You are not logged in.
Update:
The original post was not very precise: xrdp is the problem.
And I found a somewhat dirty hack to fix the keyboard layout.
The loaded keymap is:
sudo cat /var/log/xrdp.log | grep km-000 | tail -1
[20230106-12:31:06] [INFO ] Loading keymap file /etc/xrdp/km-00000409.iniWhich is "us". "de" is "km-00000407.ini".
Backup and copy:
sudo gzip /etc/xrdp/km-00000409.ini
sudo cp /etc/xrdp/km-00000407.ini /etc/xrdp/km-00000409.iniAfter logging in, it is still a "us" keymap, but switching to "de" now works
setxkbmap deWould prefer a cleaner solution, but at least it works by now.
Edit: Will mark it as solved. And look for another remote X session solution.
Download unofficial non-free image of Debian and convert it installed to Devuan.
Can't confirm that. The modprobe -i nvidia-current error in /var/log/dmesg is present in both cases, migration with installed nvidia drivers and devuan installation.
But, does it do something at all?
lsmod says nvidia modules are loaded, I' just ignoring it.
Hopefully it is not to late!
Billy Talent for the rescue?
https://www.youtube.com/watch?v=hNrj4reWIgY
Hi,
did you try to rebuild the locales?
sudo locale-genIt builds entries listed in /etc/locale.gen
I'm happily running daedalus, repo is
deb http://deb.devuan.org/merged daedalus main contrib non-freeThe output of neofetch regarding OS, Kernel and Shell is equal.
Not exactly sure about the OS name "daedalus/ceres".
Packages are tested in ceres, before integrated in daedalus.
(edit: typo)
Worst case is a new installation ... . Anyway.
Probably you want to migrate to daedalus?
It is closer to ceres.
Hi,
changing the release is always the same.
edit /etc/apt/sources.list - replace ceres with chimaera
apt update
apt upgrade
apt dist-upgrade To downgrade or make apt install older packages over newer:
cat /etc/apt/preferences
Package: *
Pin: release n=chimaera
Pin-Priority: 1001I messed the file up here https://dev1galaxy.org/viewtopic.php?id=5208
apt -s whatever does a dry run.
Hi,
when I'm at home, having a wire guard vpn connection to the company, via remote-desktop to xrdp (virtual session, not screen mirror): The keyboard layout switches to 'en_US'.
The reason is unknown, this does not happen when the connection comes form a local machine. Anyway.
I used to have a German keyboard, anything else is English / en_GB.
To switch back to the layout:
$ setxkbmap -layout "de" -nodeadkeys # same with or without 'nodeadkeys'This works not 100%. The [Altgr] key stays dead, which means none of {[]}\| .
What am I missing? Or
How to get the [Altgr] key working?
Thanks and regards.
Hi,
I disabled the spectre CPU-patches.
This is not recommend for e.g. cloud servers running potentially evil virtual machines. My computer is just a desktop.
The arguments are from https://make-linux-fast-again.com
# /etc/default/grub
GRUB_CMDLINE_LINUX_DEFAULT="quiet noibrs noibpb nopti nospectre_v2 nospectre_v1 l1tf=off nospec_store_bypass_disable no_stf_barrier mds=off tsx=on tsx_async_abort=off mitigations=off"
$ sudo update-grubWhat about
apt remove apparmorGuarantied to work with every init system.
Hi,
the gvfs-* packages are useful for the mounting features in pcmanfm, e.g. disk partitions, usb drives when connecting, MTP devices / androd cell phones. Not strictly needed, but quite handy. Taste is different of course.
The Librewolf Appimage has working sound using ALSA (the build from the apt repo does not, for whatever reason).
Maybe an option ... .
EDIT: Typo
What worked for me was a slight modification of the order:
apt install sysvinit-core
reboot
Then following the manual instructions.
The basic idea was to change pid 1 to init before doing something else.
That was after reading https://dev1galaxy.org/viewtopic.php?id=5147 post#6, and rrq's migration script and jaromil's ansible.
EDIT: Or if you want to follow the original procedure successfully, then consult farmatito's script for additional measures.
Thanks a lot - that new feature made me nuts!
This Issue came out other than I initially thought.
Dear admin,
suggestion is to move the whole thread from "Freedom Hacks" to "Installation" or "Other Issues". On occation.
Thanks.
Hm, how would the following look in doas style?
(Capt. Obviuos does not want to compile as root)
$ ./configure
$ make
$ sudo make installThat's better- thanks a lot!
Good to know, that apt is still working as expected.
Quite sure I have had some trys using a=oldstable, but there must be another error incorporated.
$ apt policy
Package files:
100 /var/lib/dpkg/status
release a=now
1001 http://ftp.de.debian.org/debian buster/contrib amd64 Packages
release v=10.12,o=Debian,a=oldstable,n=buster,l=Debian,c=contrib,b=amd64
origin ftp.de.debian.org
1001 http://ftp.de.debian.org/debian buster/non-free amd64 Packages
release v=10.12,o=Debian,a=oldstable,n=buster,l=Debian,c=non-free,b=amd64
origin ftp.de.debian.org
1001 http://ftp.de.debian.org/debian buster/main amd64 Packages
release v=10.12,o=Debian,a=oldstable,n=buster,l=Debian,c=main,b=amd64
origin ftp.de.debian.org
Pinned packages:This is from debian bullseye, backdate target is buster. The chimaera image showed a comparable behaviour.
