Subject: Re: HFS+

[peter]

Greg,

From: Greg Wooledge 
Date: Mon, 20 Jan 2025 22:58:51 -0500
> I looked at the package file lists for those two packages, and one of
> the things that looked interesting was "hpfsck".

Thanks for the reply.

# hpfsck -v /dev/sdc2
*** Checking Volume Header:
hpfsck: hpfsck: error writing to medium (Bad file descriptor)

Nothing further.  The Mac system where this drive was used is 
similarly baffled and offers no progress.

Lacking of knowledge of HFS, I wonder whether the file system 
directory is broken.

If the drive is left in this state, the owner will "recycle"; meaning 
it will go to a smelting furnace.  https://en.wikipedia.org/wiki/Smelting

The next obvious step for me is attempt reformatting with FAT or ext.  
Otherwise, wipe, partition and then format.

Thx,... P.
  

 


-- 
VoIP:   +1 604 670 0140
work: https://en.wikibooks.org/wiki/User:PeterEasthope

Re: HFS+

[peter]

From: didier gaumet 
Date: Tue, 21 Jan 2025 10:25:10 +0100
> I would look at theses packages:
> ...
> hfsplus/stable 1.0.4-17 amd64
> ...
> hfsprogs/stable 540.1.linux3-5+b1 amd64
> ...
> hfsutils/stable 3.2.6-15 amd64
> ...
> hfsutils-tcltk/stable 3.2.6-15 amd64

# apt-get install hfsprogs
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
Package hfsprogs is not available, but is referred to by another package.
This may mean that the package is missing, has been obsoleted, or
is only available from another source

E: Package 'hfsprogs' has no installation candidate

The others you mentioned are installed.
Ref. <20250121030050.9c3c420...@bendel.debian.org>

Thanks,  ... P.


-- 
VoIP:   +1 604 670 0140
work: https://en.wikibooks.org/wiki/User:PeterEasthope

Re: Subject: Re: HFS+

[Jeffrey Walton]

On Tue, Jan 21, 2025 at 9:21 AM  wrote:
>
> From: Greg Wooledge 
> Date: Mon, 20 Jan 2025 22:58:51 -0500
> > I looked at the package file lists for those two packages, and one of
> > the things that looked interesting was "hpfsck".
>
> Thanks for the reply.
>
> # hpfsck -v /dev/sdc2
> *** Checking Volume Header:
> hpfsck: hpfsck: error writing to medium (Bad file descriptor)
>
> Nothing further.  The Mac system where this drive was used is
> similarly baffled and offers no progress.

You should probably move this conversation to debian-powerpc.

> Lacking of knowledge of HFS, I wonder whether the file system
> directory is broken.
>
> If the drive is left in this state, the owner will "recycle"; meaning
> it will go to a smelting furnace.  https://en.wikipedia.org/wiki/Smelting
>
> The next obvious step for me is attempt reformatting with FAT or ext.
> Otherwise, wipe, partition and then format.

I suspect things will get worse if you do that. Mac's need the blessed
partition to boot. Blowing it away will ensure the machine does not
boot.

These things are usually discussed on the debian-powerpc mailing list.

Jeff

Correction; HFS+.

[peter]

From: didier gaumet 
Date: Tue, 21 Jan 2025 10:25:10 +0100
> hfsprogs/stable 540.1.linux3-5+b1 amd64

Here /etc/apt/sources.list had non-free-firmware; not non-free. With 
non-free added, hfsprogs installed and the main part of the drive 
reformatted as HFS+.

Now the friend can try it again on the Mac system.

Thanks, ... P.

-- 
VoIP:   +1 604 670 0140
work: https://en.wikibooks.org/wiki/User:PeterEasthope

Re: ISP's router being helpful

[tomas]

On Mon, Jan 20, 2025 at 10:03:23PM +, mick.crane wrote:
> On 2025-01-19 13:58, to...@tuxteam.de wrote:
> > On Sun, Jan 19, 2025 at 12:53:20PM +, mick.crane wrote:
> > > On 2025-01-19 12:01, to...@tuxteam.de wrote:
> > 
> > [...]
> > 
> > > OK. I can ping the PC with roundcube on it by name but "host
> > > " fails to resolve.
> > 
> > Aha. This means that your roundcube (whatever name it has, you
> > didn't tell us yet :)
> 
> I haven't tried to access it directly as yet.
> It is a  link to a soft link to the roundcube install.
> I did it like that to test other installations.

OK, the interesting tidbits are:

 - is the host name you use internally for your Roundcube in
  that URL? Or something else? I guess it's the first
 - if yes: what happens if you ping that host name from exactly
  the same box your browser runs in?

If the ping complains that it can't resolve the name, the problem
is in your resolver setup. If it can, I'd look for the DoH (DNS-
over-http) settings of your browser.

Cheers
-- 
t


signature.asc
Description: PGP signature

Re: HFS+

[didier gaumet]

Hello,

(I have never used HFS/HFS+)

I would look at theses packages:

didier@hp-notebook14:~$ LANG=en-US.UTF-8; apt search hfs
[...]
hfsplus/stable 1.0.4-17 amd64
  Tools to access HFS+ formatted volumes
hfsprogs/stable 540.1.linux3-5+b1 amd64
  mkfs and fsck for HFS and HFS+ file systems
[...]
hfsutils/stable 3.2.6-15 amd64
  Tools for reading and writing Macintosh volumes
hfsutils-tcltk/stable 3.2.6-15 amd64
  Tcl/Tk interfaces for reading and writing Macintosh volumes
[...]

