Nvidia and AMD both have differing APIs for video decoding. The idea in using Vulkan is that not every GPU supports the vendor specific decoding method, but they do all support Vulkan (modern GPUs, anyways). So this way you still get hardware accelerated decoding through a more widely supported means.

They didn’t give up, they pushed it back 6 months to wait out the shitstorm.

I hate playing this on a N64. I own it, I’ve tried, it’s a miserable experience. I largely chalk this up to it just being the N64 controller though.

With a proper controller, it controls very nicely imo. Inputs are precise and it takes some work to get good but it’s overall still one of my favorite Mario games, next to Odyssey.

Yes, you got it.

It’s possible that however your ISP provided router is designed, it’s got some hidden port forward configuration. If that router has an option typically referred to as “bridge mode”, you could bypass its routing features altogether and use your own router instead.

ISPs often have clauses about using their residential internet for hosting servers or exposed services, and it’s possible your has taken a different approach to mitigating traffic from those sources.

If you can, I’d recommend using your own router rather than what the ISP provides.

If I understand correctly, it sounds like you moved from an ISP that uses CGNAT to one that does not. Does your ISP provide a modem? If so, are you relying on the software features of that modem, or do you have a router inbetween?

Interviews do typically count, it just has to be citable. Videos are sufficient in that regard as well, not just articles or books. It would be different if Torvalds had edited his own wikipedia page, but an editor who updates the page and cites this video would not be in the wrong.

I mean, I feel like he outright confirms it in the video. It’s his distro of choice since it allows him to easily use his own compiled kernels in testing. Anything else is an inconvenience to his work.

How’s the overall health of the drives? You might want to get a quick SMART report.

Otherwise, this sounds like pretty normal drive activity. It could be the result of anything from indexing tools to casual background processes doing a read.

If it’s periodic in a way that’s consistent, then it’s almost definitely something in software. What docker services are you running? Do you have any auditing tools or security processes that might be actively logging activity?

It’s pretty unlikely you’ve contracted malware unless you’ve gone out of your way to expose the server to outside sources, so I think you can alay those concerns.

A couple points:

• Your website does not properly convey the technical context of Safebox. Docker is a complex platform, and asking someone to install it point blank on any OS, while also championing ease-of-access feels at odds here.
• There is a severe lack of documentation about the tool. Discord is not an appropriate means to find these documents if they exist. It is rarely okay as a support channel.
• I saw your post from a few days ago, but it was framed as a question about about gatekeeping specifically. The post also advertised Safebox. Given that the post no longer exists (but the comments sure do), I’m inclined to think you didn’t get quite the answer you were looking for.

I dove into self-hosting several years ago and ultimately I think I found the experience quite welcoming. I also don’t know that Safebox has a lot to offer over well-established alternatives these days like Unraid or TrueNAS, which have large user-bases and a depth of support articles to help admins better understand what they’re doing and how to do it. It’s true that not everyone would want to do this as a hobby. No one wants their services to break, or their data to be lost, and more tools that make it easier to prevent these scenarios are helpful. With that in mind, I am not left with a clear understanding of how Safebox is meant to provide safeguards here.

I used the word “admin” in the previous paragraph for good reason. Self-hosting makes you the administrator, and it means that you, the administrator, have the power to make mistakes. My recommendation is not to talk down to your users. Someone interested in self-hosting should be aware of the potential security implications of what they’re taking on, alongside the risk to their data and that breaking changes are something they can and will make along the way. If you really want to make self-hosting accessible, then the documentation for your tool needs to be accessible too.

Safebox runs on Linux, macOS, and Windows, supports both x86 and ARM64 (including Raspberry Pi, Banana Pi, and others), and handles domain/subdomain setup, Let’s Encrypt certificates, DNS configuration, reverse proxy (nginx), and also offers WireGuard-based remote access.

A user should be able to learn why these elements are important and how they work together. Talk about the limitations of running it on a raspberry pi vs a workstations or server. What’s a reverse proxy? Is WireGuard good? This doesn’t mean the average person needs to know how to configure detailed permissions or application configs, and if the goal is to provide a repository of pre-hardened Docker configs for use then that’s cool too, but there should never be a barrier to the information itself. Especially as it is relevant to the tool you’ve built.

I think that fundamentally, you’ve built a good tool that simplifies things someone who is already familiar with its components, and where it needs to improve is by expanding to help new users familiarize themselves. Education is as big a part of accessibility as the ease-of-setup.

Something KDE has done seems to have resolved the issues I used to have with DPI related scaling problems in Wayland. Once Plasma 6 hit, it’s been nothing but rapid improvements with Wayland as a focus and man does it feel nice.

That said, there’s virtually no downside to still using X unless you have explicit display features you need from Wayland like HDR or the per-display scaling. Xfce is stupid lightweight and still my default for anything where battery life is a benefit.

Most domain registrars make whois info private by default these days. It’s typically just a toggle. Same with DNSSEC

“Community Add-ons” leads me to think this is probably Kodi. You can generally do IPTV streaming through it, or torrent streaming even.

I don’t feel like this is a terribly recent attitude. It’s definitely one I’ve encountered repeatedly over a decade or more of dipping my toes in the pool. It’s not incorrect in a lot of circumstances, but it’s very difficult to find support when no one wants to help you improve. There’s always been a significant degree of ego in Linux user communities.

Flatpaks are basically containers, allowing applications to maintain their own dependencies separate from your system. It’s similar to a Windows program shipping with its own precompiled DLLs, helping prevent dependenct conflicts when you go to update something you installed with pacman or yay.

Arc support was added after release to Linux Kernel 6.2 and it’s steadily improved since. Older Linux distros, or “LTS” oriented distros that favour stability may still not have support for them. I know Unraid was very slow to pick up on it and I had to settle for passing the pcie device through to a VM to get it working. Intel is keen to made these viable though, and I love having the AV1 encoder from my A380.

That one sounds squarely on Nvidia. Any driver that uses undocumented workarounds to gain kernel level access or utilizes an access loophool for system hooks is a bad driver. I’d assume Debian, or likely more accurately the Linux kernel itself was updated following some matter of CVE that Nvidia was quietly abusing.

Frustrating, but a good example of why those kinds of proprietary drivers are such a nightmare. You really just don’t know what techniques they’re using.

Yeah, because at least a decent 3rd party might hand you documentation and have the sense to build something consistent or maintainable. AI has a limited context scope and frequently suffers a type of short term memory loss that results in repeated work or variations in work that confuse the end result.

Biggest difference is that wormhole will pass traffic between devices on different networks as long as both are routable. So it’s not limited to a local network connection.

Better than that, the lack of reliance on spinning disks means that asset duplication and data read order is less of a requirement to reduce load times. It can still be argued that there’s just too many polygons, since simply scaling things back would be plenty effective in reducing storage usage and load times.

I use Wayland exclusively, but unfortunately I don’t think I have an answer for you since I’m not entirely familiar with this idea. Is your concern just for the configuration of a universal set of hotkeys configured within the compositor rather than a desktop environment?

I wasn’t aware that x11 facilitated this. I’d have figured keyboard mappings are abstracted from the compositor and left to the DE to handle, aside from core binds that allow dropping back to tty