Quote from: rnicoll on September 18, 2015, 10:24:15 AM

Quote from: HCLivess on September 18, 2015, 09:42:15 AM

Quote from: BitUsher on September 17, 2015, 08:21:43 PM

Bitcoin is Byzantine resilient because of PoW and Game theory. Bitcoin follows Nakamoto consensus, but all Byzantine consensus algos are only resistant up to 51% or less.

The bigger question is how likely a 51% attack or sybil attack is within bitcoin and under what conditions can we make it less likely.

With PoW there is at least physical limitations and better signals that limit sybil attacks vs PoS. Nothing is trustless or completely immutable but we can get closer to these ideals with decentralization and the right security mechanisms.

Yes, for PoS the security is exponential to amount of users

Except that major exchanges tend to hold vastly more coins than individuals, so they replace mining pools in being the 51% risk, and (much) worse, their mining security can be anonymously and easily moved if it's accessed by a hacker. See for example MintPal (Viacoin) and Bter (NXT)

MintPal was Vericoin rather than Viacoin(PoW), just fyi. That was a legit attack that resulted in an attacker having control of a large enough amount of VRC to cause VRC users to choose to roll back as the lesser of two evils(debatable of course, but I remember the dev making a decent argument that it was pretty much the only option aside from complete death).

BTER's NXT being hacked is a completely different an unrelated situation though. It was simply a hacker stealing around 50m NXT. There was no resulting security issue with this theft in regards to the NXT network. Mounting a successful attack on NXT appears to be extremely difficult from what I can tell.