Quote from: BitRock on May 29, 2014, 02:36:15 AM

Quote from: ThePurplePlanet on May 29, 2014, 02:18:49 AM

Quote from: BitRock on May 29, 2014, 02:12:39 AM

Quote from: ThePurplePlanet on May 29, 2014, 01:50:11 AM

Quote from: BitRock on May 29, 2014, 01:16:37 AM

Quote from: DeathAndTaxes on May 29, 2014, 12:55:39 AM

Quote from: BitRock on May 29, 2014, 12:37:03 AM

Quote from: DeathAndTaxes on May 27, 2014, 02:21:04 PM

Quote from: digitalindustry on May 27, 2014, 02:11:47 PM

if PoS provides for more security how come advanced check-pointing was conceived with PoS only?

Because it isn't. PoS suffers from a "nothing at risk" problem. An attacker can use coins they no longer have but did have at one point in the past to attack the network. Actually there is no reason to not do this. If the attacker is unsucessful well they lose nothing in the attempt and if they are successful they get all their coins "back" (that they may have lost, had stolen, or sold). Checkpoints limit how far back the chain can be reorged but they don't solve the nothing at stake problem, only limit the extent of the damage.

Please tell us which PoS coin has been attacked by so called "nothing at risk" problem.

None definitively although more than one POS coin has been 51% attacked. Many of them have stake requirements which are negligible so often cheap attack is easier than a more sophisticated nothing at stake attack. No POS coin has extensive history in the field other than PPC and it avoids an attack by using 100% centralized checkpoints. Still you don't really believe that "hasn't happened" = "can't happen"? Bitcoin has never been 51% attacked therefore you believe an attack is impossible?

Could you provide any other theoretical evidence to support your claim if it had never happened in practice?

In PoS, you use your stake to determine the longest chain thus total cumulative stake on a chain is the one to make nodes follow that chain. Say currently 5% of the stake is used to secure the network. An early adopter/exchange/group of people/stake pool with more stake (more than 5%) at some point in the coins history (the people might have sold long time back) can create a parallel secret chain and reverse the current chain by showing more cumulative stake. How do you get the current chain with this decision mechanism? Cannot be decentralized... No matter what is the approach the attackers can repeat the steps of the PoS approach... Does it make sense?

It doesn't make sense at all. To do this kind of attack, you need generate a parallel secret chain using your 5% coin first. To sell your 5% coin on exchange you need to synchronize to the publick blockchain. Then your blockchain is not secret at all and you can't go back after you sell your coin.

The clients accept the longest chain. Once you sell you withdraw say bitcoins you dont need to go back to the current chain at all to make a profit....

To make a profit you need to reverse the current chain by showing more cumulative stake after you sell your PoS coin to bitcoin. Then you can get bitcoin and PoS coin at same time. However, to move 5% PoS coin to exchange, you have to publish your secret longer blockchain to public. Then your secret longer blockchain becomes to public chain and you lose the ability to reverse the public blockchain.

Why would the network without human intervention not accept my second chain? It perfectly follows the rules... The other attack you are describing is the nothing at stake problem where you can keep trying double send without any cost. I am talking about a more severe attack that will force the network to become centralized or investors/exchanges/reputation take massive losses....