I am not a BU member and was hoping someone could sponsor this BUIP proposal.
Motivation
It is possible if there is a divergence of EB settings of miners for median attacks to create chain-splits. It is in the miners best interest to coordinate their blocksize increase to the same value synchronously to eliminate the chance of their blocks being orphaned. To eliminate the chance of their blocks being orphaned from a chain-split, they need to coordinate changing their EB to the same value at the same time.
Using the block-chain to vote makes coordination for the miners much easier and increases transparency. Increased transparency makes it harder for bad actors to do a chain-split attack.
Objectives
Implement the following features:
Anticipated Challenges and Uncertainties
Any bugs in this automatic updating of EB could be used to trigger miners changing EB settings out of sync with the rest. This could be used right before a chain-split attack.
There is a risk that less than 50% of miners mining don’t have their EB settings set automatically. This could lead to not enough miners updating their EB setting at the same time and leave a median attack available to an attacker (same as currently). One possible option to minimize this risk is to have miners signal automatic control on the blockchain. An attacker could spoof this signaling however.
There is also a risk that miners will not be able to haggle and agree enough to reach 500 votes. This could potentially make it harder for a blocksize change if there is not enough consensus. It is not easily determined what the effect on the marketplace this would have.
Impact
It is hoped that these changes will make emergent consensus much less risky and easier to achieve. With less risk of a chain-split, the market can have more confidence in emergent consensus of the blocksize. This change also increases transparency of blocksize emergent consensus.
Additional information
http://www.btcgametheory.com/strategyremovechainsplitriskbu/
Motivation
It is possible if there is a divergence of EB settings of miners for median attacks to create chain-splits. It is in the miners best interest to coordinate their blocksize increase to the same value synchronously to eliminate the chance of their blocks being orphaned. To eliminate the chance of their blocks being orphaned from a chain-split, they need to coordinate changing their EB to the same value at the same time.
Using the block-chain to vote makes coordination for the miners much easier and increases transparency. Increased transparency makes it harder for bad actors to do a chain-split attack.
Objectives
Implement the following features:
- Miner voting on the blockchain that is independent of EB settings
- Setting that allows for automatic control of EB value
- When 500 of last 1000 blocks vote the same EB setting, a preparatory period is triggered
- After preparatory period of 144 blocks (24 hrs), EB setting automatically switches to new value.
- After the switch, a grace period of 1000 blocks is triggered in which the EB setting can’t be automatically switched. After the grace period, votes are again counted.
- Manual override setting that allows miner to have direct control of EB setting. There could be situations where miners want to change their EB settings immediately due to an emergency or surge of transactions.
- Conditional flag that requires 750 of last 1000 to signal BU in order for automatic control of EB to be allowed. BU needs to reach 75% of the network before any automatic control of EB should occur. Once 75% BU has been reached, automatic control of EB is allowed onward.
Anticipated Challenges and Uncertainties
Any bugs in this automatic updating of EB could be used to trigger miners changing EB settings out of sync with the rest. This could be used right before a chain-split attack.
There is a risk that less than 50% of miners mining don’t have their EB settings set automatically. This could lead to not enough miners updating their EB setting at the same time and leave a median attack available to an attacker (same as currently). One possible option to minimize this risk is to have miners signal automatic control on the blockchain. An attacker could spoof this signaling however.
There is also a risk that miners will not be able to haggle and agree enough to reach 500 votes. This could potentially make it harder for a blocksize change if there is not enough consensus. It is not easily determined what the effect on the marketplace this would have.
Impact
It is hoped that these changes will make emergent consensus much less risky and easier to achieve. With less risk of a chain-split, the market can have more confidence in emergent consensus of the blocksize. This change also increases transparency of blocksize emergent consensus.
Additional information
http://www.btcgametheory.com/strategyremovechainsplitriskbu/
Last edited:
