Reduce ETH issuance before proof-of-stake

lightuponlight commented 7 years ago

Coin vote is now underway on EIP-186:

As an FYI, Carbonvote set this up on their own. I think they did a very nice job and thank them for serving the Ethereum community in this way.

Please, whatever your opinion is on EIP-186, let your voice be heard by participating in the coin vote!

Modified EIP proposal with changes, followed by the original EIP proposal. These changes are based on feedback provided on Reddit and Github regarding this EIP regarding the issuance reduction schedule and minimum issuance level under proof-of-work.

  
  EIP: 186
  Title: Reduce ETH issuance before proof-of-stake
  Author: Matthew Light | lightuponlight9@gmail.com
  Discussions-To: lightuponlight9@gmail.com
  Status: Draft
  Type: Standards Track
  Created: 2016-12-21
  Replaces:  N/A
  Superseded-By: N/A
  Resolution: N/A

Abstract A reduction in the issuance of Ether (ETH) is very likely to be price-supportive and lead to increasing investments in the platform and to help ward off speculative attacks on the value of Ether by promoters of competing platforms who offer, or plan to offer, reduced token inflation rates. This proposal is based on a discussion on Reddit about ETH token issuance rates that many people participated in, including Ethereum Foundation member Vitalik Buterin.

In the current version of Ethereum as deployed on the network, there is already a coded schedule of reduced issuance per time period due to the "ice age" (an increase in difficulty designed to incent the adoption of hard forks). Normally, the "ice ages" in Ethereum are postponed by hard forks and the issuance is kept at 5 ETH per block. This EIP suggests that instead, the reduction in issuance from the "ice age" be implemented in a stepwise fashion within the protocol rules on a set block number schedule and with a defined lower bound (such as the "tail emission" in Monero). The proposed block reward reduction timetable has been pushed farther into the future, and the terminal issuance level has been increased in light of discussions on Reddit and github regarding this EIP.

Motivation There is widespread interest within in the Ethereum community to reduce the current rate of ETH issuance. Uncertainty about the future total ETH token supply is a significant factor in reducing the market value of ETH which has negative externalities on the Ethereum ecosystem, by reducing the capital available by current investors to make investments in new Ethereum-based projects and by reducing the Ethereum Foundation's funds available for spending on salaries. Reducing ETH issuance in advance of proof-of-stake would provide a measure of reassurance to investors that their holdings of ETH will be diluted to a much lower degree. It would be helpful in the cryptoplatform marketplace for Ethereum to reduce issuance given that various competing platforms have or are planning lower issuance levels, such as Ethereum Classic. Additionally, reducing block rewards in the near future helps reduce a sense of pressure on the Ethereum network architects and developers to rush a proof-of-stake implementation and try to deliver it more quickly than would be prudent. The terminal reward value of 2 ETH / block was chosen to reduce token supply inflation to around 4% annually, gradually reducing over time until proof-of-stake is launched. This is in line with the current Bitcoin token inflation rate of approximately 4%. In comparison, current proof-of-work ETH token inflation rate is a bit less than 13% per year. A reduction of token emission under proof-of-work could lead to a reduction in network security. However, a plausible improvement in the exchange rate of ETH because of a reduction in issuance will lead to an increase in network security. It is unclear how much the ETH token price might increase due to a lowered issuance rate, but the positive response of Bitcoin price to issuance reductions ("halvening") is encouraging.

Specification The current issuance level for ETH is 5 ETH per block, with various kinds of uncle rewards also provided. This EIP proposes that no change to the relationship between block rewards and uncle rewards be made, but instead that all uncle and uncle inclusion rewards be downsized proportionally with the provided block reward schedule. The proposed reduction in issuance is an approximation for providing stepwise issuance reduction approximately matching the "ice age" coin issuance reduction already built into the current specification of the Ethereum software. Vitalik Buterin gave an estimate of this reduction on Reddit: https://www.reddit.com/r/ethereum/comments/5izcf5/lets_talk_about_the_projected_coin_supply_over/dbc66rd/ Based on the ice age blocktime lengthening data Vitalik provided, the following reduction of block reward schedule is proposed as a reduction in issuance that is smaller than the reduction already scheduled (via longer blocktimes caused by the "ice age" PoW difficulty bomb). The reduction schedule is also significantly delayed over the "ice age" reduction already contained in the code: Block 3700000 - Block reward reduced to 4 ETH / block Block 5000000 - Block reward reduced to 3 ETH / block Block 7000000 - Block reward reduced to 2 ETH / block (minimum block reward until proof-of-stake) As previously stated, all uncle rewards should be proportionally reduced along with the main block reward. The reduction in issuance specified in this EIP should also be accompanied with a change to push back the "ice age" date into the future as best determined by the Ethereum Foundation in the context of planning for the Casper / proof-of-stake release. In the light of discussions with members of the Ethereum community and the Ethereum Foundation, this EIP recommends a coin vote should be taken to determine the level of community support before a decision about whether to implement it.

