Enforcing a production quota on block space that fights the free market

[brg444]

"will of the market" != "most of us"
 

Perhaps true.  But now you're talking about wealth inequality, which Bitcoin never claimed to fix to begin with.

I think that's irrelevant to Peter's question about the economic majority even though he did use that phrase (most of us).

It is true that the will of the market does not necessarily mean most of us.  But the will of the market wants to be at Q*.  The production quota is forcing it to be at Qmax.  

He does the market enforce a production quota against itself?

I'm sure the will of the market also wants Ferraris to be at Q* but it ain't happening. Some will have to settle for the back of the bus.

[Peter R]

Let's say ITBL technology gets implemented, which means basically that blocks don't need to be broadcasted anymore. Then the market equilibrium shifts completely to the right of your graph and now miners are incentivised to confirm as many transactions as possible (with fees>0) to max their revenue…

Thanks for the comment, JeromeL.  

Coding schemes like IBLT are exactly what we need to reduce the propagation impedance to permit miners to produce large blocks for a reasonable price.  I believe Rusty Russell said that IBLTs were achieving a coding gain of around 30.  This means that if all miners used this scheme, they could produce 30 MB blocks for the same cost that they are presently producing 1 MB for.  Win Win.  

However, note that if even 10% of the hash power doesn't go along with the mempool sync, that this has a very large affect on the propagation impedance.  

It will not be easy for miners to affordably produce 100 MB blocks, for example.

This chart explains how propagation impedance affects the total cost of producing various sizes of blocks.  

[jonald_fyookball]

"will of the market" != "most of us"
 

Perhaps true.  But now you're talking about wealth inequality, which Bitcoin never claimed to fix to begin with.

I think that's irrelevant to Peter's question about the economic majority even though he did use that phrase (most of us).

 

Precisely. Then you understand how that applies to Bitcoin right?

You realize how the "economic majority" is reflected in the distribution of power?

Please explain your point better because I don't know what you mean.

[brg444]

Let's say ITBL technology gets implemented, which means basically that blocks don't need to be broadcasted anymore. Then the market equilibrium shifts completely to the right of your graph and now miners are incentivised to confirm as many transactions as possible (with fees>0) to max their revenue…

Thanks for the comment, JeromeL.  

Coding schemes like IBLT are exactly what we need to reduce the propagation impedance to permit miners to produce large blocks for a reasonable price.  I believe Rusty Russell said that IBLTs were achieving a coding gain of around 30.  This means that if all miners used this scheme, they could produce 30 MB blocks for the same cost that they are presently producing 1 MB for.  Win Win.  

However, note that if even 10% of the hash power doesn't go along with the mempool sync, that this has a very large affect on the propagation impedance.  

It will not be easy for miners to affordably produce 100 MB blocks, for example.

This chart explains how propagation impedance affects the total cost of producing various sizes of blocks.  

Oh but I beg to differ. All it takes is some good ol cooperation.

[brg444]

"will of the market" != "most of us"
 

Perhaps true.  But now you're talking about wealth inequality, which Bitcoin never claimed to fix to begin with.

I think that's irrelevant to Peter's question about the economic majority even though he did use that phrase (most of us).

 

Precisely. Then you understand how that applies to Bitcoin right?

You realize how the "economic majority" is reflected in the distribution of power?

Please explain your point better because I don't know what you mean.

To put it simply Peter insists that the shouts of the poor economic minority begging for cheaper transactions somehow is representative of "the market". This likely hails from the broken American ideology that "the consumer is always right" or, as Mircea puts it well, that "people who have never interracted with any other aspect of economy besides the supermarket counter may genuinely imagine that's what economy is.".

iCEBREAKER aptly points out that Bitcoin being an economy, the distribution of its power (or wealth) answers to Pareto's law. Bitcoin being the most capitalist thing on this face of the earth acknowledges and exists to give the wealthy complete exercise over their power.

