Gold collapsing. Bitcoin UP.

[Peter R]

This is not true. As I keep explaining, your theory is to allow supply to be driven by this orphan risk. The economic cost of the orphan risk would therefore be almost the same as fee revenue. The "advantage" is not 0.2% but could be 20%

It is true (if you still think it's not true, then pinpoint where in the math the problem occurs). Remember, I'm assuming 1% orphan rates, in line with the historical norms. Like I said, the advantage would be larger at higher orphaning rates.

You keep contradicting your initial point. Yes we all agree that miners will probably be able to propagate their blocks just fine. The point is your theory rests on the assumption that this is not the case and therefore orphan risk can drive fees. Please stop making this mistake again and again.

No I'm not. The problem is that you keep equating "fast propagation" to "instant propagation". I said that small miners will be able to propagate their blocks as quickly as large miners. I did not say that they will both be able to propagate their blocks infinitely fast. (The purpose of this comment was to explain that the "large miner advantage" is only due to the fact that large miners are more likely to solve two blocks in a row and the advantage has nothing to due with block propagation times to other miners.)

That is not my point. My point is that if larger blocks do cause propagation issues (a necessary condition for Peter's fee theory) then we have large issues in the network.

But larger blocks already cause propagation issues (i.e., large blocks propagate slower and are more likely to be orphaned as the data from a few pages back showed). So the necessary condition for the fee market theory to hold is satisfied (and has always been satisfied).

Wasted work is a significant problem, as I explained to Peter a while ago. Wasted work makes the network less secure and easier to attack. (less equilibrium difficulty level)

"Wasted work" results in a reduction in equilibrium difficulty like you said. At a 2% orphaning rate, the equilbrium difficulty would be 1% lower than if the orphaning rate were 1%. Now compare that to the variance in the difficulty each adjustment period: it is not uncommon for difficulty to change by 10%. So the "wasted work" is insignificant to the variance we see month over month.

Also, @AdrianX's argument.

a) That is what strong consensus means, if a significant minority is not happy then we do not go ahead. That is veto power. This power ensures that other people's money cant be changed without their consent. That is what makes Bitcoin unique. This characteristic must be defended at all costs as it makes bitcoin viable money.

I disagree. Bitcoin is governed by the market and will ignore a minority if it judges that to be the best course of action. So far, it seems your theory has been the correct one (I'm surprised we're still beholden to Blockstream/Core). I doubt it will appear as the correct one this time next year.

b) I am not worrying about trying to control anyone. It is just that in the event of attackers trying to hardfork without consensus, the incentives are structured such that the economic majority rallys behind the existing rules to defeat the attack. Even those that agree with the rule change (like me) rally behind the existing rules to protect the system. This looks like it is working and Classic is being defeated. If the defeat against Classic is resounding it should deter further attacks.

Let's say Core released a version that allowed the operator to OPTIONALLY show support for BIP109 (2MB HF at 75% activation). Would you consider that an attack?

 

[jonny1000]

"Remember, I'm assuming 1% orphan rates, in line with the historical norms. Like I said, the advantage would be larger at higher orphaning rates."

You submit an idea, people then criticize your idea based on the theory that orphan rate and risk would be driven up, then you refute that by assuming orphan rates remain unchanged.

"The purpose of this comment was to explain that the "large miner advantage" is only due to the fact that large miners are more likely to solve two blocks in a row and the advantage has nothing to due with block propagation times to other miners."

The advantage is related to propagation times to other miners.

"Wasted work" results in a reduction in equilibrium difficulty like you said. At a 2% orphaning rate, the equilbrium difficulty would be 1% lower than if the orphaning rate were 1%. Now compare that to the variance in the difficulty each adjustment period: it is not uncommon for difficulty to change by 10%. "

I do not agree with assuming the 1% level or comparing it to the 10% figure. I do not think you appreciate all the disadvantages of wasted work. It makes attacking the network easier to an extent greater than the corresponding fall in difficulty, as selfish mining is easier.

What I find really disturbing is that you seem to think higher orphan rates and wasted work are fine or are only a minor inconvenience. In contrast to this view, we need to ensure we make the system as robust, effective, convergent and efficient as possible, which means keeping wasted work and orphan rates as low as possible.

"So far, it seems your theory has been the correct one"

Thank you, I really appreciate this comment. I hope you begin to understand how important this is to many people. If I am shown to be wrong here, then I will have no particular interest in Bitcoin and will exit the ecosystem. This issue is so important to me that I feel it is necessary to demonstrate that it is correct by showing it in practice. The more people can be convinced of this the more true it is.

In my view, if Bitcoin succeeds, then attacks on the consensus rules in the future are inevitable. If we lose to this Classic attack now, then we are likely to lose more attacks in the future. Setting the precedent that these attacks fail is vital. Therefore I hope that you become convinced of this property or leave the ecosystem. As the higher the proportion of users who value this characteristic is, the stronger the system is. This is why I have gone to such excessive lengths to make sure Classic fails, despite being totally fine with 2MB (or even 8MB) blocks.

"Let's say Core released a version that allowed the operator to OPTIONALLY show support for BIP109 (2MB HF at 75% activation). Would you consider that an attack?"

That would be totally fine, even if it were not optional that would also be fine. I have no issue with a client that flags support or "shows support" for anything. Just like all the 8MB or BIP100 votes miners put in their headers, they are fine. My issue is a client that activates a hardfork without consensus. Classic has a methodology that means it activates with exactly 25% miner opposition. Classic is therefore clearly an attack client. Make it activate at 95% and I would not regard it has an attack. Make a client that flags support for 1,000GB blocks tomorrow, (that mentions it only activates when there is 99% opposition or more), or that "shows support" for murdering all the children in the world, I would not regard these as attacks, these are not attacks on Bitcoin.

 

[Zarathustra]

O Pieter ....

 

[Zarathustra]

hey @Zarathustra you'll like this one:

... Saudi Arabia can also add liquidity worries which just spilled out into the open, because Bloomberg reported moments ago, Saudi Arabia has told banks it is considering paying some outstanding bills to contractors with government-issued bonds, citing people with knowledge of matter say.

http://www.zerohedge.com/news/2016-05-18/saudi-arabia-admits-full-blown-liquidity-crisis-will-pay-government-contractors-ious

Yes, feudalist economies cannot produce as much debt as democratic ones which are able to manage as much as 500 percent debt in relation to gdp. Those feudalist governments cannot reduce interest rates to zero. All their capital would flee.

 

[Zarathustra]

"Let's say Core released a version that allowed the operator to OPTIONALLY show support for BIP109 (2MB HF at 75% activation). Would you consider that an attack?"

That would be totally fine, even if it were not optional that would also be fine. I have no issue with a client that flags support or "shows support" for anything. Just like all the 8MB or BIP100 votes miners put in their headers, they are fine. My issue is a client that activates a hardfork without consensus. Classic has a methodology that means it activates with exactly 25% miner opposition. Classic is therefore clearly an attack client. Make it activate at 95% and I would not regard it has an attack. Make a client that flags support for 1,000GB blocks tomorrow, (that mentions it only activates when there is 99% opposition or more), or that "shows support" for murdering all the children in the world, I would not regard these as attacks, these are not attacks on Bitcoin.

The opposite is true. A 95% rule (which is a 5% veto rule) is an attack on the Nakamoto 51% rule.

 

[jbreher]

1) As you allude, no alt coin has exceeded about 15-20% of Bitcoin's market-cap (IIRC), even during manias.

While some may not consider Ripple an altcoin, its market cap briefly exceeded that of bitcoin.
[doublepost=1463642825][/doublepost]On the thesis of 'Gold Collapsing'...

Certainly, everything that the central bankers have tried to encourage inflation (QE, ZIRP, NIRP, twist, ...) since 2008 has failed. But governments need inflation in order to be able to service their debt. Among the few things not yet tried are outright purchases of gold at high valuations.

Guess what the CB think tanks have started to talk about?

https://www.pimco.com/en-us/insights/viewpoints/viewpoints/rumpelstiltskin-at-the-fed

We may soon see a sea change to 'Gold up, Bitcoin up'. Or rather 'fiat down'. In many ways the same thing.

 

[freetrader]

The more people can be convinced of this the more true it is.

I have here a potion which, if imbibed, will help you to be convinced more easily, and thus help you to understand the truth.

<passes the kool-aid>

It is just that in the event of attackers trying to hardfork without consensus

The term "hardfork without consensus" has no meaning, since "consensus" is an obviously malleable concept that applies only within the circle of Core and its supporters / financiers, and the hardfork which you speak of does not occur or originate within that circle. Your "consensus" does not apply to it, nor limit its potential.

 

[jonny1000]

The opposite is true. A 95% rule (which is a 5% veto rule) is an attack on the Nakamoto 51% rule.

The Nakamoto consensus rule is for deciding on the one "true" valid version of the blockchain and enforcing any necessary existing rules. The whitepaper makes no mention of this mechanism being used to eliminate an existing rule.

Existing rules are enforced by all nodes, that is how the system was designed. You may not like it, but its a fact.

It appears large blockists seem not to agree with that, however please do not mischarecterise the small blockists by implying we do not agree with Nakamoto consensus. Everyone agrees on Nakamoto consensus for what is says it is for in the whitepaper.

 

[bluemoon]

O Pieter ....

I fear Pieter is no Zarathustra.
[doublepost=1463646706][/doublepost]

Guess what the CB think tanks have started to talk about?

https://www.pimco.com/en-us/insights/viewpoints/viewpoints/rumpelstiltskin-at-the-fed

I wonder how long the 1MB limit would last were CBs to contemplate buying bitcoin at vastly inflated values?

 

[bluemoon]

The Nakamoto consensus rule is for deciding on the one "true" valid version of the blockchain and enforcing any necessary existing rules. The whitepaper makes no mention of this mechanism being used to eliminate an existing rule.

The white paper made no reference to bitcoin being repurposed from electronic cash to a settlement layer, even supposing that to be possible.

And Satoshi made no reference to the 1MB blocksize limit, which he introduced only in September 2010 to protect against a network DoS attack, being repurposed as a means of preventing bitcoin becoming a "peer-to-peer electronic cash system".

It is in any case hard to imagine Satoshi ever regarded the 1MB limit as a "necessary existing rule". It was to be flexible. Disposable even.

E: Which means BScore have little justification for demanding 99% 'consensus' before agreeing to accept change to such a rule, a rule so close to their heart that they choose to found a business on it.

 

[jonny1000]

"The white paper made no reference to bitcoin being repurposed from electronic cash to a settlement layer, even supposing that to be possible."

When did I ever say that? This is a malicious rumor of what small blockists want, created by large blockists. The purpose of this is to help them win the argument as if the fight on a level playing field they lose.


"It is in any case hard to imagine Satoshi ever regarded the 1MB limit as a "necessary existing rule". It was to be flexible. Disposable even."

Yes, I know you do not like it or want it to be an existing rule. However what you need to accept is the fact that it is a rule. I do not like it either.

 

[bluemoon]

@jonny1000

"When did I ever say that? This is a malicious rumor of what small blockists want, created by large blockists. The purpose of this is to help them win the argument as if the fight on a level playing field they lose."

I have no idea what you have said, other than here.

But it is no rumour: it is clear BScore seek to repurpose bitcoin as a settlement layer for off-chain solutions such as LN, as @Jihan has described:
"Yes, I know you do not like it or want it to be an existing rule. However what you need to accept is the fact that it is a rule. I do not like it either."

You are conflating rules. The rule we do not like: 1MB blocksize; and the rule you seek to invent: the 95% 'consensus' requirement.

Jeff Garzik joined Classic on the basis that artificially constraining the blocksize would itself give rise to significant and unintended changes to the ecosystem (changes which your interpretation of consensus would enable a mere 5% 'consensus' to effect).

- https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2015-December/011973.html
- https://bitcoinmagazine.com/articles/bitcoin-economics-are-changing-1451315063

You have not provided any reason for stipulating a 95% 'consensus' requirement, still less why 5% should be able to change fundamentally the nature of bitcoin. The truth is the 51% Nakamoto consensus rule is for deciding whatever it can decide upon: you simply wish to see it restricted in scope; but no other consensus rule is defined.

