Is there a "why not Tor"?
Discussion
tor is slow, doesnt scale and its not flexible. A native nostr discoverablity layer (and anonimity layer) could be used in many ways.
One of them is a better tor with lightning/ecash micropayment to actually scale the service
Tor is slow because it works and people use it. Bouncing requests through one or two nodes won't provide reliable anonymity, and using relays as TCP/IP will be unbearably slow if more than two people actually use it.
tor is slow because therse no anti-spam (actually they added a sort of proof of work client side, but it is not so effective), theres no incentive to run nodes that are the facto moslty feds and volountary. If you add micropayments as spam filter and as economic incentive for people to run relays, you can have a tor network that scales.
Also its trivial to implement in a way that provide reliable anonimity, it just need to reimplement onion routing, all problems in theory are solved.
Tor doesn't work in a browser without additional software. This will.
Furthermore, you can't trust Tor servers and a huge part of them are run by governments.
The difference here is that we can choose the relay servers and avoid the ones we don't exactly trust.
nostr isn't protected against that either
Tor does work in the browser if you're prepared to run a proxy. That is additional software.
Correct, I wrote "without additional software". A proxy is additional software.
NWS main components
Entry node: It forwards tcp packets to the exit node using a SOCKS proxy and creates encrypted events for the public key of the exit node.
Exit node: It is a TCP reverse proxy that listens for incoming Nostr subscriptions and forwards the payload to the designated backend service.
nostr:note1jaclnk5n5xf4fvnzuectlqtf57t30xkmv7kn8gx8rzya3nftr95q8qrhju
a year later and the answer seems to be "yes, tor" 🤘 🚀
it's a hard problem, and we'd be better off with a healthy tor network than trying to re-invent it