It works like that: spammers are going to operate using powerful servers with lots of computing resources and minimal energy constraints—that's the nature of their activity. In contrast, legitimate users often rely on smartphones or low-powered devices.
This creates a fundamental problem for PoW as a spam mitigation strategy: there’s no viable threshold that can effectively hinder spammers without also significantly impairing regular users.
Bitcoin is very different, as its use of PoW is fundamentally different: it’s a competitive system where miners race to solve a game in a winner-takes-all model. That schema does not make sense in a microblogging protocol like Nostr.
Login to reply
Replies (2)
You're not actually explaining how this is true.
On email, you're not showing where it goes wrong if I try to set up a proof of work email system. You keep saying legitimate users are "hindered" but not explaining how.
On Tor and Bitcoin, you didn't even mention Tor, and again, you didn't have an explanation of why it works for Bitcoin but not this, you simply stated you believe that
Ok, let me make an example.
In the paper the cost is in $, but let's simplify and use time. Let's say that we want a high PoW barrier, like 60s (average) to send an event to Nostr relays using a smartphone. Let's say that the same message costs on a server something like 10s, as the server is more powerful.
So the Nostr user will be pissed ok by waiting 60s to send a message, which will also drain his smartphone batteries.
On the other hand, the dedicated server of the spammer will send 8640 spam messages per day, flooding Nostr relays. And that's assuming the spammer has a single machine, but in reality could be a srvrfarm.
You can reduce the cost and make it even easier for the spammer.
For Tor I am not sure.
Bitcoin is very different: PoW is a competitive game to sign a block, it is not an antispam. There is no parallel to a messaging system.