Gold collapsing. Bitcoin UP.

[adamstgbit]

I did some rough calculations, trying to project the amount of fees we can reasonably expect without fee pressure. The avg fee / kb we are currently raking in, over the last 40 blocks, is surprisingly high, 0.000227 BCH / kb which translate to about 7cents pre TX

so a 1GB block would produce ~150,000$ worth of fee each block, ~21million dollar a day of fees paid to sercure the blockchain, sounds very good!

But i find it Odd that currently without fee pressure we're seeing 7cent fees

 

[NewLiberty]

The fee discussion mostly ignores economics.

Fees are not fees. These are voluntary contributions to miners. We, who desire Bitcoin to exist, use bitcoin. We make donations to mining when we do this to insure the existence of Bitcoin. Full Sotp.

 

[adamstgbit]

@NewLiberty

thats fine and dandy until someone comes up with some crazy blockchain app that produces 1GB of data every 10mins and is only willing to pay 500$ of fees pre GB.

Will bitcoincash incress blocksize or completely remove the cap, only to fill the blockchain will GB's of spam!?

this has been a strong point from core, this idea of "unbound TX demand", i think core has a point, but its not justification to allow fee to go up to >1$. some balance (blockspace Vs TX-demand) is required.

 

[satoshis_sockpuppet]

@adamstgbit

miners will either set a reasonable limit that produces optimal fees, or they will be unprofitable.

And why on earth would they ever do the second?

It has been said over and over, no miner has an interest in producing data packages they can't distribute.

Miners have been socially engineered into hurting their own business (and the wealth of the rest of the world) by the same old tactics.. won't somebody think of the poor ..children.. eh full nodes!

 

[NewLiberty]

I don't see it as a strong point from Core. It is possibly the weakest argument I have seen from them.

I see it as more of a fundamental mis-characterization of humanity which ignores the enlightened self-interest in miners who as custodians of the commons ought to exhibit the mechanisms of greed and profit motive. Those which do not, do not survive.

The argument affirms the consequent, a logical fallacy. It presumes a necessity of governance in the absence of any evidence for it.
https://www.logicallyfallacious.com/tools/lp/Bo/LogicalFallacies/14/Affirming-the-Consequent

Miners have the tools to decide what transactions to include in blocks.
Furthermore arguments are given for both the supply and demand sides!

A fee market is needed because otherwise there will be no support for mining.
Too much electricity is used which is unsustainable.
Either mining is too profitable, or not profitable enough?

Laffer curve handles such to the extent markets are efficient, which through recursive competition leads to ever more efficient markets. The attempts to create artificial inefficiencies in the market via arbitrary caps distorts this. Avoiding these distortions is why there is BU and now Bitcoin Cash.

There is a Nash equilibrium at play here. It optimizes via an elegant social construction. It is only by ignoring this that an attempt to destroy it might have succeeded if not for a brave few, who deserve our gratitude.

Satoshi gave birth to Bitcoin, but Bitcoin Cash and BU saved its life.

 

[Zangelbert Bingledack]

@adamstgbit The orphan risk analysis still applies. Why risk a significant chance of losing $100,000+ in an attempt to gain $500?

But if that's not convincing, I thought of a new angle on all this.

To echo @awemany's old observation, somehow putting software between a bunch of humans making decisions fools us into thinking it isn't just a bunch of humans making decisions.

And how do humans make decisions on things like what blocks they will mine and mine atop? They communicate to get a feel for more detailed guidelines than just those for the size of blocks. Most miners may have a "reality check" hard limit of 1GB (for now) hardcoded into their software, but through normal communication via miner consortiums and standards committees, a miner can be quite confident of the conditions that will allow their block to be accepted. Miners should be running stress tests with each other constantly to determine what parameters are definitely safe and when, below any hardcoded limits.

It could be something like a consortium of 80% of the hashpower agrees to orphan any blocks that are very large compared to their fee total, or the average size / fee total recently. Miners not deterred by "hard" network orphan risk have to contend with this additional soft orphan risk. They also might be excommunicated. They also might be concerned with price drops due to people being up in arms (rightly or wrongly) about their "full node" being bumped off.

There are a lot of factors, and as I've said before it works like a Keynesian beauty contest, where the ladies are blocks: you've got to choose the block you think the majority is most likely to accept. This folds back in on itself with the result that miners are rather conservative even within the softer guidelines and won't do things like try to screw over poorly connected miners (what pool is so tiny as to not be able to afford fast connectivity anyway?).

The same applies to removing the cap completely. Again it's not like devs set policy; miners do. BU makes the point clearest, as usual, because it lets any miner remove any "reality checks" if they choose to. Bitcoin survives by mining incentives, so no need to babysit them anyway.

