You are not logged in.
Lol. No, not really I suppose but I have noticed that pipewire insinuated itself into my Debian box when I went from buster to bullseye and now I can't remove it. Bugger.
Apologies for the clickbait thread title, I copied it from the blog post:
systemd core — the building block framework abstracting basic functionalities
A universal package manager with a single reposotiry where anyone can push
flatpak/appimage/snap — containarized applications
polkit and apparmor — granular security
systemd-homed — movable home with an immutable base system
wayland compositors — the graphical environment
pipewire — media pipeline
https://venam.nixers.net/blog/programmi … stros.html
Sounds great. I presume everybody is thrilled about this direction of development? ![]()
Proxmox is also fundamentally broken if the _many_ support threads we get over at fdn are anything to go by. The fuckers even refer users to our forum in case of problems ![]()
if runit cant handle a process , and it uses sysvinit script to start it , that means that this process hasnt a supervisor and if it fails it wont start automatically ?
I think runit relies on the same PID file hack that sysvinit uses for processes not supervised by runsv.
GLib-GIO-Message: 23:45:04.055: Using the 'memory' GSettings backend. Your settings will not be saved or shared with other applications.
Is the dconf-gsettings-backend package installed?
And how did you install MATE? I would use the desktop task:
# apt install task-mate-desktop^ That installs everything required for the full MATE desktop, just as if it had been selected from a netinstall ISO image.
Why is it so so complicated?
Because wpa_supplicant is generally used as a back-end for more "user-friendly" networking solutions such as NetworkManager, connman & wicd.
You can try wpagui but that will still need a basic configuration file that declares the control interface file and allows the configuration to be changed. Also note that the wpa_supplicant package does not configure the routing table (so you will need to either do that manually or run a separate DHCP client) and it does not supply a sysvinit script (so you will have to create one yourself if you want to run it without a front end).
There is iwd, which is a more modern alternative to wpa_supplicant that also has a command line interface (iwctl) but doesn't require you to create a configuration file beforehand, but that also doesn't supply a sysvinit script. It does include DHCP functionality and it has fewer dependencies than wpa_supplicant.
But I would recommend ifupdown instead: https://wiki.debian.org/WiFi/HowToUse#W … d_WPA2-PSK
It's always best to enter the error message into a search engine, here's the top result:
https://raspberrypi.stackexchange.com/q … t-wlan0-re
This is also mentioned in the ArchWiki:
https://wiki.archlinux.org/index.php/Wp … th_wpa_cli
And I'm curious as to why your interface is using the predictable nomenclature — are you using eudev?
apt policy eudevThanks rolfie. So the Debian and Arch packages both supply the old WebKit version whereas upstream have moved to Electron and now use a different versioning.
# aptitude hold midoriBut I don't understand why the versions are so different in the Devuan package compared to upstream and I also don't understand why the Arch package has a different version again and is drawn from a different (GitHub) repository. Very confusing ![]()
i use this
Well done, you have pulled in the libc6 package from sid and prevented it from being subsequently upgraded. Good luck with any future vulnerabilities in that package, your system will be left wide open to them.
@all: don't try this at home.
And just for the record experimental has a pin value of 1 (one) by default so you don't need a preferences entry for that and if you did want to pin a package from ceres (which you don't) then it would only need a pin value of 500 because APT would favour the higher package version automatically.
EDIT: hold on, are you using testing? The above would apply to beowulf users.
can you please tell why you did not give your genuine advice to other users ?
Because I like you best :-X
The file should be at ~/.kodi/temp/kodi.log but I think you have to enable debug mode for it to be created:
New Linux port for the M1:
https://corellium.com/blog/linux-m1
An Ubuntu variant, unfortunately, but it's a start.
so why would dylanaraps go to all the trouble of building iwd without dbus?
Dylan is utterly fanatical about minimalism and will go to almost any lengths to reduce the code base.
EDIT: I removed all of the xinerama ifdefs from my custom dwm packages for Alpine, Debian & OpenBSD for a similar reason. It saved a few hundred SLOC IIRC.
iwd doesn't have a dbus dependency and works just fine without it, dbus activation is only needed if iwd is being used as a backend for NetworkManager or connman.
@OP: sorry for the diversion, I'll stop now.
A bit perplexed now
Don't worry, Andy is just having a laugh. As am I :-)
I suppose I should make it clear that I have the utmost respect for the MX Linux developers, they make a fine distribution. I just don't get along with their forum admins, that's all. The admins here are _much_ better.
Ill probably switch to iwd when/if it is ready for devuan
The iwd maintainer wouldn't add the init scripts to the package without a proper diff but they did modify it so that non-systemd dbus activation should work. This means that the version in testing/unstable should work as a backend for NetworkManager or connman but an init script would still have to be added for standalone usage.
See https://bugs.debian.org/cgi-bin/bugrepo … bug=966518 for the details.
just needed to delete those foreign repositories. Replace the buster backports entries with devuans backports and update, upgrade should have fixed it
I don't think that would have removed any foreign packages that were already installed.
Are you referring to DistroWatch's bullshit "popularity" rankings? You know they're bullshit, right?
I'm pretty sure MX have a bot clicking on their page repeatedly (they copied the idea from Manjaro).
The user review rankings are probably a better source and Devuan does well in those:
https://distrowatch.com/dwres.php?resource=ranking
Third place is pretty good, no? MX Linux is lingering down in 17th place...
Try it and see. Backup first though.
If you have any more queries about rescuing your broken system then please open a new thread devoted to that topic. Thanks.
Yeah, kernel regressions are unfortunately common in Linux. The developers are too busy cramming in new features so old hardware support can suffer. You should probably file a bug report with the kernel developers because they can't fix it if they don't know about it and you won't be able to keep using old kernels indefinitely.
I've not tried it myself but it looks like /usr/bin/pipewire-pulse & /etc/pipewire/media-session.d/with-pulseaudio are only supplied by the pipewire-bin package that is currently in ceres (pulled in by the pipewire metapackage).
Unfortunately though it looks like it uses socket activation via a systemd unit so I'm not sure if it will actually work in Devuan.
as root user do the following.
wpa_passphrase ESSID PASSPHRASE > /etc/wpa_supplicant/wpa_supplicant.conf
It is possible to use ifupdown without creating a wpa_supplicant configuration file.
Just define the SSID and password directly in /etc/network/interfaces, like this:
auto wlan0
iface wlan0 inet dhcp
wpa-ssid $ssid
wpa-psk $passwordReplace $ssid with the name of the access point and replace $password with the password (or use the obfuscated psk line from the wpa_passphrase output but note that the hashing is performed unsalted so it should not be considered secure).
See also https://wiki.debian.org/WiFi/HowToUse#W … d_WPA2-PSK
But having said that a wpa_supplicant configuration file would allow for the use of wpagui, which needs these lines added to the top of the configuration file:
ctrl_interface=/run/wpa_supplicant
update_config=1The GUI is fairly "friendly" and may be suitable for the OP's needs.
I prefer iwd over wpa_supplicant because it is lighter, more secure, has fewer dependencies and it can be used as a complete wireless networking solution by itself (it has built-in dhcp functionality). The only drawbacks are that the Devuan package lacks a sysvinit script and the version in beowulf is pre-release.
I've just tested your sample program using the ASCII live ISO image and it compiles and runs just fine for the non-privileged user.
From this I can deduce that your system has been subtly broken by your reckless and excessive use of backported packages.
EDIT: I told you so ![]()