Damus
edouard · 19h
You are right ; we should continue. 20% of knots and 20% core 30 so 80 pct network running 80 capped op_return. I would have preferred consensus to have capped op_return but the conservativeness in ...
Tauri profile picture
You’re missing the point. Models suggest that even if 99.9% of nodes run restricted policy, a toxic op_return and a motivated attacker will still find a way to embed a large file on the network. Core 30 changed everything. Prior to it was possible to limit this attack vector with stricter policy. Consensus change is the only mitigation right now. That’s why it’s more important than ever to realise the risk associated with compromised dev team. We should stop pretending consensus is effective defence against human governance. It’s good enough, but not a painkiller.
1
edouard · 14h
I don’t understand why core 30 changed that. Even before core 30 a motivated attacker could (one of the arguments of core for changing) and it was indeed a situation where 99.9 pct of nodes were a running restrictive relay policy. And if you believe that most nodes having a restrictive or relax re...