You are not logged in.
#5, #6, #7
all of these can be fixed by: apt install dbus-x11
will make a fix for this and update the netinstall soon - I'll publish it in the normal spot so the iso mirrors pick it up, probably tomorrow
#3 (openrc shutdown hang) I have not been able to replicate yet (not for lack of trying).
#2
Technical Details: The name org.freedesktop.fwupd was not provided by any .service files
This can be resolved by removing plasma-discover-backend-fwupd. Next iso will include this fix.
First, since I neglected to say this earlier, thanks for your willingness to help ![]()
a) in default installer cases, qemu-guest-agent and spice-vdagent should be automatically installed in a VM,
for the desktop end-user. (it is the expectation that it should just work; no friction)
That sounds like a reasonable expectation. The more things that "just work," the better, IMO. Obviously, advanced users can remove / customize things however they want. So for this particular example, there would need to be some way for the installer to detect that it's being run inside a vm. Some questions: would it make sense to install qemu-guest-agent inside a gui-less guest (no desktop)? What about spice-vdagent? Personally I prefer virtualbox -- I tried the switch to qemu for about 3 weeks and then switched back. Unfortunately the two do not like to play together either -- when running vb I can't use kvm with qemu. Anyway, that's a bit beside the point, other than to say I don't have a lot of qemu experience.
I would like explore the notion (not proposal) of Devuan 7 shipping with KDE 6.7.6, even if Debian 14 does not.
Is such a thing even viable?
I believe it would be viable IF someone were to fork kde 6.7.6 and maintain it for devuan. That someone is not me.
I would also like to explore the notion of Devuan 7 shipping some form of SonicDE, even if Debian 14 does not.
Is it ridiculous, just too soon to adopt, or just simply too great of a weight?
This one seems like "too soon to adopt." From https://github.com/orgs/Sonic-DE/discussions/28 "we don't currently have packagers for the Debian family of distros"
I'll upload a new freia netinstall to https://files.devuan.org/user/tempforever/ soon (after it's done building), if you want to test/report any problems with KDE specifically that's fine.
The desktop options are part of tasksel/tasksel-data. It should be possible to patch in additional things such as SonicDE, I'm not all that familiar with it.
[edit]
plasma in freia is currently at 6.7.4
sonicde is not yet in the debian repo
It'd be possible to add an outside repo to the installer, but that seems a bit unconventional
[edit2]
following https://github.com/sonicde-debian/debian I am not even able to install sonicde on freia (testing), as it depends on sonic-win, which depends on an unknown package qt6-base-private-abi
It may be that the "img" file is of the device, but you're trying to write it to a partition.
What does this show fdisk -l /media/storage/isos/chimaera/rpi_chimaera.img
sudo chown -R root:root /usr
That'll probably fix a lot, but not everything.
On my daedalus system, under /usr: (find ! -group root -ls | grep -v ./local/)
407952 16 -rwxr-sr-x 1 root utmp 14416 Feb 26 2023 ./lib/x86_64-linux-gnu/utempter/utempter
523692 52 -rwsr-xr-- 1 root messagebus 51272 Sep 24 2023 ./lib/dbus-1.0/dbus-daemon-launch-helper
690224 4 drwxrwsr-t 2 root lpadmin 4096 Dec 27 2015 ./share/ppd/custom
393038 4 drwxrwsr-x 12 root staff 4096 Oct 22 2023 ./local
436499 636 -rws--x--- 1 root operator 649688 May 25 2022 ./bin/cdrdao
411257 16 -rwxr-sr-x 3 root mail 14608 Sep 3 2021 ./bin/mail-touchlock
398561 32 -rwxr-sr-x 1 root shadow 31184 Dec 14 2025 ./bin/expiry
411257 16 -rwxr-sr-x 3 root mail 14608 Sep 3 2021 ./bin/mail-lock
396906 24 -rwxr-sr-x 1 root mail 23040 Feb 4 2021 ./bin/dotlockfile
401820 100 -rwsr-sr-x 1 root mail 101664 Mar 1 2022 ./bin/procmail
399266 44 -rwxr-sr-x 1 root crontab 43648 Mar 2 2023 ./bin/crontab
400977 60 -rwsr-sr-x 1 daemon daemon 59768 Oct 15 2022 ./bin/at
397902 80 -rwxr-sr-x 1 root shadow 80376 Dec 14 2025 ./bin/chage
402924 120 -rwxr-x--- 1 root operator 121840 Aug 6 2019 ./bin/growisofs
410565 456 -rws--x--- 1 root operator 463776 Jun 24 2022 ./bin/wodim
401442 20 -rwxr-sr-x 1 root mail 18768 Mar 1 2022 ./bin/lockfile
400077 476 -rwxr-sr-x 1 root _ssh 485760 Apr 4 2026 ./bin/ssh-agent
411257 16 -rwxr-sr-x 3 root mail 14608 Sep 3 2021 ./bin/mail-unlockObviously, the list will vary depending on what's installed. Stuff in /usr/src seems to be owned by $USER:$USER, and stuff in /usr/local is root:staff
And if you've got a usr-merged system (excalibur), there's probably going to be more...
[edit]
here's another list from an excalibur system, not a lot installed there
/usr$ find ! -group root -ls | grep -v ./local/
66326757 44 -rwxr-sr-x 1 root shadow 43256 Jun 29 2025 ./sbin/unix_chkpwd
66331837 24 -r-xr-sr-x 1 root postdrop 22840 Aug 10 12:35 ./sbin/postqueue
66331826 24 -r-xr-sr-x 1 root postdrop 22888 Aug 10 12:35 ./sbin/postdrop
66337980 60 -rwsr-sr-x 1 daemon daemon 59768 May 4 2025 ./bin/at
66324522 32 -rwxr-sr-x 1 root shadow 31256 Apr 19 2025 ./bin/expiry
66332196 412 -rwxr-sr-x 1 root _ssh 420224 May 6 08:33 ./bin/ssh-agent
66322953 112 -rwxr-sr-x 1 root shadow 113848 Apr 19 2025 ./bin/chage
66322893 52 -rwxr-sr-x 1 root crontab 51936 Jun 13 2025 ./bin/crontab
66322658 24 -rwxr-sr-x 1 root mail 23104 Dec 31 2024 ./bin/dotlockfileThis file is generated during the installation of eudev and if at this time libsane1 is not yet installed, that content is missed.
This sounds like a reasonable explanation. The better fix, then, as you mentioned, is to re-create the hwdb. run this as root:
udevadm hwdb --update
service eudev restartThis works with my hp g3010 scanner, no need for custom udev rules.
First, you'll want to get the vendor-id and product-id of your device. the lsusb command should give this info:
lsusb
...
Bus 005 Device 024: ID 03f0:4205 HP, Inc ScanJet G3010In this case, vendor id is 03f0 and product id is 4205.
To create a udev rule, see this post, replacing with your vendor-id and product-id.
[edit]
it appears that /usr/lib/udev/rules.d/99-libsane1.rules is supposed to take care of this, by running /bin/setfacl -m g:scanner:rw on the device file. For some reason, that doesn't appear to be triggering here. If run manually,
setfacl -m g:scanner:rw /dev/bus/usb/005/024then the scanner is accessible to the users in the scanner group.
There will be some upgrades on devuan infrastructure beginning at 20:00 UTC today September 25 2026. Services may be unavailable during this time. If everything goes smoothly, downtime should be minimal.
You can try the (exprimental) netinstall iso at https://files.devuan.org/user/tempforever/, it has a different booting mechanism
Which version of devuan - daedalus (oldstable), excalibur (stable), freia (testing)?
To elaborate on Ralph's answer, that value parameter is in seconds. So to get your desired 1-week setting, the url to bookmark would be https://dev1galaxy.org/search.php?actio … lue=604800
Thanks for the report. Fixed in September 19 netinstall iso.
if `initscripts` or other software is upgraded or updated, it may overwrite my scripts with the newer versions
I believe that this can be handled via dpkg-divert?
This isn't "how to avoid the problem" - but if you have difficulty with their picture captchas, this may be of help.
After spending about a half hour clicking on buses, traffic lights, crosswalks, bicycles, stairs, fire hydrants, cars, bridges (and probably more!), in a never-ending cycle, eventually I tried the audio captcha instead. This took only about 5 tries to get it right, and then I could proceed. All the sound samples had static, most were entirely unintelligible to me. Just load a new one until you get one that you can understand.
