Question regarding the role of miners vs nodes in securing the network

[fernanvito]

Hello everyone,

Recently, I've been reading a lot about how the bitcoin governance/development process works.
I read the book "The blocksize War", and one thing that caught my attention was how central in the blocksize debate was the role of miners vs nodes.

So I've been trying to understand what's the role of miners vs nodes in securing the network, and I have a couple of questions/doubts that I'd love to get answered.

As far as I understand, miners are in charge of confirming transactions, they secure the network by making it costly to falsify transactions (create unforgeable costliness, as Nick Szabo would put it). On the other hand, nodes are in charge of validating, propagating, and keeping a copy of transactions, they make sure that miners act according to the protocol rules, not acting maliciously (i.e. introducing false transactions, or try to double spend).

My question is, in order to make a successful 51% attack: One would need to have 51% of mining power, 51% of nodes, or both?
For example, in the case that someone gets to have 51% of mining power, could nodes protect the network from that attack by not validating transactions? In other words, could the network be secure if the nodes are decentralized even if the mining pool is not?

IMO, in the answer to this question lies the key of what keeps the network decentralized.

I'm new to this forum. So I'm sorry if this question would better fit into another topic. In that case, just make me know and I will move it to the right place.

[BlackHatCoiner]

My question is, in order to make a successful 51% attack: One would need to have 51% of mining power, 51% of nodes, or both?

If you were to acquire 51% of the total computational power, it'd be enough for you to attack the network. You don't need 51% of the nodes, you just need one node who'll broadcast the chain with the most work.

For example, in the case that someone gets to have 51% of mining power, could nodes protect the network from that attack by not validating transactions?

No, they couldn't, because that would make it susceptible to censorship. The nodes just verify what they receive based on the consensus rules they follow. They can't deny a chain if it contains more work than the one they follow. Even if the blocks are empty, even if there are transactions reversed in prior blocks.

The whole system relies on the fact that there'll be honest miners outpacing anyone who'll try to attack it.

[Wind_FURY]

My question is, in order to make a successful 51% attack: One would need to have 51% of mining power, 51% of nodes, or both?

If you were to acquire 51% of the total computational power, it'd be enough for you to attack the network. You don't need 51% of the nodes, you just need one node who'll broadcast the chain with the most work.

For example, in the case that someone gets to have 51% of mining power, could nodes protect the network from that attack by not validating transactions?

No, they couldn't, because that would make it susceptible to censorship. The nodes just verify what they receive based on the consensus rules they follow. They can't deny a chain if it contains more work than the one they follow. Even if the blocks are empty, even if there are transactions reversed in prior blocks.

The whole system relies on the fact that there'll be honest miners outpacing anyone who'll try to attack it.

OP, the bolded part is a very important point. The miners that have successfully controlled 51% of total network hashing power still cannot spend coins that they do not have, or “print them out if thin air”, or send invalid transactions/blocks, or change the rules. They can merely censor transactions.

[BlackHatCoiner]

They can merely censor transactions.

And why would the bold part be important? It seems insignificant to me. The word “merely” should be meant ironically here. If they can “merely” censor transactions, they can also delete the whole transaction history and make the whole system useless.

Yes, they indeed cannot force anyone to change the total amounts of coins ever issued, or increase their wealth from other people's money, but they can force the nodes to accept their defeat; to accept that their system is no more secure; that they essentially dug their own holes with the rules they all agreed to follow.

[DannyHamilton]

They can merely censor transactions.

And why would the bold part be important? It seems insignificant to me. The word “merely” should be meant ironically here. If they can “merely” censor transactions, they can also delete the whole transaction history and make the whole system useless.

It depends on what type of malicious activity we are talking about.

If we are talking about a malicious actor with the intent to destroy the Bitcoin system at any cost, then you are correct. The entity with more than 50% of the global hashpower can do that regardless of what the nodes can or can't enforce.  This would be a VERY expensive attack, but is theoretically possible.

On the other hand, if we are talking about a malicious actor with the intent to profit from their power, then the entity with more than 50% of the global hash power is MUCH more limited in what they can accomplish.  Any action that would destroy trust in the system as a whole would simultaneously destroy their ability to profit from it.

[ranochigo]

Yes, they indeed cannot force anyone to change the total amounts of coins ever issued, or increase their wealth from other people's money, but they can force the nodes to accept their defeat; to accept that their system is no more secure; that they essentially dug their own holes with the rules they all agreed to follow.

If that ever happens, then the game theory of it would've failed. Where there is a far more incentive to be dishonest than to be honest, and they can't force the community or the nodes to accept their defeat. Any attacks of this sort will always be one-off. 51% attacks or majority attack is a byproduct of a feature and the game theory aspect of it would've failed. It would be less of a 'defeat', but rather we can acknowledge that someone was willing to sacrifice a huge, huge sum of money to attack the network.

You cannot really delete the chain either, that is possible in theory, given that nodes only function the way that do. It is impossible for a whole chain of 300GB to be propagated over the network, it would literally take hours.

[BlackHatCoiner]

If we are talking about a malicious actor with the intent to destroy the Bitcoin system at any cost, then you are correct. The entity with more than 50% of the global hashpower can do that regardless of what the nodes can or can't enforce.  This would be a VERY expensive attack, but is theoretically possible.

In any case, it's in the entity's decision, not in the community's. We'd all be dependent on what they'd decide and we could do nothing in order to prevent it. The system would suddenly have a central point of failure.

You cannot really delete the chain either, that is possible in theory, given that nodes only function the way that do. It is impossible for a whole chain of 300GB to be propagated over the network, it would literally take hours.

An empty chain wouldn't weight that much. If we assume that each empty block is 80 bytes and the new chain's height is 700,000 whose work is greater than this one's, then it's a matter of 56MBs to completely destroy it.

[ranochigo]

In any case, it's in the entity's decision, not in the community's. We'll all be dependent on what they'll decide and we can do nothing in order to prevent it. The system will suddenly have a central point of failure.

That is the whole point of PoW, isn't it? If someone is willing to spend that much resources to do so, then I concede. PoW didn't fail and it has function as it should. The whole point of PoW is for those who wield the most power (PoW in this case) to decide whatever they'd like to do with the chain. Even if you throw out the economics of doing so, the effects of the attack is easily nullified by switching to a different algorithm.

An empty chain wouldn't weight that much. If we assume that each empty block is 80 bytes and the new chain's height is 700,000 whose work is greater than this one's, then it'll be a matter of 56MBs to completely destroy it.

Ahh okay. Though the node will throw loads of warning if a major re-org happens. It is very difficult to re-org those blocks though, your commulative proof-of-work is proportional to the amount of time that has passed, which means it is extremely expensive and time consuming to do so. If you're going to start today, then re-organizing the entire blockchain would probably take years at the very least (assuming >100% of the current mining hashrate). The impacts of it would be pretty limited, and honestly no one would even think about doing it.

[BlackHatCoiner]

That is the whole point of PoW, isn't it?

That's the point of PoW, but Bitcoin is more than just a mechanism. I find its point lost if we assume that someone holds more than the majority of the hash rate. Whether they attack it or not; whether they replace the current chain with an empty one; whether they reverse a bunch of their transactions or not. Just because someone has more power than everyone else combined is enough for me to stop using it.

And I repeat, saying that you can “merely” censor the transactions is ironic the least.

[Wind_FURY]

They can merely censor transactions.

And why would the bold part be important? It seems insignificant to me. The word “merely” should be meant ironically here. If they can “merely” censor transactions, they can also delete the whole transaction history and make the whole system useless.

Compared to what newbies might assume if a cartel of miners, or a state-backed entity, successfuly controls more than 51% of the total hashing power, yes “merely” that’s all they can do. The miners can’t change the consensus rules, they can’t control coins from wallets other than their own.

[pooya87]

The miners can’t change the consensus rules, they can’t control coins from wallets other than their own.

They can change the rules, but the important thing is not about their ability to make those changes but it is about getting others to accept that change. For example there was a change which introduced the ability to spend the old unspent coins (falsely referred to as Satoshi's coins) by the miners. Obviously no bitcoiner followed that change so it became a shitcoin that failed in its early days.

[witcher_sense]

They can change the rules, but the important thing is not about their ability to make those changes but it is about getting others to accept that change. For example there was a change which introduced the ability to spend the old unspent coins (falsely referred to as Satoshi's coins) by the miners. Obviously no bitcoiner followed that change so it became a shitcoin that failed in its early days.

In other words, the moment they change the rules, they stop being bitcoin miners, instead, they're becoming the miners of "better version of bitcoin", which is also referred to as bitcoin fork or shitcoin.

And I repeat, saying that you can “merely” censor the transactions is ironic the least.

Paradoxically, but the "censorship-resistance" feature of the bitcoin blockchain implies that as long as you are following protocol rules you can do whatever you want including censor of the transactions you don't like. For example, you can put into blacklist the node that, in your opinion, has been propagating only FATF compliant transactions. You can't censor the censors, however, and you can't drive them out of the network if you both are part of the network.

[Ucy]

The 51% attack shouldn't be possible without all participants involved in running and taking part in network consensus... I sometimes wonder what happened at that particular point in time when miners became more powerful than all participants to the extent of having the ability to attack the network with higher hash power.
I think 51% of all network participants should have that kind of power to change things. That will likely make attack more difficult especially if you have clear rules how the power can be used by 51%.

Well, Bitcoin isn't perfect, I have noticed vulnerabilities that can be exploited but they're few compare to other cryptos I know about well. Infact it's probably the only true cryptocurrency as far I can tell, because it's well built (according to worlds standard), with principled developers.

