Agreed. For example, we have a gazillion reposts and shares adding the asknostr tag around here. That’s interesting, and I’m curious to see how well community self-moderation of tags, à la Stack Exchange, will work in the context of general social media.
This might be a point of contention, but in the underbelly of the beast, I also like the idea of DHT + pkarr. For now this is a somewhat uninformed opinion like my earlier take on IPFS. But assuming it works, it sounds great. Not necessarily as a be-all and end-all solution to every distributed social media problem, but I can certainly see use cases where we need aggregators and wonky bootstrapping logic which could benefit from DHT-based alternatives.
Replies (1)
Using a DHT to find content like in IPFS doesn't work because there's no point in trying to host your data in multiple places whose locations are decided algorithmically. The user needs to have direct control over which relays/servers they read/write to. However, I like the idea of using a DHT to host just the data about where someone posts their data, similar to what's done with PKDNS.
My only issue with the system is that it still feels over complicated to me. If the DHT network is truly decentralized, then anyone can spin up nodes and potentially perform a sybil attack. This isn't possible in PKARR (BitTorrent Mainline) only because the node's identifier is based on its IPv4 address, which is a centralized identifier. So, in the end, it's essentially a centralized group of servers.
So why not just have a federation of relays that host only NIP-65 (kind 10002) events
