|
Title: Swirlds hash grid consensus discussion Post by: monsterer2 on July 18, 2017, 11:18:15 AM Hi there,
I'd like to open discussion about Swirlds hash grid consensus design as it pertains to cryptocurrencies. http://www.swirlds.com/downloads/SWIRLDS-TR-2016-01.pdf http://www.swirlds.com/downloads/Swirlds-and-Sybil-Attacks.pdf At first glance this design has a lot of similarities with the way Ripple's consensus works, with the key difference that voting history is preserved. This means it stands a better chance at reducing vulnerability to long range attacks and bootstrap-poisoning. However, as the author notes in that second paper above, in it's pure form it offers no protection from sybil attack. The author suggests using stake from another blockchain in order to weight votes to provide some protection again sybil attack, but does not talk about the inherent problems in doing so, such as the transient nature of the balance of any particular staking address in another blockchain. In addition the author talks about using PoW to acquire voting weight, which does not suffer from the same transient problems associated with stake, but it is still largely PoS in nature since voting weight persists or decays at a finite speed, unlike the way PoW functions in bitcoin, for example. All in all, I think it has applications outside of cryptocurrency, but IMO is poorly suited to the requirements of the consensus mechanism within one. Cheers, Paul. |