Topic: Bitcoin is decentralised network? Hmm... not really... The idea how to fix it. (Read 1067 times)

And also true of many other pools too (increasingly so).

Pay per Share is either dying or dead, as if the pool operators charge the sort of fee that takes account of their risk with PPS, it's too unattractive to miners. Some type of PPLNS or Geometric Share awards appear to be the only model that's both sustainable to pool operators AND attractive to miners.

The second option is to make plenty more pools

The third option is to stop promoting BTCGuild and 50 BTC completely.

There is nothing stopping anyone from solo mining... except their expected life span. No one is forced into a pool.

I only ever point my hash power at p2pool. Decentralized. The way mining should be.

I can assure you p2pool is quite stable. What you probably heard was that payouts are irregular; which is true.

+1 and I'm quite familiar with p2pool inner workings (see my sig).

Carry a 51% attack is not so easy.
The pools must also have an interest to do so.

And it is necessary that the miner follow even in such a situation the corrupted pools.

I agree, but the proposal has an innovative idea: the p2pool subsidy system be THE subsidy system.

no no no... p2pool pays out more than a centralized pool even with the .5% development donation.
The work of one transaction gets split into thousands of little pieces for you to work on, you send it to the p2pool network and it gets credited as if you were mining solo by Bitcoin.

The shares are calculated differently, so you have to mine for at least 24 hours straight or until the block is found to generate a good reward, you got to remember that it's like solo mining... if you get an orphaned block, your out of luck. the big pools pay by each share regardless if they make it into the blockchain or not... there are weeks when the luck on p2pool is going to be really really bad.. and other when each block is being found every few hours.

Almost everything you've said about p2pool is totally incorrect, including the above statement. I am also not especially familiar with p2pool.

As far as I understand P2Pool does not allow to miners to choise their own difficulty.
But this is especially very important for ASIC miners.

I think it has more to do with impatience and a lack of understanding the differences between how p2pool and regular pools work, but I may be wrong.

Probably, I have heard about that and I have heard that P2Pool is quite unstable and as result of that gives to miners even less income than a big stable pool even with pool's fee.
If this feature will be included into the core of the network and provides clear benefits for miners, like avoid pool's fee then the miners will migrate from pools to solo mining.

I am not well-versed in the pool side of bitcoin, but I think you have described something that already exists: P2Pool

I am concerning about a threat to the biggest Bitcoin's feature: decentralisation.

Actually it is mining pools.
Look at the Bitcoin's chart: we have just two entities: BTC Guild and 50BTC with in summary give us more that 51% of hashing power.

So, it is very easy to somebody with enough money/power to take control over just these entities and through them to take control over whole network.

I am sure that this problem was discussed here, but did anyone suggest a solution?

I was thinking quite a long about it and I came into conclusion that the solution of this problem exists.

The root of idea is simple: to allow ordinary miner to do solo mining without any pool.
The problem is: how to do that ?
Ok, there is some rough idea about the algorithm:

First of all: we need additional blockchain, let's call it alt-blockchain.
Each block in alt-blockchain will contain only one transaction: put out of thin air some amount to the miner's address.

The algorithm:
1. The miner chooses difficulty which he want.

2. He advertise this difficulty to network together with his bitcoin address. Since this moment the address is linked to this difficulty.

3. The miner mines block in the main blockchain. There is an additional field in the block: a reference to the last block in the alt-blockchain, let's call it ALTBLOCK. This field can be empty. If it is empty then miner does usual old-style solo-mining (like now).

4. If the miner found a hash for main blockchain with difficulty less then current one but more or equal then he advertised, then a new block is added to alt-blockchain. This block contains: hash for main blockchain as proof-of-work and one transaction: put amount of chosen difficulty to the miner's address.

5. If the miner found a block with difficulty more or equal than current one then the block is added to the main blockchain and alt-blockchain. Since this moment all of miners who put blocks to the alt-blockchain have rights to transfer coins to the any of address on the main blockchain. The amount of coins which can be transferred calculated like:

 (25+fees)/AllAccumulatedDifficulty*DifficultyAccumulatedByMiner.

Where AllAccumulatedDifficulty and DifficultyAccumulatedByMiner are the sum of difficulties put in to the alt-blockchain in the period since previous ALTBLOCK till current ALTBLOCK registered in the main blockchain.


So, actually this alt-blockchain looks like one big mining pool is built into the core of the network.
Of course implementation of this algorithm demands hard-fork upgrade.

I think that the current mining pools will be against this implementation because with that they will eventually loose their business.

But we have to do something with that. Because now I would not say that bitcoin is the truly decentralized network.