configure asus amd uefi bios for netinst

[Russell L. Harris]

I have not yet managed to figure out the proper BIOS configuration to
get an AsusPRO (desktop computer) to boot from a USB flash stick for
netinstall of Debian 10.  The USB flash stick has Debian 10.2; I used
it about a week ago for a netinstall on another machine.

The machine was shipped from the factory with Windows7.  A year or
more ago I managed to install Debian 9 on the machine, overwriting W7;
but I no longer remember how I managed to boot the installer.  I wish
to do a complete installation, and not just an upgrade from Debian 9
to Debian 10.

The BOOT screen of the Asus UEFI BIOS utility has three sections which
appear applicable:

(1) Compatibility Support Module
(2) Secure Boot
(3) Boot Options Priorities
(4) Boot Override

In both "Boot Options Priorities" and "Boot Override", the USB flash
stick appears twice:

 = UEFI: Patriot Memory PMAP (7643MB)
 = Patriot Memory PMAP (7643MB)

The "Boot Override" section appears to be simply a shortcut to "Save
Configuration and Reset" for each of the boot options.

Comment enlever facilement la PUB d'un enregistrement tv ?

[machinSuite]

Bonjour à tous !

J'enregistre de temps en temps des films et reportages sur clé usb avec un
petit adaptateur tnt en h264 (mpeg4) pour la video, eac3 pour le son et ts
pour le format du conteneur et je veux enlever la pub au début, pendant et
après l'enregistrement sans faire aucun transcodage comme c'est le cas avec
handbrake.

Comment faire cela simplement et rapidement en notant les temps de début et
de fin de chaque passage de pub à l'aide de mpv puis en les entrant en ligne
de commande avec un utilitaire simple avec comme argument le fichier de 4 à
5 Go enregistré.

Merci d'avance pour le coup de main.
A plus.





--
Sent from: http://debian.2.n7.nabble.com/debian-user-french-f1152225.html

Re: cron error - date command

[Keith Bainbridge]

On 2/2/20 5:22 am, songbird wrote:

   it looks questionable to me just in that you
may be writing to that file at the same time as
you when you're doing something else that is also
trying to write to that file.



Thankyou Songbird and Roberto.

The solution was to escape the % signs in crontab (to \%)   Thanks again 
Teemu



Songbird, I note your concerns, but to date all is OK

I am simply trying to keep tack of what time I am doing stuff, and 
putting the time between the commands does that how I like it. Hope I 
don't prove your concerns correct.   I will own up IF I notice anything, 
though.



--
Keith Bainbridge

keith.bainbridge.3...@gmail.com
0447 667 468

Re: dislocker package broken, fix in 'sid'. How long 'til in backports?

[Keith Bainbridge]

On 2/2/20 4:27 pm, Michael Bonert wrote:

I am running Debian stable (buster).

The dislocker backport doesn't work for files from Windows 10 1903 (as 
noted here: 
https://www.mail-archive.com/debian-bugs-dist@lists.debian.org/msg1710592.html 
)


I noted that the latest version of dislocker in 'sid' (unstable) is 
0.7.1-5 ( https://packages.debian.org/sid/utils/dislocker )


The release notes for 0.7.1-5 ( 
https://metadata.ftp-master.debian.org/changelogs//main/d/dislocker/dislocker_0.7.1-5_changelog 
)
say:  "Added new patches to support Support Windows 10 1903 and fix 
build issues.Thanks to haobinnan " - sounds like 
exactly what I'd like to have.


How long will it take for 0.7.1-5 to appear as a backport?

Thanks,
Michael

PS -
I will trying making my own backport ( 
https://wiki.debian.org/DontBreakDebian / 
https://wiki.debian.org/SimpleBackportCreation ) but prefer to avoid 
that... if isn't too long a wait.





Another suggestion I read somewhere is, include the src repo for sid in 
your sources-list. After an update, seems you will have the code to 
build it yourself, and ensure that you don't need anything untoward. 
Never tried it my self, but sounds interesting.


To answer you question directly, not long.I used a vbox to test 
upgrading from bullseye to sid, a couple of weeks ago, to see how much 
difference there is. About 20 packages.


--
Keith Bainbridge

keith.bainbridge.3...@gmail.com
0447 667 468

dislocker package broken, fix in 'sid'. How long 'til in backports?

[Michael Bonert]

I am running Debian stable (buster).

The dislocker backport doesn't work for files from Windows 10 1903 (as  
noted here:  
https://www.mail-archive.com/debian-bugs-dist@lists.debian.org/msg1710592.html  
)


I noted that the latest version of dislocker in 'sid' (unstable) is  
0.7.1-5 ( https://packages.debian.org/sid/utils/dislocker )


The release notes for 0.7.1-5 (  
https://metadata.ftp-master.debian.org/changelogs//main/d/dislocker/dislocker_0.7.1-5_changelog  
)
say:  "Added new patches to support Support Windows 10 1903 and fix  
build issues.Thanks to haobinnan " - sounds like  
exactly what I'd like to have.


How long will it take for 0.7.1-5 to appear as a backport?

Thanks,
Michael

PS -
I will trying making my own backport (  
https://wiki.debian.org/DontBreakDebian /  
https://wiki.debian.org/SimpleBackportCreation ) but prefer to avoid  
that... if isn't too long a wait.

Re: Fresh install UEFI debian-10.2.0-amd64-xfce-CD-1 will not boot

[David Christensen]

On 2020-02-01 04:51, didier.gau...@gmail.com wrote:


Hello David,

- probably, you were not asked about grub during installation because you did 
not chose the expert installation mode
- possibly, your problems originate, at least in part, from the fact that 
albeit creating a new ESP (EFI) partition on your USB key, the installer try to 
use also the ESP partition(s) of the disk(s) on the computer used for the 
installation. I would suggest to verify that none of the partitions on the 
disks in the computer (ESP, swap, /boot, etc...) are used during partitionning.



Thanks for the reply.  :-)


When I installed from a USB stick onto a USB stick, the only other drive 
in the machine was an optical drive with no disc loaded.



David

Re: guys

[Keith Bainbridge]

On 2/2/20 7:53 am, Ben Lavender wrote:
Even though RTM always applies, sometimes things do get more complicated 
with lesser experienced users regardless and sometimes other peoples 
issues aren't always related to FAQ/DOC/MAN etc


This is a users mailing list after all.

On 01/02/2020 19:31, mick crane wrote:

I probably shouldn't post this.
I see all these questions people trying to get their installations to 
work.

It is supposed to be files with documentation what they do.
Is there a reason things seem to get more complicated ?

mick





Phew. I'm not feeling quite so guilty about NOT finding something while 
trying to RTflippingM, now.Turns out I was looking in the wrong 
flipping Manual.And I'm a long time (15 years at least, plus time  I 
spent looking for autokey).


Installation is getting easier - especially with live iso's. So why 
doesn't the project promote them more?


--
Keith Bainbridge

keith.bainbridge.3...@gmail.com
0447 667 468

Re: guys

[songbird]

mick crane wrote:
...
> I have heard that expression RTFM before, I dunno, I'm probably out of 
> order yet feeling is things are getting more complicated when you'd 
> think things should be getting simpler.

  do it for 40yrs and get back to us...


  songbird

Re: guys

[David Wright]

On Sat 01 Feb 2020 at 15:45:07 (-0500), Stefan Monnier wrote:
> On Sat 01 Feb 2020 at 19:31:31 (+), mick crane wrote:
> > I probably shouldn't post this.
> > I see all these questions people trying to get their installations to work.
> > It is supposed to be files with documentation what they do.
> > Is there a reason things seem to get more complicated ?
> 
> What's with all these "I"s?

Just his monthly rhetorical question?

https://lists.debian.org/debian-user/2020/01/msg00062.html

Cheers,
David.

Re: no mouse in midnight commander

[Frank McCormick]

On 2/1/20 2:39 PM, Joe wrote:

On Sat, 1 Feb 2020 13:46:52 -0500
Frank McCormick  wrote:


I noticed this morning that my mouse no longer works in MC. When I
use it, it leaves keycodes on mc's command line at
the bottom of the screen.

Because I use Rox-filer to put icons on the desktop, I call
MC from a small script file in $home which loads an xterm, with some
parameters to put up a bigger MC on the screen.

This is the script at the moment

xterm -g 140x40+140+20 -fn 10x20 -e mc -x

I am running Debian Sid fully updated.




How fully? My mc (in Gnome Terminal) looks OK, but I haven't upgraded
today or yesterday, as apt wants to remove 84 packages. There looks to
be some kind of logjam involving KDE and/or Qt.



  Updated as of this morning...I don't run KDE or any QT stuff so I 
didn't experience that.


--
Frank McCormick

Re: guys

[deloptes]

Nate Bargmann wrote:

> * On 2020 01 Feb 13:48 -0600, Joe wrote:
>> Whether you read or not, Debian is getting more complicated, and
>> therefore more problems are occurring.
> 
> True to a point but then we have more complicated systems to deal with,
> including but not limited to, EUFI and Secure Boot, which I use on this
> desktop.  I did not have Secure Boot enabled when I installed Debian
> again in late 2018 but I did enable it later on.  Now the only trip up
> is that I must sign any VirtualBox kernel modules which I have a script
> that will do so.
> 
> Other things have become vastly less complicated as a user and
> administrator of my own systems.  Xorg is completely self configuring.
> Gone are the days of mode lines and crossing your fingers.  USB devices
> are trivial these days as is networking.  Sure, sometimes things go
> astray and trouble ensues, but there is still plenty of help to be found
> when issues arise.  Perhaps it is easier for me having experience from
> the mid '90s forward with Slackware and almost 20 1/2 years experience
> with Debian.
> 
> Some point to systemd as needless complication.  As a user/my own admin,
> I find it really no worse than init scripts (I'm no shell scripting
> guru).  In fact, there are features I do find useful.  I also trust Debian
> to provide a rather well behaved implementation and so far I have found
> that to be true.
> 
> Modern computers are nothing but complicated.  As a result they require
> a lot of complicated software.  Fortunately, much of what I've learned
> over the years is still applicable.  Sure, some things are no longer
> needed and learning new skills is required, but remaining static holds
> no future in today's world.
> 
> - Nate
> 

absolutely agree with you

Re: guys

[Nate Bargmann]

* On 2020 01 Feb 13:48 -0600, Joe wrote:
> Whether you read or not, Debian is getting more complicated, and
> therefore more problems are occurring.

True to a point but then we have more complicated systems to deal with,
including but not limited to, EUFI and Secure Boot, which I use on this
desktop.  I did not have Secure Boot enabled when I installed Debian
again in late 2018 but I did enable it later on.  Now the only trip up
is that I must sign any VirtualBox kernel modules which I have a script
that will do so.

Other things have become vastly less complicated as a user and
administrator of my own systems.  Xorg is completely self configuring.
Gone are the days of mode lines and crossing your fingers.  USB devices
are trivial these days as is networking.  Sure, sometimes things go
astray and trouble ensues, but there is still plenty of help to be found
when issues arise.  Perhaps it is easier for me having experience from
the mid '90s forward with Slackware and almost 20 1/2 years experience
with Debian.

Some point to systemd as needless complication.  As a user/my own admin,
I find it really no worse than init scripts (I'm no shell scripting
guru).  In fact, there are features I do find useful.  I also trust Debian
to provide a rather well behaved implementation and so far I have found
that to be true.

Modern computers are nothing but complicated.  As a result they require
a lot of complicated software.  Fortunately, much of what I've learned
over the years is still applicable.  Sure, some things are no longer
needed and learning new skills is required, but remaining static holds
no future in today's world.

