You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
DGW appears to have a serious vulnerability due to the powLimit method it inherited from BTC that allows a selfish miner to get an unlimited number of blocks in 2 hours. Dash is very large so it may not be vulnerable to 51% manipulations, but it's something to be aware of and understand if Dash clones start suffering this attack. It is not limited to DGW, but applies to all simple moving averages. Maybe there is some aspect to Dash that prevents this attack that I'm not aware of.
This normally requires a selfish mine by a >51% attacker if the MTP (median time past) is reasonably small like the BTC default of MTP = 11 instead of Cryptonote's default MTP=60. Some immediately say "we can't do anything about >51% attacks, so this is not meaningfully addressable." However, this greatly amplifies the blocks acquired, which pulls 51% attacks into existence. The attack may do things to profit by forcing a chain revert instead of seeking profit from the blocks gained.
For T=120 and N=24 (DGW), the attack takes less than 2*N*T < 2 hours to get unlimited number of blocks. The attack uses a very complicated sequence of timestamps. I've described it here in the "powLimit attack" section. The shortest description of the attack is that it uses the powLimits against themselves. The attack is possible is because a limit on timespan is not really a POW limit, so it's a fundamental logic / reasoning error that goes back to BTC. The limits should have been placed on next_target instead, or just removed because they do not provide any benefit (that I am aware of) unless a coin has such a bad difficulty algorithm that it is suffering bad oscillations (and needs to fork anyway....see current situation in Ravencoin).
The text was updated successfully, but these errors were encountered:
DGW appears to have a serious vulnerability due to the powLimit method it inherited from BTC that allows a selfish miner to get an unlimited number of blocks in 2 hours. Dash is very large so it may not be vulnerable to 51% manipulations, but it's something to be aware of and understand if Dash clones start suffering this attack. It is not limited to DGW, but applies to all simple moving averages. Maybe there is some aspect to Dash that prevents this attack that I'm not aware of.
This normally requires a selfish mine by a >51% attacker if the MTP (median time past) is reasonably small like the BTC default of MTP = 11 instead of Cryptonote's default MTP=60. Some immediately say "we can't do anything about >51% attacks, so this is not meaningfully addressable." However, this greatly amplifies the blocks acquired, which pulls 51% attacks into existence. The attack may do things to profit by forcing a chain revert instead of seeking profit from the blocks gained.
For T=120 and N=24 (DGW), the attack takes less than 2*N*T < 2 hours to get unlimited number of blocks. The attack uses a very complicated sequence of timestamps. I've described it here in the "powLimit attack" section. The shortest description of the attack is that it uses the powLimits against themselves. The attack is possible is because a limit on timespan is not really a POW limit, so it's a fundamental logic / reasoning error that goes back to BTC. The limits should have been placed on next_target instead, or just removed because they do not provide any benefit (that I am aware of) unless a coin has such a bad difficulty algorithm that it is suffering bad oscillations (and needs to fork anyway....see current situation in Ravencoin).
The text was updated successfully, but these errors were encountered: