Do I think the problem is bigger than it is? I find it kind of scary to see who wrote to whom and when.
Ialso remember that you saw this as an issue about a year ago
Do I think the problem is bigger than it is? I find it kind of scary to see who wrote to whom and when.
Ialso remember that you saw this as an issue about a year ago
It is fucking scary. Wait until you see the screenshots.
Its a good reminder that this is why things like simplex exists. This is the nature of communication online. employees, governments, and companies buying your data can already see this information on existing social media platforms. nostr just exposes this reality. This doesn’t mean we can’t do better. We definitely can. Giftwraps are one approach, but we can do more targetting DMing to specific relays and use AUTH for dms. I think there is a wine relay for this, but some clients like damus still need to implement it.
with fully working NIP-42 and enabling keysend/AMP payments to top up balances we could onion route messages... but without payment how you gonna stop it being abused?
Keysend only works for the few who really run their own node for that or am I mistaken?
well, yes, so actually to really make it work there has to be some way to run small channels as well
lightning is such a bastard like this... source routing is a huge network gossip overhead cost and because privacy they strangle that gossip down to a minimum and then clients are supposed to be able to figure out valid routes while the topology of the network is partly in a fog of war
it's a really difficult problem, one that is avoided in the bitcoin architecture by using a broadcast which has other, orthogonal privacy implications that are mitigated by UTXOs
i had to spend quite a while studying LN before i understood how it might be used and one of the key things that i came up with - when designing Indranet, which is/was an attempt to make a source routed onion relay network funded by micro accounts paid for using AMP - was the idea of changing the way LN does its topology gossip to make it more open so it's more reliable, and extending AMP to enable multiple redundant paths so that stuck payments can't happen, and the amounts involved are tiny, like at most thousands of sats at a time to reduce the fingerprinting and counterparty risk
basically it would be more or less a fork of lightning to really work, and the same basic principles would apply to an onion routing scheme on top of nostr, except with an overall lower total traffic volume
i think in the long run it will have to be achieved somehow but it's a really thorny problem
They might, but a centralised but highly secure Server is still a bigger hinderance for the average blackhat who can just run a script and see who is talking to who and cross reference them all. You can even see zap history in dms, which is literally people's monetary tx history.
What does zap history in dms mean?
certain clients let you zap dms.
you see the zap above the dm
oh that pretty bad, why would clients do that 🤦♂️ i hope they at least use private zaps in that case
Why can’t this be obscured by creating a new npub/nsec for each DM while having the conversation tracked by some arbitrary Tag in the Event?
Even though this information is publicly broadcasted by relays, the npubs published wouldn’t have any identity since they are newly created with no other history. The only traceable thing would be the amount of messages that mention the tag.