Censoring miner tx fees - Is this a worthwhile initiative?

[checkmate7]

I do expect this to have some controversy as it goes against the ethos of openness and neutrality, however I believe it may be necessary to have some tools for deterrence of bad actors in addition to the ones already available in bitcoin core.

As we see greater emergence of publicly listed miners and institutionalisation of miners, there may be an increasing trend of trying to apply controls on what txs get included or not, as we have seen earlier this year by 1 publicly listed US miner. Understandably there will always be some miner some where that will be agnostic - however reciprocal check and balance can be useful.

Currently bitcoin nodes have the ability to disconnect from a block height or to blacklist a peer which are useful tools to ensure the network operates fairly. But as block subsidy reduces and the above miner trend increases, it may be worth introducing a feature that allows transmitters to somehow limit which addresses/coinbase stamps would be eligible to realise network fees. It seems like a sensible countermeasure. Of course this feature can be bypassed by modifying stamps or creating new addresses, but it certainly makes administration more difficult.

How such proposal would be implemented obviously requires block award rules to be overhauled which introduces complexity and obviously a lot of initial resistance from the dev community.

Complexity aside, how does the dev community feel about such a modification? Is it useful as a "deterrent"? 

I'm curious on people's thoughts.

[1715170235]

1715170235

[1715170235]

1715170235

[1715170235]

1715170235

[NotATether]

Basically the current way it works is that transactions include a txfee in addition to all their outputs, which the Bitcoin protocol awards to the miner who includes them in a block.

By making rules that prevent miners from receiving fees, you now have the problem of where are you going to store the txfees that haven't been claimed yet due to censorship. They are not real outputs so they cannot be stored in some address somewhere, and you can't make it so that the transaction gets a free fee because such a model will encourage users to get transactions (somehow) included in a no-fee block which will starve miners in the long run.

[pooya87]

it may be worth introducing a feature that allows transmitters to somehow limit which addresses/coinbase stamps would be eligible to realise network fees.

What you are suggesting is essentially centralizing bitcoin and will kill it. Nobody should ever decide who is eligible to mine bitcoin and/or receive the network fees. The protocol works fine and is based on the working assumption that the honest actors will always be dominating the network even if there were some dishonest ones (like MARA pool).

[checkmate7]

Certainly not a straightforward implementation, and I don't profess to have a solid understanding on the workings of the protocol, but I am sure there are brilliant people here that could find a way, assuming it is actually worth it.
If I just spitball from my limited technical knowledge it could payout to the miner address as part of a multisig with some sort of rules.
A HUGE overhaul obviously, but really if we just park the "HOW to do it" for a moment.

I am more curious on if "is this something we need ?"
i.e. is the potential threat worth introducing such a feature or is the threat exaggerated?
Does such countermeasure create massive fractures in the community that it's simply not worth it?

Is a more elegant way simply to block peer IP/tor ID, to choke a miner's mempool?

@pooya87 Today miners can soft fork and censor, but users cannot, that is the main reason for the question -but I agree with you on a philosophical level this is NOT a good idea and the working assumption of a minimum # of honest actors is the right mindset.

[NeuroticFish]

First, the tx fees alone is not something the pools care that much of, if you'd make such a rule, they'd just mine empty blocks and go on. Which may be even worse, since it'll keep the mempool full.
Second, if we try to fight censoring with censoring, we clearly do something wrong.

Is a more elegant way simply to block peer IP/tor ID, to choke a miner's mempool?

If one node didn't block the "rogue mining pool", the new mined block will still be broadcasted, isn't it? So.. this won't work.

[checkmate7]

Good discussion going here.
Allow me to re-frame and re-articulate:

1. Today fees make up ~15% of block award, in a decade or so we reach a tipping point as subsidies shrink- this becomes a fee driven market, hence also an attack vector.
2. Hash moving away from private entrepreneurs in the east to public Cos in the west, with shareholders, compliance, ESG, regs, and nations obsessed with OFAC, sanctions, KYC/AML, and weaponising capital flows.
3. Huge amount of bitcoin is getting trapped / sucked into ETF/ETP products which allows the owners to have a bigger say potentially

I do believe openness and neutrality should prevail in case of hard conflict / risk of hard fork... but the question is posed - do we need additional resilience in light of a changing dynamic?

[pooya87]

1. Today fees make up ~15% of block award, in a decade or so we reach a tipping point as subsidies shrink- this becomes a fee driven market, hence also an attack vector.