Rationale This design is deliberately simplified from the issuance function Vitalik used to calculate his predicted blocktime schedule with relatively few block reward amount transition changes in order to minimize implimentation difficulty across the various Ethereum clients. No change in the relationship between block and uncle rewards is made to avoid introducing unforseen game-theoretic changes in mining strategy. Minimum planned proof-of-work issuance is set to 2 ETH / block to provide a baseline level of security rewards for proof-of-work close to Bitcoin's current issuance rate until the transition to proof-of-stake.

Implementation This EIP must be implemented in all Ethereum validating nodes by a hard fork, either in a currently-planned hard fork such as Metropolis, or in a separate hard fork.

Original EIP proposal below

  
  EIP: 186
  Title: Reduce ETH issuance before proof-of-stake
  Author: Matthew Light | lightuponlight9@gmail.com
  Discussions-To: lightuponlight9@gmail.com
  Status: Draft
  Type: Standards Track
  Created: 2016-12-21
  Replaces:  N/A
  Superseded-By: N/A
  Resolution: N/A

Abstract A reduction in the issuance of Ether (ETH) is very likely to be price-supportive and lead to increasing investments in the platform and to help ward off speculative attacks on the value of Ether by promoters of competing platforms who offer, or plan to offer, reduced token inflation rates. This proposal is based on a discussion on Reddit about ETH token issuance rates that many people participated in, including Ethereum Foundation member Vitalik Buterin.

In the current version of Ethereum as deployed on the network, there is already a coded schedule of reduced issuance per time period due to the "ice age" (an increase in difficulty designed to incent the adoption of hard forks). Normally, the "ice ages" in Ethereum are postponed by hard forks and the issuance is kept at 5 ETH per block. This EIP suggests that instead, the reduction in issuance from the "ice age" be implemented in a stepwise fashion within the protocol rules on a set block number schedule and with a defined lower bound (such as the "tail emission" in Monero). The proposed block reward reduction timetables and the issuance floor are open to alternative suggestions by the Ethereum community if desired.

Motivation There is widespread interest within in the Ethereum community to reduce the current rate of ETH issuance. Uncertainty about the future total ETH token supply is a significant factor in reducing the market value of ETH which has negative externalities on the Ethereum ecosystem, by reducing the capital available by current investors to make investments in new Ethereum-based projects and by reducing the Ethereum Foundation's funds available for spending on salaries. Reducing ETH issuance in advance of proof-of-stake would provide a measure of reassurance to investors that their holdings of ETH will be diluted to a much lower degree. It would be helpful in the cryptoplatform marketplace for Ethereum to reduce issuance given that various competing platforms have or are planning lower issuance levels, such as Ethereum Classic. If a large amount of opposition to this EIP for issuance reduction is seen in the community, this EIP could be submitted for a coin vote. Additionally, reducing block rewards in the near future helps reduce a sense of pressure on the Ethereum network architects and developers to rush a proof-of-stake implementation and try to deliver it more quickly than would be prudent. The terminal reward value of 1.5 ETH / block was chosen to reduce token supply inflation to around 3% annually, gradually reducing over time until proof-of-stake is launched. This is in line with but slightly less than current Bitcoin token inflation rate of approximately 4%. In comparison, current proof-of-work ETH token inflation rate is a bit less than 13% per year. A reduction of token emission under proof-of-work could lead to a reduction in network security. However, a plausible improvement in the exchange rate of ETH because of a reduction in issuance will lead to an increase in network security. It is unclear how much the ETH token price might increase due to a lowered issuance rate, but the positive response of Bitcoin price to issuance reductions ("halvening") is encouraging.

