The boot record of the 
main GRUB's partition is a safer place, but it prevents embedding, so 
the risk of moving blocks comes back. However this partition should not 
be mounted and modified often, so the risk is low.



/dev/sda1: boot/rescue system

/dev/sda2: Windows 7
/dev/sda3: Windows 10

/dev/sda5: /boot - Linux system 1
/dev/sda6: / - Linux system 1
/dev/sda7: /var - Linux system 1
/dev/sda8: /var/log - Linux system 1

/dev/sda9: /boot - Linux system 2
/dev/sda10: / - Linux system 2
/dev/sda11: /var - Linux system 2
/dev/sda12: /var/log - Linux system 2

/dev/sda13: swap
/dev/sda14: LVM


From the missing /dev/sda4, I guess it is an extended partition and 
/dev/sda{5..14} are logical partitions in a DOS/MBR scheme. 10 logical 
partition is quite a lot. The linked list extended partition structure 
is quite fragile and rigid. A GPT partition table would be more robust, 
but it requires to reinstall reinstall Windows in EFI mode, and of 
course that the motherboard firmware is a UEFI. Otherwise, using LVM for 
most filesystems and swap is an option to reduce the number of partitions.

*WHY* does Debootstrap wiki page point Users to Ubuntu?

[Cindy-Sue Causey]

Hi, All :)

Please forgive me if I have simply missed the memo where we bought
Ubuntu or Ubuntu bought Debian to where this is an appropriate move.

So what had had happened was... I've been attempting to debootstrap
DEBIAN Stretch for a few weeks now. It's been failing MISERABLY. This
is the first time it has been like this in the couple years I've been
debootstrap'ing Debian.

#1 is chroot becomes a user called "I have no name!" instead of root.

#2 is apt-get fails, says it doesn't exist at all, so nothing else is
possible re growing the new setup. I a-sume that may be a secondary
issue that would clear up when user "I have no name!" gets over
his/her identity crisis.

After yet another failed debootstrap attempt earlier today, I decided
that maybe the debootstrap process has changed in some way. That meant
a rational next step was to visit our Debian Wiki to refresh my now
very stale personal notes on the subject.

Within seconds of hitting up the Debian Wiki debootstrap page, I found
myself reading the destructions for how to debootstrap in... an
operating system that is not Debian.

For a split second I thought maybe it was something about a Debian
package. NOPE. Those instructions on the debootstrap page include a
repository link leading off and completely away from anything even
remotely Debian.

https://wiki.debian.org/Debootstrap

Toward the bottom of the page there...

My sincerest apologies if I have in fact missed the memo where that's
the direction we're going. If it's an executive decision that was
made, that's what it is what it is... what it is.

#ThankYou to *EVERYONE* who contributes. Debian has been a MAJOR part
of my abject poverty level computing for several years now.

Cindy :)

-- 
Cindy-Sue Causey
Talking Rock, Pickens County, Georgia, USA

* [redacted for Civility's sake] :) *

Re: before first use of a brand new flash drive

[rhkramer]

On Sunday, November 27, 2016 03:02:19 PM Brad Rogers wrote:
> f3 is in the Debian repos.

For anybody that goes looking for it, it is apparently a "current" package for 
Jessie and Sid, but it is in backports for Wheezy.

https://packages.debian.org/wheezy-backports/f3

Re: firewall para novatos

[Alejandro Gutiérrez]

Algo simple e intuitivo podría ser Gufw, la interfaz gráfica de ufw.

El 27/11/16 a las 11:09, Ricardo Adolfo Sánchez Arboleda escribió:
> Mira
> https://es.wikipedia.org/wiki/PfSense
> https://pfsense.org/
> 
> 
> 
> *Saludes;*
> 
> rasa.
> 
> El 26 de noviembre de 2016, 12:57, Laotrasolucion
> > escribió:
> 
> 
> 
> El 26/11/16 a las 10:35, divagante escribió:
> > Hola gente!
> >
> >  Bueno, por conveniencia para quien quiera ayudarme con algun dato
> dire
> > que soy un usuario de debian de 7 años de antiguedad, pero que aun no
> > encontre el tiempo o las ganas de conocer profundamente la
> > administracion de sistemas unix o redes...
> >  No hago scrips ni entiendo sobre redes mas alla de configurar con
> > alguna guia de ayuda /etc/network/interfaces. Algo que hace rato
> ni hago
> > debido a los gestores como wicd o gnome.
> >
> >  Si bien no descarto en un futuro leer y meterme con iptables,
> quisiera
> > empezar al montar un futuro servidor de radio streaming, con un
> firewall
> > intuitivo, facil de manejar y con interfaz grafica.
> >
> >  Nota: recuerdo que hace ya algunos años usando windows y el antivirus
> > kaspersky instale el firewall de este ultimo, y la verdad me resulto
> > super intuitivo y manejable. Se veian claramente las peticiones de
> algun
> > programa freeware hacia internet y como este las denegaba si uno con
> > algunos clicks lo determinaba asi.
> >
> >  Muchas gracias por su ayuda.
> >
> >
> 
> Hola,
> 
> Yo la verdad que no he usado mucho las GUI para iptables, pero la pagina
> de iptables tiene un apartado con algunos.
> 
> http://www.iptables.info/en/iptables-gui.html
> 
> 
> Yo personalmente he usado alguna vez fwbuilder, que esta bastante bien.
> Hay bastante documentación y videos explicativos.
> No se si algunos de los que existe pueda complacer tus requerimientos
> tal como lo hizo don kaspersky, pero dejame decirte que armar un
> firewall no es una tarea tan complicada, siempre y cuando entiendas los
> conceptos.
> 
> 

Re: downloading mail

[Joe]

On Sun, 27 Nov 2016 14:37:23 -0700
Bob Holtzman  wrote:

> fetchmail
> fetchmail: no mailservers have been specified.
> 
> which is pretty strange since 
> 
> less fetchmail
> poll "pop.west.cox.net"
> protocol pop3  
> username "holtzm"   
> #password "4vr4mz4v3l"   
> password ""
> mimedecode  
> mda "/usr/bin/procmail -f -"
> 
> Tried a few other cominations including .fetchmail and .fetchmairc
> with no better results.
> 
> I'mmissing something obvious, but what?
> 

fetchmailrc is the incantation you're looking for. The colour of the
goat is optional.

-- 
Joe

Re: Jessie upgrade without systemd [was: Debian *not very good]

[Patrick Bartek]

On Sat, 26 Nov 2016 10:02:50 +0100  wrote:

> On Fri, Nov 25, 2016 at 03:57:53PM -0800, Patrick Bartek wrote:
> > On Fri, 25 Nov 2016 21:26:06 + (UTC) Latincom
> >  wrote:
> [snip]
> > > Is there a step by step guide or How to on line?
> > > I have 1 Wheezy without Systemd, and i would like to upgrade it.
> > > Thanks.
> > 
> > If you do a normal dist-upgrade Wheezy to Jessie, sysvinit will be
> > replaced with systemd.
> 
> Not forcefully.

Depends on what "forcefully" means to you.  Systemd will still be
installed because of dependencies.  Perhaps, not the init part, but the
rest of it will be even if you choose sysvinit. 

> > And probably screw everything up..
> 
> Now this is an unnecessarily loaded statement. Given the smoking holes
> the last flame war has left[1], I'd tread carefully if I were you ;-)

Yes.  "Everything" is too all encompassing, but there'll be enough
problems to consider a clean install instead of a dist-upgrade when
going from  sysv to systemd, particularly if it's customized with
custom start up scripts, etc..  I've never trusted "upgrading,"
regardless of the distro anyway.  There were way too many problems
early on.

> > I suggest
> > you do what I did: a clean install of a terminal only Jessie system,
> > replace systemd with sysvinit, then build the system up from there.
> > Just remember GNOME3 has systemd as a dependency.  Other utilities
> > do, too. I used LXDE which doesn't to keep things simple.  I
> > usually just run a window manager Openbox and a single LXPanel, but
> > that involves a lot more configuration.  Too much for an initial
> > test.
> 
> While possible, this isn't really necessary. FWIW I managed a clean
> Jessie upgrade without touching systemd, by just following the
> instructions.

In my case, my initial test (in Virtualbox) upgrade Wheezy to Jessie
with my customized Openbox/LXPanel set up with sysvinit conversion
wasn't enitrely successful.  For that reason and others, a clean
install of Jessie, terminal only to begin with, worked much better.  In
any case, after all was said and done, I decided to look for an
alternative to Debian/systemd for my next OS after long term support
ceases for Wheezy. Next year, I think.

> If you want a straight upgrade without systemd, apt-pinning seems
> to be the agreed upon way:
> 
>   
> https://www.debian.org/releases/stable/amd64/release-notes/ch-information.en.html#systemd-upgrade-default-init-system

That has its own caveats as actually noted in the above instructions.

> Note that many things (Gnome, I'm looking at you) *require* systemd
> these days: it'll be much more difficult to avoid systemd if you
> want a "modern" desktop environment.
> 
> Myself, I'm on Fvwm. I don't even need DBUS :-D

FWIW, I've ended up with systemd (not the init part) installed on my
Wheezy system when it wasn't there before.  Don't know when it happened.
And didn't intentionally do it myself.  Probably due to a dependency of
some app I installed.

I fear systemd will become pervasive.  It's already unwelcomed by many.

> [1] Sorry for the somewhat grumpy tone, but I'm pretty tired of people
>whining about systemd and borderline disrupting otherwise
> functional mailing lists with their rage. I strongly dislike systemd,
> yes, but i see no reason to hate systemd proponents let alone to
> harrass them. On the contrary, they are doing free software, FFS!

No apology necessary.  I wasn't offended.  I was merely stating what my
experience with systemd has been, not trying to start a flame war.
True, I'm not a fan of systemd for various rational reasons, but then
I'm not particularly fond of sysv either.  It's showing its age.  I
prefer runit for a "modern" init system: small, light, does the job,
and stays out of the way.

>If all that energy spent on foaming at the mouth and hatred had
> been spent on keeping viable alternatives to systemd afloat and
> running, we'd be in much better shape these days (and perhaps
> MikeeeUSA would have found other coattails to ride on).

True.

B

downloading mail

[Bob Holtzman]

fetchmail
fetchmail: no mailservers have been specified.

which is pretty strange since 

less fetchmail
poll "pop.west.cox.net"
protocol pop3  
username "holtzm"   
#password "4vr4mz4v3l"   
password ""
mimedecode  
mda "/usr/bin/procmail -f -"

Tried a few other cominations including .fetchmail and .fetchmairc with
no better results.

