You are not logged in.
More of that I understand that commercial LLMs are trained mostly, always exclusively on open source code. So, to my mind it must be (as it will be fair) free to all open source projects, but the world is not perfect.
Yes. I have mentioned that Cinnamon X11 handles HIDPI well. KDE X11 does so as well. I then explained why I chose KDE over Cinnamon:
1. Lower CPU usage.
2. Support for both X11 and Wayland, where Wayland provides better automatic configuration (resolution, brightness, and refresh rate) on multi-display setups.
"...It is written in Rust..."
By LLM.
If you have some ideological restrictions – it's no problem. I've no affiliation with this free and open source product and share this with others who need it. Personally I don't care about language and have no problem with wibecoding.
More of that I understand that commercial LLMs are trained mostly, always exclusively on open source code. So, to my mind it must be (as it will be fair) free to all open source projects, but the world is not perfect.
I personally prefer and recommend https://github.com/pkgforge-community/i … x-appimage. This is an AppImage of the VirtualBox KVM version. It includes VB extensions that do not require installation and use KVM to run VMs without needing to install additional kernel modules from Oracle that previously had critical vulnerabilities. All features of VBox are included, such as shared folders and USB support; they are out of the box. Another advantage is that it can be run in parallel with any other KVM virtualization tool.
KDE.
I have a HiDPI display and an external monitor; the settings for the internal and external displays must be distinct.
Cinnamon handles HiDPI reasonably well but its multi‑display configuration is less effective than Wayland.
All other X11 desktop environments (except Cinnamon) look terrible.
Also KDE uses fewer resources than Cinnamon when Baloo is disabled and unnecessary components are not installed.
In terms of lighter usage, I am concerned about CPU usage rather than the 100–200MB difference in RAM.
This is a new and interesting project. Its aim is to replicate the functionality of all Adobe and Microsoft Office products. The project is still in alpha, so it is not 100% completed, but it works. It is written in Rust, and each application is under 100 MB, running extremely quickly.
https://github.com/storytold/photocraft
https://github.com/storytold/filmcraft
https://github.com/storytold/lightcraft
https://github.com/storytold/wordcraft
https://github.com/storytold/gridcraft
https://github.com/storytold/deckcraft
Etc.
Enjoy!
Wrong, Debian is the originator of most hickups.
May be. But not all. I was able to install all qemu(s) on Freia from Ceres, but not libvirt and virt-manager. I got the same error.
I've used Debian's testing repo with Freia and the problem disappeared.
Thanks rolfie, good to know.
freia is testing - this happens from ti … ts better.
I did not experienced such kind of problems with Excalibur (while testing) and was surprised. I had to believe that Devuan's repo is a mirror of Debian's repo with some mods. They don't have such problems usually and there are many distributions based on testing.
Firejail from Ceres is OK. I thank it was intentionally removed.
I can't install Spaced Linux on my laptop - the installer is broken. Sometimes it shows options, sometimes not and encrypted install on separate partitions with unencrypted boot fails to boot. The live session is almost OK: it shows desktop scaled as expected automatically detecting HDPI display, then in 1-2 sec switches to nonscaled ![]()
I'm trying Freia and
First:
I'm not able to install virt-manager, qemu, libvirt, guestfs etc
The following packages have unmet dependencies:
libguestfs-tools : Depends: libguestfs0t64 (= 1:1.60.1-2) but it is not going to be installed
Depends: guestfs-tools but it is not going to be installed
Depends: guestmount but it is not going to be installed
Depends: guestfish but it is not going to be installed
libosinfo-bin : Depends: libosinfo-1.0-0 (>= 1.10.0~) but it is not going to be installed
libvirt-clients : Depends: libvirt-common (= 12.4.0-1devuan1) but it is not going to be installed
Depends: libvirt0 (= 12.4.0-1devuan1) but it is not going to be installed
libvirt-daemon-system : Depends: libvirt-common (= 12.4.0-1devuan1) but it is not going to be installed
#... etc
qemu-utils : Depends: liburing2 (>= 2.3) but it is not going to be installed
Recommends: qemu-block-extra (= 1:11.1.1+ds-1) but it is not going to be installed
virt-manager : Depends: gir1.2-gtk-3.0 (>= 3.10) but it is not going to be installed
Depends: gir1.2-gtk-vnc-2.0 but it is not going to be installed
#...etcpython3-libvirt: requires libvirt0 >= 12.5.0 but libvirt0 available in repo seems to be 12.4.0-1devuan1 etc
Full mess of package versions.
Second:
Firejail & Ko are not in the repo at all. Why? Are they banned?
Does bilibop use a lot of RAM?
How many is too many? All hard disk writes are happen in RAM. You can use noatime, nodiratime options for FS. Then all logs, /var. /tmp are in the RAM.
By default tmpfs is limited by 1/2 of RAM and you can change this value. To give you an idea, I'm using it even on very old laptop with 8 Gb RAM and no SWAP! But ... I prefer to keep only one browser with around 30 pages open and not 2 or 3 browser while watching video
on this machine.
32 Gb of RAM, no swap, default tmpfs everything works perfectly: couple of browsers, video and ... virtual machine with 8 GB of virtual RAM or small local LLM.
I am using the traditional "full disk"-encryption with luks, that leaves /boot unencrypted.
+1
Best balance simplicity/security.
Q: anyone using overlayroot?
Altoid, I'm using bilibop from devuan/debian repository. The whole / is on overlayfs on top of tmpfs. Works fine with chimaera, daedalus, excalibur sysvinit and openrc, also with ceres dinit. All software that I use works. Some minor issues like virt-manager requests log file is simply solved by creating or copying empty log file on boot, vmware complains :"no space" - solved by one line in the config to ignore space check. There are no other issues as I remember.
There is also one issue with reboot/shutdown - 30 seconds delay if luks encryption is used together with overlayfs on top of tmpfs. Only ceres dinit (from what I've tested) has no this issue.
I've modified rc-local.sh to not look for other scripts and put commands for ufw in rc.local. It's not clean solution but it works for now.
/etc/dinit.d/scripts contains some artifacts like
cat rc.local
#!/usr/bin/shwith regular user permissions. Almost all scripts in this folder have regular user permissions. I did not create them, strange ![]()
bluetooth is strated
dinitctl status bluetoothd
Service: bluetoothd
State: STARTED
Activation: start due to dependent(s)
Process ID: 893
but I've no bluetooth: no icon for it, no bluetooth in KDE settings.
ufw
type = scripted
command = /usr/sbin/ufw enable
stop-command = /usr/sbin/ufw disable
depends-on = mount-all
logfile = /tmp/ufw.log
the same:
dinitctl status ufw
Service: ufw
State: STARTED
Activation: start due to dependent(s)but
ufw status
Status: inactiveI can add ufw enable to rc.local to bypass this issue but rc-local service does not start. Log file:
ls: cannot access '/etc/dinit.d/scripts/rc-local/*.start': No such file or directoryservice:
/etc/dinit.d# cat rc-local
type = scripted
command = /usr/bin/sh /etc/dinit.d/scripts/rc-local.sh start
stop-command = /usr/bin/sh /etc/dinit.d/scripts/rc-local.sh stop
depends-on = login.target
logfile = /tmp/rc-local.logscript:
cat /etc/dinit.d/scripts/rc-local.sh
#!/usr/bin/sh
case "$1" in
start)
sh -c "/etc/dinit.d/scripts/rc.local" && ls -1 /etc/dinit.d/scripts/rc-local/*.start | while read script; do sh -c $script; done;
;;
stop)
ls -1 /etc/dinit.d/scripts/rc-local/*.stop | while read script; do sh -c $script; done;
;;
*)
exit 1
;;
esacthe folder called by script is almost empty:
ls /etc/dinit.d/scripts/rc-local/
READMEConfigure Timeshift to take one snapshot per week. I have not had any problems after updates.I've reinstalled MXD again and had no issue with today' s update.
I've few questions about MXD and dinit (I'm new in dinit)
1. The distro has no doas preinstalled but I see preconfigured doas.conf in /etc/
2. I created ufw in /etc/dinit.d:
type = scripted
command = /usr/sbin/ufw enable
stop-command = /usr/sbin/ufw disable
restart = falseand not sure if I need depends-on, before, logfile? Also the service strarts but ufw is not enabled after reboot.
3. Service rc-local was in place already:
type = scripted
command = /usr/bin/sh /etc/dinit.d/scripts/rc-local.sh start
stop-command = /usr/bin/sh /etc/dinit.d/scripts/rc-local.sh stop
depends-on = login.targetit' s enabled and started but rc.local is ignored. What can be wrong?
4. It seems that bluetooth is installed:
dpkg -l | grep blue
ii bluetooth 5.85-4 all Bluetooth support (metapackage)
ii bluez 5.85-4 amd64 Bluetooth tools and daemons
ii bluez-alsa-utils 4.3.1-3.1+b1 amd64 Bluetooth Audio ALSA Backend (utils)
ii bluez-firmware 1.2-13 all Firmware for Bluetooth devices
ii libasound2-plugin-bluez:amd64 4.3.1-3.1+b1 amd64 Bluetooth Audio ALSA Backend (plugins)
ii libbluetooth3:amd64 5.85-4 amd64 Library to use the BlueZ Linux Bluetooth stack
ii libkf6bluezqt-data 6.23.0-2 all data files for bluez-qt
ii libkf6bluezqt6:amd64 6.23.0-2 amd64 Qt wrapper for bluez
ii papirus-mxblue 25.06.02mx25 all papirus blue folders with tabs for MX
ii qml6-module-org-kde-bluezqt:amd64 6.23.0-2 amd64 QML wrapper for bluezbut I don' t see any traces of bluetooth to set it up.
You can't install that particular distro to the same drive you are booting from in ventoy. I have had that in MX Linux since even the file managers don't let us access that drive either.
Thanks for info. Anyway it's over - kernel panic after update, not this distro fault but ceres one.
My experience: MX_Devuan_KDE_dinit is great but as it's based on unstable it does not survive un update ![]()
Maybe your SSD is not visible in Live, look before installing it in Gparted and try to mount it manually.
The installer works as expected when I use a Ventoy flash drive. The aforementioned strange behavior only occurs when booting the Live system from this second SSD using Ventoy. In such cases, I see all partitions on the first drive and only Ventoy-related partitions on the second. As a result, I am unable to install the distribution to the second drive.
I found an interesting feature of this distribution - the installer refuses to see the second disk. Both disks are SSDs. Why?
I2P presence
+++
If I were in the place of bureaucrats, instead of resorting to intimidation, I would introduce substantial grants for distributions that implement verification. For the EU or the USA, one million dollars per year per distribution is a negligible amount. It would be interesting to see how many developers would still oppose it then.
Candy, like Whip and Candy Principle? It's no longer necessary. The Cow-ID test for compliance has been passed successfully. Stay home in your personal prison, don't hug your loved ones, don't communicate with them except through the internet, and don't ask what is in the syringe. Now, the whip is largely sufficient.
https://www.phoronix.com/forums/forum/s … sion/page4
SysV init 3–16 was released, but the real debate is about mandatory ass fingerprints in Linux.
Anyway, the kid rapist will keep their policy—nobody dares stop them. The FBI says it’s impossible to arrest them because it’d destroy political stability. As for "regular" UID1000 users? They’re just slaves and love their masters.
So how would I rate it? Or will I need to pay some censor authority to rate it like they would a movie film? If I rate it myself wrongly I could be bankrupted by a fine in £millions. In the UK it has been reported by someone who monitors this stuff that thousands of websites have already been pulled by their owners rather than risk a misstep.
Don't worry. I believe it's not for a long time. The world war III iis coming.
The become-ageless.sh script allows any already installed Debian system to be converted into Ageless Linux. The script replaces the /etc/os-release file and includes information about non-compliance with age verification laws (COMPLIANCE_STATUS="refused", API_STATUS="refused", VERIFICATION_STATUS="flagrantly noncompliant"). It also adds a /etc/ageless/ directory containing reports on violated laws, as well as implements a prototype of a dummy API for applications to obtain user age information.
Enjoy ![]()
It seems like I'll stop using internet except some official sh*t ![]()
It seems also that someone is already working on random credentials, but it's still like playing somebody else's game restricted by some rules which are not respected by the game owner.