What some picture when they think of having no hardcoded blocksize cap is something like, "This software will *force* you to accept a block no matter what the size." No software can ultimately force you to do anything, but BU (and ABC) doesn't even try to do that. It lets the user decide, because it doesn't bake in a "devs are part of the governance structure" philosophy into its code, unlike Core.

 

[adamstgbit]

@NewLiberty

if storing and transmitting 1GB of data for 500$ is worth it. then the blockchain WILL be filled with le spam! is that what you want!?

MAYBE miners will understand that filling GB blocks right now for only 500$ is profitable from them in the short term but is harmful in the long term, so they wont do it.

this just proves my point that miners need a blocksize limit, if there is none then they will be forced to make one up. why not give them the tools to come to an agreement on this issue Via BU's EC?

 

[NewLiberty]

The notion that software should force a business to be profitable rather than guide it to profitability with advance decision-making tools, is the means to destroying the very metrics needed to determine profitability.

 

[adamstgbit]

@adamstgbit The orphan risk analysis still applies. Why risk a significant chance of losing $100,000+ in an attempt to gain $500?

the orphan risk establishes a cost per byte which might be very very low, which will probably be something like 1GB of data needs at least 450$ of fees with it, inoder to be worth the risk.

orphan risk alone will not stop the spam attack that kills bitcoin decentralization
[doublepost=1511466819][/doublepost]

The notion that software should force a business to be profitable rather than guide it to profitability with advance decision-making tools, is the means to destroying the very metrics needed to determine profitability.

i would argue that BU's EC is a advanced decision-making tool, guiding miners to profitability.

 

[Zangelbert Bingledack]

the orphan risk establishes a cost per byte which might be very very low, which will probably be something like 1GB of data needs at least 450$ of fees with it, inoder to be worth the risk.

orphan risk alone will not stop the spam attack that kills bitcoin decentralization

That looks self-contradictory. Low orphan risk means most other miners accept it, which means most other miners believe almost all other miners will be able to handle it (Keynesian beauty contest) and that investors will not be spooked, which practically guarantees it isn't hurting miner decentralization.

By the time we reach a blocksize that starts to make some small mining pools infeasible, which I'm guessing will realistically be much higher than a gigabyte, people will start to consider the tradeoffs (unless new mining pools are coming online faster than the ones that are lost, which seems likely). Anyone have quick and dirty numbers on how much money one of the smallest mining pool operators might be making per month and the cost of whatever connections enable gigabyte, terabyte, and petabyte blocks?

 

[NewLiberty]

@NewLiberty

if storing and transmitting 1GB of data for 500$ is worth it. then the blockchain WILL be filled with le spam! is that what you want!?

MAYBE miners will understand that filling GB blocks right now for only 500$ is profitable from them in the short term but is harmful in the long term, so they wont do it.

this just proves my point that miners need a blocksize limit, if there is none then they will be forced to make one up. why not give them the tools to come to an agreement on this issue Via BU's EC?

Interestingly, I see the same proposal as proving the opposite of your point.

As there does exist today, apps that can attempt to store and transmit 1GB of data for $500. If it were worth it, the blockchain would be filled with these.
The blockchain is not filled with these, ergo it is not worth it to attempt.

By virtue of this, we do not need a blocksize limit.
===========================================

There may come a time when the Core folks may try to flood the Bitcoin Cash network with cheap transactions.
I will welcome it.
When they do, the profitability for mining Cash will soar, the attackers will be impoverished, hashpower will flock to Cash, and prices will flip.

 

[Zangelbert Bingledack]

I remember in 2015 some Chinese miners objected to 20MB citing China's slow internet and asked for 8MB instead. This just says to me that hashrate isn't the only factor that matters for the mining network, and that small blocks are what is centralizing mining in China. Big blocks would switch the geographical mining map from an energy cost map to an energy cost map superimposed on a connectivity map, implying much better geographical distribution, ironically all the more because of the Great Firewall.

 

[adamstgbit]

That looks self-contradictory. Low orphan risk means most other miners accept it, which means most other miners believe almost all other miners will be able to handle it (Keynesian beauty contest) and that investors will not be spooked, which practically guarantees it isn't hurting miner decentralization.

By the time we reach a blocksize that starts to make some small mining pools infeasible, which I'm guessing will realistically be much higher than a gigabyte, people will start to consider the tradeoffs (unless new mining pools are coming online faster than the ones that are lost, which seems likely). Anyone have quick and dirty numbers on how much money one of the smallest mining pool operators might be making per month and the cost of whatever connections enable gigabyte, terabyte, and petabyte blocks?

non-minning-full-nodes Are important
if a small-bitcoin-businesses needs a 20K machine to run a node, that HURTS.
i guess SPV could help, depending on how secure they are... idk... hasn't been proven to work so i'll ignore it.