Thanks for looking!
$ cat /etc/apt/preferences{,.d/*}
Package: *
Pin: release a=buster
Pin-Priority: 1001
cat: '/etc/apt/preferences.d/*': No such file or directory
$ apt policy
Package files:
100 /var/lib/dpkg/status
release a=now
500 http://ftp.de.debian.org/debian buster/contrib amd64 Packages
release v=10.12,o=Debian,a=oldstable,n=buster,l=Debian,c=contrib,b=amd64
origin ftp.de.debian.org
500 http://ftp.de.debian.org/debian buster/non-free amd64 Packages
release v=10.12,o=Debian,a=oldstable,n=buster,l=Debian,c=non-free,b=amd64
origin ftp.de.debian.org
500 http://ftp.de.debian.org/debian buster/main amd64 Packages
release v=10.12,o=Debian,a=oldstable,n=buster,l=Debian,c=main,b=amd64
origin ftp.de.debian.org
Pinned packages:
# some details and commands:
$ cat /etc/debian_version
11.4
$ cat sources.list
deb http://ftp.de.debian.org/debian/ buster main non-free contrib
$ apt update
Hit:1 http://ftp.de.debian.org/debian buster InRelease
All packages are up to date.
$ apt upgrade
Calculating upgrade... Done
0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.
$ apt list -a | grep linux-image-amd64/
linux-image-amd64/now 5.10.136-1 amd64 [installed,local]
linux-image-amd64/oldstable 4.19+105+deb10u15 amd64
$ apt install linux-image-amd64
linux-image-amd64 is already the newest version (5.10.136-1).
0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.
# downgrade single package:
$ apt -s install linux-image-amd64=4.19+105+deb10u15
0 upgraded, 1 newly installed, 1 downgraded, 0 to remove and 0 not upgraded.
Inst linux-image-4.19.0-20-amd64 (4.19.235-1 Debian:10.12/oldstable [amd64])
Inst linux-image-amd64 [5.10.136-1] (4.19+105+deb10u15 Debian:10.12/oldstable [amd64])Hi,
About ~18 month ago it was still possible to downgrade package versions by setting the releases pin priority to 1001. Applied with apt update - nice and easy.
Now, apt happly tells you, that the newest version of the package is installed. Have a cookie!
Is it possible to restore the previous behavoir?
Supplemental (not really part of the question, just how it happened)
The pin priority in /etc/apt/preferences is not applied directly. Checking with apt-cache policy, resulted in a values of "990 : 500 : 100" instead of "1001 : -1 : -10". Remarkable may be: Pin priority 1001 for 'force downgrade' is gone.
What's the purpose?
Distro downgrade (to try lcarsde). Sounds complicated, but it wasn't. The process was straight forward. At the time I moved a box from testing (bullseye) back to stable (buster):
Adjust /etc/apt/sources.list, pin the release to 1001 in /etc/apt/preferences, apt update; apt upgrade; apt dist-upgrade; reboot with the "new" 4.19 kernel.
That really impressed me! (later the box became bullseye again and is now chimaera)
Yesterday I tried the same with chimaere to beowulf and bullseye to buster. Countless variants of pinning and sources.list - nope.
Use option '--no-install-recommends' when you install the nvidia driver. This will prevent nvidia-persistenced from being installed
Now I'm feeling foolish - all the time wasted in the last two months on a pointless package.
Edit: There might be a take away for the OP:
If it's not working, just uninstall the whole nvidia-* to use the open souce driver nouveau. It's not worth the effort.
if you are lucky, it is just the package status:
/etc/init.d/nvidia-persistenced stop
apt-get upgrade
Hi,
depending on the partition sizes, it could be an option to mount sda3 on "/home". Minimal effort and recommended anyway.
@alexkemp
The vt's have nothing to do with X, they are steady. The underlying process is a getty.
e.g. you can switch from X to vt3 with Alt-Ctrl-F3, but from a vt to vt3 with Alt-F3 (with or without Ctrl).
Another point of view is to see a vt as text console and a X as graphical console.
$ ps ax | grep getty
1781 tty1 Ss+ 0:00 /sbin/getty --noclear 38400 tty1
1782 tty2 Ss+ 0:00 /sbin/getty 38400 tty2
(...)
1786 tty6 Ss+ 0:00 /sbin/getty 38400 tty6"I also still do not know which program is supplying the Alt-F7 command."
Good question - never crossed my mind - vt's were present as far as I can think.
root@chimaera:~# cat /etc/pihole/setupVars.conf PIHOLE_DNS_1=127.0.0.1#5335
I have the same line for PIHOLE_DNS_1. There is an unbound running on my machine, configured for port 5335 to resolve DNS (setup as described in https://forum.kuketz-blog.de/viewtopic.php?f=42&t=3067 (german)).
Did you try a "common" DNS server, like rbit suggested?
Or what is your local DNS resolver on port 5335 and is it working?
Has anyone here at Dev1 installed/run Pi-Hole?
I'm running a pi-hole on chimaera, installed before migration.
(beside the adaption of the changed network device name) no further issues so far, but I'm not frequently updating.