Post-Mining Era - Susceptible to 51% manipulation attack?

[Inedible]

Once all bitcoins have been mined and only transaction fees are available, I'm guessing there'll be a huge drop in processing power in the network.

Won't this allow someone with enough resources to then attempt the 51% power attack?

I see most miners shutting down their rigs and selling off their Bitcoins once they see there's no profit to be made making this a perfect time to harm Bitcoin.

[1713982064]

1713982064

[1713982064]

1713982064

[Inedible]

That doesn't really address the issue but wouldn't mining on another alternative currency also devalue the one you've just finished mining?

It would certainly confirm the theory that there'd be a huge drop in processing power.

[Meni Rosenfeld]

Yes, this is a problem that has been discussed (for example here) and deserves more discussion. Basically the solution will boil down to:

1. Making the transaction fees high enough to have significant network power
2. Modifying the protocol to allow more security for a given level of processing power

[Inedible]

Oh, thanks - when I did a search on Google, that thread did come up but I hadn't paid enough attention to realise it discussed my question.

[hugolp]

Once all bitcoins have been mined and only transaction fees are available, I'm guessing there'll be a huge drop in processing power in the network.

Won't this allow someone with enough resources to then attempt the 51% power attack?

I see most miners shutting down their rigs and selling off their Bitcoins once they see there's no profit to be made making this a perfect time to harm Bitcoin.

Please, can we stop asking the same question again and again? Specially when they are nonsensical.

There will be no "post mining era". The block reward does not disappear suddenly, it goes down progressively over time. The reward will disappear completely by 2040. If by that time Bitcoin does not produce a lot of activity and therefore a lot of fees, it might be that it was not needed or that the government has managed to keep imposing the monopolly on money. The idea of the initial reward is a way to create and distribute currency. Since Bitcoin is decentralized, there is no central authority that can create new bitcoins and give them to whoever they want.

But back to your question and hopefully shutting up the nonsense, the process of adapting the hashing power (whether is up or down) will be gradual because the reduction on the reward is gradual.

[matsh]

This has been analyzed extremely well by @zooko in the comments to this blog post:

https://plus.google.com/u/0/110295650700104208522/posts/FiLnEia46qM

Executive summary: No need to worry.

[Inedible]

There will be no "post mining era". The block reward does not disappear suddenly, it goes down progressively over time. The reward will disappear completely by 2040. If by that time Bitcoin does not produce a lot of activity and therefore a lot of fees, it might be that it was not needed or that the government has managed to keep imposing the monopolly on money. The idea of the initial reward is a way to create and distribute currency. Since Bitcoin is decentralized, there is no central authority that can create new bitcoins and give them to whoever they want.

But back to your question and hopefully shutting up the nonsense, the process of adapting the hashing power (whether is up or down) will be gradual because the reduction on the reward is gradual.

I've probably misunderstood what you've just said or you've missed my point.

At some point there will be no more Bitcoin to mine.

Regardless of whether that happens in a split second or happens over a period of decades, at some point it will happen.

At this point, I see a smaller proportion of available CPU power being being spent on processing transactions. (e.g. 800GH reducing to 100GH as people decide it's no longer profitable, move onto mine alternative currencies, decide to sell their rigs, etc.)

With a much reduced volume of processing power, isn't Bitcoin vulnerable at this period? The amount of money required to perform this attack now is significantly reduced and could be performed by a determined attacker.

If it helps clarify things, as we near the end of the availability of mineable Bitcoin, people will leave so the amount of hashing power will decrease gradually. Additionally, computing power per dollar will also increase. Overall, this leads to a scenario where mounting an attack will become cheaper and cheaper.

Have I missed something you've said or have I fundamentally misunderstood something? Also, if this has already been explained in another thread, I'm happy to go read it if you happen to know where it is (I did Google for an answer first but nothing seems to quite address the issue - possibly because I've misunderstood something - lol)

[Inedible]

This has been analyzed extremely well by @zooko in the comments to this blog post:

https://plus.google.com/u/0/110295650700104208522/posts/FiLnEia46qM

Executive summary: No need to worry.