I'mmissing something obvious, but what?

-- 
Bob  Holtzman
"Bother", said Pooh, as he chambered another round...

ULTIMO DIA Black Friday Aprovecha este Descuento! --Vento--

[Volkswagen Alra]

#BlackFridayOportunidades
¡Descuento Inmejorable!
Aprovecha Black Friday para obtener tu Volkswagen Vento. Podrás tenerlo con 
$25.000 de descuento. Obtené el beneficio, no dejes inmovilizado tu capital.  
¡Es muy simple!
header-image
CUPOS LIMITADOS
No dejes pasar esta oportunidad en Alra S.A. (cupos limitados)
CONSULTAR AHORA
Imágenes solo ilustrativas. Bases y condiciones disponibles en nuestras 
sucursales
© 2016 Alra VW Derechos Reservados
Web

Re: stunnel or sibling?

[Thomas Schmitt]

Hi,

Peter E. wrote:
> accept  = localhost:80
> connect = en.wikipedia.org:443
> ... Firefox continues to access http servers as if stunnel 
> isn't present

I get a bit further with

  accept  = localhost:30080
  connect = en.wikipedia.org:443

and in iceweasel as URL

  http://localhost:30080

But then i get a Wikimedia Foundation error page
  "This domain points to a Wikimedia Foundation server, but
   is not configured on this server."
IIRC, the http protocol lets the client transmit the domain of the
targeted server. "localhost:30080" is not convincing, i assume.

We need some proxy functionality. But google brings me only info
how to connect SSL aware clients through a proxy to an SSL aware server.


Have a nice day :)

Thomas

Re: How to get ssh to run in daemon script

[Russell Gadd]

On 27/11/16 13:00, Nemeth Gyorgy wrote:

...
What user runs the daemon? If it is a separate user then run the ssh
with that user interactively. If the user does not have the
~/.ssh/known_hosts file or the destination host is not in it, then you
will have problems.


I've now discovered that the daemon is run by root.

As I said in my post the ssh command does run successfully when run 
interactively "but if I run the ssh command interactively from the usual 
graphical terminal (as root) it shuts down the NAS". 
/root/.ssh/known_hosts does exist. As it does work interactively, 
presumably root must have the correct entries in the known_hosts file.


One problem is that when the daemon runs the script the ssh command 
fails and I don't know if there was any error message, and if so where 
to find any error messages after I reboot.

ULTIMO DIA Black Friday Aprovecha este Descuento!

[Volkswagen Argentina Alra]

Ver Online
facebook
logo-top
#BlackFridayOportunidades
Sinceramiento Fiscal 2016
Aprovecha Black Friday para obtener tu Volkswagen Suran . Podrás tenerla con 
$40.000 de descuento. Obtené el beneficio, no dejes inmovilizado tu capital.  
¡Es muy simple!
header-image
CUPOS LIMITADOS
No dejes pasar esta oportunidad en Alra S.A. (cupos limitados)
CONSULTAR AHORA
Imágenes solo ilustrativas. Bases y condiciones disponibles en nuestras 
sucursales
© 2016 Alra VW Derechos Reservados
Web

Re: stunnel or sibling

[peter]

*   From: "Thomas Schmitt" n.wikipedia.org:443

... Firefox continues to access http servers as if stunnel 
isn't present and the http-only client freezes when a connection 
is expected.  I don't understand this case.

Anyone have any other ideas to offer?

Thanks again Thomas.  Good to have POP3 secured,

   ... Peter E.
   
-- 

123456789 123456789 123456789 123456789 123456789 123456789 123456789
Tel: +1 360 639 0202  Pender Is.: +1 250 629 3757
http://easthope.ca/Peter.html  Bcc: peter at easthope. ca

Re: before first use of a brand new flash drive

[Brad Rogers]

On Sun, 27 Nov 2016 19:43:52 -0600
Doug  wrote:

Hello Doug,

>That's a very old page. Does any reader know of an up-to-date Linux 
>program that serves this purpose?

f3 is in the Debian repos.

-- 
 Regards  _
 / )   "The blindingly obvious is
/ _)radnever immediately apparent"
You're not so safe in the safety of your room
Nasty - The Damned


pgpJg1foa8FM6.pgp
Description: OpenPGP digital signature

Re: Dual Boot

[Jan Bakuwel]

Hi Pascal,

Thanks for responding.

On 26/11/16 11:01, Pascal Hambourg wrote:
> Le 25/11/2016 à 19:31, Jan Bakuwel a écrit :
>>
>> For reasons beyond my understanding grub complains about being
>> installed in a partition instead of the MBR ("embedding is not
>> supported and a BAD idea", yet it works fine).
>
> The GRUB BIOS boot loader is split off in three main parts :
> - the boot image stored in the MBR or in a partition boot sector (PBR)
> - the core image stored in a special area outside a filesystem (called
> "embedding") or as a regular file in a filesystem
> - modules, config files and so on stored in /boot/grub/
>
> The only purpose of the boot image is to load the core image. The boot
> image is a very small program which must fit into a single sector, so
> it does not understand any partition table or filesystem format. It
> reads a hardcoded list of physical sectors ranges ("blocklists") which
> contain the core image.
>
> Embedding of the core image is only (but not always) possible when the
> boot image is in the MBR and there is a big enough "embedded area"
> between the MBR and the first partition on a DOS/MBR partition table
> or a "BIOS boot" partition on a GPT partition table, or when the boot
> image is in a partition boot sector and the partition format contains
> a suitable area for embedding.
>
> I have yet to find such a partition format, so when installing the
> boot image in the PBR of a partition with any usual contents type
> (ext4, LVM physical volume, RAID member...), embedding is not possible.
>
> When embedding is not possible, the core image is stored as a regular
> file in /boot/grub. Then /boot/grub must be on the same drive as the
> boot image. However blocklists are not reliable with files, because
> the filesystem may move blocks containing a file around.

I find it extremely useful (to the point where it once literally saved
me from disaster which is a story beyond the context of this thread) to
have multiple OS-es installed on any machine. I've been doing that for
years and even with grub it has - so far - always worked fine. But I
hear you that that's not a guarantee, and thus I started using extlinux.

Which brings me to the following question: what is the recommended way
to boot multiple OSes with grub, for example with a partition layout as
below. Or is there simply no sane way to do this with grub?

/dev/sda1: boot/rescue system

/dev/sda2: Windows 7
/dev/sda3: Windows 10

/dev/sda5: /boot - Linux system 1
/dev/sda6: / - Linux system 1
/dev/sda7: /var - Linux system 1
/dev/sda8: /var/log - Linux system 1

/dev/sda9: /boot - Linux system 2
/dev/sda10: / - Linux system 2
/dev/sda11: /var - Linux system 2
/dev/sda12: /var/log - Linux system 2

/dev/sda13: swap
/dev/sda14: LVM

best regards,
Jan

No Front Panel Audio On Asus ROG Ranger VIII

[Zack]

Hello Debian Users!  Before I start out, I want to mention that I have
already emailed the alsa-user list, but haven't gotten any replies.

I have an ASUS ROG Ranger VIII mother board.  The rear panel audio
works, but the front panel audio output does not (though input works).
The system recognizes when headphones are plugged into the front
panel, but no sound comes out.  Here is the output of the alsa-info
script: 
http://www.alsa-project.org/db/?f=61a8cd9f7a174a7c53a96266a125b7eca9c85c86
.

Things I've done:
1. I have gone into the alsa mixer and ensured that the front panel
audio was not muted and that the volume level was up.

2. I have opened up pulse's audio control and seen that the volume bar
is jumping around with the headphone port listed as the output sync.

3. I have gone into Windows and ensured the port worked.

4. I have done a couple of random tricks with aplay.conf

So far, nothing has worked.  You can see most of the stats of my
system in the link above, but I am running Debian Testing with the 4.8
kernel.

Thanks,
Zack

Re: before first use of a brand new flash drive

[Doug]

On 11/27/2016 11:28 AM, Richard Owlett wrote:

On 11/27/2016 10:07 AM, Felix Miata wrote:

Richard Owlett composed on 2016-11-27 07:02 (UTC-0500):

...

It may be an interesting exercise, before first use of a brand
new flash drive, to examine it via Gparted and "wipefs
/dev/sdX"...


I first test that the claimed capacity is valid:
https://fightflashfraud.wordpress.com/2008/11/24/h2testw-gold-standard-in-detecting-fake-capacity-flash/ 





Thank you. The page was interesting reading. It got me thinking.
I haven't done any significant programming since CPU's heated 
buildings with 12AX7's ;/


That's a very old page. Does any reader know of an up-to-date Linux 
program that serves this purpose?


--doug

Re: Installation impossible sur Acer Travelmate 2400

[didier gaumet]

Le 16/11/2016 à 23:16, Samy Mezani a écrit :
> Bonjour,
> 
> C'est la première fois en 10 ans que ça m'arrive. Impossible d'installer
> une Debian sur un portable Acer TraveMate 2403WXMi... 
[...]
> Même sanction à chaque fois, un message d'erreur (ci-dessous) répétitif,
> non bloquant mais ensuite fonctionnement de l'installateur très lent et
> franchement inutilisable :
[...]
> Par contre j'ai essayé un vieux live CD Kubuntu 10.04, et là ça marche.
> Le noyau est un 2.6.32 en i686.
> 
> Un problème avec le noyau 3.16 ou avec le vieux processeur Intel Celeron
> M ?
> 
> Samy

A priori au moins certains Celeron M et Pentium M seraient compatible
PAE sans que cela apparaisse dans les flags du processeur. Cela pose
problème pour les distributions récentes qui ont généralement basculé
vers des noyaux PAE par défaut pour l'architecture X86. Dans ce cas deux
solutions:
- forcer le PAE (en passant au noyau le paramètre --forcepae) au boot de
l'installateur et modifier la config grub une fois Debian installé pour
y inclure le paramètre --forcepae
- installer le noyau nonpae linux-image-3.16.0-4-586 au lieu de celui
par défaut linux-image-3.16.0-4-686-pae

(je me permets exceptionnellement de mettre Samy en copie au cas où il
ne serait pas abonné à la liste vu que ça fait déjà un petit moment que
ce sujet a été posté)

Re: before first use of a brand new flash drive

[Richard Owlett]

On 11/27/2016 10:07 AM, Felix Miata wrote:

Richard Owlett composed on 2016-11-27 07:02 (UTC-0500):

...

