You are not logged in.
Curious as to what your definition of "not F***ed" is OP? Not that I disagree, I don't like most of them either, that's why I made my own.
As g4sra mentioned, Excalibur has not been declared stable by Devuan, you're liable to run into multiple issues.
Makes perfect sense, and your idea is more elegant to ensure it doesn't get clobbered via the user-intervention you mentioned.
For the lease files I added /var/lib/NetworkManager/*.lease to both the snapshot and installer excludes files, cleans 'em when you snapshot and cleans the one generated during the live-session too, so first boot into the installed system will only generate one new correct one tied to the new connection file.
I added lines to clean out the dbus/session-bus files too, the new init script I use cleans them at shutdown anyway but still didn't see a point in letting old useless files from the live-session get into the new install.
10-02-2025 Updated Vuu-do OB-Z's uploaded
A few package updates including a new kernel, and some big improvements under the hood. Added a new init script to do some cleaning at shutdown, dbus was bloating up folders in the user and root's home due to the way Devuan works, it's one of just a few distros that change the machine-id at every boot, a nice security feature but unfortunately was not clearing out old session-bus files, so that's fixed in this version. Changed some settings in Refracta Snapshot as it was deleting icon-them-cache files, and re-built the caches so that will reduce some overhead, also added a line to clear the session-bus files in root. Also added to the Refracta Installer's post-install script to seed a generic system-connection file into the new install, Snapshot clears the old files as it should, but not seeding a new one was causing Network Manager to generate new lease files on every boot, and not deleting old ones, so another instance of bloating, so that's fixed now as well. Tweaked the excludes files to really fine-tune what gets copied/installed. Should see some performance gains in this version, ram usage at idle is definitely looking better and better with all the tweaks of these last couple of versions. Updated the theme css for LightDM and created some new icons, looks nicer now. Fixed some inconsistencies in the root account and more tweaks and cleanup.
Okay, two folders at issue here:
/etc/NetworkManager/system-connections/(connection files i.e. "wired connection 1" "*.nmconnection", either way is valid)
If NM is used to create either an ethernet or wifi connection, it will append it with the ".nmconnection" and place it in this folder.
/var/lib/NetworkManager/*.lease file(s)
This is where the lease files appear, wlan0.lease or eth0.lease.
Your .nmconnection file was created by NM at some point, and was tied to a single specific .lease file, this is what you want.
If on boot, NM does NOT find a connection file in etc/NM/ system-connections, it will generate a new .lease file, but NOT a new connection file, so it will keep doing this every boot. You'll always connect just fine, but you'll also get neverending new .lease files every boot.
This only becomes a thing if at some point your eth0 connection file gets wiped out in /etc/NetworkManager/system-connections, such as when doing a snapshot that excluded those network files. We just need to put one back into the new system on install. Or the user can do it manually with some simple commands in terminal and force NM to make a new one. This just seemed more elegant.
To test, now that you moved that connection file, go count your /var/lib/NetworkManager/*.lease files, then reboot a couple times, you'll have new ones each time.
I can see the difference as it boots on the partitions i've fixed, the NM applet used to blink and hang up a bit before appearing in the panel, now it's just flick and it's on. I think I can lower my conky delay now, lol.
This looks really nice, i'm not into TDE or KDE but that desktop looks sharp, the wallpapers are gorgeous. Not a fan of having the panel on the side but I bet it would look great at the bottom. ![]()
And in 3 different languages? That's one helluva LOT of work, kudos my friend and also big props for giving Xlibre a shot! I'll have to download this later and take it for a spin. ![]()
So I tried the modded script and it worked perfectly, no more .lease file bloat! I also fine-tuned the excludes files for both snapshot and installer to exclude the .lease files and also the dbus session-bus files on both, they all get auto-genned both in live-session and in a newly installed system, so this way you get a perfectly clean new install with no extra old files hanging around.
Been able to do that since I first installed Daedalus in the fall of 2023...always been that way hasn't it?
Don't know why it would be a no-go.
Distro is Refracta 11(Chimaera). Please scroll down on the code text for system info(neofetch).
My bad, sorry. Yeah, my guess is you're missing the mediatek firmware package I mentioned, the misc firmware package got updated and the new one no longer has the mediatek and ralink firmware in it that it used to have.