[Wind_FURY]

The miners can’t change the consensus rules, they can’t control coins from wallets other than their own.

They can change the rules, but the important thing is not about their ability to make those changes but it is about getting others to accept that change. For example there was a change which introduced the ability to spend the old unspent coins (falsely referred to as Satoshi's coins) by the miners. Obviously no bitcoiner followed that change so it became a shitcoin that failed in its early days.

“They can”, is different from They Truly Can from the standpoint of Bitcoin. But you’re right, technically any miner or group of miners “can” change the rules, but it’s not Bitcoin. From Bitcoin’s standpoint nothing changed. Plus they would go back to Bitcoin because it’s more profitable. Why change the rules to fork to shitcoin?

[hatshepsut93]

Miners secure the network by keeping the cost of attack high. If many miners suddenly leave, the hashrate drops and it becomes cheaper to launch a 51% attack.

Nodes keep the network big, so it's harder to launch sybil-based attacks and not one can monopolize the network. In big block scenario regular users can't run their full nodes and have to rely on big servers that could be run by mining companies and be easily regulated, like for example demanding KYC.

[Wind_FURY]

Miners secure the network by keeping the cost of attack high. If many miners suddenly leave, the hashrate drops and it becomes cheaper to launch a 51% attack.

Nodes keep the network big, so it's harder to launch sybil-based attacks and not one can monopolize the network. In big block scenario regular users can't run their full nodes and have to rely on big servers that could be run by mining companies and be easily regulated, like for example demanding KYC.

KYC will be demanded by centralized entities regardless if network security is overshoot, or undershoot. The point of debating for smaller blocks is to overshoot network security, to keep the network more decentralized, and to let the network scale out.

[hatshepsut93]

KYC will be demanded by centralized entities regardless if network security is overshoot, or undershoot. The point of debating for smaller blocks is to overshoot network security, to keep the network more decentralized, and to let the network scale out.

Right now the network is a decentralized entity. But in big block scenario it's a centralized entity that can be regulated, which means Bitcoin itself gets regulated. Today we have some tiny percent of hasing power that wants to enforce some government blacklists, but in case of centralized network it can easily be more than 50%.

[witcher_sense]

Right now the network is a decentralized entity. But in big block scenario it's a centralized entity that can be regulated, which means Bitcoin itself gets regulated. Today we have some tiny percent of hasing power that wants to enforce some government blacklists, but in case of centralized network it can easily be more than 50%.

Many people demand regulation for the decentralized bitcoin network, but they still fail to realize that "regulated bitcoin" is nothing else but an oxymoron. If there is only one node or several nodes in the bitcoin network, it won't become a centralized entity, nor it is going to be regulated. The first thing governments will do when they know that bitcoin is no longer decentralized is destroy it out of existence. Bitcoin is a symbol of freedom, it is a means to express one's opinion and disagreement regarding the government's monopoly on money issuance. It is a symbol of disobedience to tyranny. If it loses its ability to maintain decentralization, it will stop being a symbol, it will stop being.

[BlackHatCoiner]

In big block scenario regular users can't run their full nodes

I think the correct phrasing is that it discourages them to run a full node. You should always have the right to run a full node, but if you're making it hard and expensive for a person, which would happen in other block chains if they had the same amount of transactions as Bitcoin's, calling it decentralized is highly ironic.

Bitcoin is a symbol of freedom, it is a means to express one's opinion and disagreement regarding the government's monopoly on money issuance.

And I wonder... How can one see it as a symbol of freedom since it's regulated by the governments? They may not achieve on destroying the network, but making its purpose meaningless is worse and they already do it by regulating it. The crypto-anarchists' dream doesn't seem to come true. 

With Bitcoin you're neither free or besieged. You're a free besieged. 

[witcher_sense]

And I wonder... How can one see it as a symbol of freedom since it's regulated by the governments? They may not achieve on destroying the network, but making its purpose meaningless is worse and they already do it by regulating it. The crypto-anarchists' dream doesn't seem to come true. 

With Bitcoin you're neither free or besieged. You're a free besieged. 

I don't quite understand what you are trying to say. Can you enumerate some of the government regulations that succeeded in making bitcoin's purpose "meaningless"? In your view, what are the main purposes of bitcoin? Can the government effectively prevent you from transacting with anyone they consider a bad person? Can the government meddle with the rules of the network by forcing miners to comply with the law? Wouldn't miners go to another place if the law they were coerced to comply with considerably reduced their profits or even made their business lossmaking? Can the government overwrite the rules of the network by running nodes with incompatible software? Because I personally consider consensus rules are what governs and regulates the network. If you can't change them, you can't regulate bitcoin. The only power you have is verify for yourself transactions and blocks comply with the rules, which you previously agreed with when joined the network for the first time, by running your own full node. What's your view on that?