It may be an interesting exercise, before first use of a brand
new flash drive, to examine it via Gparted and "wipefs
/dev/sdX"...


I first test that the claimed capacity is valid:
https://fightflashfraud.wordpress.com/2008/11/24/h2testw-gold-standard-in-detecting-fake-capacity-flash/



Thank you. The page was interesting reading. It got me thinking.
I haven't done any significant programming since CPU's heated 
buildings with 12AX7's ;/

Re: before first use of a brand new flash drive (was: Error message from GParted...)

[Felix Miata]

Richard Owlett composed on 2016-11-27 07:02 (UTC-0500):

...

It may be an interesting exercise, before first use of a brand
new flash drive, to examine it via Gparted and "wipefs /dev/sdX"...


I first test that the claimed capacity is valid:
https://fightflashfraud.wordpress.com/2008/11/24/h2testw-gold-standard-in-detecting-fake-capacity-flash/
--
"The wise are known for their understanding, and pleasant
words are persuasive." Proverbs 16:21 (New Living Translation)

 Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!

Felix Miata  ***  http://fm.no-ip.com/

Re: [HS] Géolocalisation et déontologie :

[Grégory Reinbold]

Salut,

A l'ouverture de la page demande confirmation à l'utilisateur sa 
GEO-localisation (depuis nginx).

* S'il accepte, tu exploite les données (PHP) et tu lui demande si la 
géolocalisation correspondant à son emplacement.

- S'il confirme : jackpot
- S'il ne confirme pas, invite-le à saisir son emplacement.

* S'il refuse, tu n'exploite pas les données et tu lui demande de saisir son 
emplacement pour continuer.

Cdt 

À dim. nov. 27 15:37:49 2016 GMT+0100, Ph. Gras a écrit :
> Bonjour à toutes et à tous !
> 
> actuellement, je travaille sur une application Web destinée à tracer des 
> bagages.
> 
> Un voyageur se connecterait sur le site où se trouve l'appli au moment de 
> fermer
> sa, ou ses valises, génère un QR Code, l'imprime et le colle dessus.
> 
> À chaque rupture de charge, un intervenant peut scanner le code et participer 
> au
> suivi des bagages : à l'aéroport, en prenant un taxi et à l'hôtel à son 
> arrivée.
> 
> J'ai commencé à développer ça en javascript, et derrière il y aura une base 
> SQL
> pour récupérer les données, celle-ci sera manipulée avec PHP.
> 
> L'implémentation de la géolocalisation m'a posé quelques problèmes. Avec JS,
> le navigateur rouspète à chaque fois qu'on l'interroge à ce sujet. Je 
> comprends
> pourquoi et je l'admets parfaitement. Le problème est que ces précautions sont
> déclenchées au chargement de la page, quand le navigateur analyse le script.
> 
> Cela ne me laisse pas le temps d'expliquer le pourquoi du comment, et je crois
> que la plupart des touristes en partance n'en n'ont rien à foutre de toute 
> façon.
> 
> Je pense donc me tourner vers le module de géolocalisation de NginX, qui est
> chargé avec le paquet Debian nginx-full :
> http://nginx.org/en/docs/http/ngx_http_geoip_module.html
> 
> Je suppose que je pourrai ensuite récupérer les données avec :
> http://php.net/manual/fr/book.geoip.php
> 
> D'un côté, ça me gêne un peu de faire ça à l'arrache (même si je peux avertir 
> le
> client d'un service qu'il devrait apprécier), et de l'autre ça me gave de 
> l'embêter
> les avertissements du navigateur, que je ne peux pas gérer.
> 
> Étant bien entendu que si le gars est derrière un proxy, et que les 
> coordonnées
> récupérées supposent la plus grande prévention, ce sont ses bagages et donc
> son problème.
> 
> Comment aborderiez-vous cette question de votre côte ?
> 
> Merci pour vos réponses et bon dimanche,
> 
> Ph. Gras
>

-- 
Envoyé depuis mon Jolla

[HS] Géolocalisation et déontologie :

[Ph. Gras]

Bonjour à toutes et à tous !

actuellement, je travaille sur une application Web destinée à tracer des 
bagages.

Un voyageur se connecterait sur le site où se trouve l'appli au moment de fermer
sa, ou ses valises, génère un QR Code, l'imprime et le colle dessus.

À chaque rupture de charge, un intervenant peut scanner le code et participer au
suivi des bagages : à l'aéroport, en prenant un taxi et à l'hôtel à son arrivée.

J'ai commencé à développer ça en javascript, et derrière il y aura une base SQL
pour récupérer les données, celle-ci sera manipulée avec PHP.

L'implémentation de la géolocalisation m'a posé quelques problèmes. Avec JS,
le navigateur rouspète à chaque fois qu'on l'interroge à ce sujet. Je comprends
pourquoi et je l'admets parfaitement. Le problème est que ces précautions sont
déclenchées au chargement de la page, quand le navigateur analyse le script.

Cela ne me laisse pas le temps d'expliquer le pourquoi du comment, et je crois
que la plupart des touristes en partance n'en n'ont rien à foutre de toute 
façon.

Je pense donc me tourner vers le module de géolocalisation de NginX, qui est
chargé avec le paquet Debian nginx-full :
http://nginx.org/en/docs/http/ngx_http_geoip_module.html

Je suppose que je pourrai ensuite récupérer les données avec :
http://php.net/manual/fr/book.geoip.php

D'un côté, ça me gêne un peu de faire ça à l'arrache (même si je peux avertir le
client d'un service qu'il devrait apprécier), et de l'autre ça me gave de 
l'embêter
les avertissements du navigateur, que je ne peux pas gérer.

Étant bien entendu que si le gars est derrière un proxy, et que les coordonnées
récupérées supposent la plus grande prévention, ce sont ses bagages et donc
son problème.

Comment aborderiez-vous cette question de votre côte ?

Merci pour vos réponses et bon dimanche,

Ph. Gras

Re: firewall para novatos

[Ricardo Adolfo Sánchez Arboleda]

Mira
https://es.wikipedia.org/wiki/PfSense
https://pfsense.org/



*Saludes;*

rasa.

El 26 de noviembre de 2016, 12:57, Laotrasolucion 
escribió:

>
>
> El 26/11/16 a las 10:35, divagante escribió:
> > Hola gente!
> >
> >  Bueno, por conveniencia para quien quiera ayudarme con algun dato dire
> > que soy un usuario de debian de 7 años de antiguedad, pero que aun no
> > encontre el tiempo o las ganas de conocer profundamente la
> > administracion de sistemas unix o redes...
> >  No hago scrips ni entiendo sobre redes mas alla de configurar con
> > alguna guia de ayuda /etc/network/interfaces. Algo que hace rato ni hago
> > debido a los gestores como wicd o gnome.
> >
> >  Si bien no descarto en un futuro leer y meterme con iptables, quisiera
> > empezar al montar un futuro servidor de radio streaming, con un firewall
> > intuitivo, facil de manejar y con interfaz grafica.
> >
> >  Nota: recuerdo que hace ya algunos años usando windows y el antivirus
> > kaspersky instale el firewall de este ultimo, y la verdad me resulto
> > super intuitivo y manejable. Se veian claramente las peticiones de algun
> > programa freeware hacia internet y como este las denegaba si uno con
> > algunos clicks lo determinaba asi.
> >
> >  Muchas gracias por su ayuda.
> >
> >
>
> Hola,
>
> Yo la verdad que no he usado mucho las GUI para iptables, pero la pagina
> de iptables tiene un apartado con algunos.
>
> http://www.iptables.info/en/iptables-gui.html
>
> Yo personalmente he usado alguna vez fwbuilder, que esta bastante bien.
> Hay bastante documentación y videos explicativos.
> No se si algunos de los que existe pueda complacer tus requerimientos
> tal como lo hizo don kaspersky, pero dejame decirte que armar un
> firewall no es una tarea tan complicada, siempre y cuando entiendas los
> conceptos.
>
>

Re: How to get ssh to run in daemon script

[Nemeth Gyorgy]

2016-11-27 11:55 keltezéssel, Russell Gadd írta:
I am trying to add a command into a script to shut down my NAS when 
the UPS detects a power loss. There is a daemon apcupsd which gets a 
signal from the UPS and runs various scripts which can be modified by 
the user.


I have a script doshutdown as follows

#!/bin/sh
 ... various messages and sleep delays
/usr/bin/ssh root@nasbox poweroff
 ... more messages and sleep delays

The ssh command in my script does not run in a power off test 
(everything else in the script works i.e. the messages and delays), 
but if I run the ssh command interactively from the usual graphical 
terminal (as root) it shuts down the NAS, so the NAS isn't stopping it 
working assuming its getting the appropriate handshaking from ssh.


My knowledge of how to get daemon scripts working and how to use ssh 
is almost nil. Also I can't find a way to get any error output which 
the ssh command might be issuing. I've tried enclosing the body of the 
script in brackets to redirect output to files:

{
} > /data/temp/shutdown-out.txt 2> /data/temp/shutdown-err.txt

I've looked in syslog, kern.log and auth.log without finding anything.

So I'm scrambling around in the dark. Currently my vague ideas as to 
what might be wrong are:


script doesn't know what nasbox is (it is defined in /etc/hosts)
ssh is being run without being associated as root
ssh cannot find root's keys
nas doesn't recognise login as authorised user

I'm also starting to think I am stressing the UPS too much by running 
too many power off tests, so I'd like to resolve this without too much 
more experimentation.


Grateful for any suggestions. I'm running Linux Mint 17.3 (based on 
Ubuntu 14.04, based on Jessie). Uses openssh (1:6.6p1-2ubuntu2.8)


What user runs the daemon? If it is a separate user then run the ssh 
with that user interactively. If the user does not have the 
~/.ssh/known_hosts file or the destination host is not in it, then you 
will have problems.

Re: Package configure problem during Installation.

[tomas]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

On Sun, Nov 27, 2016 at 04:33:58AM +0530, amir khan wrote:

(Please keep the Debian user mailing list in the loop. First of
all, there are lots of people there far more knowledgeable than
me, thus raising the probability for you to get good help, second
the results of your quest are there in the archives to help other
people

There is already a longer thread in the mailing list about your
question. If you aren't subscribed, you can see it here:

  https://lists.debian.org/debian-user/2016/11/msg00988.html

And oh, if you are not subscribed, please tell us to keep you in
copy, so you don't miss any annswers).