- Nate

-- 

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

Web: https://www.n0nb.us
Projects: https://github.com/N0NB
GPG fingerprint: 82D6 4F6B 0E67 CD41 F689 BBA6 FB2C 5130 D55A 8819



signature.asc
Description: PGP signature

[solved] [sid] xorg keymap modifying

[Grzesiek Sójka]

On 2/1/20 6:11 PM, Siard wrote:

Grzesiek Sójka wrote:

I do have sid running on  tablet pc. I decided to use one of the
physical keys as F11. Its keycode is 125 (the one with windows icon). To
do it on console i put the following in /etc/rc.local:

echo "keycode 125 = F11" | loadkeys

And it works as expected. Unfortunately, I did not find a way to do it
on xorg.


xmodmap is used for that.

$ xmodmap -e "keycode 125 = F11"


Actually, I did try using xmodmap. What I didn't know was that the 
keycodes for console and x-environment may be different. The proper 
keycode is 133 (when running xorg).


Thanks!

--
Greg

Re: cpu frequence

[Gerard ROBIN]

On Sat, Feb 01, 2020 at 08:11:17PM +0100, Jörg-Volker Peetz wrote:
> Date: Sat, 1 Feb 2020 20:11:17 +0100
> From: Jörg-Volker Peetz 
> To: debian-user@lists.debian.org
> Subject: Re: cpu frequence

> Then, take a look at the available governors:
> 
> $ cat /sys/devices/system/cpu/cpufreq/policy?/scaling_available_governors
> 
> or using cpupower, if available. As the name says, "powersave" would be the
> better choice.
> Take a look at https://wiki.debian.org/CpuFrequencyScaling as how to change 
> the
> cpufreq governor permanently even when rebooting. I suppose, you somehow 
> changed
> the default behavior.

Thanks so much I selected performance powersave (I installed
linux-cpupower) and now the frequency oscillates between 800 MHZ
and 2.8 GHz. as with Buster. :)

-- 
Gerard
___
***
Created with Mutt  1.13.2 
under Debian Linux BULLSEYE
***

Re: sudo security flaw

[Reco]

Hi.

On Sun, Feb 02, 2020 at 07:56:48AM +1100, Keith Bainbridge wrote:
> Details can be found in the upstream advisory at
> https://www.sudo.ws/alerts/pwfeedback.html .

It worth noting that to exploit CVE-2019-18634 in Debian one has to
configure sudoers a very specific way.


> For the stable distribution (buster), exploitation of the bug is
> prevented due to a change in EOF handling introduced in 1.8.26.

In another words, stable users are not affected.

Reco

sudo security flaw

[Keith Bainbridge]

Good morning All

I guess most people here are on debian security announce.

But just in case, it refers to a vulnerability in sudo:

Details can be found in the upstream advisory at
https://www.sudo.ws/alerts/pwfeedback.html .


For the oldstable distribution (stretch), this problem has been fixed
in version 1.8.19p1-2.1+deb9u2.

For the stable distribution (buster), exploitation of the bug is
prevented due to a change in EOF handling introduced in 1.8.26.

I'm on 1.8.29.

--
Keith Bainbridge

keith.bainbridge.3...@gmail.com
0447 667 468

Re: guys

[Ben Lavender]

Even though RTM always applies, sometimes things do get more complicated 
with lesser experienced users regardless and sometimes other peoples 
issues aren't always related to FAQ/DOC/MAN etc


This is a users mailing list after all.

On 01/02/2020 19:31, mick crane wrote:

I probably shouldn't post this.
I see all these questions people trying to get their installations to 
work.

It is supposed to be files with documentation what they do.
Is there a reason things seem to get more complicated ?

mick

Re: guys

[Stefan Monnier]

> I probably shouldn't post this.
> I see all these questions people trying to get their installations to work.
> It is supposed to be files with documentation what they do.
> Is there a reason things seem to get more complicated ?

What's with all these "I"s?


Stefan

Re: guys

[Brian]

On Sat 01 Feb 2020 at 19:31:31 +, mick crane wrote:

> I probably shouldn't post this.

Certainly not with the subject line you have given it, no. Adding
"and dolls" would have been led to a lengthier thread with more
discussion. :)

> I see all these questions people trying to get their installations to work.
> It is supposed to be files with documentation what they do.
> Is there a reason things seem to get more complicated ?

Just an illusion. Debian is getting simpler and more accessible.
Having users overcomplicates the process.

-- 
Brian.

Re: guys

[mick crane]

On 2020-02-01 19:35, Reco wrote:

Hi.

On Sat, Feb 01, 2020 at 07:31:31PM +, mick crane wrote:

I probably shouldn't post this.
I see all these questions people trying to get their installations to 
work.

It is supposed to be files with documentation what they do.
Is there a reason things seem to get more complicated ?


I'm not implying anything, but have you ever heard the saying: "if all
else fails, try reading documentation"?
Pretty close to how things are IMO.

I have heard that expression RTFM before, I dunno, I'm probably out of 
order yet feeling is things are getting more complicated when you'd 
think things should be getting simpler.


mick

--
Key ID4BFEBB31

Re: guys

[Joe]

On Sat, 1 Feb 2020 22:35:49 +0300
Reco  wrote:

>   Hi.
> 
> On Sat, Feb 01, 2020 at 07:31:31PM +, mick crane wrote:
> > I probably shouldn't post this.
> > I see all these questions people trying to get their installations
> > to work. It is supposed to be files with documentation what they do.
> > Is there a reason things seem to get more complicated ?  
> 
> I'm not implying anything, but have you ever heard the saying: "if all
> else fails, try reading documentation"?
> Pretty close to how things are IMO.

Whether you read or not, Debian is getting more complicated, and
therefore more problems are occurring.

