You are not logged in.
Weirdly, while I am unable to get a list of apps (including flatpaks) to install via Discover, I get prompted to update the system and am able to update the system in practice using Discover. So in some fashion, Discover is network aware, but not when it comes to software that hasn’t yet been installed…
Thanks for the post Lorenzo. Things worked as described, but Discover still can’t find a connection, despite the ethernet connection clearly working.
I just tried two more installs. What works best (relative term) on bare metal is the following:
# apt install xorg xinit xterm
# apt install kde-standard
# reboot
There are a few extra apps installed that I am not interested in, however for the most part, the system behaves as I expect it to. The exceptions are as follows:
SDDM is installed, but doesn't appear at boot. This is probably due to the second exception. Interestingly, SDDM appears when I choose one of the commands from the application menu (logout, restart, shutdown).
An ethernet connection is available and working, but not with NetworkManager, which is also likely the reason that Discover doesn't detect an internet connection.
Cups doesn't appear to be working either.
I can live without SDDM and Discover not working properly, but it would be nice if they did...
---
Not sure how, but Discover looks like it is working now. Only lacking SDDM...
The Q4OS script solution, unfortunately still looks like the only way to get a fully functioning KDE Plasma install on a runit init base. I would have preferred a fully Devuan install, but it looks like the Q4OS script or a Peppermint OS runit install are the only options for getting a fully working runit Devuan based system, the later with XFCE not KDE…
I followed the instructions exactly, but after having tried two more installs to get sddm working, I installed lightdm instead. Lightdm starts into a kde plasma gui that has a number of issues:
Unable to resize windows.
The kde application launcher kicked up some errors as well:
Errors:
file:///usr/share/plasma/plasmoids/org.kde.plasma.kickoff/contents/ui/Kickoff.qml:157:34: Type FullRepresentation unavailable
file:///usr/share/plasma/plasmoids/org.kde.plasma.kickoff/contents/ui/FullRepresentation.qml:80:22: Type NormalPage unavailable
file:///usr/share/plasma/plasmoids/org.kde.plasma.kickoff/contents/ui/NormalPage.qml:43:13: Type Footer unavailable
file:///usr/share/plasma/plasmoids/org.kde.plasma.kickoff/contents/ui/Footer.qml:155:5: Type LeaveButtons unavailable
file:///usr/share/plasma/plasmoids/org.kde.plasma.kickoff/contents/ui/LeaveButtons.qml:14:1: module "org.kde.kitemmodels" is not installed
My wired ethernet connection functions, but there is no NetworkManager directory, and no .conf file:
# cat /etc/network/interfaces
# This file describes the network interfaces available on your system
# and how to activate them. For more information, see interfaces(5).
source /etc/network/interfaces.d/*
# The loopback network interface
auto lo
iface lo inet loopback
# The primary network interface
auto eth0
iface eth0 inet dhcp
# cd /etc/NetworkManager
bash: cd: /etc/NetworkManager: No such file or directory
So far, so good. I followed your lead and installed a minimal system before adding xorg and kde-standard, with no recommendations. (I personally view this as counterintuitive, when I am using an “expert install”, but ok.)
Everything seems to have worked so far, but the system isn’t booting into sddm at start. Since this is a runit init install, I assume something needs to be done manually, but I am unsure if sddm is launched by runit or by one of the remaining sysv init scripts 🤔
—-
sddm appears in a list when I run the following:
# cd /etc
# ls sv
This would suggest that sddm should be started by runit. Running the following command would suggest that sddm does get started:
$ ps ax | grep run
returns “runsv sddm” and “runsv elogind”
I should mention that I continue to need to press the return key after boot is finished, to get a command line login.