> Yes Tomas, I tried all those listed for India. But it didn't work. Hope to
> get it done the other way round(by using the mirror redirect service).
> And also, I'm new to the debian and linux world and I'm going to use it for
> Virtualization purposes(my final year project based on it). So , I just
> want to know that, why to install debian linux distribution if I can get it
> done on Ubuntu too. I'm a bit confused, please guide me. Thanks !

Well, those are two questions:

Archives:

the other archives in India: I don't know about the reachability of
those archives. I asked just to find out wheter it could be that this
one specific archive is down *or* whether you have more generic
network problems. Since now you tried several archives, if I were in
your place I'd try to find out whether your freshly installed VM has
access to the Internet at all. Pick one of the archives which failed:
can you access it from your browser?

Debian vs Ubuntu

Ubuntu is a Debian derivative, specially aimed at desktop users who
have less experience with Unix/Linux. Some tools are more polished,
some packages are newer. If you really want to tinker, Debian offers
more possibilities. That said, the differences are not that big, and
there are many tinkerers using Ubuntu and many "end-users" using
Debian. It's free software, that means that each project can (and
does) pick good things from the other. Since Debian is an open
distribution, there are many Debian derivatives.

regards
- -- tomás
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.12 (GNU/Linux)

iEYEARECAAYFAlg6z9IACgkQBcgs9XrR2kb5cgCfTHRko+ANt7pepuCVDHLUxib9
EHYAnj4ZnjUZxgN8b7P3pHOvaXvsLCFC
=aghL
-END PGP SIGNATURE-

Re: Package configure problem during Installation.

[Sven Hartge]

Brian  wrote:

> The only thing I vaguely understand about the difference between the
> two is that httpredir.debian.org uses the traditional mirror network
> whereas deb.debian.org uses the Fastly CDN (Content Delivery Network)
> and there is some caching going on. 

Fastly and Cloudfront are used

> Because Fastly has local peering arrangements mirror choices are
> supposed to be local (and hence faster). I've not noticed any
> difference in a few short tests.

You are correct.

Additional bonus: deb.debian.org can also be accessed via HTTPS if you
want/need to further obfuscate what you pull from mirrors.

Grüße,
S°

-- 
Sigmentation fault. Core dumped.

Re: Error message from GParted, means what?

[Richard Owlett]

On 11/27/2016 3:15 AM, Pascal Hambourg wrote:

Le 26/11/2016 à 23:42, Richard Owlett a écrit :

I've been getting a strange error messages when looking at some
8 GB
flash drives from different manufacturers purchased months apart.

There is a title bar at top saying "Libparted Warning (as
superuser).
The body of the message box has:
  1. an Exclamation point surrounded by orange and yellow
triangle.
  2. a message text saying "The driver descriptor says the
physical block
 size is 2048 bytes, but Linux says it is 512 bytes.
  3. There are two buttons present.
 One saying "Cancel"
 The other saying "Ignore"


I have seen this behaviour when running parted on a USB flash
drive containing a Debian installer ISOhybrid x86 image which
seems to combine a DOS partitition table, an Apple Mac partition
table, a GPT partition table and an ISO 9660 header. I guess the
2048-byte block size is set to match the sector size of optical
media.

I have found that historically partitioning tools based on
libparted (parted, gparted, partman [Debian installer]) did not
handle well this non-standard format. The parted version in
Wheezy just could not read it.

Another problem I observed is that erasing the partition table
does not erase the ISO 9660 header, and programs based on
libblkid (udev, desktop environment file managers) still find it
and wrongly apply the identifier string in it to label the drive.

Before reusing a drive which had contained an ISO image, I advise
to erase any metadata with wipefs.


*THANK YOU*  !
I think you may have cleared up several hard to isolate problems.
It may be an interesting exercise, before first use of a brand 
new flash drive, to examine it via Gparted and "wipefs /dev/sdX"


Yesterday I had purchased a blister pack of four 8GB that were on 
clearance at a local office supply store [labeled "Lexar by 
Micron" "Jumpdrive S50" USB 2.0].


One was used without any apparent problem to be the source for 2 
preseed.cfg files. I did installs with each which ran as expected.


One was plugged in before I ran a script to prepare a file that 
would later be written to the flash drive. There was some [not 
yet identified] problem with the script indicated by flashing of 
drive activity light. That was the drive that prompted my post.


After reading your message I examined the other two drives.
  "wipefs /dev/sdb" reported a dos partition table at 0x1fe.
  Gparted showed:
1st drive
  4.03 MiB unallocated
  7.45 GiB FAT32 (1.91 MiB used) (no Flags set)
2nd drive
  7.46 GiB FAT32 (4.03 MiB used) (lba Flags set)

One of them had been plugged into one of my machines yesterday. 
Which and what action taken - unknown.


I assume that running "wipefs --all --force /dev/sdb" followed by 
using Gparted to create a partition table followed by 
partitioning will make all Debian tools "happy".


Thank you.

Re: sudo + gpg - howto?

[Michael Luecke]

On 11/27/2016 10:10 AM, Kamil Jońca wrote:

2. cannot run pinentry (my guess is that tty is owned by user kjonca and
pinentry should be as user backup)


That's the problem. The owner of ttyX must match with the user for which 
pinentry is running.


You could change the permissions of tty to the backup user with
# chown backup /dev/ttyX


So my questions are:
1. is it possible to use passphrase-fd/passphrase-file options?
2. if not, how can I configure gpg/gpgagen/sudo/whatever in order to
enter passphrase as sudo target  user?



