BIP: ?? Gradual Changing Block Rewards

[hazek]

It wouldn't be right to change any aspect of the subsidy rules in Bitcoin. These rules are a "contract" that all Bitcoin users have agreed to.

The nature of Bitcoin is such that once version 0.1 was released, the core design was set in stone for the rest of its lifetime.

It's technically very difficult, anyway. You don't need just 51% of miners: you'd need essentially 99% of all users.

This post right here just ended this thread and gged this BIP idea.

[1713586877]

1713586877

[1713586877]

1713586877

[1713586877]

1713586877

[Vladimir]

[serious]

@Blitzboom:
  Thank you for spelling out the "slippery slope", it is what I have implied in my first post.

@hazek:
  Indeed, it would be silly to ruin simple, magical, and elegant in so many areas design of Bitcoin protocol.

@theymos:
  Yes a "social contract" exists and it would be wrong to break it. Nigh impossible too, just as you have said.

@allten:
  Please do not be upset that your proposal was so quickly and so harshly criticised. You are not the first one who became excited about Bitcoin and thought that one should or can improve the protocol by introducing some simple changes. Some people, myself included, thought about it for a while and moved on, doing something more productive. Others thought about it, got their ideas ignored and laughed at and then went on creating shitcoins. One thing I know for sure, your proposal is not going to fly.

[/serious]

[Revalin]

Bitcoin is the new religion:  everyone quoting what the founder said; no one agreeing on what he meant.

[Maged]

The nature of Bitcoin is such that once version 0.1 was released, the core design was set in stone for the rest of its lifetime.

It's technically very difficult, anyway. You don't need just 51% of miners: you'd need essentially 99% of all users.

Not really, actually. You'd just need 100% of miners. However, it'd be even riskier than BIP 16/17. To do it like this (only requiring miners to upgrade instead of every single user), you'd have to throw away a LOT of bitcoins by having miners not claim the maximum block reward (we already know that this can be done, since it was done accidentally in the past). Done like this, it becomes a more restrictive block acceptance rule, qualifying for reverse-compatibility for users (not miners, though).

I highly doubt that miners would ever be willing to cut their own pay, however.

[Transisto]

The first reward halving will be the biggest puzzle speculators will face.

Aside the technical challenge parts of the equation are :

How much better is this proposal VS. What negative effect will this breach of contract have on the perception of users, especially those who've planned long term for this.

Changing the reward every two week could be cause accounting problems for some pools.

4 year is a good time frame for new mining technologies to develop and be deployed.

Abrupt reward changes will times where miners show their cards (figuratively as in what proportion of hashing has which efficiency / costs)


... Just some thoughts.

[allten]

  BIP: ??
  Title: Gradual Changing Block Rewards
  Author: Slim Pickens
  Status: Draft
  Type: Standards Track
  Created: ??-??-2012
  Post-History: ??-??-2012

Abstract
This BIP proposes for block rewards to decrease more frequently.

Motivation
Every four years the block reward for bitcoin is cut in half. There is much uncertainty surrounding bitcoin prices and total network hashing power before and after these reward reductions. This author feels very confident that bitcoin will traverse its first block decrease with success; however, months before and after will be full of large price swings with a potentially modest decrease in global hash rate. Stability has yet to be one of Bitcoin’s virtues; yet, many hope (including the author) that this will change over time. The goal of this BIP is to avoid the frenzy in price fluctuations and global hash rate during these rare events by making them much more frequent with smaller decreases.

Specification
Block rewards will begin decreasing every two weeks starting at the end of the second year of the 25 BTC block reward (What’s the specific block number?). (Note: How frequent the changes are and when it goes into effect are open for debate). Unfortunately, this author is not into coding Bitcoin so the details of this BIPs implementation will rely on feedback from experienced Bitcoin programmers.

Variables:
•   Frequency of bitcoin decreases (Initial proposal is for every two weeks)
•   Date, more specifically the Block #, at which these smaller more frequent decreases take effect. (Initial proposal is for this BIP to take effect at the end of the second year of 25 BTC rewards)
•   The Percent of the decrease. Need to find a percentage that will in effect release the same amount of coins over the long term as the original block reward system.

Rationale
Two weeks were chosen for the frequency of decreases because it coincides with the difficulty adjustment. Seems like a good idea to do them together. However, this is not critical for the success of this BIP. Depending on the general consensus from the Bitcoin community and the simplicity of coding implementation, this may change.

The date for the change is proposed at the end of the second year of the 25 BTC block rewards. This gives the community 6 months to observe the market behavior of the initial block decrease and have a software implementation 18 months prior to the change going into effect. Again, depending on the general consensus from the Bitcoin community and the simplicity of coding implementation, this may change.

Based on Moore’s law and the history of computing innovation, it is commonly quoted as 18 months for computing power to double. 18 months should be enough time for miners to adjust their capital strategy before this BIP goes into effect. It will also give the community plenty of time to adopt the new code.