Here is a more in depth look at the result of running the q4os script on a Devuan KDE runit install:
$ ps ax | grep run
1 ? Ss 0:00 runit
1542 ? Ss 0:00 runsvdir -P /etc/service log: .....................................................................................................................................................................................................................................................................................................................................check: 3: socklog-check: not found check: 3: socklog-check: not found
1593 ? S 0:00 /usr/sbin/smartd --pidfile /var/run/smartd.pid
1685 ? Ss 0:00 runsv getty-tty4
1686 ? Ss 0:00 runsv cups
1687 ? Ss 0:00 runsv getty-ttyS0
1688 ? Ss 0:00 runsv dbus
1689 ? Ss 0:00 runsv dhclient
1690 ? Ss 0:00 runsv getty-tty3
1691 ? Ss 0:00 runsv anacron
1692 ? Ss 0:00 runsv cron
1693 ? Ss 0:00 runsv sddm
1694 ? Ss 0:00 runsv acpid
1695 ? Ss 0:00 runsv elogind
1696 ? Ss 0:00 runsv getty-tty6
1697 ? Ss 0:00 runsv getty-tty5
1698 ? Ss 0:00 runsv rsyslog
1699 ? Ss 0:00 runsv dbus.dep-fixer
1700 ? Ss 0:00 runsv getty-tty2
1701 ? Ss 0:00 runsv default-syslog
1702 ? Ss 0:00 runsv getty-tty1
1706 ? S 0:00 svlogd -tt -b 2048 /var/log/runit/dbus
1709 ? S 0:00 svlogd -tt -b 2048 /var/log/runit/sddm
1712 ? S 0:00 svlogd -tt -b 2048 /var/log/runit/dhclient
1716 ? S 0:00 svlogd -tt -b 2048 /var/log/runit/anacron
1719 ? S 0:00 svlogd -tt /var/log/runit/acpid
1919 ? S 0:00 avahi-daemon: running [host_computer.local]
1924 tty7 Ssl+ 0:00 /usr/lib/xorg/Xorg -nolisten tcp -auth /var/run/sddm/{5bc3cb32-6e3c-4fe0-ae1b-8f42736e3c47} -background none -noreset -displayfd 17 -seat seat0 vt7
1980 ? S 0:00 /sbin/wpa_supplicant -u -s -O /run/wpa_supplicant
2112 ? S 0:00 /usr/lib/x86_64-linux-gnu/sddm/sddm-helper --socket /tmp/sddm-auth5d07c714-2378-448c-b18f-637ebc26e64d --id 1 --start /usr/lib/x86_64-linux-gnu/libexec/plasma-dbus-run-session-if-needed /usr/bin/startplasma-wayland --user brian
2117 tty8 Ss+ 0:00 dbus-run-session /usr/bin/startplasma-wayland
2142 tty8 Sl+ 0:03 /usr/bin/kwin_wayland --wayland-fd 7 --socket wayland-0 --xwayland-fd 8 --xwayland-fd 9 --xwayland-display :1 --xwayland-xauthority /run/user/1000/xauth_AViBbU --xwayland
2163 tty8 Sl+ 0:00 /usr/bin/Xwayland :1 -auth /run/user/1000/xauth_AViBbU -listenfd 49 -listenfd 50 -displayfd 41 -rootless -wm 44
2328 tty8 S+ 0:00 /usr/bin/dbus-daemon --config-file=/usr/share/defaults/at-spi2/accessibility.conf --nofork --print-address 11 --address=unix:path=/run/user/1000/at-spi/bus_1
2644 tty8 Sl+ 0:00 /usr/lib/x86_64-linux-gnu/libexec/baloorunner
2782 ? Ss 0:00 fusermount3 -o rw,nosuid,nodev,fsname=portal,auto_unmount,subtype=portal -- /run/user/1000/doc
3993 pts/1 S+ 0:00 grep run
/etc/runit$ ls -1F runsvdir
current@
default/
default.run/
single/
solo/
svmanaged/
/etc/runit$ ls -1F runsvdir/default
acpid@
anacron@
cron@
cups@
dbus@
dbus.dep-fixer@
default-syslog@
dhclient@
elogind@
getty-tty1@
getty-tty2@
getty-tty3@
getty-tty4@
getty-tty5@
getty-tty6@
getty-ttyS0@
rsyslog@
sddm@
$ ls -F /etc/init.d
acpid* cups-browsed* mountkernfs.sh* saned*
alsa-utils* dbus* mountnfs-bootclean.sh* sddm*
anacron* elogind* mountnfs.sh* seatd*
apparmor* eudev* networking* sendsigs*
avahi-daemon* gdomap* network-manager* single*
bluetooth* halt* openntpd smartmontools*
bootlogd* hostname.sh* procps* speech-dispatcher
bootlogs* hwclock.sh* pulseaudio-enable-autospawn* stop-bootlogd*
bootmisc.sh* keyboard-setup.sh* rc@ stop-bootlogd-single*
brightness* killprocs* rc.local* sudo*
checkfs.sh* kmod* rcS@ umountfs*
checkroot-bootclean.sh* lm-sensors* README@ umountnfs.sh*
checkroot.sh* mountall-bootclean.sh* reboot* umountroot*
console-setup.sh* mountall.sh* rmnologin* unattended-upgrades*
cron* mount-configfs* rsync* urandom*
cups* mountdevsubfs.sh* rsyslog* x11-common*
OK, I now have a working runit system, but I had to do some serious thinking outside of the box:
Comment out CD in /etc/apt/sources.list
If directories are lacking:
$ xdg-user-dirs-update
$ wget q4os.org/downloads/q4os-setup.sh
$ su (or sudo -s)
# sh q4os-setup.sh
Run the following after reboot to confirm running init system:
$ cat /proc/1/comm
The fundamental problem appears to have been that I could not get any desktop. Nouveau (if installed) didn't give me a GUI, and installing nvidia-driver ended with a dpkg error #1. Likewise, a guided disk formatting with a separate /home partition gave me a system with a home directory without any sub-directories (both sysv and runit). After having run the q4os script and rebooted, I now have a functioning system (KDE Plasma) using runit - confirmed by $ cat /proc/1/comm
Current status is that while I have a running system, /var/log/boot contains the following error message:
"Starting Avahi mDNS/DNS-SD Daemon: avahi-daemon failed!"
which could be why Discovery doesn't appear to be able to access a wired internet connection. Synaptic works fine.
I as of yet have not installed the Nvidia drivers as I am trying to avoid the dpkg error again. It will be interesting to see how long I can do that... The computer has an NVIDIA TU116 [GeForce GTX 1660 SUPER] gpu.
I have since the last post tried installing Devuan with LXQT, as well as another KDE Plasma directory configuration at install - both with runit init. The LXQT install could not complete a first system boot, pausing at "bluetooth". The KDE install paused at the same place. In both cases, pressing the return key returned a cli prompt to log in. Both installs appear to be malformed, as there are no directories in user /home. Likewise, an ls in /root returns nothing, however I was able to get into /usr and complete an ls of the directories inside. commands such as # poweroff, halt -h now, halt -r now, etc. did not work, despite a verification that they were there when doing an # ls in /sbin.
This is starting to look more and more like an issue of malformed installs...
Devuan system new KDE Plasma install with runit init:
First boot pauses at “Starting network connection manager: NetworkManager.” Refuses to continue. From what I have read, Devuan's runit implementation is a hybrid form of some runit bits, while most startup items continue to use SystemV. I am not really sure how to address the above problem to successfully get a first boot. Any ideas? I know that a PeppermintOS based Devuan install with XFCE and runit work without problems, so my working hypothesis is that NetworkManager isn't getting loaded. While I assume that NetworkManager is still SysV, I have no idea how to solve this on a hybrid runit system.
Help?
