Re: Changing the block reward reduction to happen per block rather than one day

[matthewh3]

What about changing the block reward reduction to happen per block rather than a one day event happening every four years.  The last and first block reward reduction happened when bitcoin mining was still GPU mined.  And most people were mining for the fun of it at home.  Even though they were probably paying a lot more in electricity power then they were financially generating in bitcoin.  This move would end a possible shock event in Q4 2016 - Q1 2017 when the next block reward reduction happens.  The mining scene is very different now to how it was during the last block reward reduction of Q4 2012.  As I've already said mining was mostly a hobbyists pursuit then by miners willing to mine at a short term financial loss.  To speculate long term on bitcoin and as one of the easiest and most anonymous way to acquire bitcoin at the time.  Things are very different today and the next block reward reduction.  Could quite possibly be a shock event where the hashrate may drop dramatically.  Opening up the network to an attack event a lot easier and cheaply.
This idea has already been alpha tested for the Bitcoin network by Monero.  Who already have this type of emission.  This fundamental change to the block reward does not have to affect the 21mn hard cap or any of the other bitcoin protocol fundamentals.  While in effect it could help make the bitcoin network much more secure.

[ca333]

What about changing the block reward reduction to happen per block rather than a one day event happening every four years.  The last and first block reward reduction happened when bitcoin mining was still GPU mined.  And most people were mining for the fun of it at home.  Even though they were probably paying a lot more in electricity power then they were financially generating in bitcoin.  This move would end a possible shock event in Q4 2016 - Q1 2017 when the next block reward reduction happens.  The mining scene is very different now to how it was during the last block reward reduction of Q4 2012.  As I've already said mining was mostly a hobbyists pursuit then by miners willing to mine at a short term financial loss.  To speculate long term on bitcoin and as one of the easiest and most anonymous way to acquire bitcoin at the time.  Things are very different today and the next block reward reduction.  Could quite possibly be a shock event where the hashrate may drop dramatically.  Opening up the network to an attack event a lot easier and cheaply.
This idea has already been alpha tested for the Bitcoin network by Monero.  Who already have this type of emission.  This fundamental change to the block reward does not have to affect the 21mn hard cap or any of the other bitcoin protocol fundamentals.  While in effect it could help make the bitcoin network much more secure.

This halfing of blockreward will not be good because i think because when after every block reward is halfed we never reach max. cap and in few weeks mining is not rentable anymore and people mined before make too much profits. it looses the soul of real wealth distribution which also must be decentralized. i think this only bring benefit for early adopter and miners. market will become too unstablized with this sort of reward halfings. and also i think when the next halfing in 2016 comes, the value will be good and hashrate goes up then because value goes up = so when demand is same but amount of new units is smaller. but also this is only my opinion and it s very interesting thoughts. so i wonder what other people or btc-devs think about this idea.
thank you!

[spin]

What about changing the block reward reduction to happen per block rather than a one day event happening every four years.  The last and first block reward reduction happened when bitcoin mining was still GPU mined.  And most people were mining for the fun of it at home.  Even though they were probably paying a lot more in electricity power then they were financially generating in bitcoin.  This move would end a possible shock event in Q4 2016 - Q1 2017 when the next block reward reduction happens.  The mining scene is very different now to how it was during the last block reward reduction of Q4 2012.  As I've already said mining was mostly a hobbyists pursuit then by miners willing to mine at a short term financial loss.  To speculate long term on bitcoin and as one of the easiest and most anonymous way to acquire bitcoin at the time.  Things are very different today and the next block reward reduction.  Could quite possibly be a shock event where the hashrate may drop dramatically.  Opening up the network to an attack event a lot easier and cheaply.
This idea has already been alpha tested for the Bitcoin network by Monero.  Who already have this type of emission.  This fundamental change to the block reward does not have to affect the 21mn hard cap or any of the other bitcoin protocol fundamentals.  While in effect it could help make the bitcoin network much more secure.

I'm guessing you mean that the block rewards get's reduced by a small fraction every block rather than having the step changes.  I'm also guessing you want it to be reduced at an equivalent rate.

