Quote from: BombaUcigasa on November 12, 2014, 06:56:04 PM

Quote from: gatra on November 12, 2014, 04:46:52 PM

Quote from: achimsmile on November 06, 2014, 06:55:42 AM

Quote from: bornil267645 on November 06, 2014, 05:44:28 AM

Quote from: Kimowa on November 06, 2014, 04:21:37 AM

Quote from: kokojie on November 06, 2014, 04:15:14 AM

If you reverse PoS and PoW in your post, then your post would be 100% accurate and correct

Doesn't it cost nothing to attack a PoS coin? While you must use actual resources to attempt to attack a PoW coin?

Exactly, I thought attacking PoS coin is much more cost free (next to nothing) than attacking PoW coin.

You just have to buy 51% of the currency or track down majority of stakeholders and compromise their private keys. Much cost free

lots of people seem to believe this (I think it's even mentioned in Sunny King's PPC paper), but it's not accurate: you need 51% of the actively staking coin-age. That's much, much, less than 51% of the currency.
I had some ideas to help fix this, I'm working on it.

How do you fix the past?

If someone had 51% stake share at one point, what can you do to prevent him forking the chain?

Relating to Nxt only:

There is nothing to fix. The current inclusion of the retargeting algorithm makes it computationally very expensive to do a history attack. i.e. much much more hash power the bitcoin network currently has. The better chain needs to almost mirror the honest one in terms of certain properties. And you only have 720 blocks to calculate this in.

Nxt has interlocking, layered security. A lttile bit being added at a time (economic clustering next IIRC). Transparent Forging isn't a switch that will be flipped one day, >50% of Transparent Forging is 'on' now.

/Nxt

But since you insist that POS = POS = POS, I don't think you'd be interested.