Re: ICMP router advertisement (ipv4)

[Anssi Saari]

Dan Ritter  writes:

> Tunnelbroker.net can use a dynamic update feature so that the
> static IPv6 address is only momentarily interrupted when the
> underlying v4 address is changed out. This is quite reliable if
> you hook it to the the DHCP client's post-up function.

I don't understand the relevance. Does tunnelbroker.net offer some other
kind of tunnels than 6to4 which, AFAICT, don't work without a public
IPv4 address? Which the OP does not have due to his IPv4-in-IPv6
connection. Same goes for 6in4 tunnels. I can't seem to find out what
Hurricane Electric offers without sigining up though but at least
https://forums.he.net/index.php?topic=4195.0 from last year seems to
indicate WireGuard tunnels are not supported.

If "public IPv4 address" is not understood by you it can also be
formulated as "the OP has an RFC1918 IPv4 address which is not routable
on the public internet and hence a 6to4 tunnel can't work for him."

Re: Mailing list usage questions

[Thomas Hochstein]

Tom Furie wrote:

> It occurs to me that I'm not even aware of what list management
> software these lists are managed with anymore.

|  All original Debian mailing lists are run on a special server, using an
|  automatic mail processing software called SmartList. 


-thh

Re: how to change default nameserver?

[Max Nikulin]

On 12/04/2023 00:36, Andrew M.A. Cater wrote:

If you're deafblind and using a Braille reader - it looks for a serial tty.
That's the only way it can work. It can't ask you first necessarily.


Since new class of users has been identified, namely those who have 
serial port adapters, but no brltty, perhaps, a new screen should be 
added to the installer:


Assistive technologies have been enabled because serial port adapter has 
been detected, so Braille terminal may be connected.


(*) Continue assuming presence of brltty.
( ) Disable assistive technologies and proceed as regular installation.

Re: update-initramfs

[Michel Verdier]

Le 11 avril 2023 davidson a écrit :

> # update-initramfs -u
> # update-initramfs: Generating /boot/initrd.img-5.10.0-21-amd64-kg
> W: missing /lib/modules/5.10.0-21-amd64-kg

Of course : /lib/modules/ is installed via package. You
have to do it manually to get rid of this error. And without it you can't
achieve a bootable kernel.

> You didn't make backup copies of your most recent kernel, *give them
> funny names*, and keep them in /boot. That is the distinction here, I
> think.

I don't have to do it *manually* I give funny names during build so the
.deb include all is needed. And I only need backup of the package, not
the /boot files. Obviously /boot is for operational files not for backup
ones.

> I believe the OP just wants an extra entry in his grub menu that will
> boot a redundant copy of his latest working kernel. (But that is only
> my understanding, which might be wrong. OP can speak for himself on
> this point.)

Ok to cover grub menu you just have to had it in /etc/grub.d.
You simply copy a block menuentry from /boot/grub/grub.cfg and put it in
something like /etc/grub.d/40_custom. In the copy you can change kernel
params, etc. update-grub will include it in generated grub.cfg.

> It seems to me that building and packaging like you suggest is more
> work than warranted, just to make a backup copy. According to OP's
> report, that simple practice used to work for him.

Building a kernel is much less work than believed. And much much less
complicated too. It requires apt install kernel sources, apt install deps
for build, then do the build. I build with my custom make commands but I
heard there is a dedicated debian stance to even more simplify this
point. Overall it is much less work than manually following updates on a
breaked /boot.

Re: Mailing list usage questions

[The Wanderer]

On 2023-04-11 at 16:22, zithro wrote:

> Hello all,
> 
> I have two questions about the Debians ML usage.
> 
> 1. when subscribing, the confirmation message says "By default, copies 
> of your own submissions will be returned."
> What is the meaning of "default" and "returned" here ?
> I understand that I should get my own replies. But I never get them.

Are you A: running your own mail system, or B: using an
externally-hosted mail service, either Ba: directly or Bb: under a
personal domain?

Some mail services apparently treat this "discard incoming messages that
look like duplicates of ones you already have a copy of" behavior as a
feature; Gmail is the best-known example. That has problems when (as
with this mailing list) the incoming copy is not identical to the one
that was sent, even though it has the same Message-ID, but AFAIK they
don't seem to care.

If you're using an external mail service which does have that behavior,
then that would probably explain what you're seeing.

If you aren't using an external mail service, or are using one which
doesn't have that behavior (i.e., if e.g. messages sent to other mailing
lists to which you are subscribed do come back to you), then there's
probably a different problem and it would be warranted to dig deeper.

-- 
   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: update-initramfs

[David Wright]

On Tue 11 Apr 2023 at 22:14:18 (+0200), zithro wrote:
> I thought :
> 
> - you can install as many kernel packages as you want, whether built
> or downloaded
> - updates don't automatically remove old kernels/initrd by default
> 
> So I wonder, why handling it manually ?
> What is the advantage, except for adding -confusion- ?

There is one case where a kernel is silently removed, and that is when
they tweak it without changing the minor number (or whatever that
number is now called). IIRC it hasn't happened for quite some time;
perhaps the last was in June 2020, when 4.19.118-2+deb10u1 replaced
4.19.118-2 for linux-image-4.19.0-9-amd64.

Found in APT's history log with /linux-image.*\(([-0-9\.]+), \1

Cheers,
David.

Bookworm lockup: Report bug?

[Christian Gelinek]

Dear group,

This morning I encountered my PC being frozen. Here are the last 
journalctl messages, just before I forced it to power down and reboot:


Apr 10 07:31:07 gar systemd[1]: Started anacron.service - Run anacron jobs.
Apr 10 07:31:07 gar anacron[4875]: Anacron 2.3 started on 2023-04-10
Apr 10 07:31:07 gar anacron[4875]: Will run job `cron.daily' in 5 min.
Apr 10 07:31:07 gar anacron[4875]: Jobs will be executed sequentially
Apr 10 07:36:07 gar anacron[4875]: Job `cron.daily' started
Apr 10 07:36:07 gar anacron[4882]: Updated timestamp for job 
`cron.daily' to 2023-04-10

Apr 10 07:36:07 gar anacron[4875]: Job `cron.daily' terminated
Apr 10 07:36:07 gar anacron[4875]: Normal exit (1 job run)
Apr 10 07:36:07 gar systemd[1]: anacron.service: Deactivated successfully.
Apr 10 07:50:01 gar kernel: snd_hda_intel :04:00.0: Unable to change 
power state from D3hot to D0, device inaccessible
Apr 10 07:50:03 gar kernel: [drm:fw_domains_get_with_fallback [i915]] 
*ERROR* render: timed out waiting for forcewake ack to clear.
Apr 10 07:50:03 gar kernel: i915 :03:00.0: [drm:add_taint_for_CI 
[i915]] CI tainted:0x9 by fw_domains_get_with_fallback+0x20c/0x230 [i915]
Apr 10 07:50:07 gar kernel: [drm:fw_domains_get_with_fallback [i915]] 
*ERROR* render: timed out waiting for forcewake ack to clear.
Apr 10 07:50:07 gar kernel: i915 :03:00.0: [drm:add_taint_for_CI 
[i915]] CI tainted:0x9 by fw_domains_get_with_fallback+0x20c/0x230 [i915]
Apr 10 07:50:11 gar kernel: [drm:fw_domains_get_with_fallback [i915]] 
*ERROR* render: timed out waiting for forcewake ack to clear.
Apr 10 07:50:11 gar kernel: i915 :03:00.0: [drm:add_taint_for_CI 
[i915]] CI tainted:0x9 by fw_domains_get_with_fallback+0x20c/0x230 [i915]
Apr 10 07:50:15 gar kernel: [drm:fw_domains_get_with_fallback [i915]] 
*ERROR* gt: timed out waiting for forcewake ack to clear.
Apr 10 07:50:15 gar kernel: i915 :03:00.0: [drm:add_taint_for_CI 
[i915]] CI tainted:0x9 by fw_domains_get_with_fallback+0x20c/0x230 [i915]
Apr 10 07:50:16 gar kernel: i915 :03:00.0: [drm] *ERROR* CT: 
Corrupted descriptor head=4294967295 tail=4294967295 status=0x
Apr 10 07:50:28 gar kernel: [drm:fw_domains_get_with_fallback [i915]] 
*ERROR* render: timed out waiting for forcewake ack to clear.
Apr 10 07:50:28 gar kernel: i915 :03:00.0: [drm:add_taint_for_CI 
[i915]] CI tainted:0x9 by fw_domains_get_with_fallback+0x20c/0x230 [i915]
Apr 10 07:50:28 gar kernel: [drm:fw_domains_get_with_fallback [i915]] 
*ERROR* gt: timed out waiting for forcewake ack to clear.
Apr 10 07:50:28 gar kernel: i915 :03:00.0: [drm:add_taint_for_CI 
[i915]] CI tainted:0x9 by fw_domains_get_with_fallback+0x20c/0x230 [i915]
Apr 10 07:50:28 gar kernel: [drm:fw_domains_get_with_fallback [i915]] 
*ERROR* vdbox0: timed out waiting for forcewake ack to clear.
Apr 10 07:50:28 gar kernel: i915 :03:00.0: [drm:add_taint_for_CI 
[i915]] CI tainted:0x9 by fw_domains_get_with_fallback+0x20c/0x230 [i915]
Apr 10 07:50:28 gar kernel: watchdog: BUG: soft lockup - CPU#15 stuck 
for 26s! [kworker/15:2:1033]
Apr 10 07:50:28 gar kernel: Modules linked in: snd_seq_dummy snd_hrtimer 
snd_seq snd_seq_device nfsv3 nfs_acl rpcsec_gss_krb5 auth_rpcgss nfsv4 
dns_resolver nfs lockd grace fscache netfs rfkill qrtr sunrpc 
binfmt_misc nls_ascii nls_cp437 vfat fat snd_hda_codec_hdmi snd_s>
Apr 10 07:50:28 gar kernel:  intel_uncore ee1004 pcspkr snd watchdog 
soundcore intel_vsec serial_multi_instantiate intel_pmc_core acpi_pad 
acpi_tad mei_me mei sg evdev parport_pc ppdev lp parport fuse loop 
efi_pstore configfs efivarfs ip_tables x_tables autofs4 ext4 crc1>
Apr 10 07:50:28 gar kernel: CPU: 15 PID: 1033 Comm: kworker/15:2 
Tainted: G U  W  6.1.0-7-amd64 #1  Debian 6.1.20-1
Apr 10 07:50:28 gar kernel: Hardware name: Micro-Star International Co., 
Ltd. MS-7E02/PRO B760M-P DDR4 (MS-7E02), BIOS 1.00 10/21/2022