I do think that the halving is a very predictable event.  Most large mining business that are mining near such a halving should plan for it in their business plans etc.  If not then they are just stupid.   And as you say there are less hobbyists around. 

[matthewh3]

What about changing the block reward reduction to happen per block rather than a one day event happening every four years.  The last and first block reward reduction happened when bitcoin mining was still GPU mined.  And most people were mining for the fun of it at home.  Even though they were probably paying a lot more in electricity power then they were financially generating in bitcoin.  This move would end a possible shock event in Q4 2016 - Q1 2017 when the next block reward reduction happens.  The mining scene is very different now to how it was during the last block reward reduction of Q4 2012.  As I've already said mining was mostly a hobbyists pursuit then by miners willing to mine at a short term financial loss.  To speculate long term on bitcoin and as one of the easiest and most anonymous way to acquire bitcoin at the time.  Things are very different today and the next block reward reduction.  Could quite possibly be a shock event where the hashrate may drop dramatically.  Opening up the network to an attack event a lot easier and cheaply.
This idea has already been alpha tested for the Bitcoin network by Monero.  Who already have this type of emission.  This fundamental change to the block reward does not have to affect the 21mn hard cap or any of the other bitcoin protocol fundamentals.  While in effect it could help make the bitcoin network much more secure.

I'm guessing you mean that the block rewards get's reduced by a small fraction every block rather than having the step changes.  I'm also guessing you want it to be reduced at an equivalent rate.

I do think that the halving is a very predictable event.  Most large mining business that are mining near such a halving should plan for it in their business plans etc.  If not then they are just stupid.   And as you say there are less hobbyists around. 

After the block reward reduction the value of all mining gear would be drastically reduced.  Therefor soon after this singular event it would make it a lot cheaper for a potential attacker to buy and/or rent mining equipment to mount an attack.

[teukon]

After the block reward reduction the value of all mining gear would be drastically reduced.  Therefor soon after this singular event it would make it a lot cheaper for a potential attacker to buy and/or rent mining equipment to mount an attack.

Do you believe that the market value of any given mining machine will drop sharply at the moment of the halving?

[DeathAndTaxes]

It probably would be better to have a gradual reduction but don't waste time thinking about it.  It will never happen.  A hard fork requires a super majority of users including miners.   So for the sake of the argument lets assume everyone except miners already supports this by a massive majority.  For the forked chain to be secure it needs a majority of the hashrate so the developers put into the code a phase-in once a certain percentage of miners support the change.  To avoid confusion and chaos it needs most of the hashrate something like 80% to 90% not just 50%.

So imagine the code is deployed right now.  Miners can either mine ver 3 blocks (25 BTC) or they can mine a ver 4 block which will start the subsidy decay.  Once 80% of the last 1000 blocks are ver 4 the change in subsidy will be implemented (say 24.9 BTC).  You are asking the miners to vote on if they would prefer to receive 25 BTC or 24.9 BTC.  Same work, same effort, same cost less Bitcoins and they need to vote to make it happen.  If they don't vote for it they keep getting 25 BTC if they do they get less.   It hardly matters that in the long run it will all even out.  The miner who would need to 'vote' for the change (by mining the version which reduces the block reward) may not even still be mining by the next subsidy halving.  

Who votes to give themselves a paycut when they don't need to?  Nobody.

[gmaxwell]

And most people were mining for the fun of it at home.

This isn't at all an accurate reflection of how mining was, how weird and distorted things our view of the past is. Go count up the hashrate at that point in terms of ordinary GPUs (and a significant fraction of it was FPGAs)... mining was very much industrialized then.

By making the reward decrease more frequently you increase the points where it locally more profitable for a miner to orphan the prior miners block than it is to extend it.

This idea has already been alpha tested

by other things which have had serious, network ending flaws over and over again, which for the most part haven't been attacked.  Anything is secure if no one cares to attack the thing its protecting.

Could quite possibly be a shock event where the hashrate may drop dramatically.  Opening up the network to an attack event a lot easier and cheaply

A "shock event" that people have known about for years? I think you must have a weird defintion of a shock event. Indeed, the hashrate may go down-- it's gone down this year, in fact, but this itself doesn't magically make the network any more open to attack then it would be if the reward changed slowly and ended up with the same difficulty.

