Would you support moving to a system with controled inflation?

[Ian Maxwell]

I voted yes even though I dislike inflation. I like the idea of a stable currency that is neither deflationary nor inflationary in the long run, where I can expect to buy the same product with one bitcoin today or next year. That's what a "store of value" is.

If we believe that the economy will grow exponentially, as it always has in the past, then a linear growth in the number of bitcoins is not even stable, let alone inflationary: in the long run, it is quite deflationary. An asymptotically constant number of bitcoins is wildly deflationary.

(And if we don't believe the economy will grow exponentially in the future, we have far bigger problems than inflation and should be dropping all this Bitcoin stuff to go solve them.)

That said, when compared with exponential growth the difference between "linear" and "constant" is really unimportant by comparison, so it's not that big a deal.

[1713892005]

1713892005

[1713892005]

1713892005

[1713892005]

1713892005

[1713892005]

1713892005

[ShadowOfHarbringer]

I voted yes even though I dislike inflation. I like the idea of a stable currency that is neither deflationary nor inflationary in the long run, where I can expect to buy the same product with one bitcoin today or next year. That's what a "store of value" is.

That kind of currency (50 BTC block reward forever) would also be very deflationary in practice while inflationary by design, exactly as Bitcoin.
That is because the number of goods & services offered for it would rise faster than the number of coins minted (assuming it became popular).

So it would be very similiar to Bitcoin, but simply worse (because of weaker practical deflation) and IMHO people would quickly abolish it.

Pointless. Strong No.

[Ian Maxwell]

Wait, did you just tell me that 50 BTC per block forever is worse because it leads to slower deflation?

Not because it's inflationary, but because it's deflationary-but-less-so?

So either (a) we should be looking for modifications that will increase deflation even more, like going from 50 - 25 - 12.5 to 50 - 5 - 0.5, or (b) wonder of wonders, Satoshi has gotten it exactly right on attempt #1. What luck!

[grondilu]

Making a 50 BTC reward for ever is easy.  Basically you just have to comment one line in the code.

[ShadowOfHarbringer]

So either (a) we should be looking for modifications that will increase deflation even more, like going from 50 - 25 - 12.5 to 50 - 5 - 0.5, or (b) wonder of wonders, Satoshi has gotten it exactly right on attempt #1. What luck!

There is a saying among programmers.
"If it works, don't change anything. Chances are you may break something".

Seriously, what every currency needs is stability.
I would say that following the rules set by somebody way smarter than you (satoshi) is very reasonable, and leads to the most stable situation for the currency.

So unless another genius comes along, i don't see any point of changing something that works nicely.

Shortly, I will follow the one who I know is smarter than me, not some random guy i spotted on a forum.
Current deflation rate is the best because Satoshi said so. That is good enough for me.

[wb3]

Shortly, I will follow the one who I know is smarter than me, not some random guy i spotted on a forum.
Current deflation rate is the best because Satoshi said so. That is good enough for me.

Wow,

Do you bow down at the shrine and pray? What day is the day of worship?




Don't follow people that are "smarter" than you. Smart is a relative term and a matter of perspective.

Satoshi says delfation rate is good.

Satoshi says day of rest on Wednesday.

Satoshi says Give the BitCoin to Ceasar, Give the Credit to the one.

 

[ShadowOfHarbringer]

Shortly, I will follow the one who I know is smarter than me, not some random guy i spotted on a forum.
Current deflation rate is the best because Satoshi said so. That is good enough for me.

Wow,

Do you bow down at the shrine and pray? What day is the day of worship?

Well, I do have portable ceremonial shrines... 99,99 BTC if you are interested too.
But this is a special offer that is valid only today !

[caveden]

The most important reason why I personally is interested in bitcoin is because it is deflationary long term. Otherwise, I would read about it and dismiss it as another pointless fad.

Predictable initial inflation turning into deflation with time is THE most important feature of bitcoin IMO.

+1

[caveden]

3. I am not convinced that the "transaction fees for mining" is a sustainable equilibrium that will sustain the network.

Was in an attempt to deal with this issue and also the need to eventually grow the block size limit that I once proposed this.

If, for example, at each difficult factor adjustment, the block size limit also gets adjusted to, say, 110% or 105% of the average size of the last X blocks, we not only allow the network to grow to suit demand, but we also create an artificial scarcity of block space that may drive fees up.

[casascius]

Every argument I have seen, which purports to 'fix' Bitcoin for some economic reason or other seems also to make a presumption that there will be only one; and therefore must be 'perfected' to protect it from one limitation or other. This is the case with arguments about fees, and therefore block size limits, as well as the cases made for various inflation rates, or rates of extinguishing inflation.

I think it's fair to say that an important control that WILL need to be tweaked at some point is that maximum block size or else growth is unsustainable - there is no "free market" control on that number, something has to be done about it eventually.  Shall we procrastinate it until it becomes a huge problem?

This is not an economic problem, it is a technical problem partly tied to economics.  There exists no rational economic incentive to continue mining beyond the minting of new coins, because miners have no way to compel payers to pay a fee that makes mining worthwhile.  There is no way to achieve equilibrium.  All it takes is a few anonymous miners who will accept transactions at any fee to drive the transaction fee to near zero and ruin mining for everybody else, which will make them leave, which will make the network weak.  The only real control on the transaction fee is the maximum block size, which has no free-market economic controls - it is hardcoded into the software.  Changing it will require consensus on what change is right.  If finding consensus is hard now, just wait until the technical crap hits the fan!

A situation where the selfish acts of one individual give a greater gain to that individual in spite of long-term harm to the community is a textbook tragedy of the commons, not the free market at work.  Since mining is pretty much anonymous, it's practically guaranteed to be an issue.

[FooDSt4mP]

All it takes is a few anonymous miners who will accept transactions at any fee to ruin mining for everybody else.

Not really.  If 10% of the computational power drops it's fee requirement, then yes you can send free transactions, but you have to wait (on average) 10 times as long for it to be included in a block.  This is a big problem for a lot of uses.

[ryepdx]

Haven't read the entire thread yet and don't really have time to explain my ideas in detail, but I feel they're important enough for me to at least give you my conclusion ASAP:
If we move to an inflationary model with miners always getting 50btc for solving a block, this will lead to a single interest controlling the majority of the network. If that interest proves at all hostile or even unscrupulous, it could prove to be the end of bitcoins. At least with the current model we're rewarding the people who put the hard work in to get the ball rolling in the first place.

[casascius]

This is a big problem for a lot of uses.

Any uses in particular you can name?

It is pretty well established that long term, Bitcoin won't be able to sustain the volume of handling microtransactions, and that those will need to go through a BTC "bank" like mybitcoin.com.

And without a BTC "bank", Bitcoin can't offer instant confirmation of any payment, as might be needed at point of sale.  Aside from that, what kind of transaction is acceptable with an average 5 minute wait that would not be acceptable with a 50 minute wait, and will this type of transaction be the most common on the network?  Someone who needs rapid confirmation (like a point-of-sale merchant) would use a BTC bank anyway, which could offer an instant confirmation regardless of the mining environment.

[casascius]

If we move to an inflationary model with miners always getting 50btc for solving a block, this will lead to a single interest controlling the majority of the network.

Is there any particular reason for this conclusion?

[casascius]

Miners will ultimately charge a transaction fee that allows them to continue operating profitably

The problem is that miners can't make anybody pay it, and if they don't pay it, operating profitably is not a choice.  Their only choices are to operate non-profitably, or to stop operating.

The buzz has been that Bitcoin is good for micropayments, but then there is the whole 'snack-machine' problem having to do with the amount of time it takes for confirmaitons, etc

Bitcoin's block chain won't be suitable for snack machines or micropayments long term.  It's simply not resource effective.  You don't have a kilobyte of disk space or internet bandwidth to dedicate to each pack of gum sold everywhere in the world, let alone every snack.  A billion people probably buy a snack every day... that's a terabyte per day every full Bitcoin client must download and dedicate just to storing snack transactions.  So considerations made towards speeding up confirmations on the block chain are just preparation for something that won't happen because it's impossible on its face.  Only a "mybitcoin"-type bank can enable snack machine transactions and micropayments, by accounting for them internally, isolating them from the block chain except in the aggregate as wholesale settlement transactions.

[ryepdx]

If we move to an inflationary model with miners always getting 50btc for solving a block, this will lead to a single interest controlling the majority of the network.

Is there any particular reason for this conclusion?

I have a second, so I'll go into a little more detail:
With miners getting 50btc per block ad infinitum we end up creating an exponential earnings curve. Whoever is ahead on that curve will eventually see their lead greatly exaggerated such that the majority of bitcoins created each day will be created by their cluster. This will be accompanied by their overshadowing the power of the rest of the network. They will comprise over 50% of the total network's computing power which, as I understand, is all that is necessary to create a hostile takeover.

Okay, girlfriend is pulling me away. Let me know if I'm barking up the wrong tree here.

Thanks!

[mndrix]

This is a big problem for a lot of uses.

Any uses in particular you can name?

CoinPal is one such use.  A 10 minute confirmation is fine.  However, if a transaction misses a block, I get customer support emails asking "where are my coins?"  I've considered paying transaction fees because they're far cheaper than my customer support time.

CoinCard customers are another case.  Many of them want their PayPal payment immediately.  If their transaction doesn't make it into the next block, they email asking what's happened.  So far, all those with large (size) transactions have been happy to pay a transaction fee the next time.

Dealing with customers in day-to-day business has persuaded me that transaction fees will work great in the long run.

[MoonShadow]

If we move to an inflationary model with miners always getting 50btc for solving a block, this will lead to a single interest controlling the majority of the network.

Is there any particular reason for this conclusion?

I have a second, so I'll go into a little more detail:
With miners getting 50btc per block ad infinitum we end up creating an exponential earnings curve. Whoever is ahead on that curve will eventually see their lead greatly exaggerated such that the majority of bitcoins created each day will be created by their cluster. This will be accompanied by their overshadowing the power of the rest of the network. They will comprise over 50% of the total network's computing power which, as I understand, is all that is necessary to create a hostile takeover.

Okay, girlfriend is pulling me away. Let me know if I'm barking up the wrong tree here.

Thanks!

Too many outside variables to know if this would actually happen.

[MoonShadow]

Miners will ultimately charge a transaction fee that allows them to continue operating profitably

The problem is that miners can't make anybody pay it, and if they don't pay it, operating profitably is not a choice.  Their only choices are to operate non-profitably, or to stop operating.

They don't have to make anyone pay it.  There will be those who are willing to pay it for the advantages that it holds.  The 'free' section of the block is very limited, and in a successful future, very few people will be willing to wait for those free transactions to process, as they could take weeks in a future that Bitcoin processes at a rate comparable to Visa.

The buzz has been that Bitcoin is good for micropayments, but then there is the whole 'snack-machine' problem having to do with the amount of time it takes for confirmaitons, etc

Bitcoin's block chain won't be suitable for snack machines or micropayments long term.  It's simply not resource effective.  You don't have a kilobyte of disk space or internet bandwidth to dedicate to each pack of gum sold everywhere in the world, let alone every snack.  A billion people probably buy a snack every day... that's a terabyte per day every full Bitcoin client must download and dedicate just to storing snack transactions.  So considerations made towards speeding up confirmations on the block chain are just preparation for something that won't happen because it's impossible on its face.  Only a "mybitcoin"-type bank can enable snack machine transactions and micropayments, by accounting for them internally, isolating them from the block chain except in the aggregate as wholesale settlement transactions.

Bitcoin is actually pretty crappy for micropayments, but variations on the online wallet service isn't the only way to do instant small value sales.  The double spend attack is relatively difficult to pull of, due to timing issues, so it's very unlikely that anyone is going to attempt it for a free soda and a bag of chips.  Even so, a vending machine client can reduce it's own exposure to such an attack in a couple of different ways.  Blockchain confirmations aren't even important in this context, as the transaction (once verified for validity) can be accepted at face value with as much confidence as it can accept a quarter dropped in the change slot isn't a slug.  Honestly, it's just a machine, and can't know; but the vending owner includes a risk premium to cover his own backside.

I personally know someone who owns a number of vending machines and video games at a few local bars, I need to have a talk with him about Bitcoin.  It might be time to put theory into action.

[ryepdx]

With miners getting 50btc per block ad infinitum we end up creating an exponential earnings curve. Whoever is ahead on that curve will eventually see their lead greatly exaggerated such that the majority of bitcoins created each day will be created by their cluster.

Too many outside variables to know if this would actually happen.

What kind of outside variables are there? Mining scales exponentially. Furthermore, we've seen this sort of action happen in the free market before. First there was IBM, then there was Microsoft. They own well over half the market share. If we want to keep the network from falling under the control of a single person or a small cabal, we will have to take steps to ensure that a Microsoft or IBM never emerges in the bitcoin network. The problem would be even more insidious for us because there would be only one product and no for clusters to differentiate themselves. The various products of mining clusters are qualitatively the same. There would be no room for a cluster with a sustainable competitive advantage to emerge. It's good that the incentive for an individual to own a large cluster will slowly diminish. The ideal network model would consist of a large number of small clusters working together to create a secure network. A system that encourages monolithic clusters would only hurt us.

I think that if we stick to a deflationary model, transaction fees will prove to be profitable enough. There still exists the potential for an exponential earnings curve there, but growth from transaction fees will likely be so gradual that it shouldn't matter much. If it turns out that transaction fees aren't enough to keep people mining, then we come back to the concept of security audits. If an audit shows that the network is vulnerable to a cluster that likely exists, then it will be up to those with the most to lose (i.e, those with the most bitcoins) to allocate boxes to "mining." (Though at that point the term "mining" will be a misnomer.)

I honestly think that the cause of keeping our currency network safe will be compelling enough to attract a sufficient number of miners, especially if the data on our network's security is easily accessible. There are massive networks dedicated to such tasks as searching for extraterrestrial life and folding proteins whose participants receive no incentive beyond the knowledge that their idle cycles are being used to make the world a somewhat better place.

An inflationary model just seems rather unattractive to me. If given the choice between holding a currency in an inflationary trend or holding a currency in a deflationary trend, I would rather hold the latter. If we succeed in our goals, the early adopters will end up spending their bitcoins. Yes, they will be fabulously wealthy, but I have a hard time seeing that as a good reason for decreasing the wealth of the community as a whole via inflation.

Just my 2btc. What do you think?