You could also try putting the option "allow-loopback-pinentry" to 
~/.gnupg/gpg-agent.conf and restarting the agent.


Then you could start gpg with '--pinentry-mode loopback' or put 
"pinentry-mode loopback" to ~/.gnupg/gpg.conf.


That worked for me with GnuPG 2 some time ago when I had a similar 
problem with duplicity.


Best Regards,
Michael

How to get ssh to run in daemon script

[Russell Gadd]

I am trying to add a command into a script to shut down my NAS when the 
UPS detects a power loss. There is a daemon apcupsd which gets a signal 
from the UPS and runs various scripts which can be modified by the user.


I have a script doshutdown as follows

#!/bin/sh
 ... various messages and sleep delays
/usr/bin/ssh root@nasbox poweroff
 ... more messages and sleep delays

The ssh command in my script does not run in a power off test 
(everything else in the script works i.e. the messages and delays), but 
if I run the ssh command interactively from the usual graphical terminal 
(as root) it shuts down the NAS, so the NAS isn't stopping it working 
assuming its getting the appropriate handshaking from ssh.


My knowledge of how to get daemon scripts working and how to use ssh is 
almost nil. Also I can't find a way to get any error output which the 
ssh command might be issuing. I've tried enclosing the body of the 
script in brackets to redirect output to files:

{
} > /data/temp/shutdown-out.txt 2> /data/temp/shutdown-err.txt

I've looked in syslog, kern.log and auth.log without finding anything.

So I'm scrambling around in the dark. Currently my vague ideas as to 
what might be wrong are:


script doesn't know what nasbox is (it is defined in /etc/hosts)
ssh is being run without being associated as root
ssh cannot find root's keys
nas doesn't recognise login as authorised user

I'm also starting to think I am stressing the UPS too much by running 
too many power off tests, so I'd like to resolve this without too much 
more experimentation.


Grateful for any suggestions. I'm running Linux Mint 17.3 (based on 
Ubuntu 14.04, based on Jessie). Uses openssh (1:6.6p1-2ubuntu2.8)

sudo + gpg - howto?

[Kamil Jońca]

I have gpg-encrypted backups made by amanda (gnutar).
Decrypting was done by call gpg with --passphrase-fd option.

Recently I have to see some backup contensts
and as usual I wrote:
--8<---cut here---start->8---
sudo amrecover ...
--8<---cut here---end--->8---
but without success.
Some investigation suggests that gpg-agent started as 'backup' user
1. completely ignores passphrase-fd/passphrase-file option.
2. cannot run pinentry (my guess is that tty is owned by user kjonca and
pinentry should be as user backup)


So my questions are:
1. is it possible to use passphrase-fd/passphrase-file options?
2. if not, how can I configure gpg/gpgagen/sudo/whatever in order to
enter passphrase as sudo target  user?
KJ

-- 
http://stopstopnop.pl/stop_stopnop.pl_o_nas.html
można nie mieć nawet żadnego sensownego argumentu, a dzięki odpowiedniej
gestykulacji stworzyć poczucie wiarygodności.
Puls Biznesu

Re: Error message from GParted, means what?

[Pascal Hambourg]

Le 26/11/2016 à 23:42, Richard Owlett a écrit :

I've been getting a strange error messages when looking at some 8 GB
flash drives from different manufacturers purchased months apart.

There is a title bar at top saying "Libparted Warning (as superuser).
The body of the message box has:
  1. an Exclamation point surrounded by orange and yellow triangle.
  2. a message text saying "The driver descriptor says the physical block
 size is 2048 bytes, but Linux says it is 512 bytes.
  3. There are two buttons present.
 One saying "Cancel"
 The other saying "Ignore"


I have seen this behaviour when running parted on a USB flash drive 
containing a Debian installer ISOhybrid x86 image which seems to combine 
a DOS partitition table, an Apple Mac partition table, a GPT partition 
table and an ISO 9660 header. I guess the 2048-byte block size is set to 
match the sector size of optical media.


I have found that historically partitioning tools based on libparted 
(parted, gparted, partman [Debian installer]) did not handle well this 
non-standard format. The parted version in Wheezy just could not read it.


Another problem I observed is that erasing the partition table does not 
erase the ISO 9660 header, and programs based on libblkid (udev, desktop 
environment file managers) still find it and wrongly apply the 
identifier string in it to label the drive.


Before reusing a drive which had contained an ISO image, I advise to 
erase any metadata with wipefs.

Re: stunnel or sibling?

[Thomas Schmitt]

Hi,

Peter E. wrote:
> For example, a 
> local elementary browser is not SSL capable but a connection to https 
> is required.
> Can stunnel or something else do that?

I use stunnel for connecting my old POP3 client to a SSL POP3 server.

Configuration file:
---
client=yes
foreground=yes
debug=7
pid=
sslVersion=all
[gmx_pop3]
accept=30027
connect=pop.gmx.net:995
---

My client connects to local port 30027, stunnel then forwards to pop.gmx.net
port 995. If i restart my router i also have to stop and restart the stunnel
processes even if no connection was active.

I understand that you would need one configuration and one local port for
each https server you want to connect to. Maybe there are better solutions.


Have a nice day :)

Thomas