This is also true of the actual economic majority in the current fiat economy. It is not so much after "cheap" transaction alternatives but seeks to escape political friction. Therefore the construction that all the monies of the fiat economy will be deterred to enter Bitcoin because of its financial cost is either economical incompetence or outright lies and fabrications.

[jonald_fyookball]

To put it simply Peter insists that the shouts of the poor economic minority begging for cheaper transactions someone is representative of "the market".
 

Not sure he's claiming that.  You're putting words in his mouth.  Peter, is that what you're saying?

[Peter R]

To put it simply Peter insists that the shouts of the poor economic minority begging for cheaper transactions someone is representative of "the market".

iCEBREAKER aptly points out that Bitcoin being an economy, the distribution of its power (or wealth) answers to Pareto's law. Bitcoin being the most capitalist thing on this face of the earth acknowledges and exists to give the wealthy complete exercise over their power.

I still don't think you're answering the question asked in the OP.  Even if you disagree, assume for a moment that Q* is to the right of Qmax as shown below.  By definition then, this means that the market wants more block space than is permitted by the quota.  What I'm asking is--assuming the picture does in fact look like this--what can be done to enforce the quota?  How can the market impose a quota that the market itself does not want?

[JeromeL]

However, note that if even 10% of the hash power doesn't go along with the mempool sync, that this has a very large affect on the propagation impedance.  

It will not be easy for miners to affordably produce 100 MB blocks, for example.

If 10% doesn't go along with the mempool sync, then they will be mining from an old block and it's a the big win for the big block miner because that is 10% less competition.

IBLT was just an example to illustrate the the main point that you have not responded to, so let me sum it up without any example to get distracted on :

The maxblocksize controversy is not about economics. It's about security, censorship-resistance and permissionlessness.

The free market and Bitcoin's fundamental properties aren't necessarily aligned : your economics approach is irrelevant to the blocksize debate.

[Peter R]

If 10% doesn't go along with the mempool sync, then they will be mining from an old block and it's a the big win for the big block miner because that is 10% less competition.

Can you explain this further?  Are you suggesting that the 10% would be forked off the network even though they are producing valid blocks?

[brg444]

To put it simply Peter insists that the shouts of the poor economic minority begging for cheaper transactions someone is representative of "the market".

iCEBREAKER aptly points out that Bitcoin being an economy, the distribution of its power (or wealth) answers to Pareto's law. Bitcoin being the most capitalist thing on this face of the earth acknowledges and exists to give the wealthy complete exercise over their power.

I still don't think you're answering the question asked in the OP.  Even if you disagree, assume for a moment that Q* is to the right of Qmax as shown below.  By definition then, this means that market wants more block space than is permitted by the quota.  What I'm asking is--assuming the picture does in fact look like this--what can be done to enforce the quota?  How can the market impose a quota that the market itself does not want?

Now you're just being dense.

How many time does it need to be told? The consensus rules enforce the quota.

Unless the market the peers can agree to change it nothing will happen. No, the market is not representative of the peers.

[Zarathustra]

"will of the market" != "most of us"

Why?  Because resources are distributed by Pareto power law, not imaginary one-mouth-one-cow Rawlsian social justice.

All of your considerations of spherical blockchains fail to account for the fact Bitcoin's economic majority is composed of a very small number of very rich, very smart, and very stubborn individuals.  They don't GAF how Reddit wants to redesign their coin for the greater glory of populism (and Goldman Sachs).

I'm not surprised you lacked to the class to resist trolling the scaling workshop.  Don't be shocked when you don't get invited to speak at the next one.





The area under the line represents lulz proportional to the degree of Peter R's failure to convince anyone important that XT/101 was a good idea.

So funny. They are still fighting their proxy war against XT while they are realising that the market will raise the limit.

Are you not tired fucking this strawman ?

Are you not tired to spam the threads with 50 and more 'posts' a day?

https://bitcointalk.org/index.php?action=profile;u=251751;sa=showPosts;start=0

[brg444]

If 10% doesn't go along with the mempool sync, then they will be mining from an old block and it's a the big win for the big block miner because that is 10% less competition.

