I wouldn’t recommend Docker for a production environment either, but there are plenty of container-based solutions that use OCI compatible images just fine and they are very widely used in production. Having said that, plenty of people run docker images in a homelab setting and they work fine. I don’t like running rootful containers under a system daemon, but calling it a giant mess doesn’t seem fair in my experience.

If I may ask: how practical is monitoring / administering rootless quadlets? I’m running rootless podman containers via systemd for home use, but splitting the single rootless user into multiple has proven to be quite the pain.

That’s not entirely accurate. Google’s influence on the web has grown even beyond the web browser engine majority share (which is bad enough in itself). They offer one of the most popular web frameworks and run several of the most popular websites. There is almost no way to compete when the market leader is simultaneously the developer and the major user of new features. Of course everyone else is going to switch to using your browser engine. What else are they gonna do? There are even websites now that just check the user agent string and refuse service if you don’t use a chromium based browser. Shit’s fucked.

Maybe a fixed line-height?

The whole premise of this discussion was about technological progress and growth going by your initial comment. That means refining existing models and training new ones, which is going to cost a lot of energy. The way this industry is going, even privacy conscious usage of open source models will contribute to the insane energy usage by creating demand and popularizing the technology.

Do we really need to grow our energy consumption as a society by such a disproportionate amount?

Is there a source for these haughty, cackling archeologists making fun of hairdressers or is that just to manufacture some kind of underdog victory scenario?

With bluray rips, I don’t really see any way to avoid that unfortunately, unless someone else has already added the hashes for your release. Most people use it to scan their encoded releases, which will (in most cases) have already been added to AniDB by the release group. I’m a bit surprised though, that none of your rips are recognized. Have you checked the AniDB pages for your series to see if anyone uploaded hashes for bluray rips?

Grouping seasons into a series folder doesn’t work well in some cases, because that’s not the way they are released in Japan. A new season is (most of the time) effectively an entire new show entry. Show seasons are mostly a north american thing. No matter which software you use, there’s always going to be some minor issues if you group seasons into one entry.

Shoko compares a files ED2K hash against the AniDB database. The filename doesn’t matter for automatic detection. Have a look at the log to see if there are any issues. It’s entirely possible that AniDB just doesn’t have the hashes for the raw BluRay rip. In that case you can either manually link them in Shoko, connecting the AniDB episode id to the file hash, or create new file entries on AniDB with your specific hashes.

Shoko also has rate limits. The problem is that AniDB does rate limiting in an extremely stupid way for a UDP API and doesn’t even have the decency to define clear time limits.

Pretty sure that the registry path for official images is “library” (at least it used to be). So it should be “docker.io/library/debian”, though I can’t double check at the moment.

It will probably just work, even if not officially. If any weird Windows issues crop up, Microsoft may or may not fix them. I think AMD even provided workarounds and special drivers for Windows 7, just without any official support. They may not do that this time around though, since a lot of things have changed.

Try to imagine that argument but coming from some government you really dislike. I can think of a lot of different media that might inspire violence and instability, but which would be really important for people to see or at least know about. Frankly, anyone who doesn’t see that as a potential problem is being shortsighted and really needs some historical perspective, in my opinion.

You mean hiding their public IP? I guess that’s a feature.

That’s what a firewall and a DNS service is for respectively, imho. As long as you get an IPv6 prefix from your ISP, you can expose as many devices or services to the public as you want, by just allowing incoming traffic to a listening port. That was sort of the whole point of having a large enough address space when moving away from v4. Maybe it’s just me but reading stuff about “private AI” on a website where the relation to the product is not immediately obvious, makes me question their legitimacy.

The more I look at their site, the more it reads like a sales pitch for IPv6, which sounds kind of expensive at $6-10 a month.

What problem does this solve? Do ISPs not provide IPv6 prefixes anymore?

That script is a wrapper around a single call to qrencode. I’ve been making qr codes from wireguard config files in the terminal at least since PiVPN existed. There are plenty of guides on how to do this as well.

I’m aware that any of the past attempts to classify body types are extremely pseudo-scientific and I’ve explained as much in a different comment in this thread. The point is that “body type” isn’t just necessarily just a generic way to refer to someone’s body shape. Plenty of people still believe in that made up nonsense.

What are you trying to say?