The implementation of this BIP would be a huge win for the bitcoin community. There are many articles written why Bitcoin is “fatally flawed”. They gloss over the fact that the brilliance and innovation of this new currency is in its decentralization and the fact that it is democratically governed. This system doesn’t have the flaws of other democratically governed bodies because there are no laws in effect that force anyone to use it. It is simply by free choice; therefore, no one would advocate changes to favor a few because it would self-destruct the currency. The implementation of this BIP would be a shining example that when problems arise the community will deal with it in way that will cause the currency to continue to excel. In other words, all supposed fatal issues are really non-issues. When the timing is right, mass consensus will be reached and they will be fixed.

Appreciate everyone's comments even if they totally disagree. I would like to continue to push to get this published in the WIKI BIP. Doesn't mean it will be adopted. I just want to make it an official record of what is possible and not possible with Bitcoin.
Before doing so, I would like to kindly request that you humor me and this BIP "as if" it was going to be adopted and help me nail down all 3 variables and hold off from commenting on its feasibility.

Please comment on the following:
1)Frequency
2)Block # of implementation
3)Percent decrease.

Also, what if we rounded the block reward after each decrease to four significant digits? Any thoughts.

Other request for help:
       Graphs of Bitcoin curve over time compared to the original.
       Graphs of the injection rate over time compared to the original.
       Psuedo code for the implementation.

Thanks again.

[casascius]

I think it's a non-starter.  It's a solution in search of a problem.

I wouldn't waste your time, or the developers' time with this.  The time would be better spent petitioning US Congress to require all speed limit signs in the US to be readable at 100mph so that the worst speeders will know by just exactly how much they are speeding.

[ripper234]

I think it's a non-starter.  It's a solution in search of a problem.

I wouldn't waste your time, or the developers' time with this.  The time would be better spent petitioning US Congress to require all speed limit signs in the US to be readable at 100mph so that the worst speeders will know by just exactly how much they are speeding.

+1

It's just not interesting, on a global scale. This proposal will never get off the ground.
I can't say that I object, but I won't support it either (not that I have any hash power, just saying).

Good luck getting 55% support for this.

[rjk]

The ideal test environment would be as an altchain, seriously. If you are worried about the chain being attacked by overzealous fanatics like Luke, then just create it with Scrypt as the hashing mechanism, since that is a bit more difficult to attack. Then, you would be able to tinker with the reward mechanism however you wanted, and if you came up with something good and useable you could present it to the main devs.

[cypherdoc]

It wouldn't be right to change any aspect of the subsidy rules in Bitcoin. These rules are a "contract" that all Bitcoin users have agreed to.

The nature of Bitcoin is such that once version 0.1 was released, the core design was set in stone for the rest of its lifetime.

It's technically very difficult, anyway. You don't need just 51% of miners: you'd need essentially 99% of all users.

Thank heavens for this.

[Brunic]

Certainly not.

When I've discovered Bitcoin, I've read the rules first, and took the decision of starting that mining project. I've made plans to consider the reward drop, and I'm going to live with it. You don't change the rules so late in the game. That's what we call "cheating" usually.

If you don't like the rules of Bitcoin, well, there's other crypto-currencies out there. I know Solidcoin offer a different reward mechanism, go help those guys grow their currency instead.

Changing basic rules like that will cause more harm than good. I will certainly lost a LOT of confidence in Bitcoins if you can bend the rules that way. Bitcoin was made that way, deal with it like I do.

[allten]

Alright. The key factors of this BIP have been improved and finalized.

The frequency of change is Every Block.
The Change will occur on block 315000 (Block count starting at 0)
The approximate percent decrease: 0.00031746 %

Over 100 years, the original will have approximately injected a few bitcents more than the proposed improvement.
However, these small details will be worked out during the implementation of the code. What's important is the amount injected
is approximately the same or a few satoshis less than the original.


Total Sum
     

Injection Rate

Again, all comments are appreciated and so far I've observed a lot emotional response.
That's ok. It's all part of reveling what it will take to get a well crafted IMPROVEMENT
adopted by the majority.

[rjk]

You talk like it is definitely going to happen. Which it isn't. I'm not particularly worried.
It would be interesting to see it implemented in some kind of altcoin though.

[hazek]

Again, all comments are appreciated and so far I've observed a lot emotional response.

Are you a programmer? Do you understand how this limit of 21 million is actually enforced inside the code?

[Epoch]

I think it's a non-starter.  It's a solution in search of a problem.

I wouldn't waste your time, or the developers' time with this.  The time would be better spent petitioning US Congress to require all speed limit signs in the US to be readable at 100mph so that the worst speeders will know by just exactly how much they are speeding.

Agreed, this change is unnecessary. I would even go further and say that it is a bad idea.

The protocol is the fundamental concept of bitcoin. As the underlying foundation, it needs to remain stable for it to gain (and keep) the trust of its adopters. The protocol is the last thing anyone should be considering to change, unless there is a very good reason (i.e. fixing a flaw).

There is nothing fundamentally wrong with the concept of block reward adjustment. Why would a step change every 4 years be any worse (or, for that matter, better) than a gradual change as proposed?