Re: ISP's router being helpful

[Dan Ritter]

mick.crane wrote: 
> On 2025-01-21 08:41, to...@tuxteam.de wrote:
> 
> Not quite sure what is meant by that. The link is
> http://rapunzel.home/roundcubemail
> Are you supposed to be able to ping a service?
> 
> mick@courgette:~$ ping http://rapunzel.home/roundcubemail
> ping: http://rapunzel.home/roundcubemail: Name or service not known
> mick@courgette:~$ ping http://rapunzel.home
> ping: http://rapunzel.home: Name or service not known
> mick@courgette:~$ ping rapunzel.home
> PING rapunzel.home (10.0.0.2) 56(84) bytes of data.
> 64 bytes from rapunzel.home (10.0.0.2): icmp_seq=1 ttl=64 time=0.130 ms


ping sends an ICMP packet to an IP address (which may be
specified or looked up from a domain name) and reports the time
between sending and receiving a reply packet.

It can be blocked by a firewall, dropped by a busy router, or
not answered because the other end has decided not to do so
(rare) or is not working.

A service designated by a URL (such as http://rapunzel.home/)
uses the same domain name lookup to find the IP address, but the
specific protocol (http, here) must be used to make requests.

curl is a tool which can make requests in many protocols. There
are others.

-dsr-

About booting... and installing debian from your iso's

[CJE]

Hi,
This mail is addressing some system level debian comments from recent
changes you've made in iso structure and in installation...

Ok, right now I'm pretty angry... I will surely try to be sensible...
Who: Im from Sweden so... my english is not the best... If I'm too
blurry ask me... I can explain Im a senior systems engineer from Sweden
who's been working
in our defence industry almost 30 years with Gripen (google)... and I love
Linux and debian especially
and Ive been running unix for almost 40 years... and Linux since it came...

My first point: I see that now you have changed your iso structure and
thats a good thing, a long time coming... so good for you...
but then there is your installation software... That's a whole different
story.. My absolute best tip is to delete all that shit you are asking and
replace with LMDE6 installation, hey... its free... so use it... and
learn... The first thing is to install the Debian OS and to have it
running, even if it needs more customization later... Primary: "Get it
fucking running by all means". and then when you can customize it
further... So what do you need to get it running: A user location etc.,
hardware config etc. and installer preferences... Nothing else is needed
see LMDE installation...

My most recent experience trying to install sid was: YOur fucking script
did not work AT ALL... so I had to install LMDE6 FIRST to get a running and
boootable system... then youre intsallation scripts FINALLY understood what
was going on... and lett med also add netinst debian stabel... WHat a
complete catastrophe... I mean i have mad about several hundred debian
installations
so i wonder what the hell youre testing crew is doing... shape up I
know that you can, and what software development is all about SO GET UP AND
GET IT GOING!!!

Now Im at last upp and running sid with compiz and other sw (I love
especially wobbling windows, but also the cube), spotify, pulseffects, and
a tweaked MATE desktop although painfully buggy... my plan is ardour and
music/audio and would very much like to run Debian, my absolutely favourite
OS
So Im satified for now... but Im concerned about what the fuck debian is
doing when you let newcomers mess around with worlds best Unix/Linux system
today!!! Shape up for God sake... copy what Trump is doing currently...
responsibility is not always a bad thing... Elons meritocracy is a good
thing anywhere it is used

/best regards Johan S

Re: ISP's router being helpful

[tomas]

On Tue, Jan 21, 2025 at 10:44:22AM +, mick.crane wrote:
> On 2025-01-21 08:41, to...@tuxteam.de wrote:
> 
> >  - is the host name you use internally for your Roundcube in
> >   that URL? Or something else? I guess it's the first
> >  - if yes: what happens if you ping that host name from exactly
> >   the same box your browser runs in?
> > 
> > If the ping complains that it can't resolve the name, the problem
> > is in your resolver setup. If it can, I'd look for the DoH (DNS-
> > over-http) settings of your browser.
> > 
> > Cheers
> 
> Not quite sure what is meant by that. The link is
> http://rapunzel.home/roundcubemail
> Are you supposed to be able to ping a service?

No -- but the host, rapunzel.home. Sorry for having been unclear.

My intention is that ping does two things: first, resolve the host
name to an IP address, then check connectivity by sending IP packets
and listening to the answers.

If ping can resolve rapunzel.home, but the browser is not, then it's
the browser you'll have to have a stern word with...

> 
> mick@courgette:~$ ping http://rapunzel.home/roundcubemail
> ping: http://rapunzel.home/roundcubemail: Name or service not known
> mick@courgette:~$ ping http://rapunzel.home
> ping: http://rapunzel.home: Name or service not known
> mick@courgette:~$ ping rapunzel.home
> PING rapunzel.home (10.0.0.2) 56(84) bytes of data.
> 64 bytes from rapunzel.home (10.0.0.2): icmp_seq=1 ttl=64 time=0.130 ms

... which is actually the case.

