Reasons to add sync flags to Bitcoin

- Flags will be mined selfishly, and not published until the advantage
   gained from withholding is less than the mining reward.  This effect may
   kill the decentralization features, since big miners will be the only ones
   that can selfish-mine flags.  Indeed, collusion would be encouraged... just
   ship the flag to the miners you do business with, and no one else.   At the
   expense of loss of flag revenue, your in-group would gain a massive
   advantage in main-chain mining.

---

Is there a reason miners would be more likely to engage in selfish mining of sync flags than they are now with ordinary blocks?

On Tue, Jul 26, 2016 at 9:58 PM, Martijn Meijering via bitcoin-dev < bitcoin-dev@lists•linuxfoundation•org> wrote:

> Is there a reason miners would be more likely to engage in selfish
> mining of sync flags than they are now with ordinary blocks?
>

This proposal has the same effect as adding mandatory empty blocks.

POW targeted at 2 minutes means that the POW for the flag is 25% of the block POW. That gives a flag every 2 minutes and a block every 8 minutes.

It has the feature that the conversion rate from hashing power to reward is the same for the flags and the blocks. A flag get 25% of the reward for 25% of the effort.

A soft fork to add this rule would have a disadvantage relative to a competing chain. It would divert 20% of its hashing power to the flag blocks, which would be ignored by legacy nodes. The soft fork would need 55% of the hashing power to win the race.

This isn't that big a deal if a 75% activation threshold is used. It might be worth bumping it up to 80% in that case.

This rule would mean that headers first clients would have to download more information to verify the longest chain. If they only download the headers, they are missing 20% of the POW.