That's a fantastic explanation - should be required reading

Thanks matsh!

[hugolp]

There will be no "post mining era". The block reward does not disappear suddenly, it goes down progressively over time. The reward will disappear completely by 2040. If by that time Bitcoin does not produce a lot of activity and therefore a lot of fees, it might be that it was not needed or that the government has managed to keep imposing the monopolly on money. The idea of the initial reward is a way to create and distribute currency. Since Bitcoin is decentralized, there is no central authority that can create new bitcoins and give them to whoever they want.

But back to your question and hopefully shutting up the nonsense, the process of adapting the hashing power (whether is up or down) will be gradual because the reduction on the reward is gradual.

I've probably misunderstood what you've just said or you've missed my point.

At some point there will be no more Bitcoin to mine.

Regardless of whether that happens in a split second or happens over a period of decades, at some point it will happen.

At this point, I see a smaller proportion of available CPU power being being spent on processing transactions. (e.g. 800GH reducing to 100GH as people decide it's no longer profitable, move onto mine alternative currencies, decide to sell their rigs, etc.)

With a much reduced volume of processing power, isn't Bitcoin vulnerable at this period? The amount of money required to perform this attack now is significantly reduced and could be performed by a determined attacker.

If it helps clarify things, as we near the end of the availability of mineable Bitcoin, people will leave so the amount of hashing power will decrease gradually. Additionally, computing power per dollar will also increase. Overall, this leads to a scenario where mounting an attack will become cheaper and cheaper.

Have I missed something you've said or have I fundamentally misunderstood something? Also, if this has already been explained in another thread, I'm happy to go read it if you happen to know where it is (I did Google for an answer first but nothing seems to quite address the issue - possibly because I've misunderstood something - lol)

The profit of the attack is directly related to the value of the bitcoins stolen. Performing a 51% attack is very expensive. So the higher the value of the bitcoins the higher the incentives to attack the network and the higher the resources that can be dedicated to the attack. The same is true in the opposite direction, the lower the value of bitcoins the lower the incentives to attack the network and the lower the resources that can be dedicated to the attack.

But the same is true for mining. The higher the value of the bitcoins the more hashing power you can dedicate to mining (since the bitcoins you earn from fees allows you to buy more electricity and more hardware). Again, the same is true in opposite direction.

So you can see its not a problem.

Also, note that if you perform a 51% attack on bitcoins to steal them, as soon as its known (and it will be quick as people complain) the bitcoins you have stolen would probably go down in price quick and you would end up loosing money.

[Meni Rosenfeld]

Also, note that if you perform a 51% attack on bitcoins to steal them, as soon as its known (and it will be quick as people complain) the bitcoins you have stolen would probably go down in price quick and you would end up loosing money.

... Except if you shorted bitcoins, and making them go down in price was your plan all along.

This is a real problem, please don't try to sweep it under the rug. We don't know how to make people pay transaction fees, we don't know how to prevent miners from including low-fee transactions (a prerequisite to having people pay fees), we don't know what % fee will be considered acceptable by merchants and we don't know what % fee is required to keep the network secure (against all kinds of attack, not just vanilla double-spending).

It's easy to think the network secure now when mining is funded by 37% yearly inflation.

[hugolp]

Also, note that if you perform a 51% attack on bitcoins to steal them, as soon as its known (and it will be quick as people complain) the bitcoins you have stolen would probably go down in price quick and you would end up loosing money.

... Except if you shorted bitcoins, and making them go down in price was your plan all along.

This is a real problem, please don't try to sweep it under the rug. We don't know how to make people pay transaction fees, we don't know how to prevent miners from including low-fee transactions (a prerequisite to having people pay fees), we don't know what % fee will be considered acceptable by merchants and we don't know what % fee is required to keep the network secure (against all kinds of attack, not just vanilla double-spending).