Specification The current issuance level for ETH is 5 ETH per block, with various kinds of uncle rewards also provided. This EIP proposes that no change to the relationship between block rewards and uncle rewards be made, but instead that all uncle and uncle inclusion rewards be downsized proportionally with the provided block reward schedule. The proposed reduction in issuance is an approximation for providing stepwise issuance reduction approximately matching the "ice age" coin issuance reduction already built into the current specification of the Ethereum software. Vitalik Buterin gave an estimate of this reduction on Reddit: https://www.reddit.com/r/ethereum/comments/5izcf5/lets_talk_about_the_projected_coin_supply_over/dbc66rd/ Based on the ice age blocktime lengthening data Vitalik provided, the following reduction of block reward schedule is proposed as a very rough approximation of the existing specification's reduction in issuance (via longer blocktimes caused by the "ice age" PoW difficulty bomb): Block 3700000 - Block reward reduced to 4 ETH / block Block 3900000 - Block reward reduced to 3 ETH / block Block 4100000 - Block reward reduced to 2 ETH / block Block 4150000 - Block reward reduced to 1.5 ETH / block (minimum block reward until proof-of-stake) As previously stated, all uncle rewards should be proportionally reduced along with the main block reward. The reduction in issuance specified in this EIP should also be accompanied with a change to push back the "ice age" date.

Rationale

This design is deliberately simplified from the issuance function Vitalik used to calculate his predicted blocktime schedule with relatively few block reward amount transition changes in order to minimize implimentation difficulty across the various Ethereum clients. No change in the relationship between block and uncle rewards is made to avoid introducing unforseen game-theoretic changes in mining strategy. Minimum issuance is set to 1.5 ETH / block to provide a baseline level of security rewards for proof-of-work close to but slightly below Bitcoin's current issuance rate until the transition to proof-of-stake.

Implementation This EIP must be implemented in all Ethereum validating nodes by a hard fork, either in a currently-planned hard fork such as Metropolis, or in a separate hard fork.

lightuponlight commented 7 years ago

I think a simple change to 3 ETH / block is the best plan. It's much simpler to implement and it's the Schelling point at the Carbonvote.

On Tue, Aug 1, 2017 at 7:53 PM, Alex Miller notifications@github.com wrote:

Moving discussion to this EIP per @5chdn https://github.com/5chdn's request.

This is starting to feel very nebulous and emotional, so I'd like to see a number get finalized soon. The initial EIP (which, I'd like to remind everyone, was written in December 2016) proposed a hardcoded decay based on block numbers that are now deprecated. In the spirit of the original EIP, I will propose an updated decay curve:

Beginning on block BLOCK_METROPOLIS, the issuance halves every 2 million blocks. The first halving takes place on BLOCK_METROPOLIS, the second on BLOCK_METROPOLIS+2000000, etc. This leaves us with the following reward:

Day 0 2.5 ETH Day 348 1.25 ETH Day 694 0.625 ETH

If Casper is on track, we should only see the first halving. If it is delayed significantly, we will see the second. Hopefully we will never see the third, but I believe it is important to encode for future-proofing purposes.

For those who have been discussing the numbers in this EIP, does this seem reasonable?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/ethereum/EIPs/issues/186#issuecomment-319528701, or mute the thread https://github.com/notifications/unsubscribe-auth/ASwWYTlDpBo1AwlQ8iNCk2Xwx5m0yAUJks5sT7pjgaJpZM4LTZf6 .

MicahZoltu commented 7 years ago

@alex-miller-0 When halving block reward, one can (generally) expect hashing power to halve (not exactly due to market inefficiencies). This means block times will up to double, how long does Ethereum take to stabilize difficulty around the new hashing power?

alex-miller-0 commented 7 years ago

@MicahZoltu Good question - I would have to look into that. I'm frankly fine with 30 second block times, but probably most people aren't.

@lightuponlight That's fine. I disagree numerically, but am happy to compromise as long as something gets done. If we can close this issue with a readjustment to 3 ETH/block at BLOCK_METROPOLIS_1, I and (I suspect) much of the community will be happy.

MicahZoltu commented 7 years ago

I similarly don't want to hold up Metropolis with in-fighting over block rewards, though I think a reduction to 3ETH is a bit pointless. I personally recommend halving the reward today, then halving again every year. This puts pressure on the community to switch to Casper as the security model will continue to deteriorate with each passing year until either the system is compromised or Casper is released. Should Casper completely fall through, then we can re-evaluate and hard fork.

A reduction to 3ETH feels like it is just placating the "reduce the block reward" crowd and I personally haven't heard any compelling arguments for keeping block rewards high as we head towards Casper.

As I have stated elsewhere, I think that people forget that miners are a service provider, they are not the customer. Mining is also a very competitive business with very low barrier to entry and near zero cost to switch providers, so there is no need to keep the current set of service providers (miners) happy. We only need to pay them enough to get the target amount of hashing power to show up, and nothing more.

Block rewards are part of the crypto-economic game theory around security, and ice age / reward decreasing should (IMO) be a pressure that forces people to switch to Proof of Stake (the long stated goal) because the system will eventually not be secure otherwise. With a block reward minimum, the system could remain stable/secure indefinitely without ever switching to Casper, which means making the switch is more likely to result in a contentious fork.

If miners don't like the idea of Proof of Stake, there are plenty of other crypto-currencies that offer proof of work mining and continue to do so (including ETC).

spelgubbe commented 7 years ago

23 seconds block time already, are things really going as planned?

I would really like to know what that target amount of hashing power is. Knowing that is required to set the emission of ether to a reasonable amount, not a stupid carbonvote.

Arachnid commented 7 years ago

Consensus on recent All Core Devs calls has been for a reduction to 3 ether.

Progressively decreasing issuance on a schedule is a bad idea both because we don't expect PoW to be around for long enough for it to matter, and because there's good reason to believe that a blockchain where transaction rewards outweigh block rewards is likely to be unstable.

Given that the ice age forces period hard forks anyway, it makes far more sense to adjust rewards if needed when that happens, rather than try and devise a schedule that works indefinitely.

MicahZoltu commented 7 years ago

Consensus on recent All Core Devs calls has been for a reduction to 3 ether.

@Arachnid Was there a reason for this number or was it just pulled out of the air (like the original 5ETH block reward number)? From the discussion here (unless I missed it) it appears that the 3ETH number just kind of magically appeared and started getting thrown around.

because there's good reason to believe that a blockchain where transaction rewards outweigh block rewards is likely to be unstable.

That is the point, to set the system up for failure unless Casper is accepted. By setting it to 3ETH, you are setting up a stable system that can run indefinitely without upgrades, even though upgrades are desirable. If you are arguing that the ice age is enough to force an eventual upgrade, then that is a more compelling argument.

rather than try and devise a schedule that works indefinitely

To reiterate again, I'm specifically suggesting a system that does not work indefinitely. The purpose is to have a system that eventually breaks down if upgrades don't happen. Though, this is an overlap with the ice age, so perhaps its not worth trying to have two solutions to the same problem.

kybarnet commented 7 years ago

@MicahZoltu Asks good questions and makes good points that few can understand.

A simple way to express expenses is by using Dollars. For example, here are some historical dollar costs:

Ethereum 2016 - $106 Million

LiteCoin - $250 Million

Virtually all other coins - $100 Million.

From there you set an objective. Nick Johnson set an objective of paying more than $24 Million annually, or the cost of transaction fees. I agree with that as well. Now, who pays for this cost? Either the price of Ethereum must go down, or New Investment must exceed the cost. From here, you can establish a potential cost objective, which one can speculate to be between $24 Million and $250 Million.

Then one must convert this to percents of Market Cap, or Inflation, which would be:

$24 Million over $21 Billion = 0.1% - 94,000 ETH - 0.04 ETH

$250 Million over $21 Billion = 1.2% - 1,128,000 ETH - 0.54 ETH

$500 Million over $21 Billion = 2.4% - 2,256,000 ETH - 1.07 ETH

$1 Billion over $21 Billion = 4.8% - 4,512,000 ETH - 2.15 ETH

$1.5 Billion over $21 Billion = 7.1% - 6,674,000 ETH - 3.17 ETH

For block chain technology, this tax on accumulated wealth must be expressed in forms of ETH which I provided above, with base 94,000,000 ETH, and 2,102,400 Blocks.

It takes but a small imagination to see that if the Market Cap of ETH goes up and the percentage is kept the same, then the dollar figure would grow as well. Therefor, it is extremely wise, especially for those that are unaccustomed to understanding markets, to think in terms of Dollar Figures, rather than Percentages of future Market Cap.

Thus a 3 ETH proposal is a campaign for $1.5 Billion, minimum, and a 2 ETH proposal is for $1 Billion, and a 1 ETH proposal is for $500 Million, and so on. It is much more clear to make this expression in terms of dollars.

For example, if it was expressed as $1.5 Billion, it is clear that if the Market Cap was to grow to $42 Billion, that the expense is not likewise to grow to $3 Billion, where as when expressed simply as a percentage (or ETH), the objective becomes muddled and confused. Should the expense grow to $3 Billion? Why not pay $3 Billion now, if that was the intention? Why did Ethereum pay $106 Million in 2016 if the minimum security requirement was actually $1.5 Billion?

Hopefully this will help you to understand Micah's concern over expressing expenses and costs in Dollars, vs arbitrary percentages or Eth. Bitcoin took a guess in 2008 as to what that value should be in 2017 and resulted to simply "4%", which now equates to $2 Billion. Fortunately we get to make this decision using 2017 dollars and an established market value.

Now most of this is actually arbitrary. For those that understand markets understand that expense must equal that of demand. If the amount of ETH was 20 per block, would the price of Ethereum remain $220 while miners are bonused at $8 Billion annually (vs roughly $1.5 Billion today)? Clearly not. The price must fall until New Investment equals that of expense.

For example, if we say $21 Billion is a fair price, then we are suggesting demand is $1.7 Billion, equal to that of expense. Thus if expenses are reduced to $500 Million or 1 ETH, what happens to the price? It must escalate to meet demand. This equilibrium can be discovered through simple math, or $1.7 Billion over $500 Million times $21 Billion or $71 Billion. Historically, however, Bitcoin has shown that the demand for assets which lose 4% of value annually is more than for that which lose 8% in value, by vary substantial margins.

At 8% Tax on Accumulated Wealth paid to Miner Bonuses, Miners Bonused $594 Million annually.

At 4% Tax on Accumulated Wealth paid to Miner Bonuses, Miners Bonused $1,600 Million annually.

How could miners receive more dollars as the percent went down? To understand such an equation, one must understand percents and markets. A percentage of a bigger number will yield a higher dollar figure than the same percentage of a smaller number. Thus, to balance the same demand, the number gets bigger when using a lower percent.

However, why did the bonuses actually go UP, rather than merely stabilize? Thus, we can presume that the demand for Assets which are globally tradable but depreciate at 4% is 3x the demand for similar Assets that depreciate at 8%. Frankly, I feel it's safe to say demand for lower taxes on accumulated wealth will always be more than demand for similar assets with higher taxes, though I know that is not a commonly accepted view today.

Regardless, the actually dollar figure for Miner Bonuses is unlikely to change, or would escalate, should the percent of tax get reduced, after market adjustment. And then we must ask how long does it take for markets to adjust?

Using Bitcoin as a guide, miner bonuses exceed those at double the rate after 3 months, and stabilized after 6 months. However, the second time, from 8% to 4%, miners actually received a higher dollar bonus immediately. Why?

Simple, because Bitcoin was following a schedule for tax reductions, which allowed the market to adjust values in advance, which ended up actually increasing the dollars of tax. Thus by preparing a schedule of expenses, and establishing a plan in dollar values, the outside market can properly anticipate values and increase Miner Bonus, before the reduction actually occurs.

However with last minute adjustments the markets are unable to anticipate values, and the miner bonuses could decrease for a period of 1 to 2 months. Even with the 'next update coming any minute', I think it's fair to make at least one anticipatory adjustment, to allow markets to adjust in advance rather than retroactively, which is better for the stability of the block chain and the mining community.

Oddly enough when Ethereum Forks, "Ethereum Ice" drops the rate of tax to 5% round November with 40 second blocks, to 2% round February with 2 minute blocks, to 1% round December with 4 minute blocks. I am curious to see how that hard fork plays out.