Kinda hard to help, not knowing what version of Devuan you are using, compounded by the fact that this question has been posted in Derivatives instead of the apropos tech area, are you using Devuan or one of the derivatives? Daedalus, Excalibur or other version of Devuan?
Just off the top of my head, I see the Mediatek wifi device, and there have been some changes to where Mediatek firmware is located, previously it was part of firmware-misc-nonfree but in Excalbur (and Daedalus backports) it's been split off and is now in firmware-mediatek and the new package conflicts with older versions of firmware-misc-nonfree. So maybe something got removed or held back during your upgrade depending on what you have.
Well dang, it worked! Used some script to create a new system-connections file, a basic one for eth0 only (wi-fi doesn't have issues), and now the system is using it and not genning new .lease files at every boot!
So I am wondering if this can be added to the Refracta Installers post-install script (cleanup-install.sh) so a new one will be created on a new install, worked great in a local install. So here's how I think cleanup-install.sh would look with this proposed addition:
#!/usr/bin/env bash
# cleanup-install.sh
# You can add other commands here that you want to run at the end of
# the installation process, before script self-cleanup (while the
# installed system is still mounted.)
#
# Anything you add must come before the "exit 0"
# Send errors to the installer's error log.
error_log=$(grep error_log /etc/refractainstaller.conf | cut -d"\"" -f2)
exec 2>> "$error_log"
# This removes the installer icon from the desktop upon installation.
for file in /target/home/*/Desktop/refractainstaller.desktop ; do
rm -f "$file"
done
# Create a generic eth0 system-connections file for NetworkManager
tee /target/etc/NetworkManager/system-connections/Wired\ connection\ 1 << EOF
[connection]
id=Wired connection 1
uuid=$(uuidgen)
type=ethernet
autoconnect=true
[ethernet]
auto-negotiate=true
[ipv4]
method=auto
[ipv6]
addr-gen-mode=stable-privacy
method=auto
EOF
# Set appropriate permissions for the connection file
chown root:root /target/etc/NetworkManager/system-connections/Wired\ connection\ 1
chmod 600 /target/etc/NetworkManager/system-connections/Wired\ connection\ 1
exit 0What say you? ![]()
Making headway here, about to log out and start some testing. Easy enough to bandaid this via shutdown cleanup script. But going to test some better fixes first.
In a nutshell, the system-connections files get nuked by snapshot (as they should), but there's no auto recreating them. The regular Devuan installer that I used to create the system on sda1 must have had some type of pre-seeded thing going on, a generic eth0 config near as I can tell, so it's re-using the exact same lease file every time and not genning new ones.
It's possible to fix post-install by a user, but that's not really user-friendly, so i'm off to do some testing and if it works i'd like to propose an addition to the Refracta-Installer post-install script. @fsmithred
Off your meds again Igor?
In searching the internet previously for similar solutions, I was able to find some 32 bit appimages of FF and Chromium, perhaps one of those might work for you.
So this seems like my own problem and not one with Devuan in general, so starting a new thread hoping maybe a network guru can help.
I have 3 partitions on my machine, sda1 is a Devuan Mate first installed in 2023, never modded or snapshotted or vuu-do'ed and it doesn't have this issue.
The other 2 partitions are Vuu-do, one a build partition for isos, the other a development partition, and they both suffer this issue.
A large build-up of old lease files for eth0 in /var/lib/NetworkManager, hundreds literally. System is generating a new one at every boot, and not deleting or overwriting the old one(s).
So far the only difference I have been able to find, is that sda1 has a keyfile for the interface in /etc/NetworkManager/system-connections, and the other 2 partitions do not have one. For some reason they aren't generating the keyfile, and i'm thinking this may be why there's a new lease file generated at every boot, because if the keyfile is present it just re-uses it without genning a new lease file?
