All The Fees

[Revalin]

When you use it online it's run as a credit card with the usual credit card fees (2% or so).

[1713931090]

1713931090

[1713931090]

1713931090

[1713931090]

1713931090

[Killdozer]

The banks are evil

Yeah, very evil, they have given you the ability to use a card online at all, to pay over the internet, to have nice control over your money and not having to keep all the cash in your house. They also have given a lot of businesses a chance to grow and make progress, by giving loans to them. Very, very evil indeed.

[runeks]

The developers have acknowledged that fee size determination in Bitcoin is broken. It could work though, the infrastructure is there, but we need a way to figure out how to communicate fee size between payers and miners.

The reason why I love bitcoin is the lack of fees!  I now do most of my transactions fee-less.  I am buying and selling directly with others and have far less in the way of BTC-USD need.  When it happens it is 25 cents Dwolla and .5 % exchange fees, which is pretty low when you do $300 worth. 

Don't get too attached to that. Fees will have to be introduced at some point (they already are in some cases). Mining rigs and the electricity they consume is not free.

[PrintCoins]

The developers have acknowledged that fee size determination in Bitcoin is broken. It could work though, the infrastructure is there, but we need a way to figure out how to communicate fee size between payers and miners.

The reason why I love bitcoin is the lack of fees!  I now do most of my transactions fee-less.  I am buying and selling directly with others and have far less in the way of BTC-USD need.  When it happens it is 25 cents Dwolla and .5 % exchange fees, which is pretty low when you do $300 worth. 

Don't get too attached to that. Fees will have to be introduced at some point (they already are in some cases). Mining rigs and the electricity they consume is not free.

The amount of energy consumed grows to match the price of bitcoins. Energy is basically the margin cost of mining, and bitcoins is the revenue. Miners will start buying more cards to mine more when the payout is higher, eventually pushing them to an equilibrium where everyone starts complaining that it is not profitable any more.

If the revenue were to drop to 1BTC per block, miners would start unplugging their machines to cut costs, and the difficulty would drop to match this change. Once you win a block, the cost to add a transaction is nearly nil, and to ignore low value transactions will mean that the next miner will be able to take them and pocket them.

Keep in mind that more's law also applies to miners, and 2 years from now all that mining equipment will be obsolete, so unless miners keep buying equipment, they will be replace by a new set of miners who will get twice the hash rate at half the cost (energy consumption probably won't drop).

Unless some miners get 50% of the hashing power (or some cartel forms), market laws will apply, and the cost to add a transaction will be the limit to the actual required fee. All that power consumption and fancy graphics cards will end up being sunk cost at the point of transaction recording, and thus be irrelevant.

[cunicula]

The developers have acknowledged that fee size determination in Bitcoin is broken. It could work though, the infrastructure is there, but we need a way to figure out how to communicate fee size between payers and miners.

The reason why I love bitcoin is the lack of fees!  I now do most of my transactions fee-less.  I am buying and selling directly with others and have far less in the way of BTC-USD need.  When it happens it is 25 cents Dwolla and .5 % exchange fees, which is pretty low when you do $300 worth. 

Don't get too attached to that. Fees will have to be introduced at some point (they already are in some cases). Mining rigs and the electricity they consume is not free.

The amount of energy consumed grows to match the price of bitcoins. Energy is basically the margin cost of mining, and bitcoins is the revenue. Miners will start buying more cards to mine more when the payout is higher, eventually pushing them to an equilibrium where everyone starts complaining that it is not profitable any more.

If the revenue were to drop to 1BTC per block, miners would start unplugging their machines to cut costs, and the difficulty would drop to match this change. Once you win a block, the cost to add a transaction is nearly nil, and to ignore low value transactions will mean that the next miner will be able to take them and pocket them.

Keep in mind that more's law also applies to miners, and 2 years from now all that mining equipment will be obsolete, so unless miners keep buying equipment, they will be replace by a new set of miners who will get twice the hash rate at half the cost (energy consumption probably won't drop).

Unless some miners get 50% of the hashing power (or some cartel forms), market laws will apply, and the cost to add a transaction will be the limit to the actual required fee. All that power consumption and fancy graphics cards will end up being sunk cost at the point of transaction recording, and thus be irrelevant.

Exactly, the whole system is set up to perpetually waste vast amounts of money for no purpose. Brilliant.

[Revalin]

The purpose is to protect against double-spend attacks.  Do you know a better way?

[chrisrico]

The purpose is to protect against double-spend attacks.  Do you know a better way?

After every payment, you send your digital tokens to Paul Krugman who destroys them, creates new ones, and sends them back to you. If it's a double spend, Krugman slaps you in the face and you in turn slap your customer in the face. Voila!

[DeathAndTaxes]

The amount of energy consumed grows to match the price of bitcoins. Energy is basically the margin cost of mining, and bitcoins is the revenue. Miners will start buying more cards to mine more when the payout is higher, eventually pushing them to an equilibrium where everyone starts complaining that it is not profitable any more.

If the revenue were to drop to 1BTC per block, miners would start unplugging their machines to cut costs, and the difficulty would drop to match this change. Once you win a block, the cost to add a transaction is nearly nil, and to ignore low value transactions will mean that the next miner will be able to take them and pocket them.

So your solution is "we don't need fees" because when block reward falls 90% then hashing power will fall 90%.  Except if hashing power falls 90% then Bitcoin is vulnerable to 51% attack.

It is possible for Bitcoin to have low fees.  Lower than ACH, lower than VISA, lower than WU but something needs to pay for the network.  If the network is vulnerable to 51% attack then the value of Bitcoin will decline relative to that risk.  If it it is trivially easy/cheap to 51% attack the network then the value of Bitcoin is nothing.

Keep in mind that more's law also applies to miners, and 2 years from now all that mining equipment will be obsolete, so unless miners keep buying equipment, they will be replace by a new set of miners who will get twice the hash rate at half the cost (energy consumption probably won't drop).

Which provides no security.  New attackers will have access to the same hashing power.  Thus in 2 years 10TH is more like only 5TH now.  If the network is 20TH it is only roughly as secure as it is today.  The nominal network rate is irrelevant.  What matter is the cost of an attack.  Today 10TH is prohibitively expensive.  In 60 years the Iphone 87 may have 20TH/s of computational power.   Moore's law can't make Bitcoin "stronger" because both defenders and attackers have access to the same powerful new gear.

Unless some miners get 50% of the hashing power (or some cartel forms), market laws will apply, and the cost to add a transaction will be the limit to the actual required fee. All that power consumption and fancy graphics cards will end up being sunk cost at the point of transaction recording, and thus be irrelevant.

Didn't you say that due to more efficient gear miners will be forced to either turn off rigs or upgrade them.  Equipment has a finite lifespan both physically and economically.  The equipment cost can be computed per hash by block by taking blocks of effective lifespan divided by the equipment cost adjusted for the time value of money.

[cunicula]

I have. Just make votes proportional to committed long-term ownership of coins instead of energy wastage. This is called "proof of stake". Search for it. I'm not going through details here because it would just attract trolls. I will only respond to high-level theoretical objections.

[cbeast]

I have. Just make votes proportional to committed long-term ownership of coins instead of energy wastage. This is called "proof of stake". Search for it. I'm not going through details here because it would just attract trolls. I will only respond to high-level theoretical objections.

Help us out here. Can you cite a peer-reviewed journal?

[FreeMoney]

I have. Just make votes proportional to committed long-term ownership of coins instead of energy wastage. This is called "proof of stake". Search for it. I'm not going through details here because it would just attract trolls. I will only respond to high-level theoretical objections.

You mean if there were two miners one having 10BTC and one having 5BTC the 10BTC miner would have 2x the chance of their block being accepted? What is the mechanism for actually deciding? Still hashing, but accept lower difficulty blocks if the block is signed with a private key that holds coins?

[Revalin]

"proof of stake".

Just those words and I've got it. This might be a reasonable solution to my fears about long term sustainability of Bitcoin.  It creates a dozen new problems too, but you've got my gears turning. Thanks.

[cbeast]

"proof of stake".

Just those words and I've got it. This might be a reasonable solution to my fears about long term sustainability of Bitcoin.  It creates a dozen new problems too, but you've got my gears turning. Thanks.

The word proof is rarely used in science. It is used in theory, such as maths. Unfortunately, even Bitcoin uses the term "proof of work" when really it is only an historical block chain, which is really, really hard (but not completely impossible) to fake.

[Revalin]

which is really, really hard (but not completely impossible) to fake

I'm not aware of any way to fake a SHA256 proof of work with less work than simply brute forcing the hash.  In essence, faking it is a proof of work in itself.

On the other hand proof of stake is really only a proof of either stake or absurdly good luck.  Assuming the crypto isn't broken, though, it's a proof for any practical purpose.

[cunicula]

I have. Just make votes proportional to committed long-term ownership of coins instead of energy wastage. This is called "proof of stake". Search for it. I'm not going through details here because it would just attract trolls. I will only respond to high-level theoretical objections.

You mean if there were two miners one having 10BTC and one having 5BTC the 10BTC miner would have 2x the chance of their block being accepted? What is the mechanism for actually deciding? Still hashing, but accept lower difficulty blocks if the block is signed with a private key that holds coins?

That is it, but with rules that allow for both small- and large-scale operations. The rule you describe would degenerate in to one gigantic hashing pool operated by whoever holds the most BTC.

[cbeast]

I have one question about proof of stake. Can stake be acquired by force?

[cunicula]

I have. Just make votes proportional to committed long-term ownership of coins instead of energy wastage. This is called "proof of stake". Search for it. I'm not going through details here because it would just attract trolls. I will only respond to high-level theoretical objections.

So someone who has mined since late 2009 on has more say in the decisions than someone who joined the network today? Don't we already have enough concerns over the early adopter advantage?

If I move those coins to a new address, do I lose all of my "proof of stake".

Could you direct me to a more detailed explanation?

Don't you think this already exists to some extent simply because a large holder would be concerned over the state of the network, and can influence it with their wallet, regardless of whether they mine or not?

1) Sorry, I would prefer not to discuss fairness, etc. It is too subjective. Briefly, anyone can gain more votes by investing. Or complete control by buying up a majority. This is analogous to a publically-held joint stock company.

2) What happens when you do X depends on the specific rule set chosen. The concept of "proof of stake" is general. I don't want to answer low-level questions that might have different answers depending on the implementation. Please feel free to search the forums for "proof of stake" to learn more details. Details are important, but this is not the time and place for them.

3) Sure, large holders are already extremely concerned with the state of the network. The idea is to exploit their concern to generate "proof" for txn verification purposes. Currently, electricity waste generates "proof". Use of electricity makes the payment system excessively costly. Costs are reflected in either a) currency generation or b) txn fees. Generation and fees compensate electricity wasters for their expenditures. An alternative is to use commitments to hold coins for a long periods as "proof". Since people will hold coins long-term even without an additional inducement, you can get away with paying them very small rewards. By contrast, electricity wasters need to be paid a larger amount. Under the proof-of-work regime, extravagant waste is the basis for security. Under proof-of-stake, the intrinsic interest of long-term investors is the basis of security. If someone has committed to a substantial long-term investment in bitcoin, it is not necessary to demand further "proof" of their benevolent intentions.  

[cunicula]

I have one question about proof of stake. Can stake be acquired by force?

I recommend ignoring questions and comments made by "cbeast". He does not have constructive intentions.

[cbeast]

I have one question about proof of stake. Can stake be acquired by force?

I recommend ignoring questions and comments made by "cbeast". He does not have constructive intentions.

Post by: cunicula on July 19, 2011, 12:57:34 PM
I would propose a hybrid sytem which uses proof of work and proof of stake. Transactions would still be verified using mining but people would have to pay for mining rights.

Suppose that a 100 btc security deposit had to be placed for each bitcoin mined per year. Miners could send 100 btc to a time-locked escrow transaction. The escrowed coins could be sent to other addresses as escrowed coins. Anyone holding the private key to this time-locked transaction could mine up to one bitcoin per year. The mined BTC can be spent immediately. Miners could request a withdrawal of the escrowed btc, but the request would be subject to a 6 month time lag. During the 6 month lag, no minng would be allowed and the escrowed coins could not be sent.

Where do you get the 100 btc? Simple question.

[cunicula]

Where do you get the 100 btc? Simple question.

Many possibilities.  Here is one:

Allow for high difficulty mining by individuals without assets. At the onset, no one has assets so everyone has to mine this way. Over time people accumulate coins. Once this happens, it becomes practically impossible for stakeless miners to compete (e.g. like mining with a CPU today). After the initial period, "mining" entry requires purchase of coins on an exchange.