Apr 10 07:50:28 gar kernel: Workqueue: pm pm_runtime_work
Apr 10 07:50:28 gar kernel: RIP: 0010:pci_mmcfg_read+0xb0/0xe0
Apr 10 07:50:28 gar kernel: Code: 5d 41 5e 41 5f c3 cc cc cc cc 4c 01 e0 
66 8b 00 0f b7 c0 89 45 00 eb dc 4c 01 e0 8a 00 0f b6 c0 89 45 00 eb cf 
4c 01 e0 8b 00 <89> 45 00 eb c5 e8 b6 ae 78 ff c7 45 00 ff ff ff ff b8 
ea ff ff ff

Apr 10 07:50:28 gar kernel: RSP: :b859c1963cc0 EFLAGS: 0286
Apr 10 07:50:28 gar kernel: RAX:  RBX: 0040 
RCX: 0ffc
Apr 10 07:50:28 gar kernel: RDX: 00ff RSI: 0004 
RDI: 
Apr 10 07:50:28 gar kernel: RBP: b859c1963cfc R08: 0004 
R09: b859c1963cfc
Apr 10 07:50:28 gar kernel: R10: 0004 R11: b6ba5d50 
R12: 0ffc
Apr 10 07:50:28 gar kernel: R13:  R14: 0004 
R15: 
Apr 10 07:50:28 gar kernel: FS:  () 
GS:8a3d1fbc() knlGS:
Apr 10 07:50:28 gar kernel: CS:  0010 DS:  ES:  CR0: 
80050033
Apr 10 07:50:28 ga

Re: how to change default nameserver?

[zithro]

On 11 Apr 2023 22:18, Greg Wooledge wrote:

On Tue, Apr 11, 2023 at 09:56:05PM +0200, zithro wrote:

Do you know when resolv.conf started appearing ?
I guess after TCP/IP got invented ?
The wikipedia page does not mention it.


 says it first appeared in 4.3BSD.
I can neither confirm nor deny this, but it does match my *belief* that
BSD did it first, and then System V copied it.  It's a bit before my time,
though.



THIS was a fun rabbit hole to enter ^^

From "https://en.wikipedia.org/wiki/History_of_Unix";

Perhaps the most important aspect of the BSD development effort was the 
addition of TCP/IP network code to the mainstream Unix kernel.
[...] The network code found in these releases is the ancestor of much 
TCP/IP network code in use today, including code that was *later* 
released in AT&T System V UNIX.


So your belief was right. BSD before SysV !

"http://gunkies.org/wiki/4BSD";
4 BSD does *NOT* include any TCP/IP networking

"http://gunkies.org/wiki/4.2_BSD";
4.2 BSD follows the betas of 4.1a & 4.1b. 4.2 BSD Is special 
because it incorporates the first versions of BSD TCP/IP


Let's see. I downloaded the source code of 4.2 and 4.3BSD from The Unix 
Heritage Society (https://www.tuhs.org/Archive/Distributions/UCB/)


--
$ grep -R "resolv.conf"

4.3BSD/lib/libc/net/res_init.c:char*conffile = "/etc/resolv.conf";
4.3BSD/lib/libc/net/res_init.c:printf("MAXNS 
reached, reading resolv.conf\n");
4.3BSD/etc/named/tools/ns.lookup/man/nslookup.l:(Default = value in 
/etc/resolv.conf, abbreviation = do)
4.3BSD/etc/named/tools/ns.lookup/man/nslookup.l:/etc/resolv.conf	initial 
domain name and name server addresses.

--

So it's not in 4.2, and appeared in 4.3.

Fun one "https://github.com/dank101/4.2BSD/blob/master/include/netdb.h"; :

/*
 * Assumption here is that a network number
 * fits in 32 bits -- probably a poor one.
 */

Ok, out of the hole now ^^

Re: how to change default nameserver?

[gene heskett]

On 4/11/23 13:39, zithro wrote:

On 11 Apr 2023 19:09, gene heskett wrote:

On 4/11/23 10:04, zithro wrote:

On 11 Apr 2023 04:56, gene heskett wrote:

On 4/10/23 16:53, zithro wrote:
Why can't you follow others advice, hell, if you don't trust us, 
even the perfectly correct and up-to-date manpages ?
After reading the posts of others, I'm more and more thinking your 
simply a troll (or a RedHat fanatic wasting Debian helpers time for 
no reason) ...


That is an insult. I bailed out of fedora 15 years ago, tired to 
being an always sick lab rat for redhat. [...]


Because you answer what you want, I'll re-ask :
"Why can't you follow others advice, hell, if you don't trust us, even
the perfectly correct and up-to-date manpages ?"


Because you answer what you want, I'll re-ask :
"Why can't you follow others advice, hell, if you don't trust us, even
the perfectly correct and up-to-date manpages ?"


If you don't know who Mike Sweet is, why are you hassling me?


I don't need to know who he is. Now I know, thanks, but it changes nothing.
And if Sweet is REALLY your friend, why not asking him ?

  There's 25 years of history to computing before Linus released his 
his linux


Computer history started WAY before that.


Keeping networking working on linux has been an art, not a science.


Debatable. The Debian way is one of the easiest.
But funny, only happens to you, among the zillions Debian-based installs.
Still can't see the problem ?


looking at one of my buster machines. /etc/network/interfaces has this:


What about your resolv.conf ?


Evidence on that bo that something has been there, and put a # in front 
of the search line. IDK when.



So ATM I have no clue what I did because I've forgotten whatever I did


One of the best advice I've learnt from experienced sysadmins is :
"always take notes of what you're doing".
The other one being "RTFM".

[...] I can't ask cups, my posts to the cups list are apparently 
routed to /dev/null [...]


If you react like here, maybe you've been banned from there ?
Again, contact the list owners, don't ASSUME stuff.




[...] so I come here for help and all this dirty laundry gets drug out 
again. And again. And again. While the question I asked is very 
carefully ignored. Unreal.


Really ?! Who exactly is ignoring what ?!


My question about cups.



Again, FOR THE FOURTH TIME :
"Why can't you follow others advice, hell, if you don't trust us, even
the perfectly correct and up-to-date manpages ?"

Others have warned me. I tried despite their warnings ...
I get to the same point as them : it's pointless to try to help
someone who does NOT follow advices.
I guess I'm done here.


Your choice.

Take care & stay well.

Cheers, Gene Heskett.
--
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author, 1940)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: Mailing list usage questions

[Tom Furie]

On Tue, Apr 11, 2023 at 10:51:43PM +0200, zithro wrote:
> I already sent "help" to the ML but it did not provide any hints.
> Hence my question here ; )

Oh! Yes, I see the reply to that message is much less useful than it once
was... Previously it would supply a list of commands for subscription
management. It occurs to me that I'm not even aware of what list management
software these lists are managed with anymore.

Cheers,
Tom

-- 
Nature always sides with the hidden flaw.


signature.asc
Description: PGP signature

Re: Mailing list usage questions

[Andrew M.A. Cater]

On Tue, Apr 11, 2023 at 10:22:59PM +0200, zithro wrote:
> Hello all,
> 
> I have two questions about the Debians ML usage.
> 
> 1. when subscribing, the confirmation message says "By default, copies of
> your own submissions will be returned."
> What is the meaning of "default" and "returned" here ?

If you are subscribed to the mailing list and you post, you should see
a copy turn up in your mailing list mails - that's how I read that.
So - I'll post this to the mailing list and should see a copy in a couple
of minutes.

> I understand that I should get my own replies. But I never get them.
> Should I simply add a rule to move my sent mails from "Sent" to my ML folder
> ?
> 
> 2. Is there some kind of interface to manage my subscriptions ?
> From what I read the (new?) system is not using "majordomo" anymore.
> On majordomo you could set a few options on a web page.
> Does it exist here as well ?
>

https://www.debian.org/MailingLists/ gives some details.
There's also a web form to be able to subscribe/unsubscribe to larger
numbers of lists.

All the very best, as ever,

Andy Cater
 

Re: Mailing list usage questions

[zithro]

On 11 Apr 2023 22:39, Tom Furie wrote:

In the headers of every mail on debian lists are some "List-*" headers. In
there you can find an address and subject to get help on list commands etc.
(At least, I assume it still works. I haven't used it in a long time and
didn't bother to check before writing this reply.)


Yes, I'm using those headers to sort all the MLs I'm subscribed to.

I already sent "help" to the ML but it did not provide any hints.
Hence my question here ; )

Re: how to change default nameserver?

[zithro]

On 11 Apr 2023 22:18, Greg Wooledge wrote:

On Tue, Apr 11, 2023 at 09:56:05PM +0200, zithro wrote:

Do you know when resolv.conf started appearing ?
I guess after TCP/IP got invented ?
The wikipedia page does not mention it.


 says it first appeared in 4.3BSD.
I can neither confirm nor deny this, but it does match my *belief* that
BSD did it first, and then System V copied it.  It's a bit before my time,
though.



Thanks for the hints !
I then dug a bit more, found a github which has the sources for very old 
BSDs !

https://github.com/dank101?tab=repositories

The README suggests (IMHO) that previous versions already had TCP/IP 
working.
In the 4.3BSD-Reno, there is no resolv.conf by default in /etc (but a 
namedb).
But I found a man page 
(https://github.com/dank101/4.3BSD-Reno/blob/master/share/doc/smm/11.named/resolv.conf).

Inside, there's a comment :

@(#)resolv.conf 6.2 (Berkeley) 2/29/88

Funny, it's already the true and loved syntax we use today !
domain Berkeley\fB.\fPEdu
nameserver 128\fB.\fP32\fB.\fP0\fB.\fP4
nameserver 128\fB.\fP32\fB.\fP0\fB.\fP10

(I just don't know what the gibberish chars mean, I guess it's some 
TeX/groff syntax ?)


I'll dig more into the repositories and let you know ; )
(2BSD had none, but it only has source code, more later)

Re: Mailing list usage questions

[Tom Furie]

On Tue, Apr 11, 2023 at 10:22:59PM +0200, zithro wrote:
> I have two questions about the Debians ML usage.
> 
> 1. when subscribing, the confirmation message says "By default, copies of
> your own submissions will be returned."
> What is the meaning of "default" and "returned" here ?
> I understand that I should get my own replies. But I never get them.
> Should I simply add a rule to move my sent mails from "Sent" to my ML folder
> ?
> 
> 2. Is there some kind of interface to manage my subscriptions ?
> From what I read the (new?) system is not using "majordomo" anymore.
> On majordomo you could set a few options on a web page.
> Does it exist here as well ?

In the headers of every mail on debian lists are some "List-*" headers. In
there you can find an address and subject to get help on list commands etc.
(At least, I assume it still works. I haven't used it in a long time and
didn't bother to check before writing this reply.)

Cheers,
Tom

-- 
The District of Columbia has a law forbidding you to exert pressure on
a balloon and thereby cause a whistling sound on the streets.


signature.asc
Description: PGP signature

Re: how to change default nameserver?

[gene heskett]

On 4/11/23 13:36, Andrew M.A. Cater wrote:

On Tue, Apr 11, 2023 at 01:09:37PM -0400, gene heskett wrote:

So ATM I have no clue what I did because I've forgotten whatever I did to
make it work for bullseye, lost in the noise from doing 23 damned installs
before someone suggested I unplug all usb, but my key board and mouse are
wireless, so I left those buttons plugged it and did yet another install and
this one worked. I've since rebuilt my usb tree as it reaches like a weeping
willow to nearly every inch of the real estate here.  All that re-install BS
caused by the installer silently finding a couple usb-serial adaptors and
ASSUMING I was blind and using brltty and orca. Has any of you ever tried to
use a computer whose keyboard response was delayed by nearly 2 seconds per
keystroke because its looking for a 1200 baud tty that isn't there, and then
yells the keystroke at you in some idea of a computer voice.  Spend 2 weeks
trying to use a computer like that, and you'll have a new appreciation for
my level of frustration.



Your experience led me to fixing someone else's problem on IRC.
They were trying to install Debian for linuxcnc (but on an amd64 machine)
and had the same problem because they'd left a serial lead plugged in so
got brltty and orca.

I advised them to reinstall to fix things but this time to undo the serial
lead until the install was over. That worked fine for them and they were
very happy. (It's the simple fix that I think I suggested that to you three
times in the course and was ignored.)



Possibly you did, but I didn't glom the usb connection because it came 
with no reason until the last time.



It should have asked me if I wanted that crap but did not.



If you're deafblind and using a Braille reader - it looks for a serial tty.
That's the only way it can work. It can't ask you first necessarily.


That is a huge if. Probably less than 1% of the folks who only fit in 
one box or the other. It could have loaded orca and spoke to me, but 
didn't, it could have asked me on screen and didn't.  What it did was 
make an ass out me in everyone elses mind. Yeah, these days I'm an old 
fart with the poor short term memory that goes with 88 years on the wet 
ram. But I was carving code for an rca 1802 chip before most of the 
readers here were born, doing it by looking up the memonic, and entering 
the corresponding hex code in a quest super elf's hex monitor, code so 
useful in 1979 that it was still in daily use at KRCR-tv when the 
station burnt to the ground in the later 90's, well after Mt St Helens 
blew its top. As a retired television Chief (& usually only, I'm a CET 
too) Engineer, that 16 or so years might as well be a couple of eons, 
the tech, even in never twice same color days, the tech moved a heck of 
a lot faster than that.


My wife of 31 years passed a couple years back and I should throw my 
poles in the boat and go fishing. But I've been diabetic for nearly 40 
years & the water is too cold for my feet, so for S&G's I'm building a 
3d printer farm instead.



You have all claimed I missed it or told it to install that, but 23 damned
times? I'm sure its here someplace because all 24 times I filled in the
network details manually and was equally amazed when I still had a network
on the after install reboot. That FWIW, was a first.

Everything works except printer sharing and I can't ask cups, my posts to
the cups list are apparently routed to /dev/null, so I come here for help
and all this dirty laundry gets drug out again. And again. And again. While
the question I asked is very carefully ignored. Unreal.



There are a bunch of cups manpages: there are lots of documents online.
The people recommending you avahi/bonjour/zeroconf are recommending it
because it works - for them and for 99.9% of people.


And its both a lockin for apple, and denies the feature the printer 
maker puts into his product.  And I'm purist enough to object the both.



Your system: your
rules - but it doesn't always help us help you to solve the problems.


You betcha. Which gets my posts the TL;DR treatment just because I do
try to describe things adequately. I can't win for losing.


There are reasons why we ask folk - in general on this list and in the FAQ
- to provide information, to copy/paste log entries or what they're seeing
on screen. We can't sit on your shoulder and watch what you've typed: we're
all relying on guesswork much of the time.


Cheers, Gene Heskett.


All the very best, as ever,

Andy Cater


Back at you Andy, take care & stay well.

Cheers, Gene Heskett.
--
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author, 1940)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Mailing list usage questions

[zithro]

Hello all,

I have two questions about the Debians ML usage.

1. when subscribing, the confirmation message says "By default, copies 
of your own submissions will be returned."

What is the meaning of "default" and "returned" here ?
I understand that I should get my own replies. But I never get them.
Should I simply add a rule to move my sent mails from "Sent" to my ML 
folder ?


2. Is there some kind of interface to manage my subscriptions ?
From what I read the (new?) system is not using "majordomo" anymore.
On majordomo you could set a few options on a web page.
Does it exist here as well ?

Re: how to change default nameserver?

[Greg Wooledge]

On Tue, Apr 11, 2023 at 09:56:05PM +0200, zithro wrote:
> Do you know when resolv.conf started appearing ?
> I guess after TCP/IP got invented ?
> The wikipedia page does not mention it.

 says it first appeared in 4.3BSD.
I can neither confirm nor deny this, but it does match my *belief* that
BSD did it first, and then System V copied it.  It's a bit before my time,
though.

Re: update-initramfs

[zithro]

I thought :

- you can install as many kernel packages as you want, whether built or 
downloaded

- updates don't automatically remove old kernels/initrd by default

So I wonder, why handling it manually ?
What is the advantage, except for adding -confusion- ?

Re: update-initramfs

[David Wright]

On Tue 11 Apr 2023 at 10:51:19 (-0400), Marc Auslander wrote:
> On 4/10/2023 11:00 PM, David Wright wrote:
> > On Mon 10 Apr 2023 at 20:17:11 (-0400), Marc Auslander wrote:
> > > I'm on Buster.
> > > 
> > > In /boot I keep a copy of the current working linux named by appending
> > > -knowngood to the four files.  My idea is that if an update fails, I
> > > have a recent working linux.  This is different from vmlinuz.old which
> > > is the previous kernel version.  The updates in question are not to
> > > the kernel but to initrd.image of course.
> > > 
> > > Suddenly, update-initramfs insists in trying to first update
> > > initrd.-knowngood  which of course fails because there are no
> > > underling file with that name.  This never happened in the past,
> > > AFAIK. Once it fails it gives up.
> > > 
> > > There seems no way to force update-initramfs to update the right kernel.
[ … ]
> thanks but that's the first thing I checked - it's yes, not all.  But
> my backup names contain the current version string.
> 
> I'm not sure about the sort order hack.  My goal is to have
> update-grub see the knowngood as a bootable linux and include it in
> the boot menu. That's also why .bak of initrd isn't good enough - I
> need a complete copy.