A relevant section from Snapshot:
# Clear configs from /etc/network/interfaces, wicd and NetworkManager
# connman and netman, so they aren't stealthily included in the snapshot.
if [[ -z $netconfig_opt ]] ; then
echo "# The loopback network interface
auto lo
iface lo inet loopback
" > "$work_dir"/myfs/etc/network/interfaces
rm -f "$work_dir"/myfs/var/lib/wicd/configurations/*
rm -f "$work_dir"/myfs/etc/wicd/wireless-settings.conf
[b] rm -f "$work_dir"/myfs/etc/NetworkManager/system-connections/*[/b]
rm -f "$work_dir"/myfs/etc/network/wifi/*
rm -rf "$work_dir"/myfs/var/lib/connman/*Which is very sensible and does explain why the system-connections folder is empty to start with, but doesn't explain why NM is not generating a new connection file once you've connected in a new install.
One other difference between the working and not-working systems is in /etc/network/interfaces, which in the working system reads:
# 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 loopbackWhile in the non-working partitions reads:
auto lo
iface lo inet loopbackSo the only difference is that source /etc/network/interfaces.d/* line, but I don't see how that would make a difference, there's nothing in interfaces.d on any of the partitions.
EDIT: Grasping at straws here, but another difference between the affected and not affected systems is iptables is installed along with the newer nftables in the working partition, the partitions having the issue only have nftables.
Further testing to come, might be something on my end, but the network manager files shouldn't be accumulating like that, I don't know that it has anything to do with Snapshot, I think Caja's properties tab are confusing some dates compared to PcmanFM.
EDIT: Started a different thread about this and solved it.
Really appreciate the info @grunchy! I only have Mate and Openbox here to test on.
Dbus really gets around it seems.
So looks like whatever display manager is used, a folder is created in /var/lib/(display-manager)/.dbus/session-bus/*
But i'm still trying to figure out what purpose it serves and how they got there, after logging in and out of that partition multiple times yesterday, that folder was not added to by the system, still just 21 files that all show they were created when I started a snapshot last week.
The init script is working perfectly it seems, several days of testing now and no issues.
To clean /var/lib/(display-manager)/.dbus/session-bus/* in a similar fashion would take some extra work, have to list all available display managers, there's not that many but still.
And now I just found in /var/lib/NetworkManager some 181 files, 5 of which are config/info files of one sort or another, and all the rest are eth0-lease files, and all of those say they were created at the exact same time I started that snapshot last Wednesday.
So 176 files that all do the same thing, sample filename and it's contents:
internal-f368e190-5bd0-37e6-b42d-75f71b63df1d-eth0.lease
# This is private data. Do not parse.
ADDRESS=192.168.2.5
On KDE4 in general... Personally I don't think it was particularly "bad", as much as it was subject to the same insane development model KDE has always used:
As soon as a major version is finally usable and all serious bugs are resolved, it's time for a complete rewrite and a horribly broken .0 showcasing the newest dumb trend-chasing idea.
I remember. At the time I was on a different distro and never used KDE, but I remember every time KDE would release a batch of updates the KDE section of the forum would light up like a Christmas tree. A couple of weeks would ensue of fixes, workarounds, and bandaids for all the stuff it broke. And I just laughed because I used gnome2, and it was simple and stable and useable.
Then gnome3 hit and I stopped laughing.
Unfortunately the mindsets that create this kind of mess, are ubiquitous and deeply embedded in just about everything these days and has been for decades now.
My favorite kinds of linux programs, are those that haven't been messed with in years, things that were perfected and the dev deemed the project "finished" and stopped messing with it. There's a fair amount of them in the repo, the Obsession login/logout app is one.
Ceres system completely lacks the file at /etc/default/dbus. Nonetheless, dbus is definitely installed and running. I'm also missing /etc/machine-id.
On a new Devuan install there shouldn't be an /etc/machine-id, so that seems correct.
Another puzzle: there are exactly two files in ~/.dbus/session-id/.
I don't even have that folder on my Daedalus installs, just ~/.dbus/session-bus/, perhaps somebody has renamed it in Ceres.
Logs to /var/log/user.log
.
Well that's nicer and easier to find the entries than it is in syslog, I guess I need to read the logger man page more thoroughly. So it's logging in both places, that seems redundant, maybe I need to specify.
Users of the lightdm display manager will probably also see an accumulation of files in /var/lib/lightdm/.dbus/session-bus/
Sonofagun...nice find! I wonder if we can safely delete those on shutdown as well?
Interesting...much smaller accumulation of files in there than was in the other two folders.
