It was 33rd in 2010:

In November 2010, the Air Force Research Laboratory created a powerful supercomputer, nicknamed the “Condor Cluster”, by connecting together 1,760 consoles with 168 GPUs and 84 coordinating servers in a parallel array capable of 500 trillion floating-point operations per second (500 TFLOPS). As built, the Condor Cluster was the 33rd largest supercomputer in the world and was used to analyze high definition satellite imagery at a cost of only one tenth that of a traditional supercomputer.

https://en.wikipedia.org/wiki/PlayStation_3_cluster

https://phys.org/news/2010-12-air-playstation-3s-supercomputer.html

Good video… but a nearly 2 minute ad in a 5 minute video? ~40% ad? Wow.

Not counting yt ads as I block them. I’m considering starting to use sponsorblock, this is too much.

Nix just calls the *.nix files, it’s still go under the hood. PKGBUILD is similar to the flake.nix and package.nix files to me, but I have no experience with nix.

My general view is similar, yaml is better if it should be written by humans, json is better if it should be written and read only by a machine. but hyprspace uses json for configuration, so I don’t really understand cellardoor’s comment

what:
  is:
  your:
    - problem
    - with:
      YAML
# At least you can have comments unlike in json. Who need comments in a config file anyway.

Or port forwarding. You have to open a udp port for wireguard

AUR packages ending with"-git" or “-svn” always pull the latest commit from source. The version number means that was the last time the packager had to change something on the PKGBUILD script, not the actual version which would be installed.

Where should I look? Where were these talks? I’m interested.

Edit: I found the whitepaper about hole punching: https://research.protocol.ai/publications/decentralized-hole-punching/

It says it connects to a “Hole Punch Coordination (DCUtR - Direct Connection Upgrade through Relay)”. So for NAT traversal to work, you need a third party, this relay. As I expected. I guess you can self host this, but than you could just host a wireguard server. I guess if you are on a locked down network where you cannot connect to any relay (e.g. how the Chinese Great Firewall works technically they could block it) you can’t initiate a connection behind a NAT.

Nonetheless it seems interesting, but no magic here. Maybe the big difference that the relay servers are distributed, so no central authority to block easily.

Interesting, it’s on AUR, I will try it.

So it doesn’t need any port forwarding, and works on CGNAT? How the “NAT hole punching” works? Both clients connect to something on IPFS?

Afaik, for DHT with torrent, clients need to know at least one tracker, what is the “tracker” here? Something on IPFS? Who am I sending my IP addresses?

How much overhead does this add to speed? I love with Wireguard, that it’s barely noticeable, really close to p2p speeds, OpenVPN was awful in this regard.

Posting here as there doesn’t seem to be an active Arch Linux community.

[email protected] seems active to m

Your text says qemu-desktop is only “optional” for qemu-base. You can safely remove qemu-desktop, pacman won’t nag about optional dependencies.

For checking dependencies, I like to use pactree, it draws nice graphs in the terminal: https://wiki.archlinux.org/title/Pacman#Pactree

It’s incorrect. I have 2 AMD cards, I can detach it from linux before booting the guest. After I shut down the guest I have to log out in Gnome to make the card usable again, but no reboot required. It depends on how you set it up. I have a single 34" monitor with 2 inputs, connected to both cards.

I recommend to read about this topic, it would be quicker than waiting for people to answer, your questions were answered multiple times. I recommend the vfio wiki on the r*ddit a lot of good links are collected there: https://old.reddit.com/r/VFIO/wiki/index

Thank you for the clarification, I understand what is your issue now

It would be hard to get the context, and you could totally misunderstood the thread, e.g. from your hidden comment everyone get that replies are sarcastic, but you don’t.

Maybe an “temporarily unhide blocked user’s comment and replies” button to show the whole thread.

There are instances where downvotes are disabled, if you don’t like them you can just use an instance like that.

Downvotes are public and not anonymous, but they are hidden in Lemmy ui. Afaik you can see who downvotes your posts from Piefed or Mbin. See this thread: https://lemmy.world/post/18805474 or this: https://piefed.social/post/205362

Keyboard navigation. I know about https://github.com/vmavromatis/Lemmy-keyboard-navigation but it’s annoying that I have to use an addin/userscript for such a basic feature.

Option for default comment sorting. you can change the default sort only for posts, but not for comments, comments always sorted by Hot, and you have to manually change it each time you open comments.

In Voyager you can set this up, but it would be useful in the webui as well.

Workaround: use multiline code blocks with defined language, e.g:

systemctl --user cat emacs
pactl load-module module-switch-on-connect

Drawback: you cant use it in a list or inline

If you don’t need fancy gui and authentication, registry is easy to set up and works really well: https://hub.docker.com/_/registry

I have several addresses at cock.li. Uptime is not the best, around 98%, but free. According to their policy they don’t collect any personal data, but they comply with legal requests. https://cock.li/help

You can select from a lot of domains, some of them ar normal like firemail.cc or airmail.cc, some of them are funny like aaathats3as.com, some of them are edgy like cocaine.ninja or national.shitposting.agency, some of them are racist like nuke.africa or hitler.rocks

It’s not the browser but the websites.

Two way video streaming will be always resource intensive.

But for the other websites the choise is yours, you don’t have to use amazon or goog, there are alternatives.

Also if you doesn’t have enough ram close the tabs. My 10 years old low end laptop with 8GB ram, I know I can’t have more than 6-8 tabs open at the same time.

but I don’t see how what I am proposing would make things more difficult?

Now when a user reports a troll, the report goes to the moderators of the community. But in special cases the admins of the user instances should deal with banning. So the admins of the community instances have to deal with reports, but the solution is at the hand of the user instance admins. It’s the same as dealing with users from other instances, but an edge case.

My recommendations would be something like this: (I’m just a random user, so it’s just my point of view)

• Shut down the fully inactive instances. Noone will even even notice it
• Merge the semi active communities to a handful of instances, like sports and technology… . I’ve seen active communities move instances, it would be possible, take a look how [email protected] migrated to [email protected]. Give enough time for subscribers to notice and subscribe to the new one.
• Allow registration of moderators on these instances, so they can work around the current limitations of moderation tools. Maybe an invite only solution or something like this.
• You could find help more easily if you look for admins for 3-4 instances instead of for 18 instances.

This would be useful for you and other admins, because you would have to admin much less number of instances. They would be still considered small instances, compared to big one, so you still not at the “too big to fail” level. For users it would help community discovery, there are overlap between followers of similar topics, e.g. I have friends who follow both European football and NBA at the same time, I read both selfhosting related topics and about general tech support, etc…