[If tx limit is removed] Disturbingly low future difficulty equilibrium

[Gavin Andresen]

I know I'm not using an actual argument in this post. Feels strange to do that, too. But I think it's safe to say that a solid consensus should not produce this outcome. There is no consensus. There is no generally accepted operating model for Bitcoin after minting becomes negligible. We better find one.

Why?  Thats years and years away.  I don't think any of us can predict exactly what is going to happen 10 or 15 or 20 years from now; I don't think it is possible to get a consensus now because there has never been a system like bitcoin before, so predicting how merchants, miners, and users will interact in 10 years seems to me to be impossible.

So:  do you think this supposed problem will happen all at once?  Or will it happen slowly over time?  If you think it will happen all at once, would there be any warning signs?  If it is a problem that we can clearly see coming, then there will be time to react.

I'm much more worried on the problems I see coming in the next year or two or three-- bitcoin-specific viruses and trojans, poorly coded bitcoin web services, and maybe bitcoin service operators getting charged with financial crimes that they didn't know they were violating.

[1715503978]

1715503978

[stillfire]

• Mining costs will go down, so it will be profitable again [unless the transaction price falls again... *sighs] (stillfire)

Hrm, if I said that, I was not aware of it. I said merchants will pay for mining, even if it's at a loss versus 'verification cost', and even if it would be cheaper to wait for someone else to do it, just because waiting is not free.

No, this does not prevent difficulty from falling as profits drop, but I was specifically arguing against those saying the network would stop due to near zero profits from mining in an unlimited block size world and a 'staring contest' between merchants to see who'd blink first and burn money by turning on their rigs.

[MoonShadow]

• Mining costs will go down, so it will be profitable again [unless the transaction price falls again... *sighs] (stillfire)

Hrm, if I said that, I was not aware of it. I said merchants will pay for mining, even if it's at a loss versus 'verification cost', and even if it would be cheaper to wait for someone else to do it, just because waiting is not free.

No, this does not prevent difficulty from falling as profits drop, but I was specifically arguing against those saying the network would stop due to near zero profits from mining in an unlimited block size world and a 'staring contest' between merchants to see who'd blink first and burn money by turning on their rigs.

Exactly.  And this is why I brought up Wal-Mart and it's competitors.  If Wal-mart accepts Bitcoin, it has a strong incentive to directly participate in the security of the network.  And then so does all it's competitors.  And the reverse is also generally true.  The mom&pop  vendors may not feel motivated to participate in the mining directly, but they are going to want to see their transactions processed in a timely manner as well.

[caveden]

The side denying a problem
(...)

• Transaction fees mostly payed by banks [I don't understand, cartel again?] (caveden)

Why have you put me there?
Transactions fees being mostly payed by banks will only made them lower, as most people will make their transfers without the block chain.

I do think the block size limit must create artificial scarcity to push up the transactions fees. My comment saying that we are already paying a lot to miners in the form of inflation was exactly to argue with those who don't want to pay such fees.

I defend an automatic adjustment of the maximum block size exactly for creating such artificial scarcity. Something like, at each X blocks (maybe at the same time we have a difficult adjustment), the maximum block size is set to be 110% of the average size of the last X blocks.

On the other hand, Gavin has a point when he says that you seems too much desperate with an issue that will not manifest itself in the next 15 or 20 years at least. It's true that we'd better worry with it before going mainstream, when such a change would need a blockchain split, but there is some time yet and maybe other more important priorities to be dealt with.

[Meni Rosenfeld]

The tragedy of the commons is a dilemma arising from the situation in which multiple individuals, acting independently and rationally consulting their own self-interest, will ultimately deplete a shared limited resource even when it is clear that it is not in anyone's long-term interest for this to happen.

What is the shared limited resource to be depleted in the case of mining?

The willingness of transactioneers to pay fees. When you include transactions with low fees in a block, you consume this resource.

But I think it's safe to say that a solid consensus should not produce this outcome. There is no consensus. There is no generally accepted operating model for Bitcoin after minting becomes negligible. We better find one.

What are you talking about? This is a free market, without regulation or central planning. Bitcoin cannot work, post generation subsidy, unless it is centrally planned?

We are talking about how to design the Bitcoin protocol to ensure it achieves a healthy free market equilibrium. The same way Satoshi designed some other pieces of the protocol.

