Arjen's avatar
Arjen 2 months ago
The main argument she made was that if the adversary is observing both ends of the network. So if they see anything before the tor/mixnet and after the exit node, then they can correlate traffic. @Max prob refers to Epoxy, my project attempting to do multi-hop publishing to nostr relays. It is indeed similar to your solution @Laan Tungir . My solution was to do subscriptions to relays over multiple hops and different networks (clearnet->i2p->tor->clearnet for example) epoxy: the site for testing it: You should check out @PABLOF7z 's onion routing spec, (can't find the repo now). He included padding and sats in the route as well.

Replies (1)

Thanks Arjen. At least in my case, I don't think an adversary could correlate traffic because of timing delays, reduction of padding, and I think uniquely in what I am calling superballs, the ability to add padding along the way.