You are not logged in.
At the link I posted earlier about machine-id, the author posted an init script for sysvinit that would manually rm both /etc/machine-id and var/lib/dbus/machine-id at shutdown/reboot, thus effectively doing what Devuan is doing natively and that's changing the id on every boot.
Our issue is different, but was wondering if the same method could be used to clear the user's and root's .dbus/session-bus files when you shut down? I've never messed with making init scripts, nor modding them, so I don't know squat, but I took the author's script and made what seemed like proper changes to it. If someone more knowledgeable could take a look at it and see if it's useful that would be great, and also wondering if there's a better method?
EDIT: That was ugly, working on something better right now...moar man pages and docs to read, lol!.
Maybe deleting the file on shutdown with a script would do the trick?
That's my first thought as well, but it would be a band-aid...be nicer if we could mod the behavior of whatever is creating/updating those folders though.
Yeah that should change things Altoid, from the mountain of dbus documentation i've been reading today the presence of /etc/machine-id will stop uuidgen from generating new id's. So that seems like the issue in your case. The /etc/machine-id file is some systemd hooha, may have been a leftover from the very early days.
No not the /var/lib/dbus/machine-id file, I specifically mean /etc/machine-id which is a different file..located in /etc...lol. You shouldn't have one but if you do it might be an issue.
@Altoid on the box with the machine-id that never changes, do you have an /etc/machine-id file?
I have not seen anyone (yet) report that their system is actually working as intended. ie: creating a new machine-id at boot.
Well sure you have, that's the main problem I reported on, myself and others are getting a fresh unique one every boot, it's just not deleting the old ones when it creates new ones, thus the buildup of files.
And then there's secondary issues now, like yours where it's not creating new ones to start with.
But since I know you, I have to ask, is this one of your frankenmachines we're talking about and how much hacking have you done on it? LOL! Not judging brother, i'm a system chopper myself from way back, but for troubleshooting this thing we need some examples that are a little more mainstream I think. ![]()
^^Yeah the inconsistency is befuddling, need to find the common denominator.
In general I think swapping machine id's every boot is a good thing, but need to find out why old ones aren't getting deleted, and why for some folks it's just not happening.
Not having any of the issues with /etc/resolve.conf that were mentioned, using sysvinit and Network Manager.
But as far as deb.devuan.org goes, I get no result from pinging it at all. I've been using a direct mirror for a long time. Gnlug works awesome here in the midwest US.
Which begs the following question:
In whose interests or (better yet, to what end) does my box/system need to have a unique machine-id.
https://ianlecorbeau.com/blog/resetting … -boot.html
"The /etc/machine-id and /var/lib/dbus/machine-id files can potentially be used to fingerprint a system over the long term. Of course, some will argue that this is merely FUD, but it really isn't. The key word here is *potentially*.
There may not be any indications that this file is actively being used in a malicious capacity by anyone (or anything), but the fact that it can be used in such a manner should be enough for someone who cares about their privacy to take steps to at least minimize this potential threat."
Also from that article
-
Distributions like Devuan, tails and whonix already take care of regenerating machine-id every boot, so this script is not necessary if you're using one of them.
it could be that for all of use whom get the files piled up is because we are using plain window managers that do not purge their active dbus session address file upon session end (logoff, shutdown, reboot)
Happening here in Mate as well as Openbox.
9-18-2025 - New versions of Vuu-do Openbox 5.03 max and mini uploaded.
Some 20 package updates in the max, 7 for the mini. Added exinfo and metaview extensions to the mini to bring it more inline with the max, so now you can view exif data for image files, and metadata from music and video files in the mini as well as the max. Fixed a longtime bug in Synaptic relating to the status icons, got rid of a couple old scripts and some other detritus.
Added additional support for jpeg-xl in the form of it's command-line utilities package (v. 0.11.1). Updated some extensions to reflect new mimetypes supported. Added batch processing to the re-sizer and rotator scripts, previously you could select multiple images and they would do them, but used separate dialogs for each instance so you had to enter the size or rotation for each individual image. That's fixed now, so you can select as many as you like and it will apply whatever setting you ask for in a single dialog window. The caveat is the files you select need to be the same filetype to do a batch, it won't do a mixed batch of jpg's and png's etc.
Also fixed an issue where the Refracta tools weren't showing in VAI. Added the Gdebi package installer for local .debs, and swapped in Yahtzee for xsol in the max.
Lots of little cleanups and tweaks in this version. The performance tweaks from a week ago are really doing better than i'd hoped, i'm seeing less ram usage at idle and less cpu as well. The mini is especially light now.
So it gets rebooted at least a half dozen times a day, sometimes more.
You should have a ton of files then, for some reason yours is kicking out the old ones.
To clarify i'm running Daedalus too. weird.
One other change I made was to /etc/X11/Xsession.d/95dbus_update-activation-env - removing the --systemd argument stops a pointless warning being output to xsession-errors.
Apologies for the necro-post, but wanted to thank @boughtonp for this little nugget. That little warning had been a small thorn in my eye every time I opened .xsession-errors, but I have had a 1000 other fish to fry so never gotten around to it, but stumbled upon this post whilst searching for some dbus info. Thanks! ![]()
So both the folders mentioned in the title contain a dbus file for the session you're in, and generate a new one every time you re-boot, but it's not deleting the old ones, just checked my main partition which was installed in Nov. 2023 and it has over 750 files in that session-bus folder.
Root creates a new one too any time you log-in as root, and those build up as well.
Seems like something ought to clear the old one(s) out when you shutdown.
@fsmithred ; in the Refracta-Snapshot excludes file, /home/(user)/.dbus is excluded, and so when you boot/install the live iso it just has the new one created when you boot up. But root's version is not excluded by default, lol just found a lot of files in a new version of Vuu-do i'm making in root's .dbus directory, about 150 of 'em.
Could the /root/.dbus directory be safely excluded as well?
Bleachbit isn't set to clean the .dbus/session-bus folder because you can't remove all of them, you need the most recent one. But am wondering if something could/should be added to a shutdown script to remove the old files.
You might be better off trying an appimage, a quick search shows them available in kicad versions 6,7,8, and 9.
1. a very simple bash script (e.g., three lines of code)
If you don't know how .desktops can work, you may be convinced you need a script. ![]()
[Desktop Entry]
Type=Action
Comment=Install local .deb package
Name=Install Package
Icon=info
Terminal=true
Profiles=on_file;
[X-Action-Profile on_file]
Exec=gksu.sh apt-get install %f
MimeTypes=application/x-debian-package;You could also just go the .desktop route, use apt-get, and run it in terminal, that at least would save some typing, so not a gui method, but a quick timesaver. In my workflow lately I do a lot of local installs.
Hmmm..don't know the answer, but perhaps I have a tidbit that may help some. To alter Pluma which is the text-editor I use, at least insofar as internal colors, I had to mod a file in /usr/share/gtksourceview4/styles (classic, which is default in Pluma) to have an effect, these files seem to override gtk3 settings in the global theme. But I don't see anything about icons in that file currently, but perhaps that's another parameter that could be altered in that file.
Try "su" to root, then "apt" or "dpkg".
It seems that you do not know how to write a simple script.
Well I haven't taken a sip yet today from the waterfall of secret esoteric knowledge, so my dementia lacks the mystic sauce of enlightenment to keep it at bay, perhaps you need to wall-of-text me some more of that deep deep lore that ChatGPT has bestowed upon thee. ![]()
sudo, and here I thought you were suspicious of obvious security risks.
Vuu-do don't sudo. ![]()
another option would be to cobble a script with yad for the gui (lighter for setups that already use yad) and on the inside all it does is run apt-get commands to show the information of the .deb and installs it with apt-get install <full/path/to/package.deb>
That was my thought. gdebi is only about 1 mb so it's small, but i'm pretty sure a simple script for doing this with yad and apt-get would only be like 20-30k. Edit: Actually less than that the more I think about it, apt-get will do all the legwork checking for depends etc. Just need to use the right command, then a success dialog at the end, some error checks of course etc.
If you're just making a quick and dirty for yourself and don't need a safety net, you can do it as a right-click extension with just a .desktop, like maybe so (if you know for sure you already have all needed depends installed):
[Desktop Entry]
Type=Action
Comment=Install local .deb package
Name=Install Package
Icon=info
Profiles=on_file;
[X-Action-Profile on_file]
Exec=gksu.sh dpkg -i %f
MimeTypes=application/x-debian-package;Before somebody asks again, gksu.sh is just my pkexec translator so I don't have to type a lot of extra crap when I make one of these.
yes, i've been thinking of rolling out my own apt wrapper, after all i already have apt-ui which uses fzf to provide a fuzzy ui for installing, uninstalling packages and also queriying package information: https://github.com/eylles/devuan-script … /apt-ui.sh
The comments! # if this is not your actual config dir then get rekt ![]()
Srsly tho, this put a bee in my bonnet on a small utility i've thought about in the past, a right-click context menu entry that offers to install a local .deb package when you select it in the file manager, little script + apt-get + some yad dialogs. I think that would be handy.
Probably easier to just use gdebi, it's nice.
/usr/lib/udev/rules.d/90-alsa-restore.rules
Here's the version in alsa-utils from their git:
# do not edit this file, it will be overwritten on update
ACTION=="add", SUBSYSTEM=="sound", KERNEL=="controlC*", KERNELS!="card*",@extratest@ GOTO="alsa_restore_go"
GOTO="alsa_restore_end"
LABEL="alsa_restore_go"
ENV{ALSA_CARD_NUMBER}="$attr{device/number}"
# mark HDA analog card; HDMI/DP card does not have capture devices
DRIVERS=="snd_hda_intel", TEST=="device/pcmC$env{ALSA_CARD_NUMBER}D0p", RUN+="/bin/sh -c 'echo ALSA_CARD_HDA_ANALOG=$env{ALSA_CARD_NUMBER} >> /run/udev/alsa-hda-analog-card'"
# check for ACP hardware
TEST=="device/device/acp3x-dmic-capture", GOTO="alsa_hda_analog"
TEST=="device/device/acp6x-dmic-capture", GOTO="alsa_hda_analog"
TEST=="device/device/acp63-dmic-capture", GOTO="alsa_hda_analog"
TEST=="device/device/acp-dmic-codec", GOTO="alsa_hda_analog"
GOTO="alsa_restore_std"
LABEL="alsa_hda_analog"
# restore configuration for profile with combined cards (HDA + digital mic)
TEST!="/run/udev/alsa-hda-analog-card", GOTO="alsa_restore_std"
IMPORT{program}="/usr/bin/cat /run/udev/alsa-hda-analog-card"
ENV{ALSA_CARD_HDA_ANALOG}!="", ENV{ALSA_CARD_NUMBER}="$env{ALSA_CARD_HDA_ANALOG}"
LABEL="alsa_restore_std"
TEST!="@daemonswitch@", RUN+="@sbindir@/alsactl@args@ restore $env{ALSA_CARD_NUMBER}"
TEST=="@daemonswitch@", RUN+="@sbindir@/alsactl@args@ nrestore $env{ALSA_CARD_NUMBER}"
LABEL="alsa_restore_end"And here's the current version in excalibur/trixie:
# do not edit this file, it will be overwritten on update
ACTION=="add", SUBSYSTEM=="sound", KERNEL=="controlC*", KERNELS!="card*", TEST=="/usr/sbin", TEST=="/usr/share/alsa", GOTO="alsa_restore_go"
GOTO="alsa_restore_end"
LABEL="alsa_restore_go"
ENV{ALSA_CARD_NUMBER}="$attr{device/number}"
# mark HDA analog card; HDMI/DP card does not have capture devices
DRIVERS=="snd_hda_intel", TEST=="device/pcmC$env{ALSA_CARD_NUMBER}D0p", RUN+="/bin/sh -c 'echo ALSA_CARD_HDA_ANALOG=$env{ALSA_CARD_NUMBER} >> /run/udev/alsa-hda-analog-card'"
# check for ACP hardware
TEST=="device/device/acp3x-dmic-capture", GOTO="alsa_hda_analog"
TEST=="device/device/acp6x-dmic-capture", GOTO="alsa_hda_analog"
TEST=="device/device/acp63-dmic-capture", GOTO="alsa_hda_analog"
TEST=="device/device/acp-dmic-codec", GOTO="alsa_hda_analog"
GOTO="alsa_restore_std"
LABEL="alsa_hda_analog"
# restore configuration for profile with combined cards (HDA + digital mic)
TEST!="/run/udev/alsa-hda-analog-card", GOTO="alsa_restore_std"
IMPORT{program}="/usr/bin/cat /run/udev/alsa-hda-analog-card"
ENV{ALSA_CARD_HDA_ANALOG}!="", ENV{ALSA_CARD_NUMBER}="$env{ALSA_CARD_HDA_ANALOG}"
LABEL="alsa_restore_go"
TEST!="/etc/alsa/state-daemon.conf", TEST=="/usr/sbin/alsactl", RUN+="/usr/sbin/alsactl -E HOME=/run/alsa -E XDG_RUNTIME_DIR=/run/alsa/runtime restore $env{ALSA_CARD_NUMBER}"
TEST=="/etc/alsa/state-daemon.conf", TEST=="/usr/sbin/alsactl", RUN+="/usr/sbin/alsactl -E HOME=/run/alsa -E XDG_RUNTIME_DIR=/run/alsa/runtime nrestore $env{ALSA_CARD_NUMBER}"
LABEL="alsa_restore_end"Okay, more info. The bug is in alsa-utils, in the 90-alsa-restore.rules file. Looks like alsa fixed it months ago, but then a patch from Debian broke it again so now I guess it's their bug, and it's still an issue in Trixie as of 2 days ago.
https://bugs-devel.debian.org/cgi-bin/b … ug=1093057
The patch is also supposed to be a fix for some mics not working from what i've read.
Okay, the point release from upstream came in the day after I put up those last iso's, lol, so I updated everything and ran a couple new ones. Just a maintenance release, 87 updates in the max and 70 in the mini. They're up on the site now.