-- 
Joe

Re: bash, find and error message question

[songbird]

David Wright wrote:
...
> My question would be:
>
> Which directory is your current directory—what does pwd say?
> After all, the error messages say "reading directory '.': Protocol error",
> and not knowing where you are is unsettling for any command.

  in the script i'm using a variable but the text of that
variable is the same as /home/me/pics/camera

  i've got the new version working instead where it uses
lsusb and greps for the camera ID line when it is actually
available.

  now i have to debug the rest of it which was working ok
before i made a few changes.  my eyes are tired, gonna take
a break for a while.  :)  thanks.  :)


  songbird

Re: bash, find and error message question

[songbird]

Reco wrote:
> songbird wrote:
...
>>   i don't care if it is returned, i just wonder why it isn't
>> redirected to /dev/null like i'm asking it to do.
>
> Because you're redirecting stderr of a wrong process.
> This one-liner should do it.
>
> find . -type f -exec printf %.0s. {} + 2>/dev/null | wc -m

  i don't know why, but for some reason i thought the
stderr would be passed through the entire series of
piped commands.  i've hit this before and it never 
clicked.  haha...

  i don't really do a lot of complicated things with find
very often and when i have i haven't had errors (my file 
systems are more predictable than fuser is and i don't do 
things via remote connections or external devices hardly 
ever).

  thanks!  :)  cheers.  off to take break...


  songbird

Re: no mouse in midnight commander

[Joe]

On Sat, 1 Feb 2020 13:46:52 -0500
Frank McCormick  wrote:

> I noticed this morning that my mouse no longer works in MC. When I
> use it, it leaves keycodes on mc's command line at
> the bottom of the screen.
> 
> Because I use Rox-filer to put icons on the desktop, I call
> MC from a small script file in $home which loads an xterm, with some 
> parameters to put up a bigger MC on the screen.
> 
> This is the script at the moment
> 
> xterm -g 140x40+140+20 -fn 10x20 -e mc -x
> 
> I am running Debian Sid fully updated.
> 
> 

How fully? My mc (in Gnome Terminal) looks OK, but I haven't upgraded
today or yesterday, as apt wants to remove 84 packages. There looks to
be some kind of logjam involving KDE and/or Qt.

-- 
Joe

Re: guys

[Reco]

Hi.

On Sat, Feb 01, 2020 at 07:31:31PM +, mick crane wrote:
> I probably shouldn't post this.
> I see all these questions people trying to get their installations to work.
> It is supposed to be files with documentation what they do.
> Is there a reason things seem to get more complicated ?

I'm not implying anything, but have you ever heard the saying: "if all
else fails, try reading documentation"?
Pretty close to how things are IMO.

Reco

guys

[mick crane]

I probably shouldn't post this.
I see all these questions people trying to get their installations to 
work.

It is supposed to be files with documentation what they do.
Is there a reason things seem to get more complicated ?

mick
--
Key ID4BFEBB31

Re: cpu frequence

[Jörg-Volker Peetz]

Then, take a look at the available governors:

$ cat /sys/devices/system/cpu/cpufreq/policy?/scaling_available_governors

or using cpupower, if available. As the name says, "powersave" would be the
better choice.
Take a look at https://wiki.debian.org/CpuFrequencyScaling as how to change the
cpufreq governor permanently even when rebooting. I suppose, you somehow changed
the default behavior.

Regards,
Jörg.

Re: bash, find and error message question

[David Wright]

On Sat 01 Feb 2020 at 13:38:55 (-0500), songbird wrote:
> Reco wrote:
> > On Sat, Feb 01, 2020 at 12:36:28PM -0500, songbird wrote:
> >> the directory is a mount point of a device in limbo.
> >> it is showing up as mounted but it really isn't (the
> >> device is turned off).
> > ...
> >> 
> >>   i just want the answer to be 0 or the number of files.
> >
> > Fuse does not work that way. If a userspace backend cannot perform its
> > function (in this case - a device is disconnected) - it's free to return
> > all kinds of error even on a simple opendir(3).
> >
> > Best you can do is to execute:
> >
> > fusermount -u /home/me/pics/camera
> 
>   i don't care if it is returned, i just wonder why it isn't
> redirected to /dev/null like i'm asking it to do.
> 
> 
> >>   i am using the above in a bash script so i don't want
> >> any error messages coming from the script itself unless
> >> i print them myself.
> >
> > If you need an answer whenever a camera is plugged on or not - why don't
> > you check for the device itself? I.e. lsusb and friends.
> 
>   i check via the mount command but as i've found out it
> isn't really reliable because in fact it is mounting something
> even if it isn't there.
> 
>   as it is i have to tell the camera to be unmounted via gio
> because when i plug the camera in the automounting happens
> even if i tell the system to not do anything.
> 
>   and yes, i'm annoyed at systems that don't do what you tell
> them to do.  the reason i'm writing this whole thing is because
> i dislike all the stupid assumptions and baked in crap that
> the camera manufacturer wants someone to jump through.  no.  the
> USB connection works as it should, leave me alone.  grrr!
> 
>   so i almost have my camera script done but there's these
> last niggling bits i'd like to iron out.
> 
>   thanks for the suggestion about lsusb.  that will at least
> get around the main PITA i'm hitting.  :)

My question would be:

Which directory is your current directory—what does pwd say?
After all, the error messages say "reading directory '.': Protocol error",
and not knowing where you are is unsettling for any command.

Cheers,
David.

Re: bash, find and error message question

[Reco]

On Sat, Feb 01, 2020 at 01:38:55PM -0500, songbird wrote:
> Reco wrote:
> > On Sat, Feb 01, 2020 at 12:36:28PM -0500, songbird wrote:
> >> the directory is a mount point of a device in limbo.
> >> it is showing up as mounted but it really isn't (the
> >> device is turned off).
> > ...
> >> 
> >>   i just want the answer to be 0 or the number of files.
> >
> > Fuse does not work that way. If a userspace backend cannot perform its
> > function (in this case - a device is disconnected) - it's free to return
> > all kinds of error even on a simple opendir(3).
> >
> > Best you can do is to execute:
> >
> > fusermount -u /home/me/pics/camera
> 
>   i don't care if it is returned, i just wonder why it isn't
> redirected to /dev/null like i'm asking it to do.

Because you're redirecting stderr of a wrong process.
This one-liner should do it.

find . -type f -exec printf %.0s. {} + 2>/dev/null | wc -m

Reco

no mouse in midnight commander

[Frank McCormick]

I noticed this morning that my mouse no longer works in MC. When I use 
it, it leaves keycodes on mc's command line at

the bottom of the screen.

