I also have a preview edition.

I moved HA from my server to a HA green to separate reliability (my server is a test bed and uptime isnt great, and home automation warrants better uptime than I was giving it).

The voice services don’t work as well on the green directly, but I view it as part of the HA ecosystem and I want it running on the same hardware, but it seems very much like not a great option for that. And even on my own hardware, it still seems like it was a bit slower than I’d want and not always accurate. I definitely need a lot of tweaking (just like OP) to make it worth while.

And from what I can tell based on the callout at the end… This is a line from connector which is a compatibility layer that allows running Fabric mods on Neoforge.

Which means connector is going to be included in every stack trace, regardless of how related it is to the problem. It will be the one to raise the errors that couldn’t be caught and managed… But AI will see connector being the one probably flagging the errors and be more likely to tag it as a “suspected” mod. I wouldn’t be shocked to find out that AI has a tendency to shoot the messenger.

“I tried to keylog myself and the system doesn’t support keylogging.” is a frustrating situation. Because it’s neat from a security perspective and absolutely maddening from basically every other one.

Desktops can be installed on any ubuntu release. you could install xfce on kubuntu, or kde on base ubuntu. And what DE/WM you are using doesn’t matter in a quite a few scenarios, so it’s not worth bringing up in the same way that I don’t mention “Im using cachyOS niri wm”

I personally preferred xubuntu over kubuntu back in my *ubuntu days. etc etc. there are a lot of spins of ubuntu and when the only differentiator is DE (which you can also just install on base ubuntu), it’s easier to just talk about them all as a single overaching entity/family … ubuntu.

Talking about kubuntu inherently requires talking about KDE and ubuntu both, so people just skip the KDE discussion.

“Too many distro options” is already a complaint of a lot of non-linux users, so specifying every last spin can just drive people away.

RIP stack overflow

https://www.youtube.com/watch?v=gmRk9Undc6U

https://www.gamingonlinux.com/2026/01/steamos-3-7-20-adds-the-ntsync-driver-to-help-improve-some-game-performance/

Speaking of fsync vs esync vs ntsync.

As someone else has mentioned, proton or wine or games should not be able to cause “OS crashes”. If they do, its either an error happening somewhere else in the stack (a driver perhaps?) being triggered by the game, or you’re dealing with out of memory issues, and its struggling to clean itself up. I would suspect drivers (ie, make sure the system is updated) or memory, because those are the 2 aspects of a system that games tend to push the most.

But to your actual question: It’s a few things. The main thing that differentiates proton from wine, is that wine aims for maximizing global compatibility. Proton is focused on games, and is therefore able to make tradeoffs.

What if a new feature breaks MS Office because it changes how a function works, but doesn’t impact any games, and in fact, makes games fun 2-3x faster on average. Wine would not include this new feature. Proton would. They would put this into proton experimental.

But now they find out that while it makes MOST games faster, some games it causes to crash. Some games might use that same function MS Office does. so they don’t run properly. That might warrant a hotfix release. The point of numbered releases is to have points in time where most things are known working. If this feature had shipped into Proton 123, valve could pin the broken game to Proton 122, so now having different number versions makes sense to keep each individual game working optimally, while experimental and hotfix can be used for testing out new code (literally experimentally).

A great example of this was fsync. TBH, I have no idea what fsync is, but generally speaking, the games make a Windows function call that didn’t have a direct equivalent Linux function. Wine made a workaround to make sure this windows function did what it was expected to do, and there were several versions of the workaround. Some faster, some more reliable. And then eventually the function was actually added to the Linux kernel, so no work around was needed. For a while there, fsync vs esync was a big performance and compatibility concern. And now ntsync tends to cover this. Some games worked better with fsync and some with esync, and some with it all just turned off. But we need experimental proton versions to validate all of this.

QR codes on the net make no sense.

If I’m on the net… I’m already on the device that I want to visit a webpage from, just link to the webpage.

QR codes are for sharing something from one thing to another where there is no common interface other than a camera. The internet is already a common interface for sending me a URL, you don’t need a QR code.

For example:

If I were browsing on my phone, the only QR reader I have uses the camera. I can’t take a picture of my phone using my phone. I can’t use the QR code.

If I were on my PC, I dont have a QR reader at all except on my phone, and I’m on my PC, why would I want to switch to my phone.

Sharing a QR code on lemmy/piefed never makes any sense.

That’s really cool! It does sound like something I don’t particularly have a use for, but I see the appeal for others.

And giving it a test drive with flatpak is pretty simple, so no downsides for me there.