 

[jonny1000]

"But it is no rumour: it is clear BScore seek to repurpose bitcoin as a settlement layer for off-chain solutions such as LN, as @Jihan has described:"

Yes exactly, Jihan is one of those large blockers spreading this divisive and malicious false rumor. I was at the HK agreement and his comments are false.

"You are conflating rules."

I am not, the only thing I was saying was a rule was the 1MB limit.#

"you seek to invent: the 95% 'consensus' requirement."

I am not inventing this. All full nodes fully validate the blockchain and the rules. That is what they do by design. If the rules are breached the nodes do not follow that chain.


"You have not provided any reason for stipulating a 95% 'consensus' requirement"

You are missing the point. 95% in itself does not matter. It is about the principal of not forcing an actual rule change on a minority. Classic deliberately and publicly chose 75% rather than 95% (which is the normal level for soft forks), specifically because they anticipated there would be significant opposition. This makes Classic an attack.

"still less why 5% should be able to change fundamentally the nature of bitcoin."

5% should be able to change Bitcoin if the method they use to do this change is keeping the rules the same. That is because ensuring a minority is able to block changes to the rules goes right to the core of what Bitcoin is, how it works and what makes it unique. Is this reason sufficient for you?


"The truth is the 51% Nakamoto consensus rule is for deciding whatever it can decide upon"

sure and one of the things it clearly cannot decide upon is to hardfork.

 

[Zarathustra]

The Nakamoto consensus rule is for deciding on the one "true" valid version of the blockchain and enforcing any necessary existing rules. The whitepaper makes no mention of this mechanism being used to eliminate an existing rule.

But the code makes it possible. Your 95% consensus is a joke, because the will of the 94% can be vetoed by the 6%. The whitepaper makes no mention of such perverted consensus jokes.

 

[freetrader]

one of the things it [the 51% Nakamoto consensus rule] clearly cannot decide upon is to hardfork

Anyone can hardfork at any time. It does not require an election, and therefore not 51%, 75%, or 95%, or 99% consensus of the mining fraternity or any other grouping.

If the hard fork changes the POW, then hashing power backing it is not lost to the existing Bitcoin ecosystem. If it retains the ledger, existing stakeholders are included by default and are not "attacked".
To call resulting market action in terms of valuing such forks an attack requires one to view the existence of any competing cryptocurrencies as attacks also, a view which is thoroughly anti-competitive and frankly, counterproductive.

 

[Melbustus]

While some may not consider Ripple an altcoin, its market cap briefly exceeded that of bitcoin.

Well, even if Ripple is included, that calc was always borked. IIRC, when that happened, coinmarketcap (or wherever) was using the full theoretical 100B XRP as the supply number, not the amount actually issued and outstanding which was far lower. That'd be like a company's market-cap getting calculated on the number of shares that are on the books as "authorized but not outstanding" (the 'wealth' in the stock market would be 100x if everyone did that). Makes no sense.

 

[bluemoon]

"But it is no rumour: it is clear BScore seek to repurpose bitcoin as a settlement layer for off-chain solutions such as LN, as @Jihan has described:"

Yes exactly, Jihan is one of those large blockers spreading this divisive and malicious false rumor. I was at the HK agreement and his comments are false.

It is not enough simply to assert Jihan is spreading falsehoods.

Jihan's account explains why BScore has a financial interest in restricting the blocksize limit and why the miners made their agreement and stick with it. You need to provide an alternative explanation and describe what really happened at the HK roundtable, an account which explains why Jihan said what he did.

 

[bluemoon]

5% should be able to change Bitcoin if the method they use to do this change is keeping the rules the same. That is because ensuring a minority is able to block changes to the rules goes right to the core of what Bitcoin is, how it works and what makes it unique. Is this reason sufficient for you?

The bitcoin network follows the longest chain of blocks. Its validity is determined by the majority of hashing power. No supermajorities are involved. Far from going to the core of what bitcoin is, the white paper does not envisage minorities having any veto.

 

[cypherdoc]

This is not true. As I keep explaining, your theory is to allow supply to be driven by this orphan risk.

i don't think @Peter R was being that pedantic. as i've already pointed out to you, there are other reasons why miners will naturally restrict the size of their blocks other than orphaning even in the case of No Limit. to repeat: eliminate 0 fee tx's (which they've already done), SPV mining strategies (already done), filter spam (as defined by them during an obvious attack). they are economically driven to do so to protect profits even in the absence of a strict limit.