It's easy to think the network secure now when mining is funded by 37% yearly inflation.

I dont think Im explaining myself properly today. Let me repeat the concept again:

Everything else equal, it dos not change anything in regards to the network security or the incentive to mine if the miners get money through inflation or only through fees. The inflation is a method of creating and distributing bitcoins, not a way to finance an initial investment in mining.

The reason why is the same is because if there was no inflation the value of bitcoins would rise faster (everything else equal), and the incentive would be the same. It does not matter if you get 55 bitcoins that can buy X electricity, or you get 18 bitcoins that can buy the same X electricity.

PS: People are paying transaction fees, I dont see how miners including low-fee transactions is a problem (its what makes Bitcoin great, low fees!), and yes, we dont know the rest, nobody does, thats why Bitcoin behaves in a free market way, so it can be handled efficiently.

[Meni Rosenfeld]

Everything else equal, it dos not change anything in regards to the network security or the incentive to mine if the miners get money through inflation or only through fees. The inflation is a method of creating and distributing bitcoins, not a way to finance an initial investment in mining.

I agree, assuming that the total amount given per unit time is the same. What I'm arguing is that fees will be very low (unless a solution is found), so the reward per block will decrease until it is unacceptably low.

The reason why is the same is because if there was no inflation the value of bitcoins would rise faster (everything else equal), and the incentive would be the same. It does not matter if you get 55 bitcoins that can buy X electricity, or you get 18 bitcoins that can buy the same X electricity.

It doesn't matter regarding the incentives of miners, so the hashrate will be the same in both scenarios. But it does matter regarding the incentives of attackers - in the second scenario, Bitcoin is worth thrice as much so the incentive to attack, for whatever reason, is tripled. So the hashrate the attacker is expected to amass will also triple, and the hashrate that is the same in both scenarios will not be enough to secure the network in the second scenario.

Hence, the number of bitcoins awarded per block is a good measure of the security of the network (because the other two variables are the worth of a bitcoin and the incentive to attack, which are more or less proportional).

PS: People are paying transaction fees, I dont see how miners including low-fee transactions is a problem (its what makes Bitcoin great, low fees!), and yes, we dont know the rest, nobody does, thats why Bitcoin behaves in a free market way, so it can be handled efficiently.

The problem is that people won't pay enough transaction fees if they'll be included in a block anyway. As above, by default security is proportional to total fees paid. Of course, the lower the fees can be while keeping the network secure, the better.

On the last two blocks there were 0.4 BTC fees on average. Will this number be higher in the future? Lower? Will the amount be enough to secure the network? I don't know, and I've seen no evidence the answer must be yes.

I believe a solution will be found one way or another if a problem should arise. But I think we need to be prepared in advance. And I don't think we should consider Bitcoin as having proven its technical integrity when the real test is still ahead of us.

[hugolp]

It doesn't matter regarding the incentives of miners, so the hashrate will be the same in both scenarios. But it does matter regarding the incentives of attackers - in the second scenario, Bitcoin is worth thrice as much so the incentive to attack, for whatever reason, is tripled.

This is a good point. I had not though of this.

So the hashrate the attacker is expected to amass will also triple, and the hashrate that is the same in both scenarios will not be enough to secure the network in the second scenario.

I still dont see why, you keep saying it but there is no proof. The money you need to control 51% of the network is huge and you are up to gain not much. I see an attack by some government or something much more probable than by a thieve. If you want to make money, you could just spend the money in mining.

Hence, the number of bitcoins awarded per block is a good measure of the security of the network (because the other two variables are the worth of a bitcoin and the incentive to attack, which are more or less proportional).

True.

On the last two blocks there were 0.4 BTC fees on average.

