redjard

I use an encrypted rootfs without “an initramfs”. Just requires some advanced fuckery.

A little known fun fact is that almost all kernels have a tiny stub-initram built into the kernel file. This is added when initram support is enabled, and is loaded before dedicated files are. It is however possible to supply your own initram directory or archive during kernel build to replace this built-in initram, so you can bake it in without leaving a separate file. No juggling with partitions, no boot options. Works just like a normal kernel “without initram”, since even kernels without one usually do have that stub one anyway.

The downside is that a) you have to build the kernel, and b) the files to pack have to be available when the kernel is built, meaning you can’t pack in modules of the kernel. But when building your own kernel anyway you can simply set the needed modules for encryption built-in and only pack the userspace cryptsetup executable needed for decryption, that way you get it all in a single kernel build, and the output is a single uniform kernel binary capable of decrypting your boot drive. No flags, no extra files, no access to the esp needed.

(I use gentoo with encrypted root btw.)

Found the Gentoo user

That would make it a different problem than what I saw. Either your compositor has bugs not present in its xwayland, or firefox has bugs in its wayland implementation that don’t occur under x11. Seeing this is snap, that could also be causing the issue.

concerns me since I know distros are starting to default to Wayland

xwayland will be supported for a long time to come, so this would only affect users that don’t know about this “fix” yet. You should be able to use firefox in x11 mode until this is fixed.

It would be good if you could check if this happens with a native (non-snap) firefox installation running on wayland. Ideally also if you could try it in a different dwm, probably kde since you are presumably using ubuntu with gnome.

I recall this happening on my higher refreshrate and resolution screen on X11 and nvidia (iirc a 2070 super). Moving to wayland fixed it.
I mainly noticed frame drops in videos in forefox, but I’m sure scrolling was also affected.

It probably manifests on a per usecase basis since it depends on how exactly the gpu or x11 is loaded. Whatever it is, it’s some consequence of x11 being flawed at its core abd just not scaling well with modern usecases.

In my case I was forced onto wayland after I got my new better monitor due to how choppy x11 felt.
I had 2 1080p 60Hz and got a 4k 120/144Hz screen.

I just don’t update kde very often since it might force me to re-login or restart. I currently have 9 months uptime so that’s the age of kde connect.

(gentoo)

Desktop version too old and phone version too new to be affected

(•_•)
( •_•)>⌐■-■
(⌐■_■)

But it saves the command history to disk.
It makes the exit command redundant, so that is removed from the history, but all other commands in all cases but mistakenly sshing into the wrong machine will still be saved.

invidious, archive.org

ECC ram. Hardware stability was the main factor.

Real Linus mentions spending days hunting an instability bug in the kernel to have it turn out as bad ram on his pc.

I use my mice at over 20k dpi, and for the longest time the sensors got increasingly more accurate but still had the rare wiggle. Especially with cars driving by.
To stop my system waking I first started turning the mouse upside down, but later instead propped it half up on the rim of my keyboard. That latter is really quite convenient, might also work for you.

The year is 2046, the ntfs4g+±next driver is put into maintenance mode as Linux slowly phases out support for the legacy operating system “windows”.

As long as you don’t run out of memory, you can actually insert and lookup in O(1) time for a known space of values (that we have). Therefore we do get the quadratic speedup, that when dealing with bits of keysize or entropy means cutting it in half.
Checking to get a specific uuid takes 128bit, so 2¹²⁸ draws of a uuid. Putting all previous uuids into a table we expect a collision in 64bit, so 2⁶⁴. We also need about that much storage to contain the table, so some tens of exabytes.

Against a apecific one too. Usually you’d check for duplicates against all previous uuids

In a way that weakens the open source software world by giving companies a way to avoid the GPL license of linux.
For really no reason, this project gains nothing from being MIT instead of GPL.

MIT license :|

I think they mostly are.
Cache is already in .cache/mozilla, so usually there is no change hence I didn’t mention it. They did move the cache to the XDG_CACHE_HOME by default. And yes they are using the XDG variables including the recommended fallback values.

The line between data and config has always been blurred, Most data in the browser is “data” until a user overrides it. Even extension files presence is linked with their configured state of being installed.
For log files I’m not sure anyone follows that. Besides firefox barely has any plain logs, I think having those in the profile is fine.
Currently it’s all in XDG_CONFIG_HOME

I’ve met some from asia and europe

In practice that should mean running mv ~/.mozilla ~/.config/mozilla.
Of course the old directory will still be detected and used if present, probably for a long time to come.

It’s nice but a little janky. For me for example I have to hit ctrl+C twice after closing a session where I started a kde program.

I still use X forwarding.
It works just fine using xWayland, and X forwarding has always been so janky there is no chance to notice any difference caused from using xWayland instead of native.

It will surely take many years and well established wayland native remote tunneling before anyone thinks of ditching xWayland.