As far as the changes to sudo, what systems would make /usr/local/ writeable for a non-root user? That seems silly.
my /usr/local/{bin,sbin} are owned by group staff, and group has write permission.
I've got this script (for an example) from when I used to hibernate/suspend. It might require pmtools pm-utils (as Ralph noted above).
#!/bin/bash
# Script name: /etc/pm/sleep.d/50_network
# Purpose: Restore network settings upon thaw / resume
case $1 in
suspend|suspend_hybrid|hibernate)
:
;;
resume|thaw)
# wait 10s for network to come back
sleep 10
logger -s "/etc/pm/sleep.d/50_network: Restoring network settings..."
# restore network settings
/usr/local/sbin/netconfig.sh
logger -s "/etc/pm/sleep.d/50_network: Done."
;;
esacYou did nothing wrong. Evidently, task-lxqt-desktop has moved some of its (former) Depends packages to Recommends now. I hope this was not intentional. I can manually change it in the installer iso, if it comes to that. For now, you should be able to apt install lxqt to get what you're missing.
[edit]
Evidently, this was done intentionally
* task-*-desktop: Demote Depends to Recommends.
Looking at that commit, it has affected much more than just lxqt.
Does the netinstall iso have firmware-mediatek on it?
Yes, there's firmware-mediatek_20260810-1_all.deb
However, I found in firmware/Contents-firmware:
/usr/lib/firmware/mediatek/mt7925/BT_RAM_CODE_MT7925_1_1_hdr.bin firmware-mediatek_20260622-1_all.deb non-free-firmware
/usr/lib/firmware/mediatek/mt7925/WIFI_MT7925_PATCH_MCU_1_1_hdr.bin firmware-mediatek_20260622-1_all.deb non-free-firmware
/usr/lib/firmware/mediatek/mt7925/WIFI_RAM_CODE_MT7925_1_1.bin firmware-mediatek_20260622-1_all.deb non-free-firmwareSo it appears I need to update the Contents-firmware file to the correct version on the disc image.
Hopefully it will all be good in the next iso (check mirrors in a while)
Please, when you have an opportunity, test the new devuan_freia_7.0-202609031921_amd64_netinstall.iso - it should be on the mirrors by now. And if you don't mind, report back whether that fixes the issue.
Thanks for reporting this. I'll update the freia installer soon... it still was using the excalibur Contents-firmware file which I believe is the source of your issue.
syslog-ng will replace rsyslog as default logging in the future ?
I don't know, haven't taken a poll or anything. So far it's just a test on the netinstall. Have you had any issues with it?
The reason for the change: https://www.rsyslog.com/clarifying-ai-f … r-rsyslog/
Of course, you're free to remove syslog-ng and replace it with rsyslog after the installation is complete. And I may change the default back later, if that's what most users want.
Another option...
With qemu, use -nic user option, and set up networking as follows
ip address add 10.0.2.15/24 dev eth0
ip link set dev eth0 up
ip route add 10.0.2.0/24 dev eth0
ip route add default via 10.0.2.2 dev eth0Also, your /etc/resolv.conf should contain
nameserver 10.0.2.3From there, you should be able to apt install dhcpcd-base or whatever package(s) you want.
With virtualbox, you could change the network to bridged, and set up static networking as though it were a physical device on your LAN. For example, if your LAN uses 192.168.1.0/24 with a gateway of 192.168.1.1 (I think this is a common default setup), you could do something like this
ip address add 192.168.1.20/24 dev eth0
ip link set dev eth0 up
ip route add 192.168.1.0/24 dev eth0
ip route add default via 192.168.1.1 dev eth0Obviously, you'd want to use an address not in use by another device. That was just an example.
Whatever is going on, it's blocking gnlug.org just like the other mirrors. All I get are "could not resolve" errors for both the original and these repositories.
That's odd. Is your nameserver accessible? What's in /etc/resolv.conf?
from what I'm seeing on daedalus, nvidia-driver and nvidia-kernel-dkms are at version 535.261.03-1 (daedalus), 535.309.01-0+deb12u1 (daedalus-security), and oddly 535.216.03-2-bpo12+1 (daedalus-backports).
maybe this would work?
apt install nvidia-driver/daedalus-security nvidia-kernel-dkms/daedalus-securityOf course, you'll need to have apt working properly first.