How is POW going to stop bitchat spam, if the real users are on mobile devices while the attacker runs on cloud infrastructure or bare metal?
Discussion
With the assumption that a real person just uses one PoW request for a message and attackers are creating dozens of messages which requires dozens of PoW requests?
But creating POW that equals a users mobile device on cloud is essentially free
No, cloud computing i's not free.
And the spammer, by definition, has to send thousands of messages continually.
Growing minimum PoW with other heuristics (similar messages, new npubs acting similarly, etc) could make spam more expensive.
At least, it's worth trying it, imo.
Of course it’s not technically “free”, but if it only needs to compete with mobile devices (even hundreds or thousands) it’s practically free.
A modern MacBook Air is 1.6 million times faster at sha256 than an iPhone is (citation needed, this is GPT knowledge). The gap is so massive
But you are always supposing users would be constraint to mobile CPU, and that's not entirely true:
NIP-13 allows for Delegated Proof of Work:
https://nips.nostr.com/13#delegated-proof-of-work
We all can play the same game of spammers 😉
It's super easy. Just turn off network permission.
PoW can't stop spam because attackers can scale up their hash grinding efficiency, but especially mobile users cannot. then the PoWtards suggest making this into a paid service to hashgrind your messages on faster hardware.
why not just cut out the middleman and you pay to use the relay? subscription with zaps? isn't this like the most primary logical use case for zaps?