Oh, so it's update-grub calling update-initramfs, which could
complicate things.

Quite honestly, I don't see why you want to make what are essentially
backup files into part of the working set for both Grub and initramfs,
meaning they have to process duplicate files.

Any time you have a set of files that you're happy with, why not just
copy them to another directory, like /boot/backups/, adding suitable
suffixes. With Grub's flexibility, it's very easy to boot from those
copies instead. Press e at the blue screen, and tweak the filenames.
If you forget where they are or what they're called, press c instead
and go hunting with ls.

Cheers,
David.

Re: how to change default nameserver?

[zithro]

On 11 Apr 2023 13:25, Greg Wooledge wrote:
>> On 11 Apr 2023 08:55, Richard Hector wrote:

Well, it's not in resolver(5) (which is for resolv.conf) on Red Hat 5.0.5.


It wasn't in the man page from Red Hat 5.2 when I checked in 2017, either.


Thanks for the history digging !

Do you know when resolv.conf started appearing ?
I guess after TCP/IP got invented ?
The wikipedia page does not mention it.

Re: update-initramfs

[davidson]

On Tue, 11 Apr 2023 davidson wrote:

On Tue, 11 Apr 2023 Michel Verdier wrote:

Le 11 avril 2023 davidson a écrit :

 Experiment #2: see if I could tweak OP's practice enough so that
   update-grub would not care.


...and so that "update-initramfs -u" would not notice.

--
Sometimes it pays to have squirrels in your head running around making
you question everything.  -- Clive Robinson

Re: update-initramfs

[davidson]

On Tue, 11 Apr 2023 Michel Verdier wrote:

Le 11 avril 2023 davidson a écrit :


The first experiment simply tried to replicate your observations (as I
understood them). Basically, I added "-kg" suffix to all the files in
/boot corresponding to latest installed kernel, so that I had
unsuffixed copies and "*-kg" ("knowngood") copies, and then tried

 # update-initramfs -u


I don't understand a point.


Hi Michel.

To be clear, I am not the OP.


On my system I compiled kernel and thus build a linux-image-...deb
with a specific tag.


That is much more work than I did. I compiled no custom kernel. I just
made copies of the files in /boot installed from package
linux-image-5.10.0-21-amd64, and the corresponding initrd.

Here is a run of my Experiment #1, in full:

# knowngood=( /boot/*-$(uname -r) )
# for ((i=0; i<${#knowngood[@]}; i+=1)) ; do cp -a "${knowngood[i]}" 
"${knowngood[i]/%/-kg}" ; done
# ls -l "${knowngood[@]}"
-rw-r--r-- 1 root root   236452 21 janv. 09:35 /boot/config-5.10.0-21-amd64
-rw-r--r-- 1 root root 29199065 10 avril 21:58 /boot/initrd.img-5.10.0-21-amd64
-rw-r--r-- 1 root root   83 21 janv. 09:35 /boot/System.map-5.10.0-21-amd64
-rw-r--r-- 1 root root  7019136 21 janv. 09:35 /boot/vmlinuz-5.10.0-21-amd64
# ls -l "${knowngood[@]/%/-kg}"
-rw-r--r-- 1 root root   236452 21 janv. 09:35 /boot/config-5.10.0-21-amd64-kg
-rw-r--r-- 1 root root 29199065 10 avril 21:58 
/boot/initrd.img-5.10.0-21-amd64-kg
-rw-r--r-- 1 root root   83 21 janv. 09:35 
/boot/System.map-5.10.0-21-amd64-kg
-rw-r--r-- 1 root root  7019136 21 janv. 09:35 /boot/vmlinuz-5.10.0-21-amd64-kg

# update-initramfs -u
# update-initramfs: Generating /boot/initrd.img-5.10.0-21-amd64-kg
W: missing /lib/modules/5.10.0-21-amd64-kg
W: Ensure all necessary drivers are built into the linux image!
depmod: ERROR: could not open directory /lib/modules/5.10.0-21-amd64-kg: No 
such file or directory
depmod: FATAL: could not search modules: No such file or directory
cat: 
/var/tmp/mkinitramfs_9UHxvD/lib/modules/5.10.0-21-amd64-kg/modules.builtin: 
Aucun fichier ou dossier de ce type
find: ‘/var/tmp/mkinitramfs_9UHxvD/lib/modules/5.10.0-21-amd64-kg/kernel’: 
Aucun fichier ou dossier de ce type
W: Can't find modules.builtin.modinfo (for locating built-in drivers' firmware, 
supported in Linux >=5.2)
depmod: WARNING: could not open modules.order at 
/var/tmp/mkinitramfs_9UHxvD/lib/modules/5.10.0-21-amd64-kg: No such file or 
directory
depmod: WARNING: could not open modules.builtin at 
/var/tmp/mkinitramfs_9UHxvD/lib/modules/5.10.0-21-amd64-kg: No such file or 
directory
# ls -l "${knowngood[@]}"
-rw-r--r-- 1 root root   236452 21 janv. 09:35 /boot/config-5.10.0-21-amd64
-rw-r--r-- 1 root root 29199065 10 avril 21:58 /boot/initrd.img-5.10.0-21-amd64
   
-rw-r--r-- 1 root root   83 21 janv. 09:35 /boot/System.map-5.10.0-21-amd64
-rw-r--r-- 1 root root  7019136 21 janv. 09:35 /boot/vmlinuz-5.10.0-21-amd64
# ls -l "${knowngood[@]/%/-kg}"
-rw-r--r-- 1 root root  236452 21 janv. 09:35 /boot/config-5.10.0-21-amd64-kg
-rw-r--r-- 1 root root 5924752 10 avril 22:07 
/boot/initrd.img-5.10.0-21-amd64-kg
   ^^^
-rw-r--r-- 1 root root  83 21 janv. 09:35 
/boot/System.map-5.10.0-21-amd64-kg
-rw-r--r-- 1 root root 7019136 21 janv. 09:35 /boot/vmlinuz-5.10.0-21-amd64-kg


I install package so I have kernel and initram in /boot with the
tag. Same as your system I think ?


You didn't make backup copies of your most recent kernel, *give them
funny names*, and keep them in /boot. That is the distinction here, I
think.


 And when I update-initramfs it generate right files with right
names. And never break anything on further updates.


That makes sense, of course.


So why change only filenames and not rebuild a package with a
different tag ?


I believe the OP just wants an extra entry in his grub menu that will
boot a redundant copy of his latest working kernel. (But that is only
my understanding, which might be wrong. OP can speak for himself on
this point.)

It seems to me that building and packaging like you suggest is more
work than warranted, just to make a backup copy. According to OP's
report, that simple practice used to work for him.

Myself, I was only curious to do

  Experiment #1: replicate the OP's observations

and

  Experiment #2: see if I could tweak OP's practice enough so that
update-grub would not care.

--
Hackers are free people. They are like artists. If they are in a good
mood, they get up in the morning and begin painting their pictures.
-- Vladimir Putin

Re: how to change default nameserver?

[Nicholas Geovanis]

On Tue, Apr 11, 2023, 12:40 PM zithro  wrote:

> 

>  There's 25 years of history to computing before Linus released his his
> linux
>
> Computer history started WAY before that.
>
> > Keeping networking working on linux has been an art, not a science.
>
..

> Follow advices.
> I guess I'm done here.
>

I guess it's overdue to point out again what should be obvious: You don't
have to run NetworkManager to make use of DHCP or anything else. So you
don't need to have your name resolution config overwritten. And you can
remain under the Debian umbrella.

I will go even further: NetworkManager is for laptops, not home setups. We
don't use it on servers in a data center, period end. It has no use-case in
that environment.

Re: update-initramfs

[Michel Verdier]

Le 11 avril 2023 davidson a écrit :

> The first experiment simply tried to replicate your observations (as I
> understood them). Basically, I added "-kg" suffix to all the files in
> /boot corresponding to latest installed kernel, so that I had
> unsuffixed copies and "*-kg" ("knowngood") copies, and then tried
>
>  # update-initramfs -u

I don't understand a point. On my system I compiled kernel and thus build
a linux-image-...deb with a specific tag. I install package so I have
kernel and initram in /boot with the tag. Same as your system I think ?
And when I update-initramfs it generate right files with right names. And
never break anything on further updates. So why change only filenames and
not rebuild a package with a different tag ?

Re: apt temporary failure resolving deb.debian.org

[Tim Woodall]

On Mon, 10 Apr 2023, Tim Woodall wrote:


On Mon, 10 Apr 2023, Lee wrote:


Why are you using google as forwarders ?


To eliminate as many variables as possible.

delv talking to google works.

delv talking to bind talking to google fails.

When talking directly, delv is using udp to talk to google
When talking via bind, bind is using tcp.

And while google acks the DNSKEY request from bind, the data is not
received. The seqnence number jumps from 1 on the ACK of the query to
1636 on the FIN where google closes the connection.

Thats 1635 bytes of data gone missing.



I managed to reproduce this talking to a remote bind server that I can
control, running tcpdump on both ends.

