Brickfrog

A lot of the software targets Ubuntu compatibility so that would be the easy distro choice. But, you should be able to get this stuff running on any/most distros as long as you’re comfortable building/compiling the software as needed. Or using containers like Docker, etc. I doubt you’d have any issue doing this with Mint if that’s your preference.

I’m just tinkering with this stuff so other people here will have way more experience. But as a point of reference I have no trouble running llama.cpp / sd.cpp on current Debian Trixie (with/without an updated Linux Kernel). This is with an AMD R9700 (32 GB VRAM) in an older Intel based system. Everyone says that running with Vulkan is faster but I’ve found it’s good to test your own setup - My overall inference speed is actually faster using ROCm. ROCm gives me less tokens per second but for whatever reason the RAM-VRAM operations are much faster when using ROCm which leads to the overall time from prompt to result being shorter. There’s likely more tweaking I could try to get more speed but I’m not heavily using AI at the moment so am happy with how it works now.

Before that had the same setup with a old Nvidia gaming GPU (6 GB VRAM), it sort of worked but was slow as hell and I could only get it to run with Vulkan. That GPU supported an older CUDA version (7.5 I think) and I never figured out how to get that working with llama.cpp + Nvidia software + Debian.

Haha yeah I came to this post thinking it was about KiTTY.

You are describing a hard ratio system - a private tracker with no freeleech, no bonus points, just a strict ratio system. Most private trackers don’t operate that way.

Your math falls apart once you consider people downloading freeleech torrents without having the torrents count against their ratio, or people generating bonus points just by seeding torrents and then using those bonus points to improve their ratio (bonus points converted to upload data for example). There are even private trackers that are ratioless and basically exist on minimum seed time, seeding bonus points, etc.

Long term seeders at most private trackers end up with so many seed bonus points, or just accumulated upload data, that the private tracker becomes essentially ratioless even if it technically still tracks ratio.

I don’t care if you like private trackers or not, you do you, just mentioning your logic doesn’t make a ton of sense.

From https://emule-qt.org/

eMule Qt is a free, open-source file sharing client for the ED2K and Kademlia networks, with a full Usenet engine built in

Kind of an odd client. They should have given it a different name like maybe uMule since it’s a weird hybrid usenet client + eMule client :P

EDIT - Yeah it’s all AI code, with probably a AI copied website with AI posts. At least they are upfront about that https://emule-qt.org/2026/03/05/hello-emule-2026/

All the Gazelle and UNIT3D based trackers I looked at are down for maintenance, not surprising since PHP is used there.

You may want to update the post with some setup details, but offhand:

• Did you configure it in Mixed Mode (clicked checkbox) to torrent in both I2P and clearnet? Or you left that unchecked to keep it in I2P only mode? There are less peers in the I2P network so you will encounter more public torrents without I2P seeds, that is normal. Sometimes people prefer it in Mixed Mode for that reason but it does mean you’re still downloading/uploading without I2P involved.
• qBittorrent does not support DHT in the I2P network. For any torrents you add in qBittorrent you must also add working I2P trackers to them so they can at least attempt to find peers within I2P, otherwise nothing will upload/download.

To properly test and maybe narrow down issues you could try doing 3 different torrent tests:

1. Download a working torrent that seems to be seeded in I2P, with working I2P trackers on it, and see if it finds peers eventually via I2P
2. Download a working public torrent that seems to be seeded on the clearnet non-I2P, with working clearnet public trackers on it, and see if it finds peers eventually via clearnet
3. If you have DHT enabled: Download a working public torrent, that seems to be seeded on the clearnet non-I2P, with 0 public trackers on it, and see if it finds peers eventually via clearnet

PS - You can review https://github.com/ngosang/trackerslist for a good reference on available public and I2P trackers. There are also public I2P tracker lists floating around here on Lemmy in older posts if you search around

I’m definitely in the thinking with others that keep admin/privileged and normal users separate. I kind of look at it as a way to minimize fallout from any unexpected security compromises. Will always do my best to prevent/avoid that, but if it happens it better happen on a non-privileged account.

But the downside is it can be quite inconvenient, especially if I need to administer or debug something quickly.

A bit, but I don’t find that to be too cumbersome. The initial setup period yeah, you’re constantly jumping into admin mode, but once things are configured and settled down you hopefully don’t need to jump into admin mode often.

I’m also not sure if my homelab expands, should I share the admin user credentials with other human admins or not.

That’s a fair question and it depends on the service(s) really. I don’t know what the general consensus is on that but I’d say that other admins should have their own separate admin account and user account (e.g. call it “joe” and “joe_admin” or whatever). Sharing credentials, especially admin credentials, seems like its own security risk. But I also get that sometimes it’s simply not possible to maintain multiple admin accounts and you just have to do the best you can.