The windows kernel isn’t all that great, particularly in the realm of memory security or scheduling.

You know, to each their own. Question is really whether windows maintaining a closed source kernel even makes sense from a maintenance burden perspective when it really doesn’t give them much money in return. (Most of their money in 2025 comes from cloud services, not operating systems)

Yes, but also their cars are ugly and they were riding high on musk’s cult of personality in the first place.

Rivians are better anyway.

deleted by creator

This is nearly the plot to Ghost in the Shell 2 Innocence. Nearly the keyword, they might get there eventually.

The best path forward is that developers make their linux drivers before they release their hardware to the market. You know, like what they do for windows.

There’s no silver bullet here. You have to wait for someone to reverse engineer the drivers if the developers of the hardware don’t care enough to supply even basic linux driver support. Either that or linux becomes so popular that it becomes senseless to ignore it (let’s be real though, MacOS is popular enough for this to be true and yet there’s still new hardware made that ignores that platform too.)

I mean, my issue is that most buttons on my huion are still non bindable, and some graphical interfaces cannot be interacted with in mouse mode and only register as touch. Lastly, occasionally programs completely ignore pen sensitivity, such as blender.

This experience was when I was last on gnome. I’ve been on budgie for a while as a result of needing a tablet for my hobbies.

The lack of proper tablet support in wayland prevents me from being excited for this. I wish there was more news on that front.

So you’re telling me that I should just add the word trans to my code a shit ton to opt my code out of AI training?

Centralization is a weakness

While I agree with the basic premise of this, I think this is all-the-more a good reason for design solutions around this problem to be discussed so that more competitors can exist. If the fediverse is to expand, there needs to be easy non-technical ways for users to start up a multitude of instances. More of these types of services actually reduces centralization, in that sense.

Do you work at Boeing?

Because that would certainly explain some things.

Sure.

Valve’s operating system is read only and, when steam decides to update, any root level file changes will be lost between updates. This is partly good because the system will always be recoverable and update reliably, but comes with the downside that users have to take extra steps to install some base level packages (things like tailscale, syncthing etc. There’s always work arounds, but it’s not a guarantee that these work arounds will continue to work on new updates.)

OSTree is also a read only file system utility that allows packages to be layered, so users can install their own packages. When the operating system updates, these packages are rebased and preserved on the next update so user level changes can be preserved.

There’s more to this than that, but basically steam os is dependent on valve updating packages and generally leave all extensions either hand off or need to work around root filesystem. Ostree/silverblue/bazzite allow user modification by having a slightly more sophisticated updating process.

I would say it’s great but would strongly recommend using Bazzite over steam OS even on the deck.

My biggest gripe with the steam deck is that it’s not well equipped to handle user packages in the same way OSTree is. Bazzite solves this while still mostly adhering to the design principles of steam os, so I feel it’s actually better than the stock operating system.

The big issue they’ve had, imo, is that there’s too many bots and too many IRC bridges from stubborn oldies who refused to migrate away from IRC.

There was a time when freenode went down that many channels were experiencing an exodus, which would have been the perfect time for Matrix adoption, but people were far too stubborn with IRC despite its many flaws (privacy, lack of history, poor feature set, etc.)

I assume having your own chat room with your friends would mean self hosting?

Why would you assume this? There’s plenty of servers available and conversations can still be encrypted.

Voice chat is extremely lacking, though.

It’s very strange that you’ve made a post about bugs but chose not to list any of the bugs.

Like, how can we make a recommendation if we don’t know what types of issues you’re running into? What type of hardware you have? What expectations you have?

It just kind of screams of disgruntled user syndrome. These are community lead projects so, yes, they’ll have bugs. But if people never say what they are or what issues they had with what they used, the best the rest of us can do is just guess!

We better should’ve stayed at 640kB.

When every byte mattered. Hell, when every bit mattered even!

It’s literally crazy to say something like this on Lemmy of all places.

Don’t like moderators? Fine, try to host your own instance and your own communities. You’ll find quickly that it turns to shit because it’s actually pretty hard to do well.

Calling the Scarlett Johansson lawsuit “Manufactured Drama” is certainly a take. A bad one, that is.

Just like the lifting of a famous actress voice, one has to wonder how much LLMs are siphoning the intellectual property of the little-people of the open source world and willfully tossing the license and attribution clauses down the toilet. If they were willing to do it to a multi-million dollar actress, what makes people think that the intellectual property theft doesn’t go much further?

Anyway, I think for this reason it’s actually really important to note that Junior Devs are much less likely to cause this type of issue for large companies. The question is whether the lawsuits from improper licensing cost more to settle than it costs to hire Junior devs, which brings us roughly to where the international outsourcing phenomenon brought us. At least, IMO.

Would love for you to describe exactly how it’s more complicated.

“More” is relative, ofc, so YMMV on whether you agree with me or not on this.

But the problem with pass key is that it has all of the downsides of 2FA still – you need to use a mobile device such as a cell phone, that cell phone must be connected to the internet and you often can’t register a single account to multiple devices (as in, there’s only ever 1 device that has passkey authorization.)

This isn’t an issue with ssh keys, which is a superior design despite it not being native to the web browsing experience. SSH keys can be added or removed to an account for any number of devices as long as you have some kind of login access. You can generally use SSH keys on any device regardless of network connection. There’s no security flaws to SSH keys because the public key is all that is held by 3rd parties, and it’s up to the user in question to ensure they keep good control over their keys.

Keys can be assigned to a password and don’t require you to use biometrics as the only authentication system.

I feel like there’s probably more here, but all of this adds up to a more complicated experience IMO. But again, it’s all relative. If you only ever use password + 2fa, I will give them that it’s simpler than this (even though, from the backend side of things, it’s MUCH more complicated from what I hear.)

The problem with PassKey is simply that they made it way more complicated.

Anyone who has worked with SSH keys knows how this should work, but instead companies like Google wanted to ensure they had control of the process so they proceeded to make it 50x more complicated and require a network connection. I mean, ok, but I’m not going to do that lmao.