That is not my point. My point is that if larger blocks do cause propagation issues (a necessary condition for Peter's fee theory) then we have large issues in the network.

you should outline what these issues are and why they would become so. as i just said, miners have always had the freedom to restrict their block sizes at will to prevent "large issues in the network".

Yes "wasted work" and the disproportionate impact on smaller miners are separate things.

Wasted work is a significant problem, as I explained to Peter a while ago. Wasted work makes the network less secure and easier to attack. (less equilibrium difficulty level)

i'll bet dollars to donuts you've never experienced an orphan, have you? i have, as a solo miner in the first half of 2013 soon after i got my Avalons. have you ever even mined? i'd doubt that too. yet you speak as if you're an expert. the reason it matter is that engaging in the act of mining is a unique experience few can relate to unless they've done it. that especially is so after getting an orphan. the mining process is an intense labor of love which involves getting your hands dirty along with cuts and abrasions. as a microsurgeon, i don't like that. but it's necessary when you're the only one who understands what you're doing within several states. you're always living on the edge with profitability and it's a race against time as most units are only profitable for several months until the next gen comes along. solving blocks are incredibly profitable as a solo miner (same for pool operators) so 100% uptime is paramount. getting an orphan is incredibly discouraging and eats up time investigating whether or not you got cheated somehow by malicious software. this is part of why miners have stuck with Core to date; they don't want any disruptions and as long as they don't sense any immediate danger they won't switch.

point being, i believe there has to be a certain level of orphaning that happens within the network to act as a reminder or deterrent to what can happen if you construct block sizes that are too big in the case of when the limit is above true demand. IOW, orphans are NOT always "wasted work". i would guess that the optimum level of orphaning is where we are at today and always have been roughly; 1-2%. they serve a purpose and have value as a deterrent to recklessness.

 

[Zangelbert Bingledack]

a) That is what strong consensus means, if a significant minority is not happy then we do not go ahead. That is veto power. This power ensures that other people's money cant be changed without their consent. That is what makes Bitcoin unique. This characteristic must be defended at all costs as it makes bitcoin viable money.

1) No one's money system can be changed without their consent. If Classic succeeds, there is nothing stopping Core supporters from continuing to run Core and continuing to use the Core version of Bitcoin. Same 21M coins, same 1MB blocksize limit. I suppose you agree with this, but if not, stop me here. (If the concern is identical signing algorithms, let's assume for now that Core or Classic make a trivial change to avoid cross-chain transaction issues.)

Now assuming you agree with that, I can only suppose you mean that miners will move away and/or investors will sell off the Core BTC, which they own by default even after the fork. That is likely true. But that isn't someone messing with your money system; it's investors (including miners) no longer favoring your money system. We should keep these clear.

But if you agree with this...

b) I am not worrying about trying to control anyone. It is just that in the event of attackers trying to hardfork without consensus, the incentives are structured such that the economic majority rallys behind the existing rules to defeat the attack. Even those that agree with the rule change (like me) rally behind the existing rules to protect the system. This looks like it is working and Classic is being defeated. If the defeat against Classic is resounding it should deter further attacks.

...then this quoted line of reasoning makes no sense. It cannot be an attack on the system if you can just shrug it off. That you won't shrug it off because of the effect on price and mining interest is, again, a separate issue. There is no need to "protect" anything, because nothing is being attacked. It is a peaceful split where the ones who want to change the protocol simply make a copy of the Bitcoin ledger state and go off and do their own thing with it, with investors and miners free to follow or not.

Hard forks in general cannot be "defeated," especially if they have a lower threshold. I could launch a hard fork today with a 0.01% threshold, my friends and I could mine it, and maybe even get an exchange to support it. Would that be an attack? If not, why is a 75% miner threshold an attack? It's essentially just attracting investor interest to a different ruleset, which hopefully you already agreed above is not an attack as it doesn't affect your money system - just its market price (and if price reduction is your standard for an "attack," you may want to look at an attacker named Core!).