> I know it's probably a mess. I stopped reading when it worked.
> mick

It's not a mess. It's the browser (which actually is worse than a mess.
I'm still in search of polite words which would be adequate).

Most probably it's doing DNS-over-HTTP, thus bypassing your resolver
settings. I think here [1] they explain how to disable that. At the
same time they'll tell you that it's the best thing since sliced bread,
steeped as they are in ad industry's groupthink. Rubbing salt in the
wound -- surveillance capitalism is petty like that.

Cheer
-- 
t


signature.asc
Description: PGP signature

Re: HFS+

[David Wright]

On Tue 21 Jan 2025 at 06:26:17 (-0700), pe...@easthope.ca wrote:
> From: didier gaumet 
> Date: Tue, 21 Jan 2025 10:25:10 +0100
> > I would look at theses packages:
> > ...
> > hfsplus/stable 1.0.4-17 amd64
> > ...
> > hfsprogs/stable 540.1.linux3-5+b1 amd64
> > ...
> > hfsutils/stable 3.2.6-15 amd64
> > ...
> > hfsutils-tcltk/stable 3.2.6-15 amd64
> 
> # apt-get install hfsprogs
> Reading package lists... Done
> Building dependency tree... Done
> Reading state information... Done
> Package hfsprogs is not available, but is referred to by another package.
> This may mean that the package is missing, has been obsoleted, or
> is only available from another source
> 
> E: Package 'hfsprogs' has no installation candidate
> 
> The others you mentioned are installed.
> Ref. <20250121030050.9c3c420...@bendel.debian.org>

You need non-free. I think buster was the last time it
was in the main distribution.

Cheers,
David.

Re: HFS+

[Jeffrey Walton]

On Tue, Jan 21, 2025 at 9:42 AM  wrote:
>
> From: didier gaumet 
> Date: Tue, 21 Jan 2025 10:25:10 +0100
> > I would look at theses packages:
> > ...
> > hfsplus/stable 1.0.4-17 amd64
> > ...
> > hfsprogs/stable 540.1.linux3-5+b1 amd64
> > ...
> > hfsutils/stable 3.2.6-15 amd64
> > ...
> > hfsutils-tcltk/stable 3.2.6-15 amd64
>
> # apt-get install hfsprogs
> Reading package lists... Done
> Building dependency tree... Done
> Reading state information... Done
> Package hfsprogs is not available, but is referred to by another package.
> This may mean that the package is missing, has been obsoleted, or
> is only available from another source
>
> E: Package 'hfsprogs' has no installation candidate

It looks like the package is part of non-free. See
.

I _think_ hfsprogs is part of non-free because Apple did not release
the source code under an appropriate license, but I might be mistaken.

I usually see hfsprogs discussed on the debian-powerpc mailing list.
You might try using a powerpc-based iso on an old PowerMac. John Paul
Adrian Glaubitz does a lot of work with the tools.

> The others you mentioned are installed.
> Ref. <20250121030050.9c3c420...@bendel.debian.org>

Jeff

Re: ISP's router being helpful

[tomas]

On Tue, Jan 21, 2025 at 07:17:53AM -0500, Greg Wooledge wrote:

[...]

> E.g. 

Ah, oh -- I overlooked (or forgot) that OP's brower is Vivaldi.

Cheers
-- 
t


signature.asc
Description: PGP signature

Re: ISP's router being helpful

[Max Nikulin]

On 19/01/2025 17:21, mick.crane wrote:

The other day changed the ISP's (Sky) router to have fibre connection.


Maybe the previous router was configured to serve .home DNS zone.

If vivaldi uses the same settings page as chromium than you may try to 
disable "secure DNS"


chrome://settings/security?search=dns

Re: sunscribe

[Róbert László]

Yes, I confirm.

Andrew M.A. Cater  ezt írta (időpont: 2025. jan. 21.,
K, 22:30):

> [Also sent off list]
>
> On Tue, Jan 21, 2025 at 10:06:21PM +0100, Róbert László wrote:
> > help sucsribe
>
> Hi,
>
> You will need to send a message to
>
> debian-user-requ...@lists.debian.org with a subject of Subscribe
>
> The automated system will send a receipt asking for confirmation of
> the subscription with a subject something like CONFIRMxxx
> and text asking you to confirm that you issued a subscription
> request.
>
> A simple reply to that mail message should confirm subscription
> to the list
>
> Hope this helps,
>
> Andrew Cater
> (amaca...@debian.org)
>

Re: Backports (bookworm), upgrade to all x for kernel 6.12.x, but not upgrade to 6.13

[George at Clug]

Marco,

I apologise, but I do not understand what it is you want to achieve or what it 
is that you are asking.

Can you please give more explanation?

You said: "not leave the 6.12 (upstream LTS) branch and not upgrade to some 
higher kernel version like 6.13 when they would also become available in 
backports"

What do you think "backports" are ?  

Have you already installed any backports?

Why do you not want to install backports?

Are you saying that you only want to use Debian Bookworm packages?

And not any packages that are from new releases (e.g. Trixie or Forky) that 
have been recompiled to be used in Debian Bookworm ?  (which I believe is what 
Backports are).

If this is correct, then would removing any mention of Backports from your APT 
sources.list be what you want to do?  (others on this list could comment if my 
suggestion is correct nor not).

I believe (as with previous discussions about rsync on Debian User), that 
sometimes security updates are: "packages taken from the next Debian release 
(called "testing"), adjusted and recompiled for usage on Debian stable." 

Hence if you want to stay secure and use security updates, there will be times 
when you will be using a package from Testing, ... but it will be "adjusted and 
recompiled for usage on Debian stable", as are all backports, which is why we 
use backports and not just download a package from Testing and install it.

Maybe you are concerned that your LTS Bookworm will one day just become Debian 
Trixie?  This is not how Debian works. As long as your sources in APT stays 
"Bookworm", you should always been at the Bookworm release.  To change 
"Bookworm" to "Trixie", you have to manually change your sources and do a full 
upgrade, then this will change your Debian release.

https://backports.debian.org/


https://www.debian.org/releases/
Index of releases
Version Code Name   Release DateEnd of Life (EOL)   EOL LTS 
EOL ELTSStatus
14  Forky   TBA TBA TBA 
TBA Codename announced
13  Trixie  TBA TBA TBA 
TBA "testing" — no release date has been set
12  Bookworm2023-06-10  2026-06-10  
2028-06-30  2033-06-30  Current "stable" release

https://en.wikipedia.org/wiki/Debian_version_history



https://wiki.debian.org/SourcesList
Example sources.list

If you need the contrib, non-free and non-free-firmware components, add contrib 
non-free non-free-firmware after main. For example, for Debian 12/Bookworm:

deb https://deb.debian.org/debian bookworm main contrib non-free 
non-free-firmware
deb-src https://deb.debian.org/debian bookworm main contrib non-free 
non-free-firmware

deb https://security.debian.org/debian-security bookworm-security main contrib 
non-free non-free-firmware
deb-src https://security.debian.org/debian-security bookworm-security main 
contrib non-free non-free-firmware

deb https://deb.debian.org/debian bookworm-updates main contrib non-free 
non-free-firmware
deb-src https://deb.debian.org/debian bookworm-updates main contrib non-free 
non-free-firmware

If you need the Backports, contrib, and non-free components, add 
bookworm-backports lines. For example, for Debian 12/Bookworm:

deb https://deb.debian.org/debian bookworm-backports main contrib non-free 
non-free-firmware
deb-src https://deb.debian.org/debian bookworm-backports main contrib non-free 
non-free-firmware

# nano /etc/apt/sources.list

I usually update my Bookworm by one of two methods, which do much the same 
thing:

# apt update && apt update && apt full-upgrade --autoremove -y && apt 
autoremove && apt clean

or

# apt update 
# apt list --upgradable
# apt full-upgrade --autoremove -y 
# systemctl reboot
# apt autoremove
# apt clean

I hope some of this helps, sorry I am confused about what you wanted to ask.

George.

On Wednesday, 22-01-2025 at 07:17 Marco Möller wrote:
> Hello community!
> Could you please share with me, or point me to, a howto or receipt for 
> applying all upgrades to future kernel 6.12.x versions to appear in 
> Bookworm Backports when doing "apt update && apt upgrade", but to not 
> leave the 6.12 (upstream LTS) branch and not upgrade to some higher 
> kernel version like 6.13 when they would also become available in backports?
> Thanks a lot in advance! Talby.
> 
> 

Re: Backports (bookworm), upgrade to all x for kernel 6.12.x, but not upgrade to 6.13

[Marco Möller]

On 1/22/25 00:10, George at Clug wrote:

I apologise, but I do not understand what it is you want to achieve or what it 
is that you are asking.

Can you please give more explanation?


I want to install the currently highest version of kernel 6.12 from 
bookworm-backports to my Bookworm. Upon some "apt update && apt upgrade" 
I want this kernel to become upgraded whenever in backports becomes 
available a higher version of kernel 6.12, like having 6.12.9 and 
getting 6.12.10. But I do not want this upgrade to step up to the 6.13 
versions.
For comfortably running the upgrades in an unattended way, I expect that 
I might need some pinning to allow all newer 6.12.x versions to become 
drawn in when available, but to not allow any higher kernel version than 
6.12.x, so not allowing any 6.13 to become drawn in.


It seems to obvious to me, that I should not install from backports the 
package "linux-image-amd64", because this I expect to always point to 
the very newest kernel version, but not restricting to the 6.12 series 
of kernel versions. If there would be a "linux-image-6.12" package, then 
I would be done. Easy. But such "linux-image-6.12" meta-package is not 
available in bookworm-backports.
So, I either install now manually a particular version like 
"/bookworm-backports/kernel/linux-image-6.12.9+bpo-amd64-unsigned" and 
put it on hold and then manually check if some 
"linux-image-6.12.10+bpo-amd64-unsigned" (or higher 
6.12.maintenanceVersion number) would arrive in the backports repository 
and then upgrade to it manually, repeating this manual procedure very 
frequently, or I find out how to configure some pinning or whatever for 
automation of some 6.12 series restricted upgrades.


My question is what do I have to configure for letting apt upgrade the 
backported kernel 6.12 automatically to the newest linux-image-6.12.x 
version without upgrading beyond any 6.12 kernel, for instance not 
automatically upgrading to any backported kernel 6.13.x version.

Backports (bookworm), upgrade to all x for kernel 6.12.x, but not upgrade to 6.13

[Marco Möller]

Hello community!
Could you please share with me, or point me to, a howto or receipt for 
applying all upgrades to future kernel 6.12.x versions to appear in 
Bookworm Backports when doing "apt update && apt upgrade", but to not 
leave the 6.12 (upstream LTS) branch and not upgrade to some higher 
kernel version like 6.13 when they would also become available in backports?

Thanks a lot in advance! Talby.

Re: Backports (bookworm), upgrade to all x for kernel 6.12.x, but not upgrade to 6.13

[Andrew M.A. Cater]

On Tue, Jan 21, 2025 at 09:17:52PM +0100, Marco Möller wrote:
> Hello community!
> Could you please share with me, or point me to, a howto or receipt for
> applying all upgrades to future kernel 6.12.x versions to appear in Bookworm
> Backports when doing "apt update && apt upgrade", but to not leave the 6.12
> (upstream LTS) branch and not upgrade to some higher kernel version like
> 6.13 when they would also become available in backports?
> Thanks a lot in advance! Talby.
>

Hi

It is very likely that Debian will stick with 6.12 thoughout the lifetime
of Debian Trixie / Debian 13 so you should be fine. Since 6.12 is the LTS
version, it's unlikely that we'll move forward from it (though I'm not
the Release Team).

Hope this helps,

Andrew Cater
(amaca...@debian.org) 

sunscribe

[Róbert László]

help sucsribe

Re: ISP's router being helpful

[tomas]

On Tue, Jan 21, 2025 at 10:38:51PM +0700, Max Nikulin wrote:
> On 19/01/2025 17:21, mick.crane wrote:
> > The other day changed the ISP's (Sky) router to have fibre connection.
> 
> Maybe the previous router was configured to serve .home DNS zone.

Judging by the other symptoms (ping working, browser not) the resolver
in the box is OK (the .home names are resolved in /etc/hosts). So I'm
nearly convinced it's the browser doing DoH (it's being rolled out at
the moment).

> If vivaldi uses the same settings page as chromium than you may try to
> disable "secure DNS"
> 
> chrome://settings/security?search=dns

Exactly.

Cheers
-- 
t


signature.asc
Description: PGP signature

Re: About booting... and installing debian from your iso's

[John Hasler]

Do you understand what Sid is?  If you want to run it install Stable and
upgrade.  That's the way to "Get it fucking running by all means and
then when you can customize it further".

CJE writes:
> so i wonder what the hell youre testing crew is doing.

Testing "Testing", not "Unstable".
-- 
John Hasler 
j...@sugarbit.com
Elmwood, WI USA

Re: About booting... and installing debian from your iso's

[Andrew M.A. Cater]

On Tue, Jan 21, 2025 at 12:20:44PM +0100, CJE wrote:
> Hi,
> This mail is addressing some system level debian comments from recent
> changes you've made in iso structure and in installation...
> 

Hello Johan,


> and Ive been running unix for almost 40 years... and Linux since it came...
> 

OK - so you're an experienced user of Linux.

> My first point: I see that now you have changed your iso structure and
> thats a good thing, a long time coming... so good for you...
> but then there is your installation software... That's a whole different
> story.. My absolute best tip is to delete all that shit you are asking and
> replace with LMDE6 installation, hey... its free... so use it... and
> learn... The first thing is to install the Debian OS and to have it
> running, even if it needs more customization later... Primary: "Get it
> fucking running by all means". and then when you can customize it
> further... So what do you need to get it running: A user location etc.,
> hardware config etc. and installer preferences... Nothing else is needed
> see LMDE installation...
> 

They possibly do things differently over at Linux Mint. There are various
ways to invoke the standard Debian installer and there is also the option
to install from a Debian live media image. I can understand frustration
but please tone down your expression, maybe?

> My most recent experience trying to install sid was: YOur fucking script
> did not work AT ALL... so I had to install LMDE6 FIRST to get a running and
> boootable system... then youre intsallation scripts FINALLY understood what
> was going on... and lett med also add netinst debian stabel... WHat a
> complete catastrophe... I mean i have mad about several hundred debian
> installations
> so i wonder what the hell youre testing crew is doing... shape up I
> know that you can, and what software development is all about SO GET UP AND
> GET IT GOING!!!
> 

Which installer please and downloaded from where? We don't routinely provide
a sid installer image. An install from Testing daily image has just
succeeded for me, for example. Specifics, please.

If you installed LMDE6 first - were you working on a pure Debian system
at this point?

As someone who routinely does Debian point release testing - we do test
aspects of the Debian stable installer and also recently tested the
Debian Trixie Alpha 1 release. The testing methodology and results are
routinely noted in the Debian wiki.

> Now Im at last upp and running sid with compiz and other sw (I love
> especially wobbling windows, but also the cube), spotify, pulseffects, and
> a tweaked MATE desktop although painfully buggy... my plan is ardour and
> music/audio and would very much like to run Debian, my absolutely favourite
> OS
> So Im satified for now... but Im concerned about what the fuck debian is
> doing when you let newcomers mess around with worlds best Unix/Linux system
> today!!! Shape up for God sake... copy what Trump is doing currently...
> responsibility is not always a bad thing... Elons meritocracy is a good
> thing anywhere it is used
> 

I should suggest that "newcomers" is possibly not an adequate description
for people with ~30 years worth of experience with Debian in the Debian images
team. The political references are also probably off-topic and accordingly
unwelcome - please refer to the monthly Debian-user FAQ.

> /best regards Johan S

With every good wish, as ever,

Andrew Cater
(amaca...@debian.org)

Re: ISP's router being helpful

[George at Clug]

On Wednesday, 22-01-2025 at 02:13 to...@tuxteam.de wrote:
> On Tue, Jan 21, 2025 at 07:17:53AM -0500, Greg Wooledge wrote:
> 
> [...]
> 
> > E.g. 
> 
> Ah, oh -- I overlooked (or forgot) that OP's brower is Vivaldi.

https://help.vivaldi.com/desktop/privacy/privacy-settings/
Resolving Navigation Errors with Google DNS Service
The Google DNS (Domain Name System) Service will give you the IP address of the 
website you want to visit in case your ISP (Internet Service Provider) fails to 
do so properly (in some cases the ISP may have faulty DNSs).

The Google DNS server addresses (8.8.8.8 and 8.8.4.4) have been hardcoded into 
Chromium, on which Vivaldi is built. The setting is enabled by default, but can 
be disabled in Settings > Privacy and Security.

To set a custom DNS server in Vivaldi, you can do the following: 

Open the Vivaldi menu
Select Settings
Select Privacy and Security
Select Use secure DNS
Select Custom
Enter the URL of your preferred DNS provider 

You can also choose from other DNS providers, such as Google, CleanBrowsing, 
OpenDNS, Cloudflare, or NextDNS.

https://dns.sb/doh/vivaldi/



> 
> Cheers
> -- 
> t
> 

Re: sunscribe

[Andrew M.A. Cater]

[Also sent off list]

On Tue, Jan 21, 2025 at 10:06:21PM +0100, Róbert László wrote:
> help sucsribe

Hi,

You will need to send a message to

debian-user-requ...@lists.debian.org with a subject of Subscribe

The automated system will send a receipt asking for confirmation of
the subscription with a subject something like CONFIRMxxx
and text asking you to confirm that you issued a subscription 
request.

A simple reply to that mail message should confirm subscription
to the list

Hope this helps,

Andrew Cater
(amaca...@debian.org)

Re: Backports (bookworm), upgrade to all x for kernel 6.12.x, but not upgrade to 6.13

[Marco Möller]

On 1/21/25 21:39, Andrew M.A. Cater wrote:

On Tue, Jan 21, 2025 at 09:17:52PM +0100, Marco Möller wrote:

Hello community!
Could you please share with me, or point me to, a howto or receipt for
applying all upgrades to future kernel 6.12.x versions to appear in Bookworm
Backports when doing "apt update && apt upgrade", but to not leave the 6.12
(upstream LTS) branch and not upgrade to some higher kernel version like
6.13 when they would also become available in backports?
Thanks a lot in advance! Talby.



Hi

It is very likely that Debian will stick with 6.12 thoughout the lifetime
of Debian Trixie / Debian 13 so you should be fine. Since 6.12 is the LTS
version, it's unlikely that we'll move forward from it (though I'm not
the Release Team).

Hope this helps,

Andrew Cater
(amaca...@debian.org)



I am on current "stable" Bookworm (Debian 12) and would like to stay 
with it until Debian 13 is released as the new "stable", but I would 
like to already now take advantage of some features coming with kernel 6.12.
That's why I wonder on how to best catch the upgrades for 6.12, if some 
would become available. Watching how kernel versions usually appear in 
backports I am of good hope to find future 6.12.x versions to appear 
there. The kernel backport team is doing a great job, my special thanks 
to them!
Then, when Trixie is officially out and most likely coming with 6.12, my 
upgrade from Bookworm with backports kernel 6.12 to Trixie with kernel 
6.12 should be less of trouble than having my Bookworm with a backports 
kernel already being some 6.13 (or even higher) version.

Re: ISP's router being helpful

[Greg Wooledge]

On Tue, Jan 21, 2025 at 10:44:22 +, mick.crane wrote:
> On 2025-01-21 08:41, to...@tuxteam.de wrote:
> > If the ping complains that it can't resolve the name, the problem
> > is in your resolver setup. If it can, I'd look for the DoH (DNS-
> > over-http) settings of your browser.

> mick@courgette:~$ ping rapunzel.home
> PING rapunzel.home (10.0.0.2) 56(84) bytes of data.
> 64 bytes from rapunzel.home (10.0.0.2): icmp_seq=1 ttl=64 time=0.130 ms

OK, that's good.  So, now you'll need to look at your browser configs.
Going to the list archives and finding the first message in this thrread,
it looks like your browser is "vivaldi".  I've never used that one, so
I have no idea where to look for its name resolution configuration,
other than "your favorite web search engine".

E.g. 