Exactly.  And this is why I brought up Wal-Mart and it's competitors.  If Wal-mart accepts Bitcoin, it has a strong incentive to directly participate in the security of the network.  And then so does all it's competitors.  And the reverse is also generally true.  The mom&pop  vendors may not feel motivated to participate in the mining directly, but they are going to want to see their transactions processed in a timely manner as well.

Tragedy of the commons again. Wal-mart also has incentive to sit it out while others subsidize the mining. The only way I see to make Wal-mart pay its share is with some sort of union which will defeat Bitcoin's aspirations of being decentralized.

On the other hand, Gavin has a point when he says that you seems too much desperate with an issue that will not manifest itself in the next 15 or 20 years at least. It's true that we'd better worry with it before going mainstream, when such a change would need a blockchain split, but there is some time yet and maybe other more important priorities to be dealt with.

The Bitcoin community is fairly large, there's no reason some participants shouldn't debate one problem while others debate another. Also, having a plan for what will happen 10 years forward will encourage people to join now.

[Vandroiy]

Phew... please try to cut me some slack on your exact formulation, or provide a corrected short version. It's hard to write in exactly the style someone might want.


@gavinandresen:

I think it is important that Bitcoin protocol becomes independent as soon as possible. You're doing a great job, so no offense really, but there is no guarantee anyone will be able to control the project once it takes a truly macroscopic scale. At that point, trying to change the protocol can generate disputes, splits, loss of trust, central authorities, so many unforeseen complications could happen! If the protocol is carved in stone and commonly accepted as "finished for all eternity", we might be able to avoid such problems.

An attack would be likely to happen within the months following one of those sharp reward drops, where minting is cut in half. It might happen that half the miners leave at once when this happens.

I agree that it's unlikely to be a problem until minting is down to 12.5 BTC or less, and that the immediate dangers lie elsewhere. But please don't neglect the future the way it's common in politics. The amount of confusion if the protocol is disputed is not to be underestimated. It's mainly psychology and politics, but we should know better than to let a financial system be controlled by such forces. Imagine nodes being mainly run by miners, and they all agree they take the easy "problem solution" and re-start minting the way Timothy Lee suggested! A catastrophe in which miners can damage the entire system to their advantage. There might be many more such scenarios, maybe infinitely many, if major changes in the protocol might happen at any time. We have no safe way out of a protocol dispute!


@stillfire:

Sorry. I added you with the merchants' theory.

Frankly, I think the original statement is wrong no matter how I interpret it, so I was having difficulty re-phrasing it. You say that some merchant would rather keep a rig running than wait for his competitor to do the same... the best more general re-phrase I can think of is either "Tragedy of the commons does not apply to merchants having to pay for the security of all merchants without being forced to" or "I don't believe in Tragedy of the Commons". Both are false, I can't help that, unless there's a cartel or everybody is a convinced communist.

Try to look at the bigger picture -- either why the statement is wrong, or that even with the re-phrased statement, you don't exactly have a model that's solid or generally agreed on.


@caveden:

Oh, sorry, I did indeed fail to understand what you said. Should I put you on the other side or with the "agree, but wait first" guys?

Please keep in mind what might happen when a later try to change the protocol starts a collusion war! If bitcoin becomes a currency, we can expect bribes on the order of dozens or hundreds of millions of dollars to push a protocol change one way or the other. Better act while the system is still controlled by sane people.


@Holy-Fire: Thanks for continuing to answer so much in my place. (Unless I state otherwise, I agree with posts by Holy-Fire.)

[da2ce7]

We don't' know what will happen without a proper game theory analysis of the system... You never know, Satoshi may have been a game theory expert and knows that the miners can set their own max block size and the system will still maintain a reasonable security.

Until we (or somebody) puts the effort in for a formal game theory analysis, we will not know, and this is pointless speculation.

[gim]

The side seeing a problem (if transfer limits are removed, thus epii is on both sides now.)

+1
(see http://bitcointalk.org/index.php?topic=7496.0)

[MoonShadow]

Exactly.  And this is why I brought up Wal-Mart and it's competitors.  If Wal-mart accepts Bitcoin, it has a strong incentive to directly participate in the security of the network.  And then so does all it's competitors.  And the reverse is also generally true.  The mom&pop  vendors may not feel motivated to participate in the mining directly, but they are going to want to see their transactions processed in a timely manner as well.

Tragedy of the commons again. Wal-mart also has incentive to sit it out while others subsidize the mining. The only way I see to make Wal-mart pay its share is with some sort of union which will defeat Bitcoin's aspirations of being decentralized.