Curiouser and curiouser...on my build partition where I make and snapshot Vuu-do, there are 21 files in that folder, and according to the properties tab, every single one of them was created at the same time, date, hour, even down to the same second. And that was last Wednesday the 17th at the exact same time a snapshot of Vuu-do (mini-ob-z) was started, but off by exactly an hour (that may be a daylight savings time discrepancy).
@fsmithred, this seems strange, no?
The other partitions have some files as well, but the dates are not all the same.
Well I don't know much about Devuanpi, is it using XFCE or something else? Some DE's have their own built-in mixer.
I think I would start by making a super simple .asoundrc and going from there.
pcm.!default {
type hw
card 0
}
ctl.!default {
type hw
card 0
}Of course that leaves you with the default card 0, so try changing the 0's to 2's for the card you want.
If that works, a simple script that offers a dialog where you choose either 0, 1, or 2, and it seds the change and re-starts alsa, would allow you to change during session.
Looks like from the screenshot that you are using the default (card 0) which is your headphones, when what you want is card 2, should be as simple as changing the card number in your ~/.asoundrc to 2 instead of 0. (Assuming you have that file, if not you may need to make one.)
I noticed it mentions "Broadcom Mixer", that I am unfamiliar with. If it's a genuine mixer, then it likely has control of alsa starting with boot-up, in which case the alsamixer will likely only be good for adjusting main volume. But perhaps that's just something local to the headphones.
Well I don't want to urge anybody to install this until we can get some more knowledgeable eyeballs on it.
That being said I just ran another test on a different partition and it worked just fine once again.
The script itself should be named rm-machineid and installed in /etc/init.d/. You need to make sure root owns it and that it's executable.
Then in terminal, su-to-root and:
update-rc.d rm-machineid defaults 00
Shucks...thanks my friend for the kind words! And right back at you, it wouldn't be the same around here without you. ![]()
So I tested the script, installed it into another partition running Vuu-do openbox-mini and shutdown/rebooted multiple times and it's working perfectly so far. Seems like a nice workaround that preserves Devuan's unique functionality in this area.
vuudo rm-machineid: dbus session-bus cleanup: removing session-bus files
vuudo rm-machineid: Cleaned /root/.dbus/session-bus
vuudo rm-machineid: Cleaned /home/vuudo/.dbus/session-busOkay, this seems better, but again i'm a newb at init scripts, i'm willing to test this though so i'll see later today unless somebody more knowledgeable sees a flaw, if so please speak up! This cleans the folders, also rm's /etc/machine-id if it exists (in the case of migrations or legacy holdovers) adds logging (var/log/syslog) and is to spec for init scripts protocol as near as I can tell:
#!/bin/sh
### BEGIN INIT INFO
# Provides: rm-machineid
# Required-Start: $remote_fs
# Required-Stop: $local_fs
# Default-Start:
# Default-Stop: 0 6
# Short-Description: Remove dbus session-bus files on shutdown
# Description: Cleans ~/.dbus/session-bus and /root/.dbus/session-bus, plus /etc/machine-id for migrated systems
### END INIT INFO
set -e
PATH=/sbin:/usr/sbin:/bin:/usr/bin
NAME=rm-machineid
DESC="dbus session-bus cleanup"
do_stop() {
echo "$DESC: removing session-bus files" | logger -t $NAME
# Remove /etc/machine-id for Debian migrations or legacy Devuan installs
if [ -f /etc/machine-id ]; then
rm -f /etc/machine-id
logger -t $NAME "Removed /etc/machine-id (migration/legacy cleanup)"
fi
# Clean root's session-bus dir
if [ -d /root/.dbus/session-bus ]; then
rm -f /root/.dbus/session-bus/*
logger -t $NAME "Cleaned /root/.dbus/session-bus"
fi
# Clean all user home session-bus dirs
for userhome in /home/*; do
if [ -d "$userhome/.dbus/session-bus" ]; then
rm -f "$userhome/.dbus/session-bus"/*
logger -t $NAME "Cleaned $userhome/.dbus/session-bus"
fi
done
}
case "$1" in
start)
# No-op for start
;;
stop)
do_stop
;;
restart|force-reload)
echo "Error: argument '$1' not supported" >&2
exit 3
;;
*)
echo "Usage: $NAME {start|stop}" >&2
exit 3
;;
esac