Read my EDIT, I found a (strange) solution.

Probably a graphical polkit authentication agent. Run gdebi-gtk /path/to/.deb from a terminal or check ~/.xsession-errors for anything relevant.

Try installing policykit-1-gnome, that should supply an agent and the autostart .desktop file for it.

Disclaimer: other graphical polkit authentication agents are available. I posted a full list recently but I can't be bothered searching for it....

EDIT: no, looks like polkit-1-gnome is already installed & running in Devuan's Xfce desktop and gdebi-gtk works fine in isolation.

I can get it working in my VM by clicking on the "Open with" scroll box (where it says "GDebi Package Installer (default)") then clicking on "Other..." then selecting "Gdebi Package Installer" again. Once that is done the expected password confirmation box will open and the package can be installed directly from Firefox.

Very strange

Probably off-topic but Linux now has a native NTFS driver for v5.15 which should offer much improved performance over the old NTFS-3G FUSE version.

https://www.kernel.org/doc/html/v5.15/f … /ntfs.html

Disclaimer: I don't use NTFS.

^ /dev/pts will have been mounted under /dev/ by the --rbind option.

/etc/mtab should be a symlink to /proc/self/mounts in the running system. It is not necessary to populate it while installing but you can create the symlink once the system is independently booted. Just creating /etc/mtab will not cause the listed partitions to be mounted.

If you have /var/ on a separate partition it needs to be explicitly mounted before using chroot.

So if the root filesystem is on /dev/sdX1 with /var/ on /dev/sdX2 you should use

# mount /dev/sdX1 /mnt
# mkdir -p /mnt/var
# mount /dev/sdX2 /mnt/var

Also mount any other required partitions using the same method.

If you're doing this from a Devuan or Debian live system you can use the Arch Way to proceed:

# apt install arch-install-scripts
# genfstab -U /mnt > /mnt/etc/fstab
# arch-chroot /mnt

The arch-chroot command mounts the API filesystems and copies over /etc/resolv.conf automagically. But you probably already know that :-)

You can also use the Arch Way to generate /etc/adjtime instead of creating it manually as the Debian guide suggests:

# hwclock --systohc

^ That presumes the hardware clock is using UTC, which it should be even if you're multi-booting with Windows[0]. Make sure the system clock is correct before running that.

[0] https://wiki.archlinux.org/title/System … ft_Windows

^ This: https://pkginfo.devuan.org/cgi-bin/poli … u&x=submit

There's https://github.com/keithbowes/obmenu but it isn't packaged for Devuan. It should be fairly simple to do that with debmake(1) though. See also https://www.debian.org/doc/manuals/debmake-doc/

The evidence is pretty clear in respect of macrobiotic diets: they are not healthy. And remember: the plural of anecdote is not evidence

^ I still get confused with sysvinit when it comes to enabling and disabling services and also with checking if they are actually enabled, which is why I prefer OpenRC or systemd with their more straightforward management tools.

For my ROCK Pi S switching Debian from systemd to runit (with runit-init as PID 1) saves a significant amount of memory on a device which only has 512MiB to spare.

There are also the inherent advantages conferred by moving to a smaller code base, which is especially relevant when comparing systemd with OpenRC, runit or s6: less code == less potential for bugs to creep in. Theoretical, granted, but still a factor.

EDIT: *you're

And here's a comparison of the binary sizes for the Debian bullseye/Devuan chimaera versions:

$ ls -l lib/{systemd/systemd,runit/runit-init} sbin/{openrc-,}init
-rwxr-xr-x 1 empty empty   18760 Jul 25  2021 lib/runit/runit-init
-rwxr-xr-x 1 empty empty 1739200 Jul 13  2021 lib/systemd/systemd
-rwxr-xr-x 1 empty empty   52240 Apr 18  2021 sbin/init # sysvinit
-rwxr-xr-x 1 empty empty   22520 Apr  2  2021 sbin/openrc-init
$

Just finished Galaxias by Stephen Baxter.

The dialogue was awful, the prose stilted and the characterisation crude but Mr. Baxter does big-concept hard sci-fi better than anybody. I only got the book from the library yesterday but found it almost impossible to put down. Highly recommended.

Ouch. Perhaps I was wrong about that broken POS software...

This is generated from /lib/systemd/system/alsa-restore.service in my Arch box:

#!/bin/sh
# Generated by sysd2v v0.3  --  http://www.trek.eu.org/devel/sysd2v
# kFreeBSD do not accept scripts as interpreters, using #!/bin/sh and sourcing.
if [ true != "$INIT_D_SCRIPT_SOURCED" ] ; then
    set "$0" "$@"; INIT_D_SCRIPT_SOURCED=true . /lib/init/init-d-script
fi
### BEGIN INIT INFO
# Provides:       alsa-restore
# Required-Start: $remote_fs
# Required-Stop:  $remote_fs
# Default-Start:  2 3 4 5
# Default-Stop:   0 1 6
# Description:    Save/Restore Sound Card State
### END INIT INFO
DESC="alsa-restore"
DAEMON=none
PIDFILE=none

do_start_cmd_override ()
{
  /usr/sbin/alsactl restore
  true
}

do_stop_cmd_override ()
{
  /usr/sbin/alsactl store
  true
}

do_status_override () :

^ Save that to /etc/init.d/alsa-restore and run

# update-rc.d alsa-restore defaults

Disclaimer: untested.

Why have you placed the intelpwm script under /usr/lib/udev/? I think /usr/local/bin would be better. Is the intelpwm script executable? It might be best to open an issue in the linked GitHub repository if that doesn't make it work because I don't use Intel and so have no way of testing this.

It's not important but local udev rules belong in /etc/udev/rules.d/, /usr/lib/udev/rules.d/ is for rules supplied by packages.

The udevadm info command makes no sense at all, what are you trying to achieve with that?

Anyway if you can't get the udev rule working you can try this file at /etc/modprode.d/i915-pwm.conf instead:

install i915 /usr/local/bin/intelpwm ; /sbin/modprobe i915 --ignore-install

Make sure the intelpwm script is executable.

Probably better to use memtest86 and/or memtest86+ instead because it doesn't rely on the operating system API and it tests all of the memory rather than just that available in user space.

I'm just guessing here but as Pale Moon is based on a very old version of Firefox code rot might be a factor.

It is important to remember that FF & Chrome are developed by very large teams with massive experience and huge funding whereas Pale Moon is the work of a few developers and seems to be a hobby project rather than a professional operation.