There is a commons issue, but it's not a 'Tragedy of the Commons' for several reasons repeatedly pointed out to yourself.  In the case of Wal-Mart and it's competitors, or any other large retailer and it's direct competitors, Wal-mart does benefit from the generic security of the blockchain that it's competitors contribute to, and vice versa; but what motivates Wal-Mart to contribute directly (by running or sponsoring mining clients themselves) or indirectly (by paying transactions fees to motivate third party miners) because Wal-Mart's competitors can ignore Wal-Mart's transactions whether they are fee paying or not.  Wal-Mart can do the same to it's competitors.  Granted, it wouldn't be easy to isolate the transactions produced by Wal-Mart's competitors from those of the network at-large, but if anyone has the ability to do it, Wal-Mart does.  I can think of a number of ways that a competitor's addresses could be automaticly identified, with varying degrees of success.  Even if this were not possible, if Wal-Mart is running their own miners then they can simply ignore any transaction below a certain fee and still have an economic incentive to run miners in order to process their own transactions for free.  Wal-Mart's competitors have the same incentives.

This is not a 'cartel' problem because Wal-MArt isn't a cartel itself, and the fees that they would charge others for transactions would exist primarily to hamper the successes of it's competitors, not collude to increase the price.  Independent miners would still be able to set their transactions fees lower than Wal-Mart, collecting all the transaction fees that Wal-Mart and it's competitors forgo, but this would not really undercut Wal-MArt because they are not in the game for direct profit but to defend their business model.  Wal-Mart alone has enough transactions per second to justify running some major miners just to process their own transactions; all others aside.

And they are just an example of how this would work.  This is an example of any market player who has an outsized interest in the system, there are many others that would have similar motivations to process transactions beyond direct profit.  Another such example is the "Bitcoin Bank", which has a motivation to sponsor miners, directly or indirectly, in the same fashion that present banks have the motivation to spend a million dollars on a bank safe.  In the case of a bank, however, direct profit is also a motivation.

[stillfire]

so many unforeseen complications could happen! If the protocol is carved in stone and commonly accepted as "finished for all eternity", we might be able to avoid such problems.

I do agree with this. If anything is to be changed, it better be changed early. However, that requires there to actually be a problem to fix first.

I can think of is either "Tragedy of the commons does not apply to merchants having to pay for the security of all merchants without being forced to" or "I don't believe in Tragedy of the Commons". Both are false, I can't help that, unless there's a cartel or everybody is a convinced communist.

In a "tragedy of the commons" the self interested individual actions deplete or destroy the commons even as the group loses. This is not so in the scenario I have described. The self interested, for profit actions of the merchants actually replenish the commons.

Yes, they would perhaps had been more pleased had someone else paid, but my point is that trying to maximise your Bitcoin savings versus other merchants infinitely has a price much higher than just diving in. At some point you'd be working on savings fractions of a penny and the cost of doing so (the staring contest, loss of business) is measured in dollars.

The merchants actually pay for the existence of the commons, like a park before their storefront, rather than just the use of them. If they did not they would lose more profit than the small sums they pay for commons maintenance. They will mine at a loss because as a whole it makes their business more profitable.

[Vandroiy]

In a "tragedy of the commons" the self interested individual actions deplete or destroy the commons even as the group loses. This is not so in the scenario I have described. The self interested, for profit actions of the merchants actually replenish the commons.

Yes, they would perhaps had been more pleased had someone else paid, but my point is that trying to maximise your Bitcoin savings versus other merchants infinitely has a price much higher than just diving in. At some point you'd be working on savings fractions of a penny and the cost of doing so (the staring contest, loss of business) is measured in dollars.

The merchants actually pay for the existence of the commons, like a park before their storefront, rather than just the use of them. If they did not they would lose more profit than the small sums they pay for commons maintenance. They will mine at a loss because as a whole it makes their business more profitable.

The few biggest merchants forced to mine at a loss? What makes you think they won't rather quit Bitcoin? Whether they do one or the other is merely a question of how lossy mining is, and that's dependent on attacker size, which is practically an external factor. You risk that, depending on some external quantity, exiting Bitcoin is an intelligent decision. Very dangerous.

That's what I meant with the convinced communist solution... works only if you can force people to not evade into a different setup. Which we can't, not now and not when Bitcoin is the super world currency; people can always quit.