The DNS response was 1661 bytes split into two TCP packets with TCP
segment len of 1208 and 455 (The other two bytes are the DNS response
length itself)

My router (at least I assume it's my router) is then dropping them.

Change to use a non-standard port for the remote dns resolver and it
works.

Re: how to change default nameserver?

[zithro]

On 11 Apr 2023 19:09, gene heskett wrote:

On 4/11/23 10:04, zithro wrote:

On 11 Apr 2023 04:56, gene heskett wrote:

On 4/10/23 16:53, zithro wrote:
Why can't you follow others advice, hell, if you don't trust us, 
even the perfectly correct and up-to-date manpages ?
After reading the posts of others, I'm more and more thinking your 
simply a troll (or a RedHat fanatic wasting Debian helpers time for 
no reason) ...


That is an insult. I bailed out of fedora 15 years ago, tired to 
being an always sick lab rat for redhat. [...]


Because you answer what you want, I'll re-ask :
"Why can't you follow others advice, hell, if you don't trust us, even
the perfectly correct and up-to-date manpages ?"


Because you answer what you want, I'll re-ask :
"Why can't you follow others advice, hell, if you don't trust us, even
the perfectly correct and up-to-date manpages ?"


If you don't know who Mike Sweet is, why are you hassling me?


I don't need to know who he is. Now I know, thanks, but it changes nothing.
And if Sweet is REALLY your friend, why not asking him ?


  There's 25 years of history to computing before Linus released his his linux


Computer history started WAY before that.


Keeping networking working on linux has been an art, not a science.


Debatable. The Debian way is one of the easiest.
But funny, only happens to you, among the zillions Debian-based installs.
Still can't see the problem ?


looking at one of my buster machines. /etc/network/interfaces has this:


What about your resolv.conf ?


So ATM I have no clue what I did because I've forgotten whatever I did


One of the best advice I've learnt from experienced sysadmins is :
"always take notes of what you're doing".
The other one being "RTFM".

[...] I can't ask cups, my posts 
to the cups list are apparently routed to /dev/null [...]


If you react like here, maybe you've been banned from there ?
Again, contact the list owners, don't ASSUME stuff.

[...] so I come here for 
help and all this dirty laundry gets drug out again. And again. And 
again. While the question I asked is very carefully ignored. Unreal.


Really ?! Who exactly is ignoring what ?!

Again, FOR THE FOURTH TIME :
"Why can't you follow others advice, hell, if you don't trust us, even
the perfectly correct and up-to-date manpages ?"

Others have warned me. I tried despite their warnings ...
I get to the same point as them : it's pointless to try to help
someone who does NOT follow advices.
I guess I'm done here.

Re: how to change default nameserver?

[Greg Wooledge]

On Tue, Apr 11, 2023 at 01:09:37PM -0400, gene heskett wrote:
> Keeping networking working on linux has been an art, not a science.
> looking at one of my buster machines. /etc/network/interfaces has this:
> 
> # The primary network interface
> allow-hotplug eno1
> iface eno1 inet static
>   address 192.168.71.10/24
>   gateway 192.168.71.1
>   # dns-* options are implemented by the resolvconf package, if installed
>   dns-nameservers 192.168.71.1
>   dns-search coyote.den
> 
> And it Just Works.

The last two lines of that file are configuration elements for the
resolvconf package.  They only take effect if resolvconf is installed.

> But that was way the hell and gone too easy, so it just
> had to be changed, so that stuff no longer works for bullseye.

Most likely, your bullseye system simply doesn't have resolvconf.

This is discussed at
.

The comment right before them also tells you which package you need for
those lines to work.

Re: how to change default nameserver?

[Andrew M.A. Cater]

On Tue, Apr 11, 2023 at 01:09:37PM -0400, gene heskett wrote:
> So ATM I have no clue what I did because I've forgotten whatever I did to
> make it work for bullseye, lost in the noise from doing 23 damned installs
> before someone suggested I unplug all usb, but my key board and mouse are
> wireless, so I left those buttons plugged it and did yet another install and
> this one worked. I've since rebuilt my usb tree as it reaches like a weeping
> willow to nearly every inch of the real estate here.  All that re-install BS
> caused by the installer silently finding a couple usb-serial adaptors and
> ASSUMING I was blind and using brltty and orca. Has any of you ever tried to
> use a computer whose keyboard response was delayed by nearly 2 seconds per
> keystroke because its looking for a 1200 baud tty that isn't there, and then
> yells the keystroke at you in some idea of a computer voice.  Spend 2 weeks
> trying to use a computer like that, and you'll have a new appreciation for
> my level of frustration.
> 

Your experience led me to fixing someone else's problem on IRC.
They were trying to install Debian for linuxcnc (but on an amd64 machine)
and had the same problem because they'd left a serial lead plugged in so
got brltty and orca.

I advised them to reinstall to fix things but this time to undo the serial
lead until the install was over. That worked fine for them and they were
very happy. (It's the simple fix that I think I suggested that to you three
times in the course and was ignored.)

> It should have asked me if I wanted that crap but did not.
> 

If you're deafblind and using a Braille reader - it looks for a serial tty.
That's the only way it can work. It can't ask you first necessarily.

> You have all claimed I missed it or told it to install that, but 23 damned
> times? I'm sure its here someplace because all 24 times I filled in the
> network details manually and was equally amazed when I still had a network
> on the after install reboot. That FWIW, was a first.
> 
> Everything works except printer sharing and I can't ask cups, my posts to
> the cups list are apparently routed to /dev/null, so I come here for help
> and all this dirty laundry gets drug out again. And again. And again. While
> the question I asked is very carefully ignored. Unreal.
> 

There are a bunch of cups manpages: there are lots of documents online.
The people recommending you avahi/bonjour/zeroconf are recommending it
because it works - for them and for 99.9% of people. Your system: your
rules - but it doesn't always help us help you to solve the problems.

There are reasons why we ask folk - in general on this list and in the FAQ
- to provide information, to copy/paste log entries or what they're seeing
on screen. We can't sit on your shoulder and watch what you've typed: we're
all relying on guesswork much of the time.

> Cheers, Gene Heskett.

All the very best, as ever,

Andy Cater

Re: update-initramfs

[davidson]

On Tue, 11 Apr 2023 Marc Auslander wrote:

On 4/10/2023 11:00 PM, David Wright wrote:

On Mon 10 Apr 2023 at 20:17:11 (-0400), Marc Auslander wrote:

I'm on Buster.

In /boot I keep a copy of the current working linux named by appending
-knowngood to the four files.  My idea is that if an update fails, I
have a recent working linux.  This is different from vmlinuz.old which
is the previous kernel version.  The updates in question are not to
the kernel but to initrd.image of course.

Suddenly, update-initramfs insists in trying to first update
initrd.-knowngood  which of course fails because there are no
underling file with that name.  This never happened in the past,
AFAIK. Once it fails it gives up.

There seems no way to force update-initramfs to update the right kernel.


Perhaps check that "all" hasn't been accidentally inserted:

   $ grep update /etc/initramfs-tools/update-initramfs.conf
   # Configuration file for update-initramfs(8)
   # update_initramfs [ yes | all | no ]
   # If set to all update-initramfs will update all initramfs
   # If set to no disables any update to initramfs beside kernel upgrade
   update_initramfs=yes
   $

A workaround: change the sort order of the backup initrd files
by adding an appropriate prefix, like backup-knowngood-…
so the "real" ones get updated first.

Cheers,
David.
thanks but that's the first thing I checked - it's yes, not all.  But my 
backup names contain the current version string.


I'm not sure about the sort order hack.  My goal is to have update-grub see 
the knowngood as a bootable linux and include it in the boot menu. That's 
also why .bak of initrd isn't good enough - I need a complete copy.


Caveat: I don't know WTH I'm doing. Also, I used a bullseye system for
experiments below, not buster.

I read update-initramfs(8) and did a couple experiments, one to
replicate your observations on my bullseye system, and then a variant
of David's version-naming hack.

Results are summarised below, in case you find them interesting.

EXPERIMENT #1

The first experiment simply tried to replicate your observations (as I
understood them). Basically, I added "-kg" suffix to all the files in
/boot corresponding to latest installed kernel, so that I had
unsuffixed copies and "*-kg" ("knowngood") copies, and then tried

 # update-initramfs -u

which, as you report, targeted the "*-kg" files (and turned a 28M
initrd.img-*-kg into a 6M file... no longer warranting the suffix
"-kg").

Also (after making fresh "*-kg" copies), I tried

 # update-initramfs -u -k all

which went first for the "*-kg" files, trashed the initrd.img-*-kg as
before, but then continued on to update all the other versions.

EXPERIMENT #2

The second experiment is similar to david's suggestion, but alters the
end of the name instead. (I didn't think of using a prefix.)

The manual implies that...

 update-initramfs -u

...by default tries to update the "latest" kernel version. The
following result suggests that it determines which files correspond to
that "latest" version by just looking at their filenames.

So basically instead of only adding "-kg" to the files in /boot, I
decremented the last character of uname -r and *then* added "-kg", so
that it wouldn't look like a later version.

# knowngood=( /boot/*-$(uname -r) )
# for ((i=0 ; i<${#knowngood[@]}; i+=1)) ; do cp -a "${knowngood[i]}" 
"${knowngood[i]/%4/3-kg}" ; done
# ls -l
total 105464
-rw-r--r-- 1 root root   206413 20 déc.  17:56 config-4.19.0-23-amd64
-rw-r--r-- 1 root root   236452 21 janv. 09:35 config-5.10.0-21-amd63-kg
-rw-r--r-- 1 root root   236452 21 janv. 09:35 config-5.10.0-21-amd64
drwxr-xr-x 5 root root 4096 11 avril 01:28 grub
-rw-r--r-- 1 root root 26109076 10 avril 21:59 initrd.img-4.19.0-23-amd64
-rw-r--r-- 1 root root 29199065 10 avril 21:58 initrd.img-5.10.0-21-amd63-kg
-rw-r--r-- 1 root root 29199065 10 avril 21:58 initrd.img-5.10.0-21-amd64
-rw-r--r-- 1 root root  3418327 20 déc.  17:56 System.map-4.19.0-23-amd64
-rw-r--r-- 1 root root   83 21 janv. 09:35 System.map-5.10.0-21-amd63-kg
-rw-r--r-- 1 root root   83 21 janv. 09:35 System.map-5.10.0-21-amd64
-rw-r--r-- 1 root root  5303616 20 déc.  17:56 vmlinuz-4.19.0-23-amd64
-rw-r--r-- 1 root root  7019136 21 janv. 09:35 vmlinuz-5.10.0-21-amd63-kg
-rw-r--r-- 1 root root  7019136 21 janv. 09:35 vmlinuz-5.10.0-21-amd64

# update-initramfs -u
# update-initramfs: Generating /boot/initrd.img-5.10.0-21-amd64

# update-grub
Generating grub configuration file ...
Found linux image: /boot/vmlinuz-5.10.0-21-amd63-kg
Found initrd image: /boot/initrd.img-5.10.0-21-amd63-kg
Found linux image: /boot/vmlinuz-5.10.0-21-amd64
Found initrd image: /boot/initrd.img-5.10.0-21-amd64
Found linux image: /boot/vmlinuz-4.19.0-23-amd64
Found initrd image: /boot/initrd.img-4.19.0-23-amd64
Warning: os-prober will be executed to detect other bootable partitions.
Its output will be used to detect bootable binaries on them and create new boot 
entries.
done

So update-grub didn't s

hplip on debian 11

[Russell L. Harris]

A few weeks ago, I installed Debian 11 on a Dell Vosotro-200 belonging
to a friend.  I installed the hplip package and used hp-setup, which
successfully got an old HP ink jet printer working on the system.

The friend also has a Dell Optiplex 330, on which I installed Debian
11 a few months ago.  He wished to move the inkjet printer to the
Optiplex, so, using synaptic, I tried to install hplip on the
Optiplex.  However, the SEARCH function of synaptic could find no
match for ``hplip''.

I was able to execute RELOAD, MARK ALL UPGRADES, and APPLY on the
Optiplex and upgraded chromium.  


The repository is:

deb http://ftp.debian.org/debian stable main contrib non-free

Re: how to change default nameserver?

[gene heskett]

On 4/11/23 10:04, zithro wrote:

On 11 Apr 2023 04:56, gene heskett wrote:

On 4/10/23 16:53, zithro wrote:
Why can't you follow others advice, hell, if you don't trust us, even 
the perfectly correct and up-to-date manpages ?
After reading the posts of others, I'm more and more thinking your 
simply a troll (or a RedHat fanatic wasting Debian helpers time for 
no reason) ...


That is an insult. I bailed out of fedora 15 years ago, tired to being 
an always sick lab rat for redhat. [...]


Because you answer what you want, I'll re-ask :
"Why can't you follow others advice, hell, if you don't trust us, even
the perfectly correct and up-to-date manpages ?"

Maybe it's a bug in CUPS or w/e soft you're using. Try to find other 
people having this, or report it as a bug.


Not possible. Michael and I have known each other since the '80's when 
he was a starving college student. I'll just let it go at that.


?! First, who's Michael, second, even if it's your friend, you won't
fill a bug because of that ?
What if it's really a bug, and other people have the same one ?

.


If you don't know who Mike Sweet is, why are you hassling me?  There's 
25 years of history to computing before Linus released his his linux, 
Michael and I were both there in the trs-80 color computer days and 
while Mike was take college classes and releasing code for the color 
computers in the mid 80's, it was sort of a contest among the coco users 
running os-9 on our coco's, to see who who find and fix the bugs in code 
he released. os-9 was unix that ran on a 8 bit machine with a 65k max 
memory.


Keeping networking working on linux has been an art, not a science.
looking at one of my buster machines. /etc/network/interfaces has this:

# The primary network interface
allow-hotplug eno1
iface eno1 inet static
address 192.168.71.10/24
gateway 192.168.71.1
# dns-* options are implemented by the resolvconf package, if installed
dns-nameservers 192.168.71.1
dns-search coyote.den

And it Just Works. But that was way the hell and gone too easy, so it 
just had to be changed, so that stuff no longer works for bullseye. 
There is nothing in /etc/network/interfaces.d
So let me snoop, here cuz that entry does not exist on this bullseye 
machine. Looking to see what did make it work, brb.
I can't recall which debian, but at one time I had to put it in the last 
stanza of /etc/dhcp/dhclient.conf but I see that's not been touched, 
here or on a buster machine..

/etc/nsswitch.conf looks default.

So ATM I have no clue what I did because I've forgotten whatever I did 
to make it work for bullseye, lost in the noise from doing 23 damned 
installs before someone suggested I unplug all usb, but my key board and 
mouse are wireless, so I left those buttons plugged it and did yet 
another install and this one worked. I've since rebuilt my usb tree as 
it reaches like a weeping willow to nearly every inch of the real estate 
here.  All that re-install BS caused by the installer silently finding a 
couple usb-serial adaptors and ASSUMING I was blind and using brltty and 
orca. Has any of you ever tried to use a computer whose keyboard 
response was delayed by nearly 2 seconds per keystroke because its 
looking for a 1200 baud tty that isn't there, and then yells the 
keystroke at you in some idea of a computer voice.  Spend 2 weeks trying 
to use a computer like that, and you'll have a new appreciation for my 
level of frustration.


It should have asked me if I wanted that crap but did not.

You have all claimed I missed it or told it to install that, but 23 
damned times? I'm sure its here someplace because all 24 times I filled 
in the network details manually and was equally amazed when I still had 
a network on the after install reboot. That FWIW, was a first.


Everything works except printer sharing and I can't ask cups, my posts 
to the cups list are apparently routed to /dev/null, so I come here for 
help and all this dirty laundry gets drug out again. And again. And 
again. While the question I asked is very carefully ignored. Unreal.


Cheers, Gene Heskett.
--
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author, 1940)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: how to change default nameserver?

[zithro]

On 11 Apr 2023 00:28, Andy Smith wrote:

Hello,

On Tue, Apr 11, 2023 at 12:04:24AM +0200, zithro wrote:

So, I got curious about his claim


Well you can't say you haven't been warned. This rabbit hole goes
very deep and the bottom will not contain the answers you seek!

Cheers,
Andy



Ahah, I dunno if in this case I'd like to eat the red pill ^^

Re: update-initramfs

[zithro]

On 11 Apr 2023 16:43, Marc Auslander wrote:

On 4/11/2023 9:30 AM, zithro wrote:

The solution is in "man update-initramfs" :
update-initramfs -c -k $KERNEL_VERSION

-c creates a new initramfs
-k specifies the version of the kernel
This breaks when package update tries to update-initramfs.  My copies 
have the kernel version in their names - with -knowngood appended.




Breaks how ?

In Bullseye you have to remove old kernels/initrd manually
(with for example apt autoremove)
I just checked some logs and it also worked like this in Buster
You can keep as many versions as you want.

Also, I found this old post on this very ML:

On Thu, Dec 13, 2007 at 09:54:49PM -0500, Marc Auslander wrote:
> The problem is operator (that's me) stupidity.
>
> I have an overloaded find so I can say find foo and have it mean
> find . -name foo -print
>
> mkinitramfs uses find . | cpio to build the initrd.img
>

Re: Riddling over systemctl, pulseaudio, firefox-esr, and salsa.debian.org

[David Wright]

On Mon 10 Apr 2023 at 17:42:04 (+0200), Thomas Schmitt wrote:

> i experience a strange behavior of my Debian 11 with firefox-esr and
> pulseaudio.
> After visiting
>   https://salsa.debian.org/groups/optical-media-team/-/activity
> the most busy process on my Debian 11 is the one that was started at boot
> (or user login) automatically by
>   /usr/bin/pulseaudio --daemonize=no --log-target=journal
> It uses tenfold more CPU time than the idle firefox-esr (says top(1)).
> I cannot tell whether it really tries to make noise, because no
> loudspeakers are attached.
> Before the visit to salsa.debian.org pulseaudio is not shown by top(1)
> among the busiest processes. Afterwards it steadily uses 10 percent of
> a 4 GHz Xeon core.
> Leaving the web site and even removing the browser tab does not reduce
> this activity.
> 
> So what is salsa.debian.org doing with firefox-esr to get pulseaudio so
> excited ?

I don't use pulseaudio (I'm currently browsing the pipewire docs to
see whether it can do anything for me), but googling pulseaudio cpu usage
produces a lot of hits.