Re: About booting... and installing debian from your iso's

[Joe]

On Tue, 21 Jan 2025 12:20:44 +0100
CJE  wrote:


> 
> My most recent experience trying to install sid was: YOur fucking
> script did not work AT ALL... so I had to install LMDE6 FIRST to get
> a running and boootable system... then youre intsallation scripts
> FINALLY understood what was going on... and lett med also add netinst
> debian stabel... WHat a complete catastrophe... I mean i have mad
> about several hundred debian installations

A note: the direct sid installation is fairly recent and is a bit
experimental. Debian Stable is the usual version installed, and the
Stable installer is well-tested. There should be no problems even for
newcomers to Linux.

I use sid on my workstation, and have had to reinstall a few times over
the years. I would never consider using the sid installer, I'm aware it
may not work well, and I want to end up with a reliable system, or at
least as reliable as sid can be.

I would always install a minimal system (no GUI) from the current Stable
netinstall, then upgrade to sid. If the installation is minimal you can
usually get away with a direct upgrade to sid, though it is recommended
to go via Testing. Then I would add what I needed.

I understand your annoyance, but sid is a testing distribution which
may have problems at any time, and the installer can hardly be expected
to be better than the OS it installs. I'm sure in the past you have
installed sid via Stable, and that is still the most reliable method.

-- 
Joe

Re: ISP's router being helpful

[mick.crane]

On 2025-01-21 08:41, to...@tuxteam.de wrote:


 - is the host name you use internally for your Roundcube in
  that URL? Or something else? I guess it's the first
 - if yes: what happens if you ping that host name from exactly
  the same box your browser runs in?

If the ping complains that it can't resolve the name, the problem
is in your resolver setup. If it can, I'd look for the DoH (DNS-
over-http) settings of your browser.

Cheers


Not quite sure what is meant by that. The link is 
http://rapunzel.home/roundcubemail

Are you supposed to be able to ping a service?

mick@courgette:~$ ping http://rapunzel.home/roundcubemail
ping: http://rapunzel.home/roundcubemail: Name or service not known
mick@courgette:~$ ping http://rapunzel.home
ping: http://rapunzel.home: Name or service not known
mick@courgette:~$ ping rapunzel.home
PING rapunzel.home (10.0.0.2) 56(84) bytes of data.
64 bytes from rapunzel.home (10.0.0.2): icmp_seq=1 ttl=64 time=0.130 ms

I know it's probably a mess. I stopped reading when it worked.
mick

Re: ISP's router being helpful

[Max Nikulin]

On 21/01/2025 23:31, to...@tuxteam.de wrote:

On Tue, Jan 21, 2025 at 10:38:51PM +0700, Max Nikulin wrote:

On 19/01/2025 17:21, mick.crane wrote:

The other day changed the ISP's (Sky) router to have fibre connection.

Maybe the previous router was configured to serve .home DNS zone.

Judging by the other symptoms (ping working, browser not) the resolver
in the box is OK (the .home names are resolved in /etc/hosts).


I have read somewhere that chromium may read /etc/resolv.conf and send 
requests to the specified servers directly bypassing /etc/nsswitch.conf. 
(The statement needs verification.)


So ping uses name resolution similar to

getent hosts rapunzel.home

while vivaldi might do (if DoH is disabled) equivalent of

dig rapunzel.home

An attempt to guess if the router uses some zone

dig -x 10.0.0.2
dig @10.0.0.1 -x 10.0.0.2

The latter is to bypass local DNS caching that might do other fancy 
things (systemd-resolved, dnsmasq started by NetworkManager).


At least cloudflare and google do not resolve the host name (other DoH 
provider may behave in a different way)


curl -s -H 'accept: application/dns-json' \
   'https://dns.google/resolve?name=rapunzel.home' | jq .
curl -s -H 'accept: application/dns-json' \
   'https://1.1.1.1/dns-query?name=rapunzel.home' | jq .
{
  "Status": 3,
  "TC": false,
  "RD": true,
  "RA": true,
  "AD": true,
  "CD": false,
  "Question": [
{
  "name": "rapunzel.home",
  "type": 1
}
  ],
  "Authority": [
{
  "name": "",
  "type": 6,
  "TTL": 86400,
  "data": "a.root-servers.net. nstld.verisign-grs.com. 2025012102 
1800 900 604800 86400"

}
  ]
}

Re: Backports (bookworm), upgrade to all x for kernel 6.12.x, but not upgrade to 6.13

[Max Nikulin]

On 22/01/2025 03:17, Marco Möller wrote:
Could you please share with me, or point me to, a howto or receipt for 
applying all upgrades to future kernel 6.12.x versions to appear in 
Bookworm Backports when doing "apt update && apt upgrade", but to not 
leave the 6.12 (upstream LTS) branch and not upgrade to some higher 
kernel version like 6.13 when they would also become available in 
backports?


This way you may lost some important security update.

If the risk of shooting your foot is accepted then read 
apt_preferences(5). Pinning version to a regex likely would allow to do 
what you wish. To debug use "apt policy" with package name argument.

Debian 12 and AMD Ryzen 9 9950X with X870-E chipset kernel support

[Johannes Krottmayer]

Hi!

I have planned to upgrade my (very old) system. The new system will be
a mainboard with X870-E chipset and a AMD Ryzen 9 9950X3D (the CPU isn't
currently available), so have written AMD Ryzen 9 9950X in the mail
subject.

Has somebody such a combination running with Debian 12? Are there any
troubles with the default kernel? Or, is a newer kernel required to run
this combination (9950X/X870-E)?

Thanks in advance!

Re: sunscribe

[Jeffrey Walton]

On Tue, Jan 21, 2025 at 8:02 PM Róbert László  wrote:
>
> help subscribe

Re: ISP's router being helpful

[mick.crane]

On 2025-01-21 21:34, George at Clug wrote:


To set a custom DNS server in Vivaldi, you can do the following:

Open the Vivaldi menu
Select Settings
Select Privacy and Security
Select Use secure DNS
Select Custom
Enter the URL of your preferred DNS provider


Seems this option not available through vivaldi settings but via
chrome:settings/security
mick

Re: About booting... and installing debian from your iso's

[Jeffrey Walton]

On Tue, Jan 21, 2025 at 5:12 PM John Hasler  wrote:
>
> Do you understand what Sid is?  If you want to run it install Stable and
> upgrade.  That's the way to "Get it fucking running by all means and
> then when you can customize it further".
>
> CJE writes:
> > so i wonder what the hell you're testing crew is doing.
>
> Testing "Testing", not "Unstable".

This may help CJE:  and
. The Debian Unstable page
states the following under the Installation section:

   * Use the current "stable" installer to install a minimal stable
system (recommended).
  * Change your apt sources to point to "unstable".
  * Run apt update and apt full-upgrade.

Jeff

Re: Backports (bookworm), upgrade to all x for kernel 6.12.x, but not upgrade to 6.13

[Greg Wooledge]

On Wed, Jan 22, 2025 at 01:00:49 +0100, Marco Möller wrote:
> On 1/22/25 00:10, George at Clug wrote:
> > I apologise, but I do not understand what it is you want to achieve or what 
> > it is that you are asking.
> > 
> > Can you please give more explanation?
> 
> I want to install the currently highest version of kernel 6.12 from
> bookworm-backports to my Bookworm.

The fundamental question is why you want to do this.  Is your hardware
not supported by the bookworm kernel?  Do you *need* this backported
kernel?

Backports are a set of packages that are created by hand, whenever
someone feels like doing so.  When you install one, you are cherry-picking
it from the available set, manually.

The backported kernel that you install will not necessarily receive any
security updates, or bug fixes, or anything.

> Upon some "apt update && apt upgrade" I
> want this kernel to become upgraded whenever in backports becomes available
> a higher version of kernel 6.12, like having 6.12.9 and getting 6.12.10. But
> I do not want this upgrade to step up to the 6.13 versions.

Then you're in luck.  That's how backports work.  Well, sort of.

A kernel package has its own name which contains the kernel ABI
(Application Binary Interface) identifier, which determines whether
modules can be loaded into it without recompilation.  For example,
the stable kernel in bookworm currently has the ABI "6.1.0-30-" in
its package name.

If a minor security update were to be released in which changes are
backported to the 6.1.0-30- bookworm kernel, then the new kernel
package might also have the same ABI, and in this case the kernel
package can just be upgraded to the new version.

However, that's pretty rare in practice.  Usually kernel security
updates come in waves, with a lot of upstream bug fixes grouped in
with them, and the whole mess requires a new ABI version.  So, the
next bookworm kernel might have the 6.1.0-31- ABI.  In that case,
the kernel package can't just be upgraded.  A whole *new* kernel
package with a new name has to be installed, and an old kernel
package removed (or left in place).

The kernels in backports will work the same way.  If you install
backported kernel linux-image-6.12.0-5-amd64 and there's a big
security update which causes linux-image-6.12.0-6-amd64 to be
created, then your kernel won't be upgraded automatically.  You'll
have to install the new kernel (and its corresponding headers, if
you have any DKMS modules to build).

As the user of a cherry-picked backported kernel, it'll be your own
responsibility to be aware of when a new kernel exists, and to
install it if you wish.

You *definitely* won't get a linux-image-6.13.0-1-amd64 package
automatically.

Of course, you should still run a stable kernel if you can.  It's not
yet clear to me whether you're trying to use a backported kernel because
you *need* it, or because it has a higher number and you think higher
numbers are better.

Re: ISP's router being helpful

[tomas]

On Wed, Jan 22, 2025 at 09:48:30AM +0700, Max Nikulin wrote:
> On 21/01/2025 23:31, to...@tuxteam.de wrote:
> > On Tue, Jan 21, 2025 at 10:38:51PM +0700, Max Nikulin wrote:
> > > On 19/01/2025 17:21, mick.crane wrote:
> > > > The other day changed the ISP's (Sky) router to have fibre connection.
> > > Maybe the previous router was configured to serve .home DNS zone.
> > Judging by the other symptoms (ping working, browser not) the resolver
> > in the box is OK (the .home names are resolved in /etc/hosts).
> 
> I have read somewhere that chromium may read /etc/resolv.conf and send
> requests to the specified servers directly bypassing /etc/nsswitch.conf.
> (The statement needs verification.)

Oh, goody.

[interesting stuff snipped]


> At least cloudflare and google do not resolve the host name (other DoH
> provider may behave in a different way)

But most probably not in the way the OP expects, since they can't read
(?) their local /etc/hosts...

Cheers
-- 
t


signature.asc
Description: PGP signature