[Pieter Wuille]

While re-reading this thread, I realized something. I didn't read it completely, so it may already have been implied by others, though.

When I pay a transaction fee, I pay for a level of security, like explained by Mike. If I don't fear reversal of my transaction, I have little incentive for getting it in the blockchain. However, security here consists of several things:

• 1. Gettting your transaction included in the block chain
• 2. Getting the chain your transaction is in to be extended
• 3. Getting this chain to grow as fast as possible

Your fee goes to the one who puts your transaction in a block, so you pay for #1. However, you don't pay any reward to those who extend that chain. The miner who put your transaction is a block has all incentive to keep mining on this chain, since it gave him a reward that would be reverted by a split, but this is not true for other miners. So you only pay partly for #2. Finally, you don't pay for #3 at all - as explained here earlier, one possibility is that miners turn off their hardware if not enough fees are accumulated. However, every second that mining hardware is off, gives attackers one second's worth of catching up.

I feel that this is at least part of the problem: we only pay for inclusion in a block, not for getting it buried in the block chain. The cost for burying it is carried by all who want additional transactions processed.

I wonder if this could be solved by modifiying the fee system. A miner cannot take all fees for himself, only N% of it (for example, 5% of it). Miners are however allowed to also take the difference between the previous blocks' input+coinbase and outputs as additional fee. This causes the actual fee paid by someone issuing a transaction to be distributed in an exponentially decreasing way over the miners of the first many blocks to come.

[MoonShadow]

This might be possible, but I don't see a real gain to doing this.

[gim]

I wonder if this could be solved by modifiying the fee system. A miner cannot take all fees for himself, only N% of it (for example, 5% of it). Miners are however allowed to also take the difference between the previous blocks' input+coinbase and outputs as additional fee. This causes the actual fee paid by someone issuing a transaction to be distributed in an exponentially decreasing way over the miners of the first many blocks to come.

I've also been thinking about that, but...
 - It won't completely solve the problem because people issuing many transactions still pay for the security of the whole network. It might be preferable for them to use a chain secured by inflation as they mostly don't care about inflation.
 - From the technical point of view, progressive dispatching of fees would add a huge amount of complexity to the current chain format.

[Mike Hearn]

This topic has also been discussed in this thread:

   http://forum.bitcoin.org/index.php?topic=8363.0

[Pieter Wuille]

I wonder if this could be solved by modifiying the fee system. A miner cannot take all fees for himself, only N% of it (for example, 5% of it). Miners are however allowed to also take the difference between the previous blocks' input+coinbase and outputs as additional fee. This causes the actual fee paid by someone issuing a transaction to be distributed in an exponentially decreasing way over the miners of the first many blocks to come.

I've also been thinking about that, but...
 - It won't completely solve the problem because people issuing many transactions still pay for the security of the whole network. It might be preferable for them to use a chain secured by inflation as they mostly don't care about inflation.

As they should - they're the one who need it.

- From the technical point of view, progressive dispatching of fees would add a huge amount of complexity to the current chain format.

It wasn't a serious suggestion, but just trying to find enforceable solutions to the presented problem. On the other hand, technically it's quite simple to implement actually. The difficulty would be getting the changed rule accepted by the network.

[gim]

As they should - they're the one who need it.

Everyone needs some amount of security.
If the mining rate falls too much, and an attacker has let's say twice the global hashing power...
then it's completely impossible to receive coins safely. Waiting for hundreds or thousands or more confirmations is just pointless.

So, the global value accorded to a given chain will fall very quickly if its hashing rate falls under a certain threshold.

[asdf]

50% of the fee goes to the miner that solves the block. 50% goes to the miner that solves the next block. This way, txfees approach the cost including a transaction x2.

The miner covers his tx inclusion costs by including transactions with fees that are at least double the cost of inclusion and the profits from the half the fees collected in the previous block.

problem solved?

[MoonShadow]

50% of the fee goes to the miner that solves the block. 50% goes to the miner that solves the next block. This way, txfees approach the cost including a transaction x2.

The miner covers his tx inclusion costs by including transactions with fees that are at least double the cost of inclusion and the profits from the half the fees collected in the previous block.

problem solved?

What would it change, really?  Maybe if the transaction fees are shared with a block that is 2016 blocks behind it, so that it's in a relative spot in the next difficulty retarget.

[MoonShadow]

I've started a new thread related to this one in economics section.