I believe, with Authorized Fetch (what Mastodon calls secure mode) blocking intermediaries won’t be needed, as instances will have to cryptographically “authorize” themselves to receive/send data, and you can just say “no” to any requests coming from threads.net, acting basically as a “defederation enforcement mode”.

I could be wrong though, haven’t caught up on the exact details.

Well you’re in luck because Lemmy 0.18 rips out all WebSocket code.

Not OP but I can answer with my own stats:

In just a week, With BTRFS compression (compress-force=zstd:3) & deduplication (via bees), media is at about 1GB (and I am subscribed to media-heavy communities like 196) and the postgres DB is at about 550MB (which is also currently shared with Matrix Dendrite)

At “idle” (as you can be while being connected to ActivityPub & Matrix), the immediate CPU and RAM usage breakdown per container is:

NAME        CPU %       MEM USAGE / LIMIT  MEM %       NET IO             BLOCK IO           PIDS        CPU TIME         AVG CPU %
pict-rs     0.20%       18.92MB / 4.005GB  0.47%       3.319GB / 1.105GB  17.58GB / 3.239GB  13          1h16m57.232828s  0.59%
crowdsec    1.39%       44.23MB / 4.005GB  1.10%       106.4MB / 23.46MB  25.53GB / 486.7MB  11          45m28.744419s    1.95%
caddy       0.63%       73.06MB / 4.005GB  1.82%       1.675GB / 1.977GB  3.322GB / 720MB    10          21m9.94572s      0.90%
dendrite    1.58%       197.7MB / 4.005GB  4.94%       912.8MB / 2.33GB   8.718GB / 4.761GB  12          53m26.302022s    1.43%
postgres    5.33%       82.51MB / 4.005GB  2.06%       56.22GB / 7.961GB  20.92GB / 295.7GB  23          8h20m28.078567s  2.86%
lemmy-ui    0.00%       48.71MB / 4.005GB  1.22%       3.491GB / 5.961GB  3.603GB / 5.267GB  12          31m35.884936s    0.24%
lemmy-be    2.82%       29.01MB / 4.005GB  0.72%       16.45GB / 57.85GB  7.966GB / 6.439GB  6           3h6m34.633508s   1.42%

Net IO you shouldn’t really care about as that includes inter-container networking. I’m trying to find how much outgoing data have been transferred but because the month just ended I have no idea how accurate the numbers are.

Perhaps they’re not trying to “be successful”?

Depending on how well you know your way around, my recommendation is to not use the Ansible setup but instead treat it as documentation while doing things your way. It has quite a bit of strange stuff going on (postfix? two nginx installs with only one being in a container?) and seems to be missing important things such as SSH hardening. It also assumes it’ll be the only thing running in your server just in general (horrible yet common practice, unfortunately) so if you have anything set up it may or may not clobber over it to do things it’s own way, and end up breaking something.

Also, can you log in to wefwef through your instance, or how do you access everything, specifically on mobile?

I haven’t tried wefwef in particular but all native apps I tried work just fine. An issue I can see cropping up from wefwef is that Lemmy’s CORS policies are way too restrictive by default. No idea if they do any kind of proxying to get around that but that would be the main issue I’d imagine.

There are many guides on getting started with Linux servers as a whole. I recommend installing Debian Bookworm on a virtual machine or a spare laptop at first and going through the writeups all major cloud providers have, just to get a feel for using the terminal & initial setup (SSH hardening and reverse proxy configuration and so on)

After getting an initial feel for Linux admining, start reading up on Docker, Docker Compose, and containers in general. Avoid Podman until you’re experienced with Docker as it’s just different enough to trip you up. You can also check out LXC/LXD although it’s way less popular.

Oh, and speaking of Docker: UFW AND DOCKER WILL NOT WORK TOGETHER! DOCKER BYPASSES UFW (just making sure you don’t learn this until it’s too late)

Be careful of guides that are old (even a year makes a difference) or for different “distros” than the one you have. An exception for the second case is the Arch Linux wiki, which is one of the best resources just in general, aside from a few Arch specific bits like the exact package names to install. You should also use Arch’s “man pages” reference, as they’re built from the latest versions of packages compared to other man page renderers that are frequently outdated (like die.net)

Lemmy itself is harder to get right because the instructions so far are intended for people who kinda know what they’re doing, but once you have the base Linux admin knowledge, it won’t be that hard to pick up the parts necessary to get working with something like Lemmy.

The instance I’m replying from is a 5 eur/mo box from Hetzner.

Your main concerns are gonna be active user count & storage space. Especially if you decide to allow image or god forbid video uploads. Having a bunch of inactive users aren’t going to affect costs that much as long as they don’t have, like, a milion subscriptions. (If they’re all subscribed to the same community things will “deduplicate”)

It’s generally more like “Steve’s 10 eur/mo cloud server in which they run ten other things next to Lemmy, which is written by two devs and barely held together by duct tape and prayers”

But that doesn’t change the overall point.

…so create your community on that instance. Others will still be able to access it just like you’re accessing communities elsewhere.

Some instances disallow community creation. That’s the only part where this argument has any merit. Otherwise which instance a community is on doesn’t really matter.

There are instances that disable downvotes. https://github.com/maltfield/awesome-lemmy-instances has a field saying if an instance has them enabled or not.

Alpine is completely separate by RHEL by a country mile (hell, it doesn’t even use glibc). You’re probably thinking of Rocky

I read Haiku (OS) has no future recently and I think a similar thing is happening here.

Do we want to “have a future” or do we just want to be a cool place doing cool shit for ourselves?

I don’t see why it wouldn’t. You may need to enable a config option or two though. Documentation isn’t NixOS’s strongest suit.

It doesn’t have to. I ran Sway on Nix the entire time I used it, and I know Hyprland supports Nix as well

GNOME’s stance on user customization has been “users can do whatever they feel like using 3rd party tools like Gradience or entirely custom CSS, but if you’re a distro maker then only use the Approved Ways™ to customize things”

Now, I have zero clue if that solves anything (it very likely doesn’t), but it’s actually more than most people give them credit for.

I’d say “go join in on the issue tracker and tell GNOME about this” but hearing from some people who tried that before you I’m not too hopeful that would do much of a difference. All I know is that complaining here isn’t going to solve anything.

Even in the most safe space-y parts of Mastodon (which has way more experience with trolls & assholes than Lemmy just by existing for longer), single user instances don’t get defederated without reason. So I’d be more in favor of reflecting inwards and wondering if there is a specific reason why they defed’d me.

Because of the way it works, you can try out on a VM for a bit and move your config over to real hardware trivially if you end up liking it. That’s how I did it before I realized how immature it’s rocm support is and had to switch back to arch

Definitely not Gentoo based, but if you can get by with their unique approach to basically everything, NixOS can be pretty interesting, in that while it is technically source based, binary caches are widely used to basically “pretend” to be a binary distro. And it does let you patch things shouid you want it (at the expense of recompiling everything that even slightly comes in contact with the patched package)

There are some parts that are too “baked in” to change – requiring systemd, for instance – so that may be a dealbreaker for you.