Can you explain this further?  Are you suggesting that the 10% would be forked off the network even though they are producing valid blocks?

They would be forked off by not being able to compete with the 90% of hashing power of rest of the network to propagate their block.

[Peter R]

If 10% doesn't go along with the mempool sync, then they will be mining from an old block and it's a the big win for the big block miner because that is 10% less competition.

Can you explain this further?  Are you suggesting that the 10% would be forked off the network even though they are producing valid blocks?

They would be forked off by not being able to compete with the 90% of hashing power of rest of the network to propagate their block.

So you mean they would go out of business (which is different than being forked off).  Anyways, I think a hypothetical network configuration like this is far-fetched (but the fee market would still hold, nonetheless). 

[brg444]

If 10% doesn't go along with the mempool sync, then they will be mining from an old block and it's a the big win for the big block miner because that is 10% less competition.

Can you explain this further?  Are you suggesting that the 10% would be forked off the network even though they are producing valid blocks?

They would be forked off by not being able to compete with the 90% of hashing power of rest of the network to propagate their block.

So you mean they would go out of business (which is different than being forked off).  Anyways, I think a hypothetical network configuration like this is far-fetched (but the fee market would still hold, nonetheless).  

But still you ignore Jerome's point.

Bitcoin's objective is not to maximize miners profit using a free market approach but set security limits to maintain its decentralized core.

[sAt0sHiFanClub]

... MPfag crud ....

Bingo!  

Listen.  7 transactions per second.  Repeat that to yourself a few times until it sinks in.

7 transactions per second

Global reserve currency?  @ 7tps, for who exactly?

What a joke.

Fedwire & TARGET2 run at 4 TPS  

We will see in due time if we can scale this a bit. For now, we need to decruft the turd of its inefficiencies and build the eco-system it will support.

Fedwire is a gross settlement network, not at all what bitcoin is supposed to be. A single settlement tx could reflect millions of other transactions - out of band from that network. I think that is the ultimate goal for yourself and the MPfags on here. Your dream is that all sidechain/LN activity will have to be settled with bitcoins, and of course the wealthy bankers will have to come to you on bended knee in the hope you might sell them for the equivalent of the GDP of a small country.

Sorry, aint gonna happen.  Its such a ludicrous vision.

All the talk of economics and miner behavior is a side show. There will be a use case for side chain like features in bitcoin ( implemented by alt-coins) in the future, but that is years ahead, and it will be implemented in the context of blocks and transaction rates that will grow with adoption.

[brg444]

... MPfag crud ....

Bingo!  

Listen.  7 transactions per second.  Repeat that to yourself a few times until it sinks in.

7 transactions per second

Global reserve currency?  @ 7tps, for who exactly?

What a joke.

Fedwire & TARGET2 run at 4 TPS  

We will see in due time if we can scale this a bit. For now, we need to decruft the turd of its inefficiencies and build the eco-system it will support.

Fedwire is a gross settlement network, not at all what bitcoin is supposed to be. A single settlement tx could reflect millions of other transactions - out of band from that network. I think that is the ultimate goal for yourself and the MPfags on here. Your dream is that all sidechain/LN activity will have to be settled with bitcoins, and of course the wealthy bankers will have to come to you on bended knee in the hope you might sell them for the equivalent of the GDP of a small country.

YES!

Better start buying 

[TransaDox]

But let me point out that the maxblocksize controversy is not about economics. It's about security, censorship-resistance and permissionlessness.

I agree with most of your views but I think this is stretching the importance. If you remove political reasons there is basically one reason for the block debate and that is transactions per second. As much as someone may or may not see it as a scalability issue now or in the future, the block size is the [other] easy way to increase transactions per second if there is a limit imposed.



I think you can see that if the dream of 4000 tx/sec is ever to be realised then 1 or 8 MB is rather moot if you cannot change the 10 minute interval.