Because I use Rox-filer to put icons on the desktop, I call
MC from a small script file in $home which loads an xterm, with some 
parameters to put up a bigger MC on the screen.


This is the script at the moment

xterm -g 140x40+140+20 -fn 10x20 -e mc -x

I am running Debian Sid fully updated.


--
Frank McCormick

Re: bash, find and error message question

[songbird]

Reco wrote:
> On Sat, Feb 01, 2020 at 12:36:28PM -0500, songbird wrote:
>> the directory is a mount point of a device in limbo.
>> it is showing up as mounted but it really isn't (the
>> device is turned off).
> ...
>> 
>>   i just want the answer to be 0 or the number of files.
>
> Fuse does not work that way. If a userspace backend cannot perform its
> function (in this case - a device is disconnected) - it's free to return
> all kinds of error even on a simple opendir(3).
>
> Best you can do is to execute:
>
> fusermount -u /home/me/pics/camera

  i don't care if it is returned, i just wonder why it isn't
redirected to /dev/null like i'm asking it to do.


>>   i am using the above in a bash script so i don't want
>> any error messages coming from the script itself unless
>> i print them myself.
>
> If you need an answer whenever a camera is plugged on or not - why don't
> you check for the device itself? I.e. lsusb and friends.

  i check via the mount command but as i've found out it
isn't really reliable because in fact it is mounting something
even if it isn't there.

  as it is i have to tell the camera to be unmounted via gio
because when i plug the camera in the automounting happens
even if i tell the system to not do anything.

  and yes, i'm annoyed at systems that don't do what you tell
them to do.  the reason i'm writing this whole thing is because
i dislike all the stupid assumptions and baked in crap that
the camera manufacturer wants someone to jump through.  no.  the
USB connection works as it should, leave me alone.  grrr!

  so i almost have my camera script done but there's these
last niggling bits i'd like to iron out.

  thanks for the suggestion about lsusb.  that will at least
get around the main PITA i'm hitting.  :)


  songbird

Re: cron error - date command

[songbird]

Keith Bainbridge wrote:
> Good evening All
>
> I have a niggling problem.Any suggestions, please?
>
> If I run
> echo `date +%d%b%Y` >>   /home/keith/.bash_history
> in a xterm (Mate I think always), I get the expected result: a line in 
> .bash_history reads
>
> 2020Feb01
> followed by
> echo `date +%Y%b%d` >>   /home/keith/.bash_history
>
>
>
> I have this line as a cron-job, and constantly get error messages:
>
> /bin/bash: -c: line 0: unexpected EOF while looking for matching ``'
> /bin/bash: -c: line 1: syntax error: unexpected end of file
>
> and no output to .bash_history
>
>
> Thanks for your suggests

  it looks questionable to me just in that you
may be writing to that file at the same time as
you when you're doing something else that is also
trying to write to that file.

  if you want a log put it in /var/log/.log 
using the rsyslog or in the journal that system-d 
uses via the logging interfaces.

  i haven't looked at cron jobs in ages.

  for putting something into the bash history from the
command line use the history -s command so for your
above idea i would use:

$ history -s `date +%d%b%Y`


  songbird

Re: bash, find and error message question

[Reco]

Hi.

On Sat, Feb 01, 2020 at 12:36:28PM -0500, songbird wrote:
> the directory is a mount point of a device in limbo.
> it is showing up as mounted but it really isn't (the
> device is turned off).
...
> 
>   i just want the answer to be 0 or the number of files.

Fuse does not work that way. If a userspace backend cannot perform its
function (in this case - a device is disconnected) - it's free to return
all kinds of error even on a simple opendir(3).

Best you can do is to execute:

fusermount -u /home/me/pics/camera

>   i am using the above in a bash script so i don't want
> any error messages coming from the script itself unless
> i print them myself.

If you need an answer whenever a camera is plugged on or not - why don't
you check for the device itself? I.e. lsusb and friends.

Reco

bash, find and error message question

[songbird]

  i'm trying to get the following search to give me the
answer and at the same time discard any error messages.

the directory is a mount point of a device in limbo.
it is showing up as mounted but it really isn't (the
device is turned off).

mtab says:
fusectl /sys/fs/fuse/connections fusectl rw,nosuid,nodev,noexec,relatime 0 0
/dev/sdc1 /mb ext4 rw,relatime 0 0
gphotofs /home/me/pics/camera fuse.gphotofs 
rw,nosuid,nodev,relatime,user_id=1000,group_id=1000 0 0

(more on this mount/fuser stuff as a different question)

the directory is empty as it should be
but even a simple ls gives me the protocol error, but
at least the redirect functions as expected.

$ ls
ls: reading directory '.': Protocol error

$ ls 2>/dev/null
total 0


$ find . -type f -exec printf %.0s. {} + | wc -m
find: ‘.’: Protocol error
0
$ find . -type f -exec printf %.0s. {} + | wc -m 2>/dev/null
find: ‘.’: Protocol error
0

# at least this one does what i tell it to do...
$ find . -type f -exec printf %.0s. {} + | wc -m >/dev/null
find: ‘.’: Protocol error



  even if i try to wrap it up in a command or subshell it
still gives me the error message and i don't really care.

  i just want the answer to be 0 or the number of files.

  i am using the above in a bash script so i don't want
any error messages coming from the script itself unless
i print them myself.

  thanks!  :)


  songbird

Re: [sid] xorg keymap modifying

[Siard]

Grzesiek Sójka wrote:
> I do have sid running on  tablet pc. I decided to use one of the 
> physical keys as F11. Its keycode is 125 (the one with windows icon). To 
> do it on console i put the following in /etc/rc.local:
> 
> echo "keycode 125 = F11" | loadkeys
> 
> And it works as expected. Unfortunately, I did not find a way to do it 
> on xorg.

xmodmap is used for that.

