[bitcoin-dev] Consensus fork activation thresholds: Block.nTime vs median time vs block.nHeight

Tue Aug 4 21:29:56 UTC 2015

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

On 4 August 2015 16:02:53 GMT-04:00, "Jorge Timón via bitcoin-dev" <bitcoin-dev at lists.linuxfoundation.org> wrote:
>One thing I've noticed there seems to be disagreement on is whether
>miners' upgrade confirmation (aka voting) is necessary for
>uncontroversial hardforks or not.

To be clear, without a strong supermajority of miner support the fork risks attack. Requiring 95% approval - which is actually just a 50% majority vote as the majority can squelch the minority - is an obvious minimum safety requirement.

Another option is Hearn's proposal of using centralised checkpoints to override PoW consensus; obviously that raises serious questions, including legal issues.

For forks without miner approval miners have a number of options to defeat them. For instance, they can make their own fork with a new consensus algorithm that requires miners to prove they're attacking the unwanted chain - Garzik's recent 2MB blocks proposal is a hilarious, and probably accidental, example of such a design, with the original Bitcoin protocol rules having the effect of attacking the Garzik 2MB chain.