There may come a time when the Core folks may try to flood the Bitcoin Cash network with cheap transactions.
I will welcome it.
When they do, the profitability for mining Cash will soar, the attackers will be impoverished, hashpower will flock to Cash, and prices will flip.

it wont be core, poeple will say "gr8 bitcoin cash allows for unlimited TX at near zero cost, now i can try my cool app that pumps 5million TX a day"

5million tx a day paying 1 sat / 256bytes, is a real thing that bitcoin cash will need to learn how to deal with.

if the answer is let the blocksize be 1GB right now to include this spam, i'm going to be disappointed.

 

[Erik Beijnoff]

What is this talk of spam I'm hearing? If the miners can be convinced to include a transaction, why should we question their reasons for doing so?

 

[adamstgbit]

miners are probably totally fine with including 1GB blocks pretty much right now, given the right sotfware, they have the hardware and for miners a onetime 50K opperating cost, is nothing.

however this would clearly decimate the non-minning full nodes.

is that OK?

are we willing to decimate node count, because someone came up with a use case that pumps out 5million tx a day and pays miners a minimal amount.

 

[adamstgbit]

these questions will be much tougher to answer in a year when bitcoin cash is faced with "unbound spam demand"

when we get to the 8MB limit, because of 1/10th of cent TX fees, its going to be hard to use that spam as justification to lift the 8MB limit.

 

[Erik Beijnoff]

miners are probably totally fine with including 1GB blocks pretty much right now, given the right sotfware, they have the hardware and for miners a onetime 50K opperating cost, is nothing.

however this would clearly decimate the non-minning full nodes.

is that OK?

are we willing to decimate node count, because someone came up with a use case that pumps out 5million tx a day and pays miners a minimal amount.

Absolutely. If we would go to 1GB blocks tomorrow, I'm all for. That is, if the demand was there, which it currently isn't.

Non-mining nodes that don't have an economic function is just a burden for the network. They have about the same influence over the network as a person that listens to broadcast radio have over which programs that are being broadcasted. Probably less since broadcast radio does regular polls to deem the popularity of the content.

Side note: Economic function does not include a small retailer that wants to set up a node just because he think that's the way to do it. He could just instead subscribe to a node service, existing for that specific purpose.

With that said, I value the possibility of letting high-end hobbyist follow the chain, to keep the barriers for entry low. A high-end hobbyist would be someone that is prepared to pay about 7000-10000 dollars yearly for a professional setup + network connection. This goal can easily be achieved at that price if blocks are 100MB. I'm guesstimating that 1GB blocks are almost reachable today as well with a setup at that price.
[doublepost=1511476076][/doublepost]To put it blunty, this naïve theory of unbounded demand for cheap or free transactions is nonsense. You don't see people storing genome data on Google Drive, despite the almost unlimited and free storage possibilities.

Apart from that, no sane miner would include transactions below the raw cost of doing so. In the same way that a salesman at a bazaar doesn't bargain below the price that he acquired his goods for, the miners won't include transactions that are paying less than the cost of including them.

 

[solex]

Once a free-thinker, always a free-thinker.

 

[AdrianX]

It could be something like a consortium of 80% of the hashpower agrees to orphan any blocks that are very large compared to their fee total, or the average size / fee total recently. Miners not deterred by "hard" network orphan risk have to contend with this additional soft orphan risk. They also might be excommunicated. They also might be concerned with price drops due to people being up in arms (rightly or wrongly) about their "full node" being bumped off.

There is the incentive to relay and prioritize transactions by fee, with Xthin you want all nodes to have the transactions already.

Aa cretin percent of free transaction are always relayed, but when the network is overburdened with free or extremely low fee transactions it's not in your best interest as a node to congest the network and relay all transactions. For a miner using Xthin if they want optimal block propagation they want to avoid uncertainty, they want to include transaction they know all other miners already have.

so they will construct blocks to have transaction that have propagated the network with relative certainty.
[doublepost=1511480731][/doublepost]

orphan risk alone will not stop the spam attack that kills bitcoin decentralization

@theZerg did a paper on the security of single transaction blocks, and it deals with this issue. If big blocks that under pay are propagated (I don't think it can happen the cost of running a mining operation is too high) but lets say an excessive block is mined and not orphaned, you are incentivized to mine an empty block on top for security and while that happens a fee market builds. Miners then optimize prioritizing high paying transactions in future blocks.
[doublepost=1511481282,1511480270][/doublepost]

these questions will be much tougher to answer in a year when bitcoin cash is faced with "unbound spam demand"

Bitcoin has always had this problem but no one has ever spammed the network. it's only when 1MB limit was limiting transaction volume did we see low fee transaction force up transaction fees.

Just to increase the the block size once by 1MB you need to spend around $3000 in $0.01 transaction fees. the cost to store that data for my lifetime is just pennies. this notion of Spam is a red herring.

 

[Norway]

Another great vid from Daniel Krawisz: