There’s no real / true decentralization

That’s not exactly true.

That said, you do need some form of centralized service to connect peers, but you can federate those. It’s only job would be to connect peers, and a STUN server w/ TURN fallback is usually the approach here. These instances don’t need to store any data long term, they just need to connect peers, and the client is free to choose any instance they want, or host their own.

That’s how Tor works (entry nodes), and most decentralized systems use a similar system.

One of the best parts here is that offline often just works, and you can sneakernet around firewalls (e.g. if you visit China or something), and all you need to do is connect to a local relay to find local peers.

Blockchain

My understanding is it’s only used for name resolution, so the number of data points here should be in the thousands, not millions or billions, so the resource usage should be minimal.

Basically, the blockchain is functioning as DNS here.

Lemmy is not decentralized; it’s federated. “Decentralized” and “federated” are not synonyms,

This isn’t quite accurate. Lemmy is decentralized, but it’s not distributed. It’s decentralized because the source of truth for a community isn’t your instance, but your instance caches content for that community locally.

They’re not synonyms, true, but federated systems are typically (always?) decentralized, and rarely (never?) distributed.

Plebbit seems to be a weird mix of both. Communities are centrally managed, but the data seems to be distributed, at least upon creation (everything probably makes its way back to the creator for moderation).

DHTs and distributed ledgers are notoriously difficult to design well, often suffering from syncing lags and block delivery failures

I haven’t looked into it too closely either, but it seems the blockchain is only used for name resolution (seems to be used for community names), so updates should be fairly infrequent.

I assume they’re using a DHT for data though, probably a separate one for each community, but maybe not. Those can be updated asynchronously, so if data is cached locally, latency shouldn’t be an issue.

I really don’t think that would scale at all. A reasonably popular community could have tons of simultaneous posts, and if everyone needs to sync before posting, that would suck. You could probably avoid the worst of it by having posts use uuids, but you’re going to have IO issues at scale. Also, would you need the full repo cloned? That can get big, and you generally only care about recent posts.

Also, if you’re doing the UUID thing, you’d have sort everything every time locally. That’s fine if you only have a few thousand posts, but if you get into millions or billions, it’ll get bad, especially if you’re dealing with files.

Databases solve these problems really well. Even a simple SQLite dB would be much better than a filesystem, like orders of magnitude better.

My understanding is the blockchain bit is optional and only used to establish ownership over a name. The posts and whatnot are not on a blockchain, that would be silly.

That’s a reasonable reaction if it was true. But it’s not, plebbit is FOSS.

Yeah, that’s a super uninformed take. Blockchain is perhaps the best solution for authentication in a P2P system. I assume they’re linking blockchain to cryptocurrency, but AFAIK, there’s no cryptocurrency in Plebbit.

For authentication, you need a central authority of some form, and blockchain is about as decentralized as you can get while having that central source of truth. It’s a good solution.

Bluesky. If you’re okay with Twitter/X, you should be okay with Bluesky.

Maybe in the short term, but longer term, I don’t think your vision of what the community wants to be will necessarily match what the community wants to be. I guess we’ll see how it turns out, and ib sincerely hope I’m wrong because I want an improvement on Reddit and Lemmy.

Sure, I’m just worried it’ll have similar problems as reddit, just without global admins to fix/enforce things. The creator of a community is rarely the right person make decisions long term. Moderation should be based on trust and merit, not first come first served because moving everyone to a new community is hard.

We had similar problems here on Lemmy when most of the popular communities were on Lemmy.ml and subject to their moderation.

But maybe it’s fine. It’s probably an improvement on Reddit, and maybe an improvement on Lemmy if it actually encourages more diversity in community ownership. I’ll certainly check it out!

Yeah, I’ll check it out. It’s certainly an interesting approach. I’m interested to see how the moderation system ends up working in practice.

Wow, that looks just like old Reddit, awesome!

Yeah, I was pretty stoked for it too. As someone building something like this on my own time, I really want someone to beat me to the punch, because maintaining something like this isn’t something I really want to do.

Building something like this is hard, marketing a project is hard, and getting the timing right is also hard (major usability issues solved before everyone comes to try it out).

But yeah, I’m still here until I find something better.

Awesome!

Is this something that we could have a script for? I imagine this isn’t the last time this will happen.

I don’t trust automation for restoring from backup, so I keep the restoration process extremely simple:

1. automate recreating services - have my podman files in a repository
2. manually download and extract data to a standard location
3. restart everything and verify that each service works properly

Do that once/year in a VM or something and you should be good. If done properly, it shouldn’t take long.

You could use scheduled snapshots to provide the backup portion.

That’s a great use-case for snapshots! RAID still isn’t a backup, but it can be quite robust.

Yup. My “essential” backups are well under 1TB, the rest can be reacquired.

So, you’re geocaching a bunch of USB drives then?

Tons of typical self-hosted services, like vault warden, actual budget, etc.

But for something that old, I’m thinking more along the lines of RC cars and other projects involving the GPIO. I’ll reserve my newer ones for self-hosting stuff, older stuff for things that don’t benefit much from extra processing power.

Depending on how you intend to use Kali, that’s either a bad idea or a terrible idea. Don’t use Kali bo for long running processes.

Just get a router with that feature or flash a firmware with that feature.