At the bare minimum, it makes for an interesting problem which will be studied in Block chain history books for months to come, for those that have interest in block chain technology outside Ethereum, and historical events.

But that's a lot of data :P I am all for having fun too, so that is why I enjoy 3 ETH FOREVER!

Whichever way it goes, we are in for a world of fun over the next few months!

alex-miller-0 commented 7 years ago

Given that the ice age forces period hard forks anyway, it makes far more sense to adjust rewards if needed when that happens

This effectively addresses my concern (albeit in a procedurally inefficient way), so all good from my perspective. Glad there's consensus from the Core Devs. 👍

Danieljohnsz commented 7 years ago

I know that you are all devs, and see this issue from one side of the spectrum, so if you don't mind, please allow me to express the flip side of this spectrum (the "bloated" miners perspective), in real dollars, so that you can take that information and see where it takes you in terms of when to expect network hashing power to drop off from the difficulty bomb.

Real world numbers as of 8/3/2017:

I'm going to use a 4 GPU rig of GTX 1060's as my example, as it hashes nicely at an even 100 mh/s. Extrapolate this out for larger operations.

Hashing: 100 mh/s Wattage: 400w System + 50w Fan Mid-of-the-road Electric Rate: .13 kWh Frugal build cost: $1600

Let's assume the miner started 6/1/2017, here would be his ETH earnings (Current USD $220):

June - 2.33 ETH +512.60 -42.12 = $475.16 July - 1.35 ETH +297.00 -42.12 = $254.88 Aug - .585 ETH +128.70 -42.12 = $86.58 Sept - .285 ETH +62.70 -42.12 = $20.58 Oct - .138 ETH +30.36 -42.12 = ($11.76) Nov - .067 ETH +14.74 -42.12 = ($27.38) Total $798.06

So by October you'll see about 1/3 to 1/2 of hashing power drop off, because they will have no chance of recouping their ROI and will be losing money by continuing.

By November only the individuals that live near hydro or nuclear plants that get .03 kWh power will remain and Ethereum would be a stretch to still call a "decentralized network".

Keep in mind:

Miner does not hit ROI
Miner's time is valued at zero
Does not include hotter climates where A/C is required (Easily add 25-50% Cost)

With all this being stated, I hope that you can see that miners are not rolling around in piles of cash.

Please consider this calculation in your decision to reduce block rewards and continue with current difficulty increases due to the bomb.

We all want to see Ethereum succeed. If PoS was to release in full by the end of October, then perhaps the numbers work, but if the possibility of a year and a half wait time that's been thrown around is valid, then the network just simply can't sustain. Especially with a decrease in block reward.

AndreasThom commented 7 years ago

Looking at the problem from a game-theoretic/sociological perspective it may matter a lot if the developers agree now on a progressively decreasing issuance or just decrease the issuance once and wait for the next ice age to make further decisions. The current situation is easy politically and it is unclear how it will develop if the transition to PoS is getting closer. It might well be that it will play a very important role what the "established state-of-the-art" is in one year and who is perceived as changing it or forking away. What is perceived as "classic" and what is not has mattered (unfortunately) quite a bit in similar situations in the past. Hence, one should definitely think now about how miners can be discouraged to block the transition to PoS. Having an increasingly bad economic environment would be one way - the transition to PoS should be inevitable even though the timeframe should remain flexible within reasonable bounds.

Arachnid commented 7 years ago

Was there a reason for this number or was it just pulled out of the air (like the original 5ETH block reward number)? From the discussion here (unless I missed it) it appears that the 3ETH number just kind of magically appeared and started getting thrown around.

It's roughly the issuance at the time the HF is planned.

That is the point, to set the system up for failure unless Casper is accepted. By setting it to 3ETH, you are setting up a stable system that can run indefinitely without upgrades, even though upgrades are desirable. If you are arguing that the ice age is enough to force an eventual upgrade, then that is a more compelling argument.

The ice age exists explicitly for this purpose. There's no reason to add a second mechanism that does the same thing.

@Danieljohnsz Reducing issuance will coincide with postponing the ice age, so will result in a restoration of difficulty to its original schedule.

kybarnet commented 7 years ago

@AndreasThom You make good points. I agree that Dollars are a suitable way to quantify expenses, historically and as future objectives.

