yeah thats like not feasible on a big relay. i need code that will do that.
what is nostrich.house ?
you can ping me when you see a bad note/user..
yeah thats like not feasible on a big relay. i need code that will do that.
what is nostrich.house ?
you can ping me when you see a bad note/user..
There's no reason why relays need to be big, is the thing.
they can be
But they don't have to be. And there's the point that the big ones make good hubs for a hub-and-spoke network, but there's no need to read/write directly to them.
Both of these models, in other words, don't work very well, but relays allow us to have a combination of the two. Like a logistics network, but with data.

or you can remove the indirection and directly read/write to hubs and lose absolutely nothing
You lose censorship resistance, decentralization, independence from large relay operators, and on-server bespoke administration.
You also lose a lot of redundancy, but the redundance is all behind a limited set of access points. Walled cities, rather than individual homesteads.
I prefer to have a homestead and visit the cities.
i think the outbox model is worse. not scalable to mobile phones and mobile data settings imo.
We use community outboxes. Everyone in the community reads and writes to the same relay(s), and those relays distribute and pull from other relays on the server side. And those community relays are actively moderated by machines and humans.
And we encourage people to also have a personal relay, and/or a local one.
This is why I only use one relay on mobile.
Relays can handle the outbox stuff. No actual need to have the clients do it. Relays can just talk to relays.
not if the outbox requires the user's auth for the event (eg DMs)
DMs are a special case. Should probably be P2P, to be honest.
And for the rest of the sensitive stuff, that's what the personal relay is for. That's a grand total of 2 relays, max, and one of them is already predetermined. Throw in a localhost relay, for the lulz. Costs practically nothing.
yeah, DMs are the case that makes having a rendezvous relay proxying function make sense. at most the relay only sees traffic volumes and who connects to who, the content is encrypted. without much more elaborate than that you could use tor to set up accounts on several intermediary relays and voila, you don't even need tor to isolate who's talking to who, just two intermediaries and basta.
Also, there's no rule saying that an AUTH relay can't aggregate from other relays. It can just go find stuff, from the subscribers, and pull it.
sounds VERY technical for an average user
They don't have to know.
Moderated Communitues take :90percent: of the responsibility. And that's a win-win.
Yeah, you just go to the client and use it, with its default settings.
Least-technical thing you can do.
Primal is the same way. You go to Primal and can immediately see stuff on their community relays. Their community is just not very discriminating.
Exactly.
The "just go to the client" is currently real bad advice though.
Since none of them (realize they) are community apps.
Besides our projects that is.
Users will learn or leave.
And devs will ask a robot for a new super smart #WoT bandaid that'd make decentralized Twitter work.
With both the Web and the Trust getting smaller with each attack.
WoT ghetto, where you can only talk to the people you already talk to...
The PGP-like side-effect where you can only enter of someone vouches for you is a cool idea tho.
Follows is just a really dumb way to do that 😉
The simple workable way forward 🚂 .