> Next riddle is how i could keep pulseaudio from being started automatically
> for my desktop user. I understand from web and man page of systemctl that
>   systemctl --user disable pulseaudio.service pulseaudio.socket
> should do the trick.
> But on next boot and login (even via ssh) there is again a pulseaudio
> instance running with my desktop user id.

AFAICT that command touches files under /run, which would explain why
rebooting resets it. I just tried disabling obex.service because this
machine has no wireless. Each time I ran it, the result was:

  $ strace -f systemctl --user disable obex.service |& cat > /tmp/sysudiso

As root if you want to check those inaccessible directories:

[ … ]

/run/user/1000/systemd/generator.late:
total 4
-rw-r--r-- 1 454 Apr 11 09:55 app-geoclue\x2ddemo\x2dagent-autostart.service
drwxr-xr-x 2  60 Apr 11 09:55 xdg-desktop-autostart.target.wants/

/run/user/1000/systemd/generator.late/xdg-desktop-autostart.target.wants:
total 0
lrwxrwxrwx 1 49 Apr 11 09:55 app-geoclue\x2ddemo\x2dagent-autostart.service -> 
../app-geoclue\x2ddemo\x2dagent-autostart.service

[ … ]

I couldn't make anything out from the trace: it didn't contain the
word geoclue. It's probably all implemented by dbus messages.

> The superuser can prevent the starting of pulseaudio for my user and for
> user "lightdm" by
>   systemctl --global disable pulseaudio.service pulseaudio.socket
> 
> So do i get the man page wrong and systemctl has no means to prevent
> services for particular users ?

I'm guessing from what Michel Verdier wrote elsewhere that you
have to create a symlink, pointing at /dev/null, somewhere in
~/.config/systemd/user/. Take a look at what the global command
wrote, and copy it from /etc/systemd/user/, before removing the
global one.

Cheers,
David.

Re: update-initramfs

[Marc Auslander]

On 4/10/2023 11:00 PM, David Wright wrote:

On Mon 10 Apr 2023 at 20:17:11 (-0400), Marc Auslander wrote:

I'm on Buster.

In /boot I keep a copy of the current working linux named by appending
-knowngood to the four files.  My idea is that if an update fails, I
have a recent working linux.  This is different from vmlinuz.old which
is the previous kernel version.  The updates in question are not to
the kernel but to initrd.image of course.

Suddenly, update-initramfs insists in trying to first update
initrd.-knowngood  which of course fails because there are no
underling file with that name.  This never happened in the past,
AFAIK. Once it fails it gives up.

There seems no way to force update-initramfs to update the right kernel.


Perhaps check that "all" hasn't been accidentally inserted:

   $ grep update /etc/initramfs-tools/update-initramfs.conf
   # Configuration file for update-initramfs(8)
   # update_initramfs [ yes | all | no ]
   # If set to all update-initramfs will update all initramfs
   # If set to no disables any update to initramfs beside kernel upgrade
   update_initramfs=yes
   $

A workaround: change the sort order of the backup initrd files
by adding an appropriate prefix, like backup-knowngood-…
so the "real" ones get updated first.

Cheers,
David.
thanks but that's the first thing I checked - it's yes, not all.  But my 
backup names contain the current version string.


I'm not sure about the sort order hack.  My goal is to have update-grub 
see the knowngood as a bootable linux and include it in the boot menu. 
That's also why .bak of initrd isn't good enough - I need a complete copy.

Re: update-initramfs

[Marc Auslander]

On 4/11/2023 9:30 AM, zithro wrote:

On 11 Apr 2023 02:17, Marc Auslander wrote:

I'm on Buster.

In /boot I keep a copy of the current working linux named by appending 
-knowngood to the four files.  My idea is that if an update fails, I 
have a recent working linux.  This is different from vmlinuz.old which 
is the previous kernel version.  The updates in question are not to 
the kernel but to initrd.image of course.


In addition to what David wrote, why are you not using the backup
facility of initramfs instead of doing it manually ?

$ cat /etc/initramfs-tools/update-initramfs.conf
[...]
#
# backup_initramfs [ yes | no ]
#
# Default is no
# If set to no leaves no .bak backup files.

backup_initramfs=yes
[...]



Suddenly, update-initramfs insists in trying to first update
initrd.-knowngood  which of course fails because there are no 
underling file with that name.  This never happened in the past, 
AFAIK. Once it fails it gives up.


There seems no way to force update-initramfs to update the right kernel.

Ideas?



RTFM ? :)

The solution is in "man update-initramfs" :
update-initramfs -c -k $KERNEL_VERSION

-c creates a new initramfs
-k specifies the version of the kernel
This breaks when package update tries to update-initramfs.  My copies 
have the kernel version in their names - with -knowngood appended.

Re: how to change default nameserver?

[zithro]

On 11 Apr 2023 04:56, gene heskett wrote:

On 4/10/23 16:53, zithro wrote:
Why can't you follow others advice, hell, if you don't trust us, even 
the perfectly correct and up-to-date manpages ?
After reading the posts of others, I'm more and more thinking your 
simply a troll (or a RedHat fanatic wasting Debian helpers time for no 
reason) ...


That is an insult. I bailed out of fedora 15 years ago, tired to being 
an always sick lab rat for redhat. [...]


Because you answer what you want, I'll re-ask :
"Why can't you follow others advice, hell, if you don't trust us, even
the perfectly correct and up-to-date manpages ?"

Maybe it's a bug in CUPS or w/e soft you're using. Try to find other 
people having this, or report it as a bug.


Not possible. Michael and I have known each other since the '80's when 
he was a starving college student. I'll just let it go at that.


?! First, who's Michael, second, even if it's your friend, you won't
fill a bug because of that ?
What if it's really a bug, and other people have the same one ?

Re: How to make apt-cdrom read Flashdisk?

[David Wright]

On Tue 11 Apr 2023 at 16:00:22 (+0700), SteffenTAN wrote:
> so I got a Last School Work but, They didn't give Me a CD/DVD to Install
> Debian 10.4, Samba and isc-dhcp-server
> 
> They only give Me 3 files with name of DVD 1, 2 and 3
> so, I searching for answer then, I use rufus to Install the 1st DVD (Debian
> OS) it's work. not work for CD 2 and 3 because the format then, I think I
> need to extract the files into flashdisk.
> 
> but, I can't install samba because the "apt-cdrom" stay read from CDRom
> while the File was on Flashdisk

apt-cdrom has a number of switches, documented in   man apt-cdrom
and it sounds like you may need:

  apt-cdrom --no-auto-detect -m -d /mnt add

--no-auto-detectDon't mess with the optical drive
-m  Assume the flashdrive has been mounted
-d /mnt /mnt is where the flashdrive is mounted

You might need to mount the stick first, in your usual manner,
and use -d to tell apt-cdrom where it is mounted.

If you have an automounter, then you should just determine
where it got mounted, and use that mountpoint name instead.
It might be automounted under /media.

Cheers,
David.

Re: ICMP router advertisement (ipv4)

[Dan Ritter]

Anssi Saari wrote: 
> Jeremy Ardley  writes:
> 
> > Your only option seems to be to sign up with some external IPv6
> > provider. This service (I've never used it so beware) says it gives
> > you ipv6 etc for free. What their business model is I'm not sure 
> > https://tunnelbroker.net/
> 
> I doubt that's going to work for him either due to his peculiar setup
> with DS-lite. With a quick look tunnelbroker.net only provides a 6to4
> tunnel and that's going to need a public IPv4 address which he doesn't
> have.

Tunnelbroker.net can use a dynamic update feature so that the
static IPv6 address is only momentarily interrupted when the
underlying v4 address is changed out. This is quite reliable if
you hook it to the the DHCP client's post-up function.

I use this in preference to my ISPs native v6 support because
tunnelbroker (Hurricane Electric) supplies static v6
assignments, and my ISP does not.

-dsr-

Re: update-initramfs

[zithro]

On 11 Apr 2023 02:17, Marc Auslander wrote:

I'm on Buster.

In /boot I keep a copy of the current working linux named by appending 
-knowngood to the four files.  My idea is that if an update fails, I 
have a recent working linux.  This is different from vmlinuz.old which 
is the previous kernel version.  The updates in question are not to the 
kernel but to initrd.image of course.


In addition to what David wrote, why are you not using the backup
facility of initramfs instead of doing it manually ?

$ cat /etc/initramfs-tools/update-initramfs.conf
[...]
#
# backup_initramfs [ yes | no ]
#
# Default is no
# If set to no leaves no .bak backup files.

backup_initramfs=yes
[...]



Suddenly, update-initramfs insists in trying to first update
initrd.-knowngood  which of course fails because there are no 
underling file with that name.  This never happened in the past, AFAIK. 
Once it fails it gives up.


There seems no way to force update-initramfs to update the right kernel.

Ideas?



RTFM ? :)

The solution is in "man update-initramfs" :
update-initramfs -c -k $KERNEL_VERSION

-c creates a new initramfs
-k specifies the version of the kernel

Re: ICMP router advertisement (ipv4)

[Anssi Saari]

Jeremy Ardley  writes:

> Your only option seems to be to sign up with some external IPv6
> provider. This service (I've never used it so beware) says it gives
> you ipv6 etc for free. What their business model is I'm not sure 
> https://tunnelbroker.net/

I doubt that's going to work for him either due to his peculiar setup
with DS-lite. With a quick look tunnelbroker.net only provides a 6to4
tunnel and that's going to need a public IPv4 address which he doesn't
have.

