You are not logged in.
1) Keep kernel 5.10 as a safeguard (in case of any kernel panic when compiling any new/custom kernel)
The installed kernels should always be available as boot options if GRUB was installed from the Devuan system.
You may might have to make a new boot entry for your custom kernel though. Not sure if grub-mkconfig will pick it up automatically.
2) Compile a new kernel using the config-5.10.0-8-amd64 (file located at /boot) as starting point
From the custom kernel source tree:
cp /boot/config-$(uname -r) .config
make oldconfigSee also https://wiki.gentoo.org/wiki/Kernel/Con … figuration
3) New kernel does not need to be packaged as .deb
Well yeah, nobody needs chocolate but it's really nice...
https://kernel-team.pages.debian.net/kernel-handbook/
EDIT: corrected command block.
SLiM is old and broken. Does startx work from the console?
FWIW the X.Org log looks truncated.
blackhole wrote:headstick will be nursing a semi reading this....
What does that mean? And why exactly?
Don't feed the troll ![]()
Interesting news, thanks for sharing.
wayland seems dicey, especially if any of this is true
For an alternative opinion see https://drewdevault.com/2019/02/10/Wayl … unked.html & https://drewdevault.com/2021/02/02/Anti … eshit.html
I've just debootstraped a Debian buster system to test:
$ cat /tmp/debian/etc/hosts
127.0.0.1 localhost
::1 localhost ip6-localhost ip6-loopback
ff02::1 ip6-allnodes
ff02::2 ip6-allrouters
$@OP: Was that the content prior to your modifications? localhost should work just as well as the actual FQDN (AFAIK).
why not in the dependencies of the package?
ALSA is part of the kernel. The user space utilities are already installed.
If you need help configuring it see https://wiki.archlinux.org/title/Advanc … sound_card
Note that rxvt-unicode will read ~/.Xdefaults if there are no resources set (ie, if ~/.Xresources does not exist). The file is read automatically on startup so there's no need to run xrdb --merge to apply any configuration changes.
The man pages are very good, especially section 7 (the configuration options).
Sorry OP, my posts in this thread are irrelevant — I failed to read through the thread properly before answering and didn't appreciate that you're trying to boot a USB stick rather than an installed system.
Apologies for the noise.
Head_on_a_Stick wrote:So did you check your boot entries?
efibootmgr -uvNo. Sorry if I'm being dense.
But I am running Windows 10 on this PC.
The equivalent command under Windows would be
bcdedit /enum bootmgrif I want to run Synaptic
How are you starting Synaptic?
the graphical password box reports back incorrect password - even though it works when su in terminal
Which password are you entering? I think pkexec will expect the user's password so if you're entering the root password (as you would for the su command) then it won't work.
Only just started this issue today
If the behaviour has changed recently then we should probably check your sources so please also post the output of
apt policycat <<EOF | sudo tee /etc/apt/sources.list.d/mozilla.list
deb http://downloads.sourceforge.net/projec … ozilla/apt all main
EOF
Leave that poor cat alone!
sudo tee /etc/apt/sources.list.d/mozilla.list <<!
deb http://downloads.sourceforge.net/project/ubuntuzilla/mozilla/apt all main
!And are you sure the UbuntuZilla repository is compatible with Devuan? Most PPAs aren't.
if im in UEFI mode only, the installer works normally, but i noticed that it didnt ask me about the disk to copy grub, it only says that its running grub-mkconfig or something like that, but didnt ask me the disk to implement the boot and grub, that shows when i use legacy mode.
That is the expected behaviour. UEFI systems use the EFI system partition to store GRUB's core.img (grubx64.efi) and the motherboard NVRAM to store the boot entries so there is no need to specify the target drive. The UEFI grub-install command does not expect a block device as an argument and will not recognise it if it is provided; the equivalent option for UEFI systems would be --efi-directory (pointing to the EFI system partition).
In UEFI only mode, also, when i reboot, it shows an error saying that i dont have any OS installed, as its like the disk doesnt have any boot installed
Did you try disabling Secure Boot? Devuan chimaera should support it but some motherboards require that .efi loaders be authorised in the firmware ("BIOS") settings first.
You get quite a few hits if you search for
grub "Image failed to verify with ACCESS DENIED"for example https://bugs.launchpad.net/ubuntu/+sour … ug/1528345
So did you check your boot entries?
efibootmgr -uv^ That will show the .efi loader used for each entry.
Other potential causes of a lack of UEFI variable support and ways to fix them are listed here:
https://wiki.archlinux.org/title/Unifie … le_support
if i used UEDFI only, the installation did not copied grub to the installation disk
What does that mean, exactly? What actually happens when you try? What do you see in the installer? Are there any errors? If so did you check TTY4 for the full message?
If the installer proceeds without error then what happens when you start the machine up afterwards? What do you see?
I like trolling. Trolling is fun. I left fdn because they banned trolling. Bastards.
Alternatively rename /boot/efi/EFI/devuan/ to /boot/efi/EFI/debian/ and keep the signed GRUB package so that Secure Boot will actually work.
See also https://bugs.debian.org/cgi-bin/bugrepo … bug=925309
EDIT: corrected paths.
Check the logs.
The firmware package for AMD graphics cards is called firmware-amd-graphics.
Check dmesg to see if it's loaded and check the X.Org log for errors and to see which DDX drivers you're using. lspci -k will show the loaded kernel drivers.
Debian now runs on the M1:
https://twitter.com/alyssarzg/status/14 … 1058194436
The display driver works and will be added to the mainline kernel in due course.
EDIT: and the developer runs a dwm desktop. Nice.
Yeah musl rules :-)
You can run glibc programs from a flatpak or chroot anyway:
Isn't their kernel installable in Devuan?
Remember to use their hardened malloc implementation too:
it's just cruising along at 250mb
Is that the musl libc version? I've recently installed that myself but it's at the bare console stage atm.
@Jakoline: I've just tested the Debian bullseye netinstall image and that doesn't seem to install rfkill if no desktop is selected in tasksel. Have you considered reporting this to Debian?
And for general reference:
Assuming the installer brought up the network, one way forward for you would be to start another installation but stop before partitioning; i.e., DON'T partition and instead use ctrl-alt-f2 to gain a shell.
In that shell, you first mount the previously installed file system partition, then chroot into that, and then install rfkill.
The Devuan beowulf netinstall ISO includes a "Rescue Mode" available from the "Advanced Options" boot sub-menu — this can be used to open a shell in an installed system without needing to mess around with chroot (or arch-chroot(8)).
More details on this here: https://www.debian.org/releases/stable/ … 06.en.html
rfkill WAS NOT INSTALLED!
Are you sure? Did you try
/usr/sbin/rfkillOr perhaps even
man rfkillSee also
echo $PATH