You are forgetting about the price which is the most important factor in determining whether miners need the fees. Put simply price needs to only be $60k in next halving for miners to earn the same amount they were happily earning in the past 4 months and needs to only be $120k after the next decade!
Historically speaking price rise is massively higher than this.

2. Hash moving away from private entrepreneurs in the east to public Cos in the west,

I feel like you are saying this because of the recent Chinese drama, but the east is a lot more than just China and China didn't even have that much hashrate to begin with.

3. Huge amount of bitcoin is getting trapped / sucked into ETF/ETP products which allows the owners to have a bigger say potentially

Bitcoin is not PoS so it doesn't matter how much of it you own, you don't have any say in anything by just owning bitcoin.

[NotATether]

Is a more elegant way simply to block peer IP/tor ID, to choke a miner's mempool?

maybe by setting the nLocktime to (currentblockheight+2) if you have a suspicion a mining pool you don't like (but again, why the suspicion when it would've included your transaction anyway?) will mine the next block, but this is fuzzy and not guaranteed to work.

[kaggie]

I do expect this to have some controversy as it goes against the ethos of openness and neutrality, however I believe it may be necessary to have some tools for deterrence of bad actors in addition to the ones already available in bitcoin core.

As we see greater emergence of publicly listed miners and institutionalisation of miners, there may be an increasing trend of trying to apply controls on what txs get included or not, as we have seen earlier this year by 1 publicly listed US miner. Understandably there will always be some miner some where that will be agnostic - however reciprocal check and balance can be useful.

How would you avoid creating the problem that you say you want to solve?

On a protocol level, because miners and nodes would not agree on who to censor, wouldn't this result in a large number of forked chains?
It may not be a technically infeasible problem, considering there are already centralized crypto coins out there with few miners that claim to do this. But a centralized crypto is useless unless it originates with a nation-state or major financial player.

And who decides who the bad actors are? Should every coin be labeled with the individual miner or name who owns it? That would result in a lot of $4 wrench attacks. What happens when you disagree with people in country X, so every currency transaction or miner from that country is banned? That results in an unusable currency between countries, which makes it no longer a currency.

Presently the protocol is simple with miners competing across international institutions, which simplicity you want if you want the method to go up in value and utility. The more complicated you make the rules, the more likely that there will be flaws that require arbitration when something goes wrong. If you are trying to build a currency to "change the world" so to speak, you need it to be simple to ensure fungibility. While the one miner that you mention is willing to censor transactions, there are plenty of miners who are competing to not censor transactions. If your solution to a miner censoring transactions is to censor miners, then you end up with the problem that you created. You have to consider this in an opposite way, how do you award miners who don't censor to keep the protocol open?

If you censor miners, then aren't you already censoring transactions?

Is a more elegant way simply to block peer IP/tor ID, to choke a miner's mempool?

maybe by setting the nLocktime to (currentblockheight+2) if you have a suspicion a mining pool you don't like (but again, why the suspicion when it would've included your transaction anyway?) will mine the next block, but this is fuzzy and not guaranteed to work.

Agreed with you NotA..
I'm confused at the problem. If a miner is processing transactions, then why would you want to block them? Even if they don't mine all transactions, this miner clears portions of the mempool so that other miners can be awarded for processing the attempted blocked transactions.

The protocol was built to be resilient against this kind of attack where a miner was disliked due to the number of nodes available. If you don't like a miner, then you don't have a choice but to use them. You could block their IP, etc, but they would be able to pop up again rapidly with a new IP.

[ranochigo]

A few issues that I can foresee if this proposal is ever implemented. It doesn't serve as a viable disincentive against rogue actors; introducing points of failure and trusting specific miners goes against the ethos of Bitcoin. It is extremely annoying to be forking Bitcoin every time a bad actor appears and the community attempts to censor it; you need all of the clients to upgrade immediately or risk certain nodes staying on the wrong fork.

Rogue actors do not care about mining fees, it would merely subsidize parts of their attack but would be fairly negligible as the price tanks after that. They would probably start mining blocks which censors specific transactions and with one of the legitimate miner's address being included in the coinbase. Your censoring would fail in this aspect as they don't care about collecting any part of the mining fees.

Effective censorship requires 51% of the network participating in the said attack. Any less than that, then transactions would potentially just have their confirmations delayed and would eventually make it into a block and get confirmed. If a 51% attack does happen, then I would assume that Bitcoin doesn't really have any use. Once it gets into a block, then it could potentially taint other outputs. At some point in time, all of the outputs would be tainted in some way or another. Should miners start censoring every single transaction?

No one should be dictating the core Bitcoin functions; specifically who gets to mine and who doesn't. I've never really seen any policies implemented which goes against this, perhaps it might change but I can't see it happening without major backlash.

I am more curious on if "is this something we need ?"
i.e. is the potential threat worth introducing such a feature or is the threat exaggerated?
Does such countermeasure create massive fractures in the community that it's simply not worth it?

IMO, no. Exploring the benefit/cost of an attack like this instead of trying to "attempt" and albeit fairly ineffectively fix the problem would result in far more pressing issues to solve.

Is a more elegant way simply to block peer IP/tor ID, to choke a miner's mempool?

No.

[stompix]

1. Today fees make up ~15% of block award, in a decade or so we reach a tipping point as subsidies shrink- this becomes a fee driven market, hence also an attack vector.

You are forgetting about the price which is the most important factor in determining whether miners need the fees. Put simply price needs to only be $60k in next halving for miners to earn the same amount they were happily earning in the past 4 months and needs to only be $120k after the next decade!
Historically speaking price rise is massively higher than this.

Slight problem with this.
If the price only doubles you will only keep the same level of security, and it will not scale to the amount of money "at risk".
So basically you will have  3 trillion in market cap with the same protection in hashrate as you had when they were only 750 billion.
If that number goes through 3 cycles the difference will become x8, how would you feel now if the amount of gear needed to flop the network would not cost 2 billion but just 200 million?  

This aside any idea of restricting transactions based on blacklist under any form will be the end of Bitcoin if implemented, we're fixing something that is not a problem generating a lot of problems with no fix.

I do expect this to have some controversy as it goes against the ethos of openness and neutrality, however, I believe it may be necessary to have some tools for deterrence of bad actors in addition to the ones already available in bitcoin core.

The main problem is who decided who are the bad actors, who sets the minimum and the maximum, who votes for the people who will determine the bad actors. Any system that is based on human decisions will be opened to personal biased views, let the network work how it is designed to, without a centralized supervisor who can judge people on limited information. Can I nominate Elon to be part of this council?  

[NotATether]

The main problem is who decided who are the bad actors, who sets the minimum and the maximum, who votes for the people who will determine the bad actors. Any system that is based on human decisions will be opened to personal biased views, let the network work how it is designed to, without a centralized supervisor who can judge people on limited information. Can I nominate Elon to be part of this council?  

You can't, he's quite busy advising proposals to make Dogecoin more "green"  

On a serious note though, the selection criteria would be extremely vague and would eventually boil down to a form of "If XYZ pool makes a press release that they're blocking some transactions then we will block them!!1" which gets mitigated easily by not reporting such things. Some of the bigger pools cannot keep their mouths shut though, so it's mainly going to be clusters of rogue smaller pools that evade detection.

[DannyHamilton]

If rogue mining pools gain more than 51% of the world's bitcoin hashpower, then bitcoin has failed as a concept and we can all move on.  No need to implement any pool-punishing system.  The very fact that Bitcoin is a failed concept will punish them as there will no longer be any fees for them to collect (other than those that WANT to participate in their censor system).

If rogue mining pools are unable to gain more than 51% of the world's bitcoin hashpower, then by restricting which transactions they are willing to accept the rogue mining pools force themselves to choose lower fee transactions (leaving behind the higher fee transactions for the remaining pools).  As such, they ALREADY punish themselves financially AND reward the other pools financially.

Therefore, I really don't see any need to implement a block censoring system.

[pooya87]

Slight problem with this.
If the price only doubles you will only keep the same level of security, and it will not scale to the amount of money "at risk".
So basically you will have  3 trillion in market cap with the same protection in hashrate as you had when they were only 750 billion.
If that number goes through 3 cycles the difference will become x8, how would you feel now if the amount of gear needed to flop the network would not cost 2 billion but just 200 million? 

That's true. Double the price is the minimum to have the same security, and price will always be both the solution and a problem since if it goes up like before (thousands of percentage) the hashrate will continue rising with it and any kind of attack will continue becoming more expensive and if it doesn't go up in 4 years or drops, the hashrate could also decrease and threaten the security.