A WireGuard tunnel from route64.org works through CGNAT. Using that
tunneling setup though... He'd have a wireguard tunnel for IPv6 which
goes through DS-Lite IPv4-in-IPv6 pipe through IPv4 CGNAT to the tunnel
broker? Very cool if it works.

I guess actually contacting Virgin and asking technical questions is not
a possibility?

Re: how to change default nameserver?

[Greg Wooledge]

> > > So Gene, can you tell us where you read this ?
> > 
> > In a man page from a good 20 years ago. I still have a copy of that
> > original redhat 5.0 on a shelf above me, but not a floppy drive to read
> > those disks with.
> 
> Well, it's not in resolver(5) (which is for resolv.conf) on Red Hat 5.0.5.

It wasn't in the man page from Red Hat 5.2 when I checked in 2017, either.

https://lists.debian.org/debian-user/2017/09/msg01038.html

Here's another thread from 2018:

https://lists.debian.org/debian-user/2018/08/msg01433.html

Note that this time, Gene omitted the comma, and said that the
resolv.conf file "looks something like this"; IOW he's going from
memory rather than just catting the file.

And a thread from 2019:

https://lists.debian.org/debian-arm/2019/06/msg00046.html

This time the comma is gone, and Gene may or may not have actually
catted the file -- we can't be sure, because he does not paste the actual
terminal session including the shell prompt and the command issued.

As I've said, we've been through this for literally YEARS.

Re: How to make apt-cdrom read Flashdisk?

[Thomas Schmitt]

Hi,

SteffenTAN wrote:
> They only give Me 3 files with name of DVD 1, 2 and 3
> so, I searching for answer then, I use rufus to Install the 1st DVD (Debian
> OS) it's work.

I guess you could install a base system from the "DVD-1" stick and then
put the mount point addresses of the two other "DVD" images into
  /etc/apt/sources.list
of the installed system before you install more packages.
But that's Debian sysadmin work which others can explain better.

(I see that Timothy M Butterworth proposes a similar approach in
  https://lists.debian.org/debian-user/2023/04/msg00464.html
  "Can't you install software from the online repos?"
I.e. you ignore DVD-2 and 3 and rather use Debian's normal way of
installing additional packages from the internet repositories by
help of "apt" or "apt-get". The matching entry in sources.list gets
created by the installation from DVD-1.
Better ask your teacher first, whether this method is ok.)

-
Alternative approaches:

Possibility 2 would be the one which Pete Batard, the author of Rufus,
defends with great emphasis every time that it gets broken by distro ISOs.
Recently in
  https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1031696
he wrote
  "In short, the expectation of many UEFI users -- [...]
   -- is that they should be able to [format] a USB media to FAT32 and
   then extract the full content from the Debian ISO there, to end up
   with a media that they can use to both boot and install Debian."
I would bet that this works for multiple ISOs, too.
When trying i would first unpack DVD-2 and DVD-3 to the freshly formatted
FAT32 on the USB stick and only then DVD-1. So if any files are on DVD-1
and another DVD image, then the ones from DVD-1 would survive.

-

A third possibility would be to ask the teacher for permission to use
  debian-10.4.0-amd64-STICK16GB-1.iso
which you can get from
  
https://cdimage.debian.org/mirror/cdimage/archive/10.4.0/amd64/jigdo-16G/debian-10.4.0-amd64-STICK16GB-1.jigdo
and is supposed to contain what DVD-1 to 3 contain plus some more.
You'd need the program "jigdo-lite" which is available as MS-Windows
version from Steve McIntyre, a main Debian Developer of the installation
ISO images at
  https://www.einval.com/~steve/software/jigdo/download/jigdo-win-0.8.1.zip
See also the "Downloads" paragraph on
  https://www.einval.com/~steve/software/jigdo/
Instructions how to use it are at "4.2" and "4.3" of
  https://tldp.org/HOWTO/Debian-Jigdo/downloadingyourfirstimage.html

Possibility 3 1/2 would be to follow
  https://wiki.debian.org/JigdoOnLive
which explains how to use a Debian Live system (from USB stick) to
download the ISO by help of Debian's package "jigdo-file".
This wiki age also gives a more recent example on how to use program
jigdo-lite. See
  https://wiki.debian.org/JigdoOnLive#Install_package_jigdo-file
I assume that this example works with the MS-Windows binary too.

-

Possibility 4 would be to merge the three DVD images into one. See
  https://wiki.debian.org/MergeDebianIsos
This needs a Unix-like shell environment (CygWin ?) and program xorriso,
of which .exe binaries exist in the web.
The resulting ISO will not have a Joliet tree and thus MS-Windows will show
only its dull ISO 9660 names when you inspect the result. This will not
hamper installation by Debian because as a GNU/Linux it will perceive the
Rock Ridge names which are even better than Joliet.
(The script "merge_debian_isos" can be modified to add Joliet.)


Have a nice day :)

Thomas

Re: How to make apt-cdrom read Flashdisk?

[Timothy M Butterworth]

On Tue, Apr 11, 2023 at 5:18 AM SteffenTAN <
steffentansoehiantosoeti...@gmail.com> wrote:

> so I got a Last School Work but, They didn't give Me a CD/DVD to Install
> Debian 10.4, Samba and isc-dhcp-server
>

Debian 10.4 is pretty old. Is there a reason why you have to use that
version?



> They only give Me 3 files with name of DVD 1, 2 and 3
> so, I searching for answer then, I use rufus to Install the 1st DVD
> (Debian OS) it's work. not work for CD 2 and 3 because the format then, I
> think I need to extract the files into flashdisk.
>

Are the three files iso images. Do they end in .iso?


> but, I can't install samba because the "apt-cdrom" stay read from CDRom
> while the File was on Flashdisk
> please help Me
>

Can't you install software from the online repos?


> I'm very frustated
>


-- 
⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org/
⠈⠳⣄⠀⠀

How to make apt-cdrom read Flashdisk?

[SteffenTAN]

so I got a Last School Work but, They didn't give Me a CD/DVD to Install
Debian 10.4, Samba and isc-dhcp-server

They only give Me 3 files with name of DVD 1, 2 and 3
so, I searching for answer then, I use rufus to Install the 1st DVD (Debian
OS) it's work. not work for CD 2 and 3 because the format then, I think I
need to extract the files into flashdisk.

but, I can't install samba because the "apt-cdrom" stay read from CDRom
while the File was on Flashdisk
please help Me

I'm very frustated

Re: questions about cron.daily

[Max Nikulin]

On 09/04/2023 14:54, Michel Verdier wrote:

Le 8 avril 2023 Max Nikulin a écrit :

There is ready to use one: /usr/lib/systemd/user/emacs.service Perhaps there
is no such file in buster.


/usr/lib/systemd/user is for global system running. If you want to change
something in the service you copy it in ~/.config/systemd to supersede
global one.


For minor tuning (adding new parameters or overwriting specific ones) it 
is possible to create a drop-in file, e.g. 
~/.config/systemd/user/emacs.service.d/override.conf (or in /etc for all 
users).


systemctl --user edit emacs.service

See systemd.unit(5). Examples are usually given for system, not user units:
- https://wiki.debian.org/systemd#Creating_or_altering_services
- https://wiki.archlinux.org/title/systemd#Editing_provided_units
- 
https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/8/html/configuring_basic_system_settings/assembly_working-with-systemd-unit-files_configuring-basic-system-settings#proc_extending-the-default-unit-configuration_assembly_working-with-systemd-unit-files

Re: ICMP router advertisement (ipv4)

[debian-user]

Jeremy Ardley  wrote:
> On 11/4/23 11:40, Tim Woodall wrote:
> >
> >
> > My googling suggests that a superhub or hub 5 can be switched to
> > 'modem only' mode but I've got a hub 6 which doesn't have that
> > option. 
> >>
> >>   Virgin Media: Virgin Media is the largest cable broadband
> >> provider in the UK, operating its own network separate from
> >> Openreach's infrastructure. Virgin Media's network is based on
> >> Hybrid Fibre-Coaxial (HFC) technology, which delivers high-speed
> >> internet using a combination of fiber-optic and coaxial cables.
> >> Virgin Media is not obligated to provide wholesale access to its
> >> network, meaning customers can only access services directly from
> >> Virgin Media. 
> >  
> Looks like you have HFC service. In itself that's good, but Virgin
> don't let anyone else on their HFC network which is bad. Worse is
> that they don't use an industry standard modem such as an Arris unit.
> Instead they use a proprietary NTD and router in the same box.
> 
> Your only option seems to be to sign up with some external IPv6 
> provider. This service (I've never used it so beware) says it gives
> you ipv6 etc for free. What their business model is I'm not sure 
> https://tunnelbroker.net/

Note that Dublin is in Eire and Eire is not part of the UK. The Virgin
company operating in Ireland is different to that operating in the UK.
Pages may help such as 

https://www.havevirginmediaenabledipv6yet.co.uk/
https://community.virginmedia.com/t5/QuickStart-set-up-and/IPv6-support-on-Virgin-media/td-p/35748/page/129
 
> Jeremy
> (Lists)