Where Peter comes in, this limit (whatever it may be) creates a market which he is attempting to analyse. That's fine. But it is an artificially created market due to a design decision. Where I sit (and I think you do to) is that we need to remove the temporary spam limit which would remove his market and a different analysis would come into play - one of optimum block size for best return (blocks size vs orphans). His analysis also covers that, I believe but the buzzword at the moment is the "fee market" which is artificial, to you and I, and more of a bug that needs to be fixed rather than exploited.

One way we might be able to remove the limit completely might be to not allow miners to increment the coinbase integer and instead they must include new transactions to change the hash. This would create a demand for any transaction, not just paid-for ones and make spam/dust/small transactions useful for miners to move forward. They would have to consume any and all transactions just to produce a new hash but can choose a strategy that fits their bandwidth, mining power and other physical and financial constraints but ultimately serving the network rather than the network serving them.

[brg444]

But let me point out that the maxblocksize controversy is not about economics. It's about security, censorship-resistance and permissionlessness.

I agree with most of your views but I think this is stretching the importance. If you remove political reasons there is basically one reason for the block debate and that is transactions per second. As much as someone may or may not see it as a scalability issue now or in the future, the block size is the [other] easy way to increase transactions per second if there is a limit imposed.



I think you can see that if the dream of 4000 tx/sec is ever to be realised then 1 or 8 MB is rather moot if you cannot change the 10 minute interval.

Where Peter comes in, this limit (whatever it may be) creates a market which he is attempting to analyse. That's fine. But it is an artificially created market due to a design decision. Where I sit (and I think you do to) is that we need to remove the temporary spam limit which would remove his market and a different analysis would come into play - one of optimum block size for best return (blocks size vs orphans). His analysis also covers that, I believe but the buzzword at the moment is the "fee market" which is artificial, to you and I, and more of a bug that needs to be fixed rather than exploited.

One way we might be able to remove the limit completely might be to not allow miners to increment the coinbase integer and instead they must include new transactions to change the hash. This would create a demand for any transaction, not just paid-for ones and make spam/dust/small transactions useful for miners to move forward. They would have to consume any and all transactions just to produce a new hash but can choose a strategy that fits their bandwidth, mining power and other physical and financial constraints but ultimately serving the network rather than the network serving them.

Again this ignores the nodes.

Our goal should not only be to optimizes miners revenue but to do so in context of the costs externalized to nodes.

If we remove this limit the blockchain will grow too fast for nodes to keep up and within 15-20 years no new nodes will be able to enter the network.

[TransaDox]

Again this ignores the nodes.

Our goal should not only be to optimizes miners revenue but to do so in context of the costs externalized to nodes.

If we remove this limit the blockchain will grow too fast for nodes to keep up and within 15-20 years no new nodes will be able to enter the network.

My solution is not to optimise miners revenue. On the contrary, miners would resist, I think, because they would be dependant on being fed transactions so transaction propagation would play a part not just block propagation. Nodes are a different problem and one of static storage and incentive so yes. I did ignore them. I have mentioned alternatives for them in other threads for the storage. I have no answer currently for their incentive as I am one that has turned off my node.

[sAt0sHiFanClub]

Fedwire is a gross settlement network, not at all what bitcoin is supposed to be. A single settlement tx could reflect millions of other transactions - out of band from that network. I think that is the ultimate goal for yourself and the MPfags on here. Your dream is that all sidechain/LN activity will have to be settled with bitcoins, and of course the wealthy bankers will have to come to you on bended knee in the hope you might sell them for the equivalent of the GDP of a small country.

Sorry, aint gonna happen.  Its such a ludicrous vision.

All the talk of economics and miner behavior is a side show. There will be a use case for side chain like features in bitcoin ( implemented by alt-coins) in the future, but that is years ahead, and it will be implemented in the context of blocks and transaction rates that will grow with adoption.

YES!

Better start buying  

Banks have a scarce history of giving poor people huge quantities of their money, but this time its different, right?

But I applaud your optimism, even if it is foolhardy.