Proof of Raffle

[Kevin7777777]

Proof of Work consensus protocol is a waste of computational resources. This model will not be scalable and decentralized at the same time. We propose a solution that is somewhere between the Ripple and Bitcoin protocols and we include a Raffle, we call it Proof of Raffle.

Bitcoin client can establish that miners must always try to fill up the block size limit (today 1 mega), and transactions must be ordered in alphabetical and time stamp order into the block. Then all miners could achieve almost exactly the same tx’s in the exact order into each block (it doesn’t matter if some miners try to cheat by manipulating tx’s names or timestamps…).

Tx´s propagation statistics show that in 15 seconds more than 99% of tx’s have reached all nodes.

When 15 seconds have passed, miners could obtain the concatenated hash (hash from last block and merkle hash from latest block hashed all together). From this hash then we pull only the numbers and obtain a BIG NUMBER. Also we make a circular list of active nodes (nodes that have validated transactions in this last block) and we order them alphabetically in a circular way (the last node is next to the fist node of the list). Starting from the first node of the list, we use the BIG NUMBER to count around the circular active nodes list until we pick the winner. This will be the proposed winner node.

Every node then send (and receive) their proposed last block (and proposed winner node) to all other nodes. If a node receives any more late transactions that will change their proposed block it can send their new proposed blockchain (and winner node) again to all other nodes. Finally every node can pick the block that is confirmed by most of the nodes as the last block of the blockchain (maybe wait for 80% consensus) and accept it as the last confirmed block, and confirmed winner node.

Folks, try to analyze the concept, don’t hit me in nonsense stuff, and sorry for the bad English.

[TrueBeliever]

bad English? Your English looks pretty good to me.  Are you pretending to be Japanese too?


'Proof of Raffle'?  Looks more like Proof of Luck to me.  How can you get consensus on who is in the raffle?  Due to network latency there are likely to be multiple raffles happening, so every block you have a fork.  Then how do you identify who is the final winning fork?  In Proof of Work the winner is the miner who did the most work to mine the block.  In your case how can you decide who is the 'luckiest'?

Any protocol in which all nodes need to know the entire state of the network will not work effectively and be subject to many attack vectors.

In addition if there is no cost for me to add extra nodes into the network to enter the raffle then I will spin up lots of virtual machines to increase my chances, so then we suddenly go to millions of nodes across the planet.  I would like to see how you can have consensus across millions of nodes across the planet for a raffle.

[monsterer]

Sybil attack

[tupelo]

Nodes are cheap to operate. Your system would be very easy to attack. Don't forget that hashing power secures Bitcoin's block chain.

[TrueBeliever]

Sybil attack

I am not sure how the sybil attack would work in this scenario.  I would have thought the lottery system would attract a large number of cheap nodes.  An attacker would find it difficult to get a useful proportion of nodes into the system to allow them to influence the network.  ie PoW requires more hashing power in order to get a useful proportion, whereas here actors would all be increasing their own node count.

I think the sheer number of nodes combined with inherent network latency would mean that forking would be rampant, and there would be insufficient time to identify the best chain and re-sync across the network.  The system would simply 'fork' itself to death 

[monsterer]

I am not sure how the sybil attack would work in this scenario.  I would have thought the lottery system would attract a large number of cheap nodes.  An attacker would find it difficult to get a useful proportion of nodes into the system to allow them to influence the network.

And so, what is a node?

A private key?
An IP address?

?

[erik777]

I am not sure how the sybil attack would work in this scenario.  I would have thought the lottery system would attract a large number of cheap nodes.  An attacker would find it difficult to get a useful proportion of nodes into the system to allow them to influence the network.

And so, what is a node?

A private key?
An IP address?

?

A port?  

https://www.reddit.com/r/Bitcoin/comments/2vv8m6/pseudonode_when_is_a_full_node_a_full_node/

One would have to wonder what would happen in this model if you introduced a proof of stake.  E.g., each node must own 1 coin -- max coins being 22 million.  

I would do multi-sig to ensure that the node could sign the proof with one sig but require two sigs to transact, protecting it from hacking.

[monsterer]

One would have to wonder what would happen in this model if you introduced a proof of stake.  E.g., each node must own 1 coin -- max coins being 22 million.  

The you make the network attack cost a constant (number of coins owned), rather than a being linear (number of blocks produced).

[mookid]

It is true that as long as bitcoin grow, PoW becomes more and more expensive. I think Nakamoto underestimated the amount of energy and computing power that was needed in order to fuel bitcoin at great scale. Maybe this problem can be surpassed with Technological advances? so far this hasn't been the case, it only pushes up the hash rate.