As casascius and others have already stated, this is a solution in search of a problem. There is little or no benefit to do this, but it holds a real risk of undoing much of the trust that bitcoin protocol has already earned. Let's not shake the foundations of what what has defined bitcoin since its infancy unless there is a 'really good reason' for doing so. I cannot see a 'really good reason' for doing what has been proposed.

Sure, let's consider it. Let's discuss it. But then let's bury it and give it a rest. I would be deeply concerned if this idea gained any serious traction.

[DeathAndTaxes]

Do you understand how this limit of 21 million is actually enforced inside the code?

My understanding is there is no reference to 21,000,000 anywhere in the code.

The "limit" is simply an enforcement of what subsidy can be included in a block (which is based on the block's height) and be considered valid by other nodes.  Since the subsidy on each block must follow a required decrease the total # of coins that will be minted can be calculated.

All values are stored as 64 bit integers in satoshis (1E-8 BTC).
The initial subsidy is  5,000,000,000  (50 BTC).

The subsidy is right shifted once for every 210,000 blocks in height.
At block height 210,000 a valid node will reduce the subsidy to 2,500,000,000  (25 BTC).  Any node that doesn't will have their block rejected by other nodes as invalid.

So at block 6,720,000 the subsidy will be right shifted from 2 satoshi to 1 satoshi.
At block 7,140,000 the subsidy will be right shifted from 1 satoshi to 0 satoshi and no more coins will be minted.
Any block after 7,140,000 with a subsidy of any amount will be considered invalid by other nodes.

If you add up all the block rewards for the first 7,140,000 blocks it adds up to 21M BTC (2,100,000,000,000,000 discrete units).

Actually unless my understanding is incorrect there will never be 21M coins.  Due to truncation of some rewards there will be exactly 2,099,999,997,690,000 satoshis mined ( 20,999,999.9769 BTC).

[BubbleBoy]

And precisely because there is no hard limit on the total money supply, just a side effect of the subsidy function, changing this particular aspect of the algorithm is sacrilegious for the cult of Bitcoin. Sure, we could use this curve, we could use Pieter Wuille's curve, or any other heretic curve with fix supply equal to 21m.

But what guarantees do the faithful have that it will stop here ? If miners are free to disobey the laws of the great Satoshi as they see fit, what stops them from using the same consensus mechanism and set a fixed 50 BTC block reward ? It's certainly in their interest. Or maybe, gasp ! an increasing reward ! Blasphemy !  Cursed demon of Hiperinflation, you dare show your face HERE ? Begone with your heathen BIP !

[kjj]

There is a reference to the coin maximum (MAX_MONEY) in the code.  I don't think it gets used much, mostly just a sanity check.  It would be trivial to change.

The client has no idea of how many coins have been created.  There is no running total anywhere.

By the way, how would this proposal deal with the integer representation?  The first step in the new system would be from 2500000000 to 2499992063.50.  Would this round up, or down?  When it gets down around 0.00315 BTC, the delta will be less than one unit.  How will this be handled?

[blablahblah]

Appreciate everyone's comments even if they totally disagree. I would like to continue to push to get this published in the WIKI BIP. Doesn't mean it will be adopted. I just want to make it an official record of what is possible and not possible with Bitcoin.
Before doing so, I would like to kindly request that you humor me and this BIP "as if" it was going to be adopted and help me nail down all 3 variables and hold off from commenting on its feasibility.

Please comment on the following:
1)Frequency
2)Block # of implementation
3)Percent decrease.

Also, what if we rounded the block reward after each decrease to four significant digits? Any thoughts.

Points disputing the rationale:

• It's too complicated. Instead of having nice round numbers to work with for various calculations or estimates, one would need to spend a lot of time with spreadsheets (or custom software) making sure that they get the decay curve right. Even with a simpler hypothetical situation, such as adjusting the reward every year, instead of dividing by 2, you're looking at dividing by the 4th root of 2 (if my mental math is right). Very user-unfriendly.
• The rationale assumes that the 4-year steps will be a problem. A bit of a straw man argument? Won't it get priced-in?
• What about possible benefits to having periodic jolts in the block rate?
  1) One that comes to mind could be a sustained step in the price that helps Bitcoin achieve mainstream acceptance.
  2) Compared to the fiat world, Bitcoin banking is boring - it could give an outlet for investors looking to make money on cyclic changes. (I'm not sure if that was such a great argument, but I thought I'd better mention it anyway).
  3) Having these periodic jolts could help to keep the system robust. Without such a mechanism to help maintain robustness, alternative currencies -- with unknown inflation rates -- could cause problems for miners.

[FreeMoney]

If you divided by the 4th root of 2 every 52500 wouldn't you be producing at the same rate as the old way for one block every 210,000 and producing fewer all other times leading to way fewer blocks? What is the right number?

To restate everyone, this isn't happening. Fantasy is fine.

I wonder how much thought Satoshi gave this and what his main reason was. I think it was kind of a fairness thing. Whatever it was I think this decision more than most involved his guessing about how (time line etc) Bitcoin would be adopted.