I think that I may have gotten this the wrong way round. From what I can see online, there never has been support for the sound card. I just always used it with a Behringer U-Phoria UMC202HD USB sound card, and never tried the built-in sound card until after I tried the AudioQuest sound card, making me think something happened due to that card… Unless someone has some ideas about a driver for the built-in sound card, I guess that this topic could be marked “solved”…
I would like to install Devuan, but as I have been experiencing audio issues, I started by booting from the live desktop image and testing the hardware. I should preface this by writing that I have just picked up the computer today, from my local computer tech shop. I had taken the computer there because of the audio problems, and was told that there is nothing wrong with the hardware. Sooo... I brought the computer back home, only to find that only the hdmi sound (crappy speakers in a monitor) is working. the built-in sound card still doesn't work in Linux. Here is what the hardware look like, and I emphasize that it was working before I tested the sound card:
$ inxi -SMA
System:
Host: devuan Kernel: 6.1.0-10-amd64 arch: x86_64 bits: 64 Desktop: Xfce
v: 4.18.1 Distro: Devuan GNU/Linux 5 (daedalus)
Machine:
Type: Desktop System: HUAWEI product: PUM-WDX9 v: M1010
serial: <superuser required>
Mobo: HUAWEI model: PUM-WDX9-PCB-B1 v: M1010 serial: <superuser required>
UEFI: HUAWEI v: 1.28 date: 08/22/2022
Audio:
Device-1: AMD Renoir Radeon High Definition Audio driver: snd_hda_intel
Device-2: AMD ACP/ACP3X/ACP6x Audio Coprocessor driver: N/A
Device-3: AMD Family 17h/19h HD Audio driver: snd_hda_intel
API: ALSA v: k6.1.0-10-amd64 status: kernel-api
Server-1: PulseAudio v: 16.1 status: active
$ lsmod | grep snd
snd_hda_codec_realtek 172032 1
snd_hda_codec_generic 98304 1 snd_hda_codec_realtek
snd_hda_codec_hdmi 81920 1
snd_hda_intel 57344 3
snd_intel_dspcfg 36864 1 snd_hda_intel
snd_intel_sdw_acpi 20480 1 snd_intel_dspcfg
snd_hda_codec 184320 4 snd_hda_codec_generic,snd_hda_codec_hdmi,snd_hda_intel,snd_hda_codec_realtek
snd_hda_core 122880 5 snd_hda_codec_generic,snd_hda_codec_hdmi,snd_hda_intel,snd_hda_codec,snd_hda_codec_realtek
snd_pci_acp6x 20480 0
snd_hwdep 16384 1 snd_hda_codec
snd_pcm 159744 5 snd_hda_codec_hdmi,snd_pci_acp6x,snd_hda_intel,snd_hda_codec,snd_hda_core
snd_timer 49152 1 snd_pcm
snd 126976 14 snd_hda_codec_generic,snd_hda_codec_hdmi,snd_hwdep,snd_hda_intel,snd_hda_codec,snd_hda_codec_realtek,snd_timer,snd_pcm
soundcore 16384 1 snd
snd_pci_acp5x 20480 0
snd_rn_pci_acp3x 20480 0
ledtrig_audio 16384 2 snd_hda_codec_generic,huawei_wmi
snd_acp_config 16384 1 snd_rn_pci_acp3x
snd_soc_acpi 16384 1 snd_acp_config
snd_pci_acp3x 20480 0
The USB sound card that I was testing is an AudioQuest DragonFly Black v1.5.
Any ideas about how I can get the hardware working?
Many thanks for the replies.
Thanks for the reply. I found that I am able to get to the Trinity desktop by choosing “nomodeset” in GRUB. There appears to be a problem using my Nvidia graphics card. All I can get is what looks like an 800x600 resolution. The card on my computer is an Nvidia GeForce GTX 1660 SUPER.
Trying to test Exe GNU/Linux, but I can't get past the cli login. Shouldn't it boot into a graphical environment, and why can't I find any information about user name and password? Trying to use exegnu64_daedalus-20231220
OK, the latest update is that I changed the Devuan /boot/extlinux/extlinux.conf back to the original setup:
UI menu.c32
PROMPT 0
MENU TITLE Welcome to Devuan - Debian without SystemD
TIMEOUT 100
DEFAULT devuan
LABEL devuan
MENU LABEL Devuan
KERNEL /vmlinuz
APPEND initrd=/initrd.img ro root=/dev/sdd1
by mounting the partition in Slackware and editing. When I rebooted, I was suddenly able to boot into the system again, so I uninstalled all kernels except for the low latency kernel and after having rerun the NVIDIA.run script, everything seems to be working fine. I will wait a couple of days before I mark this topic as solved, just in case there is something else.