[jl2012]

In terms of mining profitability, reward halving and price collapsing by 50% have the same effect, except that the first one is planned and highly predictable and the latter one is not.

[lyth0s]

I think if bitcoin really takes off the "Halving Events" could be considered celebrations amoung people knowing that the calculated inflation will be decreasing that year. Keeping it for social celebrations of the network is worth it all on its own IMO.

--Not even to mention DeathandTaxes's point on the fact that the supra-majority of miners would never agree to implement it.

[TimS]

It probably would be better to have a gradual reduction but don't waste time thinking about it.  It will never happen.  A hard fork requires a super majority of users including miners.   So for the sake of the argument lets assume everyone except miners already supports this by a massive majority.  For the forked chain to be secure it needs a majority of the hashrate so the developers put into the code a phase-in once a certain percentage of miners support the change.  To avoid confusion and chaos it needs most of the hashrate something like 80% to 90% not just 50%.

So imagine the code is deployed right now.  Miners can either mine ver 3 blocks (25 BTC) or they can mine a ver 4 block which will start the subsidy decay.  Once 80% of the last 1000 blocks are ver 4 the change in subsidy will be implemented (say 24.9 BTC).  You are asking the miners to vote on if they would prefer to receive 25 BTC or 24.9 BTC.  Same work, same effort, same cost less Bitcoins and they need to vote to make it happen.  If they don't vote for it they keep getting 25 BTC if they do they get less.   It hardly matters that in the long run it will all even out.  The miner who would need to 'vote' for the change (by mining the version which reduces the block reward) may not even still be mining by the next subsidy halving.  

Who votes to give themselves a paycut when they don't need to?  Nobody.

If you implement it in the first ~100,000 blocks after a halving, the reward will actually go up in the short term, so miners will prefer the change, assuming they'd rather have more BTC now. It would be cleanest to implement this at the start of a halving IMO (so it doesn't jump around so much), which would e.g. mean that miners get to mine 17 BTC blocks instead of 12.5 BTC.
To understand why, understand that since we still want it to end at the same time, with the same number of Bitcoins mined, each 4 year block will still mine just as many Bitcoins: blocks 420k-630k will still mine 2625000 BTC (12.5 BTC/block, avg) altogether. But instead of "BTC per block" being a flat line for those 210,000 blocks, it'd be an exponentially(?) decaying curve, so it will start out higher and then go lower: for the first almost-half of the blocks, the smoothly-decaying curve is higher than the flat line.

[Taras]

I wonder what satoshi's motives behind the sudden drops were. Maybe he just made those up?

[DeathAndTaxes]

I wonder what satoshi's motives behind the sudden drops were. Maybe he just made those up?

Simplicity.  Bitcoin implements the subsidy cut as an integer right shift.  This is one line of code and almost impossible to get wrong when porting to another client or language.  Anything else even if it has some advantages has the disadvantage of higher complexity.

[Nowi]

Better machines will come before the reward for found block.
But there is problem with BTC value, coz who will pay for it when bitcoin is staying about 300 dollars

[alani123]

You can create an altcoin that does this and tell us how it did. Your proposal sounds exploitable though, you can pump and dump mining power easily. Wouldn't that allow miners and multipools to game block rewards?

[TierNolan]

Simplicity.  Bitcoin implements the subsidy cut as an integer right shift.  This is one line of code and almost impossible to get wrong when porting to another client or language.  Anything else even if it has some advantages has the disadvantage of higher complexity.

You could still do it reasonably simply using something like this:

Code:

if (mint_fee > 0) 
    mint_fee -= (mint_fee >> 18) + 1

After 210,000 blocks the fee would drop to 44%.

The +1 is required because mint_fee >> 18 drops to zero before mint_fee is zero.

Assuming that the initial reward is 5 billion Satoshis, this rule would give 1310361517414979 of them (13,103,615.17414979 BTC).  This is lower than the current target due to the reward beginning to fall immediately.