Quote from: dinofelis on April 13, 2017, 03:03:50 PM

Quote from: IadixDev on April 13, 2017, 02:39:43 PM

If you are a guy in your basement, and that you want to find a solution to keep the chain coherent, and you choice between à simple well proven determinstic solution , or a solution that is completely off chart, super costly, and risky, why he would choose the second ? Why going through all this bother with pow and block reward who introduce huge complexity ? Why ?

What deterministic solution ? There isn't any that isn't centralized or permitted. Proof of Stake was a possibility, but Satoshi was facing the problem that he was the only stake holder in the beginning. He would have had to sign all blocks by himself, and unless someone actually GOT COINS FROM HIM, there was no way to get a second stake holder.

Quote

That could be just be as simple as selecting block and tx based on which have the lowest hash. Period. No pow, no reward, no mining craze.

The problem is, WHEN do you consider that transaction A is the valid one ? How LATE can transaction B be propagated and WIN from transaction A ?

Suppose I pay you 100 BTC. You observe transaction A on the network paying you. How long do you wait before you consider that this payment is secure ? Suppose I buy a car with that. How long do you wait until you let me have the car ?

Suppose that the next day, I make a new double spend payment to myself. I can modify my receiver addresses until I find a payment that has a smaller hash than transaction A. I call that transaction: B. I now transmit B on the network. As B has a smaller hash than A, the consensus tells us one should take B over A, and finally, your transaction is eliminated.

Ok, but one day later, we don't accept this any more. Ok, but how long do we have to wait ? At what point do you consider that A is definitively the accepted transaction ? After 30 minutes ? But what if B comes in after 29 minutes for Joe and after 31 minutes for Jack ? Joe and Jack will now disagree FOREVER over what was the right transaction ? If you connect to Joe, you see your transaction reversed, while if you connect to Jack, you see your transaction not reversed ?

--> this is the consensus problem. It is already difficult if most players want to play honestly. It becomes very hard if you get a sybil possibility of 90% of the nodes conspiring to game the system (90% of nodes in the hands of one entity).

Suppose that I transmit transaction B almost immediately after transaction A, but I fire up 90% of nodes that "ignore" transaction B. You will probably not see transaction B, and you think that after half an hour, you are safe. Then I switch off my sybil nodes. The rest of the network has preferred transaction B. When you try to spend your coins a few months later, your right to spend doesn't exist on most nodes, because they had rejected A, and chosen B, and forgot about A. You are the only one remembering A, thinking it was right.

Satoshi found a kind of solution with PoW. It is a clunky solution, but he needed one.

You could just remove the reward, any one can mine new block out of the mem pool, if two blocks or tx are in common, a determinstic algorithm could be used to select between the two.

With the hash of previous block in the header including timestamp for me it's enough to prevent sybil attack. Checkpoint could be made every 100 blocks and hashed in the chain.

And selecting conflicting blocks/tx within a timeframe with deterministic algorithm.

You can only emit a new block based on the last good one, including a timestamp and still need to keep checking other nodes for better blocks based on deterministic algoritgm.

Well would need to cut some corner there with the timestamp of valid chain to avoid sybil attack.

Other than this, all remain the same without the whole competition for the reward.

In other words, the amount of computational resource required to solve double spent is much < to cost of mining the proof.

Same for the complexity and time of solving the problem of double spent with classic deterministic solution, and putting the whole pow in place.

Because blockchain are a very specific case of byzantine general, and a lot can be assumed from other nodes due to the protocol , there is lot of things that can be assumed about the valid chain, and it's not like just any data can fit and there is no way to say which one is good or wrong at all.

It's like all the generals know each other thinking from before the battle. And the plan cannot be just anything.

If 90% of node want to collude on a chain, then it's the good chain, cant really prevent it anyway lol if 90% of users are against the protocol, why would they use it to begin with ?

The thing is in most case , all the chains will still contain sensibly the same txs in an order or another, so there is nothing to loose or win in choosing one or the other, it's just reordering the tx to fit the block header hash, outside of double spent that are fraud and should be removed anyway. And all non double spent tx made on a chain is also valid on the other.

So in the end, why twist it so much with the pow and reward, for me it's still someone who know well the world of investment start up and trading, and it's not too far stretched either to think there is some kind of financial plan with it, and it's never really clear what those plans are, and there are many shady area surrounding it's thinking and deployment.

Either it was planed as a boat, a space ship, a submarine, hard to even say.

But it's hard to think there no plan with it all, or that it was studied with good knowledge of economic theories, and the world of IT, and he knew at least a minimum where he wanted to get at.

But his idea seem quite oriented with free market and still thinking with buisness in mind , with some kind of philosophy or plan.