#askNostr Quick question, genuinely trying to understand this better. In cases like CVE-2018-17144 or similar consensus-critical bugs: • Since malformed transactions (eg duplicate inputs / other non-standard or invalid patterns) get rejected at the mempool/relay layer, they wouldn’t propagate across the network, right? • So my doubt is: could an attacker craft a transaction that still ends up mined by an unrelated miner (because it sits in some miner’s mempool), or would all sane mempools reject it so the only path is a miner including it deliberately in their own block template? • Meaning for a true consensus exploit you need PoW to even push it into the state transition path (ConnectBlock / UTXO updates), because without a mined block it never becomes “real”. Then the failure-mode question: • Even if a bug were exploited “silently” and the invalid chain extended beyond 6 blocks, wouldn’t external watchers (exchanges, pools, explorers, independent indexers, people running older/other implementations) notice via divergence, coinbase/supply inconsistencies, or reorg signals? • And once detected, wouldn’t the economic + social layer coordinate around the rules-correct chain (patch + reorg / rollback), so Bitcoin doesn’t “die” but instead shows antifragility under stress? Curious how you think about whether a truly silent consensus bug is realistically possible today, given the diversity of monitoring and independent reconstruction of state.