@ fsmithred
Here I thought that I had everything sorted, but I compiled a low latency kernel following the instructions here and now even though I changed the entries in the Devuan /boot/extlinux/extlinux.conf to as follows, it refuses to boot, but just cycles through the countdown from 10-0 and back again:
UI menu.c32
PROMPT 0
MENU TITLE Welcome to Devuan - Debian without SystemD
TIMEOUT 100
DEFAULT devuan
LABEL devuan
MENU LABEL Devuan
KERNEL /vmlinuz-4.9.88-lowlatency
APPEND initrd=/initrd.img-4.9.88-lowlatency ro root=/dev/sdd1
Interestingly, if I mount the Devuan partition /boot/extlinux/ and manually rename all of the files that end in ".amd64" and add ".bak" at the end, and reboot, I get the same error about the partition /dev/sdd1 no longer existing, as I did when I tried editing the Devuan /etc/inittab and changed:
# The default runlevel.
id:2:initdefault:
to
# The default runlevel.
id:1:initdefault:
sooo I am pretty sure that there is some sort of glitch in Devuan ASCII, as switching runlevels in the /etc/inittab shouldn't cause an error like that, and I can't imagine that renaming files with ".bak" should do that either.
Sorry I don't have anything useful to add. When I get a chance, I'll try extlinux with a multi-boot system and let you know what I find.
I just successfully booted into Devuan ASCII on my triple boot system. Here are my extlinux.conf entries:
Slackware64 14.2 /boot/extlinux/extlinux.conf entry for Devuan ASCII install:
LABEL devuan
MENU LABEL Devuan - Debian without SystemD
COM32 chain.c32
APPEND hd3 1
Devuan ASCII /boot/extlinux/extlinux.conf entry (Devuan ASCII only):
UI menu.c32
PROMPT 0
MENU TITLE Welcome to Devuan - Debian without SystemD
TIMEOUT 100
DEFAULT devuan
LABEL devuan
MENU LABEL Devuan
KERNEL /vmlinuz
APPEND initrd=/initrd.img ro root=/dev/sdd1
On the Devuan ASCII side, I could make do with less, as I really don't need a menu. All I want to be able to do is to boot into the system from the Slackware extlinux menu, but it is the first thing that works.
Sorry I don't have anything useful to add. When I get a chance, I'll try extlinux with a multi-boot system and let you know what I find.
Looking forward
@fsmithred
OK, So now I have what I believe to be a working Devuan system (as there were no errors during the install or when install extlinux), but I am unable to boot into it. /boot/extlinux/extlinux.conf contained the following to boot the Devuan system to start with:
LABEL devuan
MENU LABEL Devuan
COM32 chain.c32
APPEND hd3 1
As this is a matter of chain loading, I only included a minimal /boot/extlinux/extlinux.conf on the root partition of the Devuan install. It read:
KERNEL /vmlinuz
APPEND initrd=/initrd.img ro root=/dev/sdd1
However, I got the error:
No Default or UI configuration directive found!
Next, I copied the menu.c32 module from my Slackware64 install to /boot/extlinux/ on the Devuan system and edited the /boot/extlinux/extlinux.conf for the Devuan system to read:
UI menu.c32
PROMPT 0
MENU TITLE Boot Menu
TIMEOUT 100
DEFAULT devuan
LABEL devuan
MENU LABEL Devuan
KERNEL /vmlinuz
APPEND initrd=initrd.img ro roodt=sdd1
On reboot, got the error:
failed to load COM32 file menu.c32
Sooo, then I returned to my Slackware64 install, again edited the /boot/extlinux/extlinux.conf for Devuan to read:
UI chain.c32
KERNEL /vmlinuz
APPEND initrd=/initrd.img ro root=/dev/sdd1
and copied the chain.c32 module from my Slackware64 install to /boot/extlinux/. On reboot, I this time got the error:
failed to load COM32 file chain.c32
I am not really sure where to go from here. My Slackware64 /boot/extlinux/extlinux.conf has constantly had the following information for booting my Devuan install:
LABEL devuan
MENU LABEL Devuan
COM32 chain.c32
APPEND hd3 1
----
On the off chance that it had something to do with the Slackware64 extlinux modules, I mounted the Devuan system and then copied the modules as per the instructions within the Devuan system:
cp /usr/lib/syslinux/modules/bios/*.c32 /boot/extlinux
The actual command was of course slightly different as I was working from within another system, but the modules and the target were correct. There is no change in the error that I receive, so it must be a configuration issue, but I haven't been able to pin it down.
I see. How about this:
#!/bin/sh SES="$( update-alternatives --list x-session-manager )" C=1; for S in $SES; do echo "$C: $S" C=$(( $C + 1 )) done echo S=''; while true; do echo -n 'Select window manager [1]: ' read I echo "$I" | grep -q '[^0-9]' && continue if [ -z "$I" ]; then S=1 break; fi if [ "$I" -gt 0 ] && [ "$I" -lt $C ]; then S=$I break fi done CMD="$( echo "$SES" | head -n$S | tail -n1 )" echo "$CMD" > "$HOME/.wm" echo "Default WM set to $CMD."
Then you could just do something like:
alias startx='startx "$( cat "$HOME/.wm" )"'
Very cool. Thanks.
----
So in reference to a triple boot system, extlinux and Devuan, I assume that Devuan continues to use runlevels as Debian used to do with Sys V. What argument should be included in these lines
LABEL devuan
MENU LABEL Devuan - Debian without SystemD
COM32 chain.c32
APPEND hd2 1
in /boot/extlinux/extlinux.conf to boot into runlevel 1???
----
Well, alternatively it is possible to go in and edit /etc/inittab, changing
id:2:initdefault:
to
id:1:initdefault:
This is good if you need to run NVIDIA driver install scripts that are very sensitive to any X.org running, but it is only a root command line. I have discovered that it is possible to
# su - username
to get into the user account. Unfortunately, it isn't possible to run an X session in that mode, as permission is denied...
devuan_dk_fan wrote:$ startx
I wish there was something similar for Devuan, thus avoiding the login manager for those users that like to keep it simple.
I might miss something here but startx exists on Devuan?
Yes, but the process I was describing doesn't work with Debian based systems, as there is no xwmconfig. The closest to that is running
# dpkg-reconfigure gdm
but that just switches between login managers and not window managers.
If it's already dual boot and you're adding another OS, couldn't you just make a new entry in the existing boot menu?
That was the original plan. As far as I understand, while extlinux is able to read the entire file system (any file system) it still needs a boot loader at the target.
You could let it install grub and then switch to extlinux. When I did that, it was easy. I didn't set it up for multiple systems, but I do use syslinux on multiboot live-usb frequently, but those are all on the same partition. I'm not sure what happens when you want to boot from different partitions.
I followed your instructions to do that on another computer, which of course changed the MBR, but leaves /boot/grub on the system. It didn't interfere with anything on a single system boot, but I was wondering how that would play out with a triple boot system. It is of course possible to
# rm -r /boot/grub
but again was unsure if that would potentially cause problems with or affect the stability of a triple boot system.
At the slim login screen, press F1 to toggle different window managers.
Thanks. I know. I have set up the necessary i3.desktop files if they didn't exist, or if the existing ones didn't appear to be working with slim, but never had any success booting from anything other than the default WM. Having gained some experience with Slackware, I prefer starting at the command line, logging in, using the Slackware xwmconfig script, and then running
$ startx
I wish there was something similar for Devuan, thus avoiding the login manager for those users that like to keep it simple.
@ fsmithred It looks like you are right about GRUB getting installed. I thought I avoided it by not specifying a location and then pressing the "go back" button. Unfortunately, the "expert" install sane defaults don't work either, as the install fails. I will have to take a look at this again tomorrow.
Regarding the ASCII "expert" installer, I personally am of the opinion that it has gotten too complicated. I assume that this is something that Debian has decided, but I don't recognize half of the options any more, and for some weird reason, it appears that a choice of login managers is only available with LXQT. Being a happy WM switcher, depending on my workflow (XFCE/i3) the current situation is a nightmare as SLIM or SLiM is no longer developed, and provides no functional, practical means by which it is possible to easily switch between window managers.