My understanding is, that the 1.0 spec for the extension was basically finalized in 2021 and CPUs using it are already available. Now it’s just fully ratified. Also, while it might seem like RISC-V is “behind” compared to AVX-512 for x86_64 or SVE for ARM, this fundamentally differs from these SIMD Instructions. They talk more about it in this article SIMD Instructions Considered Harmful. So, this is not merely RISC-V playing catch-up, but also trying a “new” (the idea is actually old and how things used to be done) ways to make a more sustainable ISA.

I’m not sure that Proton can fix your problem. However, I feel like this project would love your help with capturing the USB traffic to get it supported and hopefully upstreamed in the kernel some day :)

Great, that sounds amazing. Let’s hope it’s also used even if it means less excises for tracking.

Could the new CHIPS functionality help websites like Microsoft Teams working without you having to enable third-party cookies for their websites? If I understood it correctly this might be exactly the kinda use case but I couldn’t find anything specific online.

I looked at some info for reporting this to the kernel developers but the process is too complicated at the time. I’m currently a bit short on time but I did report it to libinput, maybe they can give pointers where exactly to report this.

On Mastodon they said there will be a blog post outlining the changes. That will probably be out tomorrow because that’s when alpha 2 officially launches.

What about Tauri? I don’t know what exactly your app is but since you mentioned Electron as an option I guess Tauri could run it. Offers more choice for frontend frameworks hence less „language lock-in“ than Qt.

!helldivers2@lemmy.ca just dropping this here to help growing smaller communities :)

Just know that sites like this are useless if you don’t understand the results. There are anti-fingerprinting techniques that add random noise to your fingerprint. This might result in these kind of tests claiming you have a completely unique fingerprint, even though the anti-fingerprinting mechanisms randomise the fingerprint for every site, browser session, etc. (depending on the config). This would mean that you are relatively „safe“ from fingerprinting because you never have the same print twice but tests think you are very vulnerable because it’s still a random “unique“ fingerprint.

Yea but many of them were involved. The Audi CEO at the time was on the board making the decision and the first to be convicted.

Don’t worry, I don’t think you are. I just think there’s a reason they admitted so easily. Probably just another calculated fallout to save all their other brands from their own mini backlash which would ultimately cause more damage.

But yes, the whole industry is a dumpster fire when it comes to regulations and also lobbying.

I mean they also own like half the industry. So, I don’t feel particularly bad for them to be honest.

I mean they did also inject affiliate links without the users noticing which is really shady behaviour from a browser because it has one job, open the link I click and nothing else. But that’s just IMHO if that is acceptable for you personally then there is no issue with that.

Yes, BUT the first CAMM was proprietary to DELL. This is what came from them giving the standard to JEDEC. So not all hope is lost!

I never actually tried myself, but it seems like the documentation certainly could be improved. I saw that they provide a Docker compose, so perhaps that could be of help if you didn’t use that the last time around. They are currently in the process of cleaning up the projects to make things more maintainable and easier to get an overview, so let’s hope things might improve a bit. I think for me personally, this certainly seems like the most promising Discord replacement because it feels like a set and get solution for non-techy people trying to switch instead of relearning everything like with Matrix.

Revolt is self-hostable. It isn’t E2EE but if you’re controlling the users anyways transport encryption should be enough since you have control over the data anyway.

I don‘t think you understand my point. Let me be a bit more high level. It’s not about the three major outages WhatsApp had this year for like 30 mins. or whatever.

A perfectly set up Matrix server with more than enough resources allocated has issues decrypting messages when there’s a few hundred people and that’s without federation. This is still happening to today, fully updated server and clients.

As I said, I know they are working with a lot less resources than Meta. But at the moment the implementation doesn’t even do the most basic thing, deliver messages reliably. I know their new encryption library is supposed to do a better job but it’s just the cold hard truth that it’s not up there with the big messengers yet. Denying that doesn’t do the project any good.

Matrix does definitely not have the same reliability as WhatsApp or Signal. I’ve used it for around 3 years now with a group of tech savvy friends.

It’s still a regular occurrence that we get cannot decrypt errors, sometimes the app doesn’t show new messages in the chat but they are visible in the preview, also the app can be soooo slow.

Also, I know it’s not user error. If you check the Matrix development and follow their blog posts they already acknowledged the issues and are working on fixes. But for now it’s just wishful thinking when one calls them reliable alternatives for mainstream use. I’m not hating and will keep using the project because I truly think they are doing amazing work.

I don’t know how exactly it’s possible. You’d have to figure that out yourself. However, if you use a tool like chezmoi you can most probably sync the correct folders inside your profile.

May I ask what part you’re paranoid about since Firefox Sync is fully encrypted anyway?

Pop!_OS, Fedora, and I think Ubuntu as well have already done this change long ago in this sequence. So no need to switch to Arch. Also, you can edit this manually, it’s just about changing the default.