Looks suspiciously like https://github.com/marktext/marktext

Edit: Ignore please, the project is dead

Any reason for avoiding flatpak?

For me, it’s after trying a non tiling WM. I’m too deep into i3 (now Sway) + tmux.

For me, it’s after trying a non tiling WM. I’m too deep into i3 (now Sway) + tmux.

Too low on the list

Ooof. Really puts things into perspective

why would an xbox that runs an x86_64 architecture need a translation layer to run Windows binaries?

This is based on my partial understanding of the xbox architecture. I suspect that there is a difference between full fat Windows and the xbox OS. So certain function calls (required by Steam) are not supported by xbox OS. So it won’t be a traditional layer like WINE or Proton. Just enough for certain gamestore applications to run (Epic?).

going to be a fixed hardware Windows PC with a custom interface replacement and probably all the unnecessary services ripped out of it.

I hope for this too. But how will traditional Xbox games run if it’s just Windows? Will it have a built-in emulator running on top of Windows? Also, it will invite people to hack the device since it’s a standard Windows OS. I doubt MS wants their next xbox to be so open, given their history regarding MS Store.

I’m going to assume MS is going to containerize/translate Steam somehow. So you will have the standard locked-down xbox OS for playing all xbox games + Windows/W32 translation layer for using Steam (Windows Subsystem For Xbox). All games launched via Steam will use the translation layer.

The WSX will be cryptographically signed so that only authorized W32 applications can be executed which are pushed by MS only.

Agreed. OP should just backup their Linux home directory (Everything inside ~/) and do a fresh install. Your Linux install will thank you for that.

In my university (where we used proxy), we had a saying “it’s always DNS proxy”

That is a good point. But I remember launching the application from the terminal where I explicitly set the env variables as null. Still it didn’t work.

Agree.

However, I always have those “trivial” conditions be explicitly set. Usually, when the device is unplugged correctly, your primary display will be automatically be turned on. However, in very weird scenarios (incorrect voltage signals, loose wire, etc), it’s possible that the explicit else condition will be triggered.

What about the condition that when the external display is disconnected, the main display should be reenabled?

You can have a perfect distro on a USB which boots into Linux automatically when inserted into your PC and it preserves all your files and favorites. Still it will hardly increase the market share by 1 or 2%. It’s because a super minority of people will bother to get the USB drive.

The core issue we have to understand is “availability of preinstalled Linux on PCs in brick and mortar shops”. Till this is solved, we won’t get market share. The only reason people are using Linux on their SteamDeck is availability.

My brother in Christ, do you think an average person knows what BIOS/boot menu are?

It’s some sort of proprietary peer to peer algorithm

I completely lost interest for the project at this point of the text

Issue resolved!

It was swhkd. Thank you very much for your insight and extremely detailed response!

$ ls -l $(which swhkd)
-rwsr-xr-x 1 root root 2583192 Mar 10 17:16 /usr/bin/swhkd

Since we know what’s causing it, can you make a “guesstimate” of what it’s doing? Why are other applications are getting infected by it? And why is a keybind manager affecting permissions?

I will raise an issue on their github. The project is already looking for maintainers.

How do you open the shell inside sway? Keyboard binding from sway config? Launcher? Which terminal? Do any of the involved programs have setuid root bit set (looks like rws instead of x in ls -l output)?

I think you may have just pointed me to the correct direction.

My keybinds setup is a bit weird. I’m using swhkd instead of sway’s built in keybinds. swhkd is a setuid binary (https://github.com/waycrate/swhkd?tab=readme-ov-file#running) which might be causing the issue. I’ll quickly disable swhkd and check if the issue is resolved. Will keep you posted.

$ which sway
/usr/bin/sway

$ sway --version
sway version 1.9

$ file $(which sway)
/usr/bin/sway: ELF 64-bit LSB pie executable, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib64/ld-linux-x86-64.so.2, BuildID[sha1]=70fe358f7e410f618ad8a9ce0e573ed6826b2e75, for GNU/Linux 3.2.0, stripped

$ ls -l $(which sway)
-rwxr-xr-x 1 root root 600352 Apr  1  2024 /usr/bin/sway

id pre and post login

uid=1000(xavier666) gid=1000(xavier666) groups=1000(xavier666),0(root)
---------------
uid=1000(xavier666) gid=1000(xavier666) groups=1000(xavier666),4(adm),24(cdrom),27(sudo),30(dip),46(plugdev),120(lpadmin),132(lxd),133(sambashare)

A funny thing; I think this has nothing to do with gdm. I have gdm disabled now and launching sway directly from the terminal and the issue still persists.

The problem goes away (xavier666 becomes part of sudo like expected) when I type exec su - xavier666 for that terminal session only. If I open a new terminal, it problem reappears. I’ll just in case check if zsh/omyzsh is doing something funny.

I was experimenting. This might be helpful (https://lemm.ee/post/62662283/20097388)