Note: your udev filter (the ones with ==) matters!! I thought I could re-use the filter generated from loginctl (i.e. ENV{ID_FOR_SEAT}=="..."). This led to several confusing hours where I could set ENV{ID_SEAT} but not ENV{WL_SEAT}. Follow the docs above and you'll be fine.
Once I got it working in Weston, I tried it on Plasma, and got my answer: no, Plasma Wayland does not currently support multi-cursor.
It looks like the best solution in our case will be to use the Plasma X11 session. I was able to reinstall the x11 session on my pc and tested multi-cursor there, and it seemed to work as expected.
If the X11 session is too unstable, I guess Weston is there as a backup.
Retro gaming device seems pretty cool actually. What if you configure your system so that when you plug it in, it autolaunches an emulator and looks at ROMs on the USB :o
If you aren't aware, pbm represents an image with plaintext, which makes it great for when you want to easily create an image with code
I recently learned there is a whole suite of CLI tools which work with the format. Like conversion to/from png, scaling, and overlaying one image on top of another.
I would normally use KDE Connect (iOS version) but I had some issues with it. The push local clipboard didn't work, and I think receiving files to my phone didn't work either. It also had a hard time reconnecting to my computer after pairing. It might be better now though, this was several months ago.
I primarily want something simple I can bend to my will, and secondly I want a good out of the box experience. For me that's been Arch + KDE. The wiki and AUR are great!
I would say every step of the way I just wanted more and more ownership of my system. I make it, I break it. One specific experience which drove me to that camp was the time I had to jailbreak my iPhone and dig through files to disable some deprecated parental control setting... give me ownership over my stuff!
Good call with using a shared reference so now we can kiss our partner(s) more than once!!
Kissing many partners with one big smooch might be hard tho, maybe we need to change the implementation to use combinatorics so that everyone kisses each of their partners at least once.
Since there's no way to tell who kisses who, i can cheat by printing nCr times :3
rust
pub fn lovers(partners: &[&dyn Gender]) {
match partners.len() {
0 => {
println!("ghost smooch~");
}
1 => {
println!("mirror smooch~");
}
p => {
// p >= 2, safe to unwrap
for _ in 0..ncr(p, 2).unwrap() {
println!("smooch~");
}
}
}
}
pub trait Gender {}
// implement your own gender, and if u want, publish online! :3
pub struct Masculine;
impl Gender for Masculine {}
pub struct Feminine;
impl Gender for Feminine {}
// removed old api:
// fn two_lovers(one: Masculine, two: Feminine)
// srry not srry for breaking backwards compat
// new version has more flexible api!<3
fn two_lovers(one: Box<dyn Gender>, two: Box<dyn Gender>) {
println!("smooch~");
}
// todo: other functions!!
#[cfg(test)]
mod tests {
use super::*;
#[test]
fn me_and_who() {
two_lovers(Box::new(Feminine), Box::new(Feminine));
}
}
Someone can correct me if im wrong but I think nullable in C# still doesnt force you to do a null check before accessing properties or methods, however I believe it does give a type error when using a nullable type where non-nullable is expected, which is good. So it might not be 1-to-1 but it fulfills a similar purpose for sure.
Coming from a rust background I would have sure appreciated my workplace enabling the nullable feature on our main codebase from the start. I've run out of patience for null errors :(
I've had it where my wireless mouse (connected with a usb dongle) stutters when my wifi/bluetooth chip is going full-throttle. I thought it was some polling rate on my mouse, or maybe my mouse was dying, but nah lol. This is next level insane tho
That's real. It sounds like despite all the circumstances being bad you've settled in to a spot where you can be as good as you can be for now. It also sounds like your humble research with dpdr has been good in finding explanations for your lived experience, which I hope has a side effect that you feel less alone :) All of this plays a role into how you feel, and I appreciate the detail. no worries about the lil vent 🩷
Yesss repairing stuff feels amazing, not only do you get a sense of independence, but the computer goes through this hero's journey-like experience that you get to play a guiding hand in. At the computer's darkest moment you were able to step in and save it from death, and let it breathe once more >:3
Ooo I need to do this for my old Windows 8 laptop. It has a touchscreen and I'd like to use it for a low-spec + Wayland touchscreen testing device. Unfortunately, the battery has been dead for years, and I've also lost where the laptop went D:
This was my thought as well because the problem resolves itself within a matter of minutes. However it doesn't make sense that OP would have the same problem for multiple instances, and could access each instance in the browser without issue. Also, our instance showed 100% uptime when I initially checked the post so I'm not sure it's that.
My biggest hurdle to making the switch is that it takes effort.
Thats pretty understandable honestly. If my old netbook wasn't so bogged down from Windows, there's a chance I wouldn't have switched to Linux. In a way, I'm glad for it, but it woulda been easier if it just came with Linux preinstalled <3
The biggest thing I remember was driver support being awful. I guess it's a lot better now?
In my experience drivers have been pretty solid... except for NVIDIA. People seem to either have 0 issues or tons. Fingers crossed the upcoming Nova+NVK driver combo brings a more consistent and stable experience for all :)
I'm not sure, but I think the windows progress bar thingy includes the time spent actually writing to disk, whereas on Linux (or i guess cinnamon) only shows when it gets to disk-cache. If you are full on RAM or tried shutting down immediately afterwards it should take a bit longer since it has to actually write it to disk
Thats a good observation. I would think I use Ctrl more frequently than Alt, so it would make sense to have Alt where Ctrl usually is, and Ctrl where Shift usually is.
Even if you do use both shift keys correctly, it still feels awkward to me since you have to sacrifice a pinky from home row and shift your hands.
It's why I've been curious to try a layout where Shift and Alt are swapped, that way you can use your thumb to press shift and keep your fingers in place.
I got inspired and made a similar effect is desmos
f(theta, t) = 0.1 max(cos(theta + t), 0)^2 sin(8 theta)
r = 1 + f(theta, t)
r = 1 - f(theta, t)
where: 0 <= theta <= 2 pi
t = 0
where: 0 <= t < 2 pi
It's not quite right though, the sin(8 theta) part needs to phase shift, but basing it off of t messes up the perfect loop.
At that point I would reach for code, which also lets you save the animation. Using something like p5.js (editor.p5js.org) you can easily draw graphics and save each frame as its own image
I dug even deeper and installed Weston to see how seats are configured. Turns out this is done by attaching a udev property to the device called WL_SEAT, which libinput uses to make logical seats, which Weston uses for seat management. See: https://wayland.freedesktop.org/libinput/doc/latest/device-configuration-via-udev.html
Note: your udev filter (the ones with
==) matters!! I thought I could re-use the filter generated from loginctl (i.e.ENV{ID_FOR_SEAT}=="..."). This led to several confusing hours where I could setENV{ID_SEAT}but notENV{WL_SEAT}. Follow the docs above and you'll be fine.Once I got it working in Weston, I tried it on Plasma, and got my answer: no, Plasma Wayland does not currently support multi-cursor.
It looks like the best solution in our case will be to use the Plasma X11 session. I was able to reinstall the x11 session on my pc and tested multi-cursor there, and it seemed to work as expected.
If the X11 session is too unstable, I guess Weston is there as a backup.