Fees were zero at the beggining and have been increasing. Its spected that this will continue as more transactions start happening in the Bitcoin network.

Will have to see, but I doubt a 50% attack is economically viable.

[Ten98]

This is a real problem, please don't try to sweep it under the rug. We don't know how to make people pay transaction fees, we don't know how to prevent miners from including low-fee transactions (a prerequisite to having people pay fees), we don't know what % fee will be considered acceptable by merchants and we don't know what % fee is required to keep the network secure (against all kinds of attack, not just vanilla double-spending).

All true, but we do have over 20 years to think of these things. Please don't worry about it right now.

[Inedible]

Also, note that if you perform a 51% attack on bitcoins to steal them, as soon as its known (and it will be quick as people complain) the bitcoins you have stolen would probably go down in price quick and you would end up loosing money.

This suggests that it's trivial to work out what was stolen.

Is it therefore possible to return the 'stolen' coins to their rightful owners or are they just lost for good?

When you say those coins will lose their value, is there an implication that they might be worth a fraction of their original value or will it simply be lost completely (as above)?

we don't know how to prevent miners from including low-fee transactions

Let's say it's not possible to separate out transactions with fees from those without. Does this destroy the basis for transaction fees?

All true, but we do have over 20 years to think of these things. Please don't worry about it right now.

I'd rather know that the time and energy I invest into Bitcoin now will be worth something in 20 years time.

Not worrying about it now seems pretty short sighted to me.

[Ten98]

All true, but we do have over 20 years to think of these things. Please don't worry about it right now.

I'd rather know that the time and energy I invest into Bitcoin now will be worth something in 20 years time.

Not worrying about it now seems pretty short sighted to me.

Well of course you can think about these things if you wish, but please realise that there is very little you can achieve by concerning yourself with something which is not due to happen for the next 2 decades.

20 years is a ridiculously long time in computer years, think about the computer you used in 1991 compared to what you have now, if you were even born then

Who knows what new technology and computational power will become available, maybe in 10 years finding one Bitcoin block will take an entire moon of silicon, or a single quantum singularity processor. Maybe in 15 years the internet as we know it today won't exist, and the whole concept of Bitcon will have had to be redesigned. You can't predict how, but rest assured things will change drastically between now and then.

There's no point worrying about it now, because any solution you invent now will be obsolete by next year, next month, even next week. So live your life, try to hang on for the ride and stay sane

[hugolp]

This suggests that it's trivial to work out what was stolen.

Is it therefore possible to return the 'stolen' coins to their rightful owners or are they just lost for good?

When you say those coins will lose their value, is there an implication that they might be worth a fraction of their original value or will it simply be lost completely (as above)?

Well, its not trivial, but in very evident cases it could be reverted. The problem really is, if something like this happened, Bitcoin credibility would suffer a big hit.

[bitfoo]

Well of course you can think about these things if you wish, but please realise that there is very little you can achieve by concerning yourself with something which is not due to happen for the next 2 decades.

20 years is a ridiculously long time in computer years, think about the computer you used in 1991 compared to what you have now, if you were even born then

Who knows what new technology and computational power will become available, maybe in 10 years finding one Bitcoin block will take an entire moon of silicon, or a single quantum singularity processor. Maybe in 15 years the internet as we know it today won't exist, and the whole concept of Bitcon will have had to be redesigned. You can't predict how, but rest assured things will change drastically between now and then.

There's no point worrying about it now, because any solution you invent now will be obsolete by next year, next month, even next week. So live your life, try to hang on for the ride and stay sane

"Hey, it's the year 1980, lets use just two digits to represent the year on a computer. After all in 2 decades, all our hardware and software will be obsolete anyway..." - a mistake that cost roughly 300 billion dollars.

Just saying, I think it's very important to think about these things sooner rather than later.

[322i0n]

what would be the problem with bitcoin devaluing while bitcoin2.0 became more valuable (is there a word that means the opposite of devalued?)