Here is a clip of the core dev meeting on the subject, if curious. Enjoy!

https://www.youtube.com/watch?v=hRQg_lHEKl4&feature=youtu.be&t=41m51s

M41 commented 7 years ago

The world of blockchain now enjoys a large degree of resiliency and miners have a lot of options. Some miners will move on to other chains due to lack of sufficient ROI. Difficulty decreases, remaining miners will enjoy a higher ROI and are happy to stay and continue mining. We have seen it in the past, self-adjustment and habituating to new set of conditions. That's what forks did. They all worked out at the end to everyone's satisfaction more or less. Different players in the ecosystem just need to come into terms with the new set of conditions during a period of adjustment.

larspensjo commented 7 years ago

It's roughly the issuance at the time the HF is planned.

I think this is the best solution. That way, nobody can be accused of reducing the issuance to increase the price. As long as the slope of the ice age difficulty is low enough, it also gives the miners a chance to predict future issuance.

fre 4 aug. 2017 kl 17:26 skrev M41 notifications@github.com:

The world of blockchain now enjoys a large degree of resiliency and miners have a lot of options. Some miners will move on to other chains due to lack of sufficient ROI. Difficulty decreases, remaining miners will enjoy a higher ROI and are happy to stay and continue mining. We have seen it in the past, self-adjustment and habituating to new set of conditions. That's what forks did. They all worked out at the end to everyone's satisfaction or after coming into terms with it after a period of adjustment.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/ethereum/EIPs/issues/186#issuecomment-320278199, or mute the thread https://github.com/notifications/unsubscribe-auth/ACTPNgAvMP-Fd3vpN24JvZuGkKsTeHHgks5sUzg7gaJpZM4LTZf6 .

Danieljohnsz commented 7 years ago

@M41 That is ideal, but the difficulty isn't decreasing based on miners, it is increasing on a set schedule due to the difficulty bomb already set in place, to my understanding it won't fluctuate based on the current hashrate of the network, it will just continue to go up until we hit the ice age. Is this not the case?

kybarnet commented 7 years ago

Misplaced conversation from 669 moved to 186 per @5chdn request -

@Arachnid - Wrote -

The bomb is not being 'frozen', just delayed. Any attempt to determine miner profitability needs to factor in second-order effects; if some miners move to other coins because Ether is no longer profitable, the difficulty will go down and mining profitability will go up, until it reaches an equilibrium.

@kybarnet - Replied -

You make good points Nick, it is unreasonable for Ethereum to insure unlimited number of miners are profitable. At one time, I heard the goal of supporting up to 10,000 Miners, or nodes, but that seems out the window, yes? Not sure our current number of nodes, but I think I've heard up to 35,000.

Likewise, when determining the Dollars of payout, the Price of Ethereum matters. 3 ETH at $220 is Equal to 2 ETH at $330, or 5 ETH at $132 or 1 ETH at $660. The total dollar of payout is a function of ETH x Price. Thus, if the objective is to secure a fixed or approximate dollar amount, it is worthwhile to consider market factors.

However, I've already done the math, and know that this argument is mute. Miners, Holders, Programmers, Users, all benefit from increased efficiency, and the more immediate and sudden, the better. No matter how large the increase in efficiency becomes, be it 3 or 2 or 1, miners will make astronomically more money than they did at 5 ETH, even at the $400 price. The only discussion currently is miners arguing to pay themselves less, but unknowingly.

I see. I have found the unspoken Math.

Ethereum is running 24,000 Nodes Currently at $2.2 Billion.

Factoring that Down to 10,000 Nodes equates to roughly $1 Billion.

Thus, the justification for the unspoken objective of securing $1 Billion in payouts. Good to know :)

If you were to pretend $220 represented a stable price - 5 ETH Equates to $2.2 Billion. If the issuance was reduced to 3 ETH at $220, that would be about $1 Billion, forcing a $1.2 Billion buy order upon the market, raising the miner bonuses until they were once again $2.2 Billion, long term. This adjustment takes about 2 months.

@int03h - Replied -

@kybarnet nodes are not miners, nor are all miners nodes. Ethermine on its own 59782 active miners. Miner bonuses don't rise. We get a flat fee of exactly 5ETH per block, every block, whether there is one miner or 2 million. So who exactly has this $1.2B that you are so obsessed with exactly ? 5 ETH x approx 4,040 a day @ $300/ETH = $6M a day x 30 = $182M a month.

@kybarnet - Replied -

Note : His original statement had no math, and was edited after my reply.

Double Note : Also, note how he "Pretends" to not understand math, again. 5,760 Blocks are created daily under standard circumstances. In his attempt to confuse people, he suggests 5,760 x 5 = 4,040 ETH daily . The actual math is 28,800 ETH issued daily. All discussions of Eth bonus realignment are under the assumption of standard conditions, or 5,760 blocks per DAY - Not 808 as they suggest... his math is just very poor, but his $6 Mil per day figure is about correct.

"Flat rate of 5 ETH" . Holy Jesus man. 5 ETH at $20 = $100 , 5 ETH at $200 = $1,000.

For those that are observing the discussion from the outside, the Mining Union has brigaded the Ethereum Team so intently, it is now customary and acceptable to suggest there is no dollar value to Ethereum. To the outside world, it's widely accepted to be the Market Price, or about $280 currently. However, in programmer and miner discussions, they will suggest there is NO dollar value, and claim that 5 Eth at $10 is equal to 5 Eth at $400. The absurdity of such logic baffles some people, and they find it unbelievable that the anyone could actually believe such a statement, or even condone such discussion, regarding finances and Dollars of expense.

To be clear, it is acceptable within the Ethereum Discussion to claim 20,000 ETH payments at $1 ($20,000) is exactly equivalent to payments at a price of $400 ($8,000,000). The claim is that these two numbers are equal, and purchase equivalent amount of goods within the real world, such as mining equipment (however this is blatantly, obviously, false).

@int03h I would be happy to continue discussing mining rewards with you for 5,000 ETH. You have a significant misunderstanding regarding the concepts of mining, and are hurting miners and Ethereum. If you value the importance of mining, I request you send me 5,000 ETH so that your false notions can be corrected, which is equivalent to about 4 hours of mining reward at 5 ETH. I hope you value mining discussions as much as I do.

2 Million blocks are created annually, so 3 ETH = 6,000,000 Eth issued to miners bonuses, per year.

At $20 per ETH, that's $120 Million annually, at $400 per ETH that's $2.4 Billion annually. A miner will say these numbers are the same.

alex-miller-0 commented 7 years ago

@cbice66 Good question. An uncle reward has been proposed here https://github.com/ethereum/EIPs/pull/669#issuecomment-315380512. I assume it will be included as part of EIP186.

5chdn commented 7 years ago

If 649 replaces 186, then uncle reward will be updated, yes.

lightuponlight commented 7 years ago

A proportional reduction in uncle rewards is included in EIP-186.

On Mon, Aug 14, 2017 at 4:42 AM, Afri notifications@github.com wrote:

If 649 replaces 186, then uncle reward will be updated, yes.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/ethereum/EIPs/issues/186#issuecomment-322132156, or mute the thread https://github.com/notifications/unsubscribe-auth/ASwWYawSxQiDOIAKk_GRB2N7lNCL71afks5sYAiBgaJpZM4LTZf6 .

yuhong commented 7 years ago

Thinking about it, GPU mining is so important I am surprised that AMD/NVIDIA senior executives aren't talking with coin developers. I prefer they do the work in public and use a SEC filing if needed to disclose to solve any Reg FD issues.

nicksavers commented 7 years ago

@lightuponlight This EIP seems to have been superseded and implemented by #649. Does this issue still need to remain open? If not could you close it please?

Arachnid commented 6 years ago

This pull request is referenced by #649, which has already been merged. As such, it needs to either be finalised and merged, or the dependency in the existing EIP needs to be removed.

Arachnid commented 6 years ago

I've removed the reference from 649, since this was never merged as an EIP.

ethereum / EIPs

Reduce ETH issuance before proof-of-stake #186

Modified EIP proposal with changes, followed by the original EIP proposal. These changes are based on feedback provided on Reddit and Github regarding this EIP regarding the issuance reduction schedule and minimum issuance level under proof-of-work.

Original EIP proposal below