Author: Mike Hearn 2015-08-24 15:21:47
Published on: 2015-08-24T15:21:47+00:00
The proposed BIP for disabling Bloom filtering is flawed according to NACK. The first rationale that disabling Bloom filtering would improve privacy is spurious as it only adds privacy for lightweight wallets, not node operators. Furthermore, with a high false-positive rate, Bloom filtering can still provide query scrambling and low bandwidth usage. SPV usage is also increasing, contrary to Peter's data, which did not account for the fact that SPV clients tend to connect and disconnect rather than remain connected all the time. Mike's data shows that bitcoinj and breadwallet make up the bulk of P2P clients. Lastly, NACK found that the claim of a DoS attack due to Bloom filtering has not been researched, so he implemented a similar attack and found that Bloom filtering is extremely fast and that transactions and blocks continued to be relayed without delay. Therefore, even if a refined attack could deny service to someone, a fix would have to apply to regular non-filtered block fetches too. In conclusion, the proposed BIP creates more problems than it solves and is an unnecessary upgrade headache.
Updated on: 2023-06-10T20:45:20.039545+00:00