$ xmodmap -e "keycode 125 = F11"

Re: cpu frequence

[Gerard ROBIN]

On Sat, Feb 01, 2020 at 04:04:29PM +0100, Jörg-Volker Peetz wrote:
> Date: Sat, 1 Feb 2020 16:04:29 +0100
> From: Jörg-Volker Peetz 
> To: debian-user@lists.debian.org
> Subject: Re: cpu frequence
> X-Spam-Checker-Version: SpamAssassin 3.4.2 (2018-09-13) on bendel.debian.org
> X-Spam-Level: 
> X-Spam-Status: No, score=-10.5 required=4.0
>  tests=FREEMAIL_FORGED_FROMDOMAIN,
>  FREEMAIL_FROM,HEADER_FROM_DIFFERENT_DOMAINS,LDOSUBSCRIBER, LDO_WHITELIST
>  autolearn=unavailable autolearn_force=no version=3.4.2
> 
> what is the outcome of the following command:
> 
> $ cat /sys/devices/system/cpu/cpufreq/policy?/scaling_governor
performance
performance
performance
performance

> Regards,
> Jörg.

-- 
Gerard
_
*
 Created with "mutt 1.13.2-1"
 under Debian Linux BULLSEYE
*

Re: cpu frequence

[Jörg-Volker Peetz]

what is the outcome of the following command:

$ cat /sys/devices/system/cpu/cpufreq/policy?/scaling_governor

Regards,
Jörg.

Re: Can't find a way to preseed keyboard layout for all d-i questions

[Yvan Masson]

Le 31/01/2020 à 21:48, john doe a écrit :

On 1/31/2020 9:38 PM, Yvan Masson wrote:



Le 31/01/2020 à 20:54, john doe a écrit :

On 1/31/2020 8:37 PM, Yvan Masson wrote:

Le 31/01/2020 à 16:50, john doe a écrit :

On 1/31/2020 10:36 AM, Yvan Masson wrote:

Le 29/01/2020 à 18:16, MAS Jean-Louis a écrit :

Le 29/01/2020 à 14:50, Yvan Masson a écrit :


However, before loading preseed.cfg, installer asks for computer
name: I
would like this question to be asked in French and more
importantly to
have the keyboard layout configured in French.

I have tried many boot parameters (layout=fr, layout=fr(latin9),
language=fr, language=fr_FR.UTF-8…) but could not find anything
working.

After answering this question, preseed.cfg is loaded so language and
keyboard layout are properly applied.


It's a well known bug unfortunately

I have asked the same question some time ago on the French
debian-user
list, and frederic boiteux gave me some interesting clues.

You may search for this thread :

"Configurer un clavier français via preseed"

A similar bug was reported with an Hungarian keyboard, without any
fixes
so far.

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=931368

Regards


Thanks for the information. However I just checked by doing a fresh
installation with BIOS PXE boot, di-netboot-assistant and a preseed
file
served by TFTP: locale and keyboard layout are properly applied:
- during install once preseed.cfg is loaded
- after reboot


Yes, because this is delayed after the preseed file is fetched! :)

But if I'm not mistaking, you want to be able to specify the hostname
manually because you have no control over the dhcp server?


Exactly, so I want to preseed most of the question but:
- hostname
- user password

As the hostname and domain are asked before preseed file is fetched, I
tried to use boot options to set the domain (it works) and the locale
(which does not work).

Maybe preseeding has not been designed to use both file and command line
options… I will submit a bug report, please tell me if you think I
shouldn't.



Note that the debian-boot mailing list is responsible for the Debian
installer, before filing a bugreport I would first seak advice there.


Thanks, I will ask there.


Maybe (1) could help you getting what you want.


Thanks, but no, I really don't to modify the installer's initrd :-)


Actually, what you want, as kernel boot parameter could be 'install
'.

I am sorry but is it documented somewhere?




 From (1):

"The "auto" command launches the installation in the automated mode,
where the configuration of hostname, locale and keymap are postponed so
that they can be answered from the preseed file loaded from the network.
You could use "install url=..." but you'd have to answer these questions
manually, regardless of what you have in the preseed config. "

1)
https://wiki.debian.org/DebianInstaller/Preseed#Loading_the_preseeding_file_from_a_webserver


Thanks! I completely missed this… After some experiments, I have what I 
need (manual answer only for the machine name and the user password) 
with the following parameters:


install url=tftp://myserver locale=fr_FR keymap=fr(latin9) 
domain=mydomain ---


Thanks everybody for your help!

Yavn


--
John Doe

[sid] xorg keymap modifying

[Grzesiek Sójka]

Hi there,

I do have sid running on  tablet pc. I decided to use one of the 
physical keys as F11. Its keycode is 125 (the one with windows icon). To 
do it on console i put the following in /etc/rc.local:


echo "keycode 125 = F11" | loadkeys

And it works as expected. Unfortunately, I did not find a way to do it 
on xorg.


Thanks in advance for any help.
Greg

Re: cron error - date command

[Roberto C . Sánchez]

On Sat, Feb 01, 2020 at 09:42:25PM +1100, Keith Bainbridge wrote:
> Good evening All
> 
> I have a niggling problem.Any suggestions, please?
> 
> If I run
> echo `date +%d%b%Y` >>   /home/keith/.bash_history
> in a xterm (Mate I think always), I get the expected result: a line in
> .bash_history reads
> 
> 2020Feb01
> followed by
> echo `date +%Y%b%d` >>   /home/keith/.bash_history
> 
> 
> 
> I have this line as a cron-job, and constantly get error messages:
> 
> /bin/bash: -c: line 0: unexpected EOF while looking for matching ``'
> /bin/bash: -c: line 1: syntax error: unexpected end of file
> 
> and no output to .bash_history
> 
> 
Hi Keith,

I'm not sure what problem you are trying to solve, but let me suggest
that you instead set the HISTTIMEFORMAT environment variable.  For
example

$ export HISTTIMEFORMAT="%Y%b%d "
$ history | tail
25567  2020Jan31 git log --oneline --graph --branches
25568  2020Jan31 gbp buildpackage --git-tag --git-tag-only
25569  2020Jan31 git log --oneline --graph --branches
25570  2020Jan31 for base in /var/cache/pbuilder/base-* ; do sudo cowbuilder 
--update --basepath ${base} ; done
25571  2020Jan31 sudo debootstrap --variant=buildd sid /srv/chroot/sid/ 
http://ftp.us.debian.org/debian/
25572  2020Jan31 schroot -c source:sid -u root -d tmp -- sh -c 'apt-get update 
&& apt-get dist-upgrade -y && apt-get clean'
25573  2020Jan31 sysctl -n hw.logicalcpu
25574  2020Jan31 grep -c ^processor /proc/cpuinfo
25575  2020Jan31 export HISTTIMEFORMAT="%Y%b%d "
25576  2020Jan31 history | tail

Regards,

-Roberto

-- 
Roberto C. Sánchez

Re: Fresh install UEFI debian-10.2.0-amd64-xfce-CD-1 will not boot

[didier . gaumet]

Hello David,

- probably, you were not asked about grub during installation because you did 
not chose the expert installation mode 
- possibly, your problems originate, at least in part, from the fact that 
albeit creating a new ESP (EFI) partition on your USB key, the installer try to 
use also the ESP partition(s) of the disk(s) on the computer used for the 
installation. I would suggest to verify that none of the partitions on the 
disks in the computer (ESP, swap, /boot, etc...) are used during partitionning.

Re: OT red por cable con portal captivo sin trafico interno.

[Ramses]

El 31 de enero de 2020 12:04:37 CET, Antonio Trujillo Carmona 
 escribió:
>El 29/1/20 a las 17:41, Paynalton escribió:
>>
>>
>>
>> El mié., 29 ene. 2020 a las 7:40, Antonio Trujillo Carmona
>> (> >) escribió:
>>
>> El 28/1/20 a las 8:42, Antonio Trujillo Carmona escribió:
>> >     En nuestro hospital tenemos una VLan de gracia para los
>> equipos no
>> > identificados.
>> > Debido al abuso que se hace de esa vlan nos estamos planteando
>> poner un
>> > portal de validación y anular el trafico interno.
>> > No se trata tanto de bloquear o filtrar usuarios como de evitar
>> que se
>> > puedan conectar dispositivos electromédicos u OT a la red, por
>> lo que no
>> > es importante el nivel de seguridad, cualquier elección haría
>que un
>> > dispositivo automático fallara en adquirir red, que es lo que
>> buscamos.
>> > Los conmutadores (HP procurbe) solo admiten 2 de 3 posibles
>> formas de
>> > acceso y tienen activado el filtrado 802.1x y por MAC, por lo
>> que no se
>> > puede activar el acceso web.
>> > ¿Alguna idea?
>> >
>> Muchas gracias a todos por las respuestas.
>>
>> Realmente mi pregunta no iba sobre que portal usar, aunque
>> agradezco los
>> apuntes y los probare, si no por como configurar una red por dhcp
>para
>> que los equipos que estén en la misma red y en el mismo
>conmutador
>> (switch) no se vean entre ellos.
>>
>>
>>
>> Para mantener aislamiento debes usar vlans, manteniendo a la red
>> médica en una vlan y la red pública en otra.
>>
>> El mismo DHCP puede decidir a qué vlan se va cada equipo y qué
>> servicios puede tener.
>>
>> En el gateway de la red pública debes colocar un acceso por proxy
>> controlado por temporizador como te había mencionado en un correo
>> anterior.
>>
>> El DHCP debe entregar la ruta de un wpad para la configuración
>> automática del proxy.
>>
>> Debes tener un servicio web que entregue el archivo wpad, el cual
>> indicará que la salida a internet es a través del proxy.
>>
>> Así, en un caso de uso típico sucede:
>>
>> Caso A:
>>
>> -visitante llega con su teléfono.
>> -visitante se conecta a la red pública abierta
>> -teléfono solicita configuración al DHCP
>> -DHCP entrega configuración de red y una ruta para wpad
>> -visitante intenta entrar a internet
>> -navegador del teléfono consulta el wpad
>> -navegador redirige la petición al proxy
>> -proxy redirige al visitante a una página de error donde le pide
>> contraseña, o una encuesta o la foto de la enfermera Salo en traje de
>baño
>> -visitante interactúa con la página y gana el acceso temporizado
>> -proxy permite el acceso por 15 minutos antes de mostrar de nuevo el
>> pack de verano de la enfermera Salo.
>>
>> Caso B:
>>
>> -llega un interno con un novedoso aparato que no sirve para nada pero
>> que consiguió barato en amazon.
>> -interno conecta el aparato a la red pública por flojera de ir a
>> sistemas a pedir acceso
>> -aparato no tiene navegador, por lo que no puede ver las candentes
>> fotos de la enfermera Salo
>> -aparato no logra conectarse y el interno no tiene más remedio que ir
>> a pedir acceso a la red controlada.
>> -Helpdesk registra macaddress en el DHCP
>> -aparato se vuelve a conectar a la red
>> -DHCP encuentra al aparato en su waitlist y entrega IP de la vlan
>> controlada.
>>  
>
>Muchas gracias por las aportaciones.
>
>Si esto ya lo se, se trata de evitar que llegue un laboratorio e
>instale
>unos equipos sin pasar por el servicio de informática, en la
>actualidad,
>como no están identificados van a parar a la VLAN de gracia donde si se
>ven entre ellos y verifican el funcionamiento con el portatil que lleva
>el instalador, lo dan por bueno y se van, después llaman al servicio de
>informática por que la red del hospital esta mal y no se ven desde los
>ordenadores del hospital, porque ellos han verificado la instalación
>que
>hicieron.
>
>Como soy muy cabezota, tengo que encontrar la solución, me he planteado
>varios caminos:
>
>Investigar a fondo ipv6 que creo que traía algún protocolo para esto
>(forzando a levantar una comunicación punto a punto entre la maquina y
>un nodo centrar donde instalare alguno de los portales que me han
>aconsejado).
>
>Subdividir el rango de la VLAN en redes con prefijo 30, aunque en los
>conmutadores solo admiten una vlan por defecto, esto reduciria de 254 a
>64 los equipos que se permiten concurentemente en la Vlan de gracia,
>pero espero que sea un numero suficiente.
>
>Investigar el tema de la validación web para que "emule" la validación
>MAC y puedan acceder tanto los equipos con MAC autorizada (en sus
>Vlanes
>correspondientes) como los no autorizados a las Vlanes preparadas de la
>forma que he dicho antes.
>
>
>Contare como acaba la cosa, y otra vez muchas gracias por las
>aportaciones.

Antonio, buenos días, 

No veo lo de asignar /30 a los Hosts dentro de la misma VLAN. ¿Podrías 

Re: entirely keyboard based installation ...

[Albretch Mueller]

On 1/30/20, David Wright  wrote:
> I take it that Bil[d?]schirmbild means Screenshot:

 Yes, and sorry I meant "Screenshot"

>  yes, it does appear
> that TABbing doesn't reach that button (including using modifiers).

 Hmm! Does it a bit weird! Is it so by design?

> Would you be able to navigate the cursor over the button by using the
> Accessibility options (2nd item I think) to select 'moving the cursor
> with the arrow keys' (assuming that's available).

 Accessibility options? I will have to try next time I reinstall.
Could you provide me with more information about it? I can't even
understand what you are saying about "Accessibility options" right
during the installation.

 lbrtchx

Re: entirely keyboard based installation ...

[Albretch Mueller]

On 1/30/20, Nektarios Katakis  wrote:
> Did you install any proprietary drivers after installation?

 No, I didn't. The mouse pad didn't work during installation and it
started working right back afterwards

 lbrtchx

Re: entirely keyboard based installation ...

[Albretch Mueller]

On 1/30/20, Charles Curley  wrote:
> On Thu, 30 Jan 2020 11:37:11 -0600
> David Wright  wrote:
>
>> If you file a bug, make sure the model number is specific (which
>> I think yours is).
>
> You should be able to get model, serial and other numbers from
> dmidecode.

dmidecode 3.0
Getting SMBIOS data from sysfs.
SMBIOS 2.6 present.
58 structures occupying 1821 bytes.
Table at 0x000F9D00.

Handle 0x, DMI type 0, 24 bytes
BIOS Information
Vendor: LENOVO
Version: 8RET54WW (1.17 )
Release Date: 10/24/2012
Address: 0xE
Runtime Size: 128 kB
ROM Size: 2048 kB
Characteristics:
PCI is supported
PNP is supported
BIOS is upgradeable
BIOS shadowing is allowed
Boot from CD is supported
Selectable boot is supported
EDD is supported
3.5"/720 kB floppy services are supported (int 13h)
Print screen service is supported (int 5h)
8042 keyboard services are supported (int 9h)
Serial services are supported (int 14h)
Printer services are supported (int 17h)
CGA/mono video services are supported (int 10h)
ACPI is supported
USB legacy is supported
BIOS boot specification is supported
Targeted content distribution is supported
BIOS Revision: 1.17
Firmware Revision: 1.11

Handle 0x0001, DMI type 1, 27 bytes
System Information
Manufacturer: LENOVO
Product Name: 0629W3H
Version: ThinkPad X130e
Serial Number: LRF11R5
UUID: 26913601-5216-11CB-A901-CD01727DCD96
Wake-up Type: Power Switch
SKU Number: Not Specified
Family: ThinkPad X130e
. . .

Handle 0x0005, DMI type 8, 9 bytes
Port Connector Information
Internal Reference Designator: J1900
Internal Connector Type: None
External Reference Designator: Touch Pad
External Connector Type: PS/2
Port Type: Mouse Port

 lbrtchx

Re: cron error - date command

[Keith Bainbridge]

On 1/2/20 10:10 pm, Teemu Likonen wrote:


If you don't want this % effect you need to escape those characters with
backslash:

 echo `date +\%Y\%b\%d`




Thankyou

Worked just as I wanted.



--
Keith Bainbridge

keith.bainbridge.3...@gmail.com
0447 667 468

Re: cron error - date command

[Teemu Likonen]

Keith Bainbridge [2020-02-01T21:42:25+11] wrote:

> echo `date +%Y%b%d` >>   /home/keith/.bash_history

> I have this line as a cron-job, and constantly get error messages:
>
> /bin/bash: -c: line 0: unexpected EOF while looking for matching ``'
> /bin/bash: -c: line 1: syntax error: unexpected end of file
>
> and no output to .bash_history

The "%" character has a special meaning in crontab file. It is
interpreted as the end of line. The rest of the line will be sent to the
program as standard input stream. See crontab(5) manual:

http://man7.org/linux/man-pages/man5/crontab.5.html

If you don't want this % effect you need to escape those characters with
backslash:

echo `date +\%Y\%b\%d`

-- 
///  OpenPGP key: 4E1055DC84E9DFF613D78557719D69D324539450
//  https://keys.openpgp.org/search?q=tliko...@iki.fi
/  https://keybase.io/tlikonen  https://github.com/tlikonen


signature.asc
Description: PGP signature

Re: cron error - date command

[Reco]

Hi.

On Sat, Feb 01, 2020 at 09:42:25PM +1100, Keith Bainbridge wrote:

> I have this line as a cron-job, and constantly get error messages:
> 
> /bin/bash: -c: line 0: unexpected EOF while looking for matching ``'
> /bin/bash: -c: line 1: syntax error: unexpected end of file
> 
> and no output to .bash_history

An output of "crontab -l" would be helpful here.

Reco

cron error - date command

[Keith Bainbridge]

Good evening All

I have a niggling problem.Any suggestions, please?

If I run
echo `date +%d%b%Y` >>   /home/keith/.bash_history
in a xterm (Mate I think always), I get the expected result: a line in 
.bash_history reads


2020Feb01
followed by
echo `date +%Y%b%d` >>   /home/keith/.bash_history



I have this line as a cron-job, and constantly get error messages:

/bin/bash: -c: line 0: unexpected EOF while looking for matching ``'
/bin/bash: -c: line 1: syntax error: unexpected end of file

and no output to .bash_history


Thanks for your suggests


--
Keith Bainbridge

keith.bainbridge.3...@gmail.com
0447 667 468