Keep up the great work.

This intrigued me so I grabbed it but I still dont quite know what I grabbed.

Is the whole point of this to be able to combine 2 controllers into 1 controller? Some of this feels niche, and like the people who need it will REALLY want it, but for most people, it does nothing beneficial, or am I missing something here that I should keep an eye on.

I switched to vaultwarden back when it was bitwarden_rs due to the crazy overdone bitwarden docker setup… and then started using some of the licensed features. I have a home organization that I use to share passwords with my family. So now I can’t switch back to bitwarden official (even lite) unless they provide me a way to handle that.

I’m not opposed to paying them, but I am opposed to subscriptions for access to something I’m hosting on my own server. So a subscription license isn’t happening. I don’t see a reason to leave vaultwarden at this point

At my last job, they were trying to put LLMs in charge of doing data stuff to avoid typos, and it was just easily provably making up data.

Automation is not LLMs/GenAI.

Imagine having a router with an AP built in. We don’t use that consumer tier stuff around here. 😎

Amazon Sidewalk means that PiHole or AdGuard won’t stop anything, because the point of amazon sidewalk is that if a device can’t find a usable internet connection (ie, you didnt set up wifi, or you dont have wifi), it can still function by connecting to nearby open networks, or perhaps even someone just walking by with a phone that it can piggyback off of.

So setting up a pihole prevents it from phoning home on your network, which will just prompt it to jump to another network potentially.

The question is just “how aggressively does this device want to phone home?” Some devices will actively seek out ways to phone home if blocked, some devices will try to phone home and give up if blocked. Some devices don’t try to phone home at all.

Not phoning home is ideal, a pihole will (probably) keep a device that stays on your network under control, and you should just not buy things that actively will work around your intentions to do whatever it wants without your permission.

But this person has already bought the thing, so “dont buy it” isn’t an option. But “don’t use it” might be. Depending on which category it’s in. (which I do not know, just trying to illustrate the bigger problem with Amazon Sidewalk)

https://docs.bazzite.gg/Dev/

If IDEs from Flathub and CLI tools from Homebrew serve your needs, no further action is required. If deeper system integration is needed for VSCode (ie. devcontainers), Docker (ie. Podman is not sufficient), etc - then see below specialized images.

There is a whole Bazzite for Devs page that mentions Bazzite-DX for development to handle some things like devcontainers: https://github.com/ublue-os/bazzite-dx

Their main website also says:

Running a game, a development environment, a container for your Jellyfin server, or a utility only available on the Arch User Repository? You can rest assured it works here. Bazzite is developed on Bazzite.

At the end of the day, its an immutable fedora distro. Which may serve your needs. or may not. And bazzite’s primary focus is on gaming. It will most likely work (given a few criteria), but it may not given that is not their primary focus.

pre-ryzen AMD CPUs were always a bit on the budget side even when they were new. They were a bit more power efficient and cheaper, but never were amazing performance. So yeah, a 15 year old CPU is often rough, but a 15 year old low-mid end CPU is going to be even worse off.

Ive not had issues with Counterstrike 2 on linux, aside from wanting to play FaceIt or ranked matches that have stricter third party anticheat. And that’s just the anticheat being the problem.

I would venture a guess that linux is not “the problem” here and it’s more likely just aging hardware meeting increasing game demands.

The only time in recent years I’ve had specifically a problem with linux gaming performance (not related to anticheat), was playing VR on a pre-GCN AMD GPU which didn’t support async reprojection properly which caused quite a bit of stuttering.

https://komodostation.com/

Komodo Steam Deck Store rebranded to Komodo Station today. Which to me indicates that Valve’s distributor in Asia is already ready to begin selling more types of hardware. So that’s at least a positive sign.

https://github.com/FEX-Emu/FEX

My understanding is that FEX + Proton is the plan and just double up on the translations as appropriate.

Valve has supposedly just been heavily investing in fex.

Steam deck highlights that Windows -> SteamOS translation is good enough.

I’ve use my Index on Bazzite successfully with no issues, so I’m confident in SteamOS VR capabilities.

ARM-based is the only wildcard, but if fex works, then that’s not an issue either.

Then just onboard compute performance is the only factor. But like you said, even if this winds up only being a “stream everything, its a wireless index,” then I’m already excited.

There is an optional accessory for knuckle straps for these, mentioned in some of the 3rd party reviews.

Which somewhat implies the ability for custom knuckle straps and other accessories as well since they are detachable with mount points.

edit: screenshots from the LTT coverage: