I think you do still need the current kernel headers to recompile the kernel module for NVidia if you are using NVidia's own driver.

I have this working in Beowulf with lightdm and Cinnamon.

I have a GRUB_CMDLINE_LINUX="text" in /etc/default/grub.

I assume that you ran update-grub to change /boot/grub/grub.cfg?

This should have put a 'text' parameter in the linux /vmlinux.. line of your grub.cfg's menuentry.

This will start you in text mode, not sure if this will then permit you to successfully run startx - it is implemented as a simple trap in the lightdm init script (/etc/init.d/lightdm):

case "$1" in
  start)
        CONFIGURED_DAEMON=$(basename "$(cat $DEFAULT_DISPLAY_MANAGER_FILE 2> /dev/null)")
        if grep -wqs text /proc/cmdline; then
            log_warning_msg "Not starting Light Display Manager (lightdm); found 'text' in kernel commandline."
        elif [ -e "$DEFAULT_DISPLAY_MANAGER_FILE" ] && \
           [ "$HEED_DEFAULT_DISPLAY_MANAGER" = "true" ] && \
           [ "$CONFIGURED_DAEMON" != lightdm ] ; then
                log_action_msg "Not starting Light Display Manager; it is not the default display manager"
        else
                log_daemon_msg "Starting Light Display Manager" "lightdm"
                start-stop-daemon --start --quiet --pidfile /var/run/lightdm.pid --name lightdm --exec $DAEMON -b|| echo -n " already running"
                log_end_msg $?
        fi
  ;;

In my case this is on my mail server and I just run it headless and ssh -X into it from my desktop PC. Graphics mode (Cinnamon desktop) is installed as I used it when configuring the server and it's there if I were to need to work on the server with a keyboard and screen plugged in on my desk.

I can't speak about Chimaera, as I'll stick with stable/Beowulf, but at some point in my migrations I found I could no longer get uswusp to work but found that kernel suspend works OK.

I use pm (pm-utils package, pm-suspend or pm-hibernate commands) to suspend and hibernate and this has options for the power down methods:

#   kernel    The built-in kernel suspend/resume support.
#                 Use this if nothing else is supported on your system.
#   uswsusp   If your system has support for the userspace
#                 suspend programs (s2ram/s2disk/s2both), then use this.
#   tuxonice  If your system has support for tuxonice, use this.

You can also specify suspend and resume modules in the config (/etc/pm/config.d/defaults). I do this for my wireless ath9k modules as otherwise my atheros wireless card doesn't power down.

You'll also need a resume=[swap partition] parameter in your grub's linux /boot/vmlinuz line if you use hibernate.

Of course they may have removed pm-utils in Chimaera, which would mean I'd have problems upgrading too.

Glenn, If you check earlier in this thread e.g. contribution #7 you'll see that for Default-Start/stop you would normally (at least in sysvinit) use:

# Default-Start:  S
# Default-Stop:   0 6

The reason to use S rather than 2 is that it executes as early as possible, which is a desirable feature of a firewall.

Is there no S level in open-rc?

It should, if the sysvinit defaults has been set up when configuring the deb and the systemd default is also to enable. However some packages don't always enable by default - if the installation instructions explicitly tell you to enable the deamon then it won't have done it in the postinst script.
For example the nftables deb (another deb that doesn't install an /etc/init.d/nftables) also doesn't enable nftables by default even if you are using systemd.

Quite frequently when only systemd services are enabled and therefore /etc/inti.d/initscript is not included then the sysvinit postinit and postrm elements are not included either.

What I would suggest is that the OP either rerun update-rc.d, or at least check if their particular daemon has been enabled by looking to see if the synlinks are there.

You can also inspect the deb using gdebi. You'll find the postinst script in the 'installed files tab'. It should include something like this (taken from the chrony deb):

# Automatically added by dh_installinit/12.1.1
if [ "$1" = "configure" ] || [ "$1" = "abort-upgrade" ] || [ "$1" = "abort-deconfigure" ] || [ "$1" = "abort-remove" ] ; then
	if [ -x "/etc/init.d/chrony" ]; then
		update-rc.d chrony defaults >/dev/null
		if [ -n "$2" ]; then
			_dh_action=restart
		else
			_dh_action=start
		fi
		invoke-rc.d --skip-systemd-native chrony $_dh_action || exit 1
	fi
fi
# End automatically added section

# apt install network-manager{,-gnome}

FWIW Network Manager comes as default and works fine in a Beowulf + sysvinit + Cinnamon system i.e without systemd.

The renaming of ethX and wlanX was something that was introduced in Debian 10 Buster and there was a warning about it in the Release Notes. Devuan Beowulf however retained the traditional names and when I've installed it (on AMD64 systems) it's worked out of the box.

I'm unclear how this was done. However I don't think the Devuan linux-images are patched (there aren't even any +deb10u1 markers on the more recent 4.19 linux-images).
My desktop PC which used wlan0 has a udev rule (/etc/udev/rules.d/70-persistent-net-rules)

# PCI device 0x168c:0x0030 (ath9k)
SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="c0:4a:00:1e:b5:03", ATTR{dev_id}=="0x0", ATTR{type}=="1", KERNEL=="wlan*", NAME="wlan0"

My mail server, which uses eth0 doesn't.
Neither have the GRUB_CMDLINE_LINUX parameter set either.

Comparing your output, running as root with mine as local user, in addition to the lack of pulseaudio (which I have) I also note that you don't seem to have a desktop:

Linux Client Version is 5.3.465578.0920
QSG_RENDER_LOOP is
XDG_CURRENT_DESKTOP = ;   GDMSESSION =

whereas in mine:

Linux Client Version is 5.0.418682.0603
QSG_RENDER_LOOP is 
XDG_CURRENT_DESKTOP = X-Cinnamon;   GDMSESSION = cinnamon

Also you've truncated the line that gives your PC spec. How much memory do you have? I have 8GiB. Zoom only seems to be using 57 MiB idling but I don't know how much it uses in a live session. I'd expect it would also use quite a bit of graphics memory.

Probably worth checking whether when zoom is active it's not driving you into using swap (disk) memory.

Just as another data point.
I have Zoom installed, version 5.0.418682.0603 on a normal Beowulf installation. It installs in /opt. I have a old Logitech webcam (low res).
And it works fine.
I note that it seems to uses Qt. Could this be an issue?

Whatever the other issues , this looks problematic for successful hibernation since the Mem in use (491Mi), which is what needs to be written to disk on hibernation, is nearly twice as large as Swap you have (255Mi, all free).
Usually you would want to have as much swap as memory.

You can achieve hibernation either through kernel mode, which requires the appropriate kernel module, or in userspace, using uswsusp or tuxonice.

If your device https://leste.maemo.org/Motorola_Droid_4 lacks both kernel support (available kernels don't have the appropriate kernel module) and you don't have the package uswsusp in your sources then I think that means that you would have to compile one or the other to get hibernation (though even then it could be possible that your hardware lacks the appropriate hooks to make it work).

NB. Just as an update on my earlier posts I do now have hibernate running using kernel mode on Beowulf on my (pretty ordinary) AMD64 hardware, though I also still use pm-hibernate to call it as I have a few things I want to invoke in the pre-sleep and post-wake scripts.

You also need to have a resume=UUID=your.swap.drive.UUID (or similar) command in the linux vmlinux line in your grub menuentry. This tells the kernel where to store/restore your memory image.