Gold collapsing. Bitcoin UP.

[freetrader]

@jonny1000 :
That thread is highly instructive, from a retrospective point of view. Thank you for bringing it up. We need to bear in mind that those were the days when nearly everyone could still mine. A lot has changed, and a lot of arguments that applied to a possible future back then have ceased applying to current conditions at hand.

You did not answer my question about whether you view Bitcoin as being in a calamitous state right now (as regards mining centralization in the form of a cartel, wasting processing power).

EDIT: due to browser refresh after editing, I only saw your answer to that question after posting my question above. I don't quite see how the relatively high block subsidy addresses the centralization problem. All things being equal (which they are not of course), a high subsidy should still be driving decentralization. Then again, increased adoption leading to higher price could have the same effect... as evidenced by the posters on this forum and others who state that if the price were to reach X, it would become economical for them to resume mining, even using their older equipment.

How do you measure the robustness of the industry, as you call it, that being one thing you wish to ensure?

Sadly, many predictions from the past in that BCT thread have come true.
I still wonder - are you proposing any plan to address the centralization issue? I only recall that you were petitioning for a 2MB fork with a 95% consensus hurdle and a vastly longer (order of 6 months or more) grace period.
Has your thinking on that changed?

On a side note, are you being sponsored to attend the Milan conference?

 

[AdrianX]

My point is based on the idea that a miner does not need to propagate to themselves. A miner with 50% of the global hashrate only needs to propagate to the remaining 50% of the network. This should half the orphan risk costs.

@jonny1000
We know that a miner with over 50% can have far more negative impact on the network than a variable income increase due to orphan risk. Can you please remind me what the orphan risk is currently and how much the miner will benefit? While you're at it why don't you compound the benefit over time to make it bigger and see how it works out.

The variance you base your argument on is so small it's counteracted by variables in operating costs and interns speeds on any given day.

It's such an insignificant centralization factor it's less than a rounding effort.

I tell you what has a huge impact of centralization though it's the efficiency of mining equipment and while miners have this advantage they'll grow bigger and more centralized. If you truly want decentralization get them to give to stop innovating and improving the network security.

The centralization we're seeing will reverse when mining technology becomes more readily available. No miner has a monopoly on that technology. Just a short window to capitalize on the risks taken.

And in the meanwhile we can continue to rely on the incentives designed into the protocol that have facilitated the growth we've seen over the past 7 years those same incentives that protect the network and those same incentives that cause the centralization of miners who have an incentive to maintain the network they depend on for income.

 

[cliff]

I would be a little humbled if this were our jonny (just ran across my tweeter feed - happens to contain some decent rationale for decentralization, similar to what I was looking for a few days back from @jonny1000 ): http://www.forbes.com/sites/jonmatonis/2012/10/04/bitcoin-prevents-monetary-tyranny/#2c81eca16cda.

good times despite the contentious debate - I sorta agree w/ cory fields' take here: http://coinjournal.net/mits-cory-fields-contentiousness-in-bitcoin-is-sign-of-good-health/ - apathy might be a sign of stagnation.

 

[cypherdoc]

In my view we need to ensure the industry is robust in as many situations as possible, and therefore we should keep an economically relevant blocksize limit, as a defense for when this competitive industry structure emerges.

this seems is contradictory to me. you just said that miners will alternate sometimes btwn short term profit gains (adding tx's up to their marginal cost) and long term strategies (where they voluntarily limit blocksizes) and therefore we (whoever the heck central planner that is) should impose a blocksize limit. this is wrong. precisely *because* you don't know which behavior they will exhibit at any given time, a limit should NOT be imposed to allow miners to choose freely which behavior they wish to exhibit at any given time. after all they, with the direct input of large numbers of users paying tx's, are in the unique position (certainly kore dev is not) of being able to determine which strategy fits at any given time.
[doublepost=1468856872][/doublepost]

It's such an insignificant centralization factor it's less than a rounding effort.

@Jihan has already told us what it is:
https://bitco.in/forum/threads/gold-collapsing-bitcoin-up.16/page-693#post-24285

You exaggerated the situation by manipulating the data.

Right now the orphan rate risk is generally very very low, even for a small pool. if orphan rate is below 0.2%, a 76% is only 0.15% advantage, which is very minimal. 0.15% advantage compared with a 2.5%~4% PPS fee rate, or 2%-4% PPLNS tx fee revenue, is also negligible.

You have an assumption which is wrong. The assumption is wrong that information spread inside a large scale pool is larger than light speed. Inside a large scale mining pool which serves global users, nodes are distributed in different places, the propagation inside one same pool also will consume time. So a pool has 24% market share, has more than 76% of the average orphan risk level.

 

[priestc]

It doesn't need to be a single node, this just centralize miners by geography, they could end up in clusters around a few or even one location.

Just because nodes are geographically close together doesn't mean they are immune to orphan blocks. Blocks still have to transverse the network. Its less about geographical closeness, and more to do with network topology. If there are 20 network hops between you and another miner, that connection is more likely to cause more orphan blocks that another miner that only has to go through 5 hops. Geographical closeness has very little to do with it. Its completely possible that Chinese ISPs just simply suck, and connections between them are really bad. This will very well cause some chinese miners to have a higher orphan rates than miners in the rest of the world.

You're making the mistake of thinking any advantage a set of miners have over another set leads to centralization. This is like claiming that a 1% sales tax is "povertizing". Yes having to pay more for things means you have less money, but to claim paying 1% more for something puts you into poverty is absurd. Claiming one set of miners having a slight advantage over another set of miners is "centralizing" is also completely absurd.

 

[cliff]

Question - maybe this old thinking too:

Is it possible to ratchet up fees as unlimited blocks fill - like an increased load/weight fee. blocks would be theoretically unlimited in size, but after a certain point it costs more and more to ride the next block/ a certain block than to just wait for the next one/a later one. seems like this could solve a few different concerns on both sides of the blocksize debate - but I dunno, just thinking aloud more than anything.

 

[Inca]

Is it possible to ratchet up fees as unlimited blocks fill - like an increased load/weight fee. blocks would be theoretically unlimited in size, but after a certain point it costs more and more to ride the next block/ a certain block than to just wait for the next one/a later one. seems like this could solve a few different concerns on both sides of the blocksize debate - but I dunno, just thinking aloud more than anything.

Blocks can never be unlimited in size. No matter how advanced the network propagation becomes there will always be a point where trying to make a block just one piece bigger (just one more transaction) results in that block being orphaned.

The risk of a block being orphaned is what keeps miners from potentially creating a block the size of the entire mempool to recoup all the fees.

 

[cypherdoc]

Just because nodes are geographically close together doesn't mean they are immune to orphan blocks. Blocks still have to transverse the network. Its less about geographical closeness, and more to do with network topology. If there are 20 network hops between you and another miner, that connection is more likely to cause more orphan blocks that another miner that only has to go through 5 hops. Geographical closeness has very little to do with it. Its completely possible that Chinese ISPs just simply suck, and connections between them are really bad. This will very well cause some chinese miners to have a higher orphan rates than miners in the rest of the world.

You're making the mistake of thinking any advantage a set of miners have over another set leads to centralization. This is like claiming that a 1% sales tax is "povertizing". Yes having to pay more for things means you have less money, but to claim paying 1% more for something puts you into poverty is absurd. Claiming one set of miners having a slight advantage over another set of miners is "centralizing" is also completely absurd.

absolutely. and @jonny1000 completely ignores the possibilities for other parts of the network to "react". let's say we lift the limit. he claims Chinese miners somehow will exploit the fact that they live behind a GFC with crappy BW and somehow leverage this as an advantage. i think that's completely ludicrous. i, as a potential mining entrepreneur, would interpret the potential for bigger blocks and a flood of new users as a HUGE opportunity to develop mining outside of China so as to leverage my BW advantages. now, this doesn't have to be a bad thing for existing Chinese miners. i feel this would be a situation of a rising tide lifts all boats. and there are certainly strategies the Chinese could employ to "react" as well; like employ XThin, XVal to skirt the GFC. as the price skyrockets all miners should benefit from global adoption.
[doublepost=1468858235,1468857451][/doublepost]

Blocks can never be unlimited in size. No matter how advanced the network propagation becomes there will always be a point where trying to make a block just one piece bigger (just one more transaction) results in that block being orphaned.

The risk of a block being orphaned is what keeps miners from potentially creating a block the size of the entire mempool to recoup all the fees.

yes, these 2 graphs from reddit are instructive:


as you can see, miners over the last 7.5y have acted very responsibly and according to their own interests, by walking the blocksizes they've produced up gradually. no one rushed to fill blocks up to 1MB starting back in 2010 when the 1MB limit was imposed. and look at the orphan rate; it's stayed the same during the whole time period, irrespective of the bigger blocksizes. which brings up another point the small blockists like to make; this concept of infinite demand. there's no such thing. demand during the last 7.5 yrs has respected the generalized Bitcoin economic growth. this demand was well below 1MB for most of our history. and yet, users paid fees! and predictably, miners have begun setting minfees which is perfectly fine and predictable. now that we've slammed up against the 1MB for the first time in our history, you are seeing all sorts of distortions, ie, delays, high fees, unconfirms, terrible user experiences. lift the limit higher or eliminate it all together and demand will continue to rise smoothly as it has in the past in this graph.

 

[cliff]

@Inca & @cypherdoc - i totally spaced/overlooked the orphaning risk. i still think fee markets should be able to co-exist with a BU-type client & with increased risk of orphaning as block size increases (i.e., doesn't have to be one or the other)

 

[Inca]

absolutely. and @jonny1000 completely ignores the possibilities for other parts of the network to "react". let's say we lift the limit. he claims Chinese miners somehow will exploit the fact that they live behind a GFC with crappy BW and somehow leverage this as an advantage. i think that's completely ludicrous. i, as a potential mining entrepreneur, would interpret the potential for bigger blocks and a flood of new users as a HUGE opportunity to develop mining outside of China so as to leverage my BW advantages. now, this doesn't have to be a bad thing for existing Chinese miners. i feel this would be a situation of a rising tide lifts all boats. and there are certainly strategies the Chinese could employ to "react" as well; like employ XThin, XVal to skirt the GFC. as the price skyrockets all miners should benefit from global adoption.
[doublepost=1468858235,1468857451][/doublepost]
yes, this graph from reddit is instructive:


as you can see, miners over the last 7.5y have acted very responsibly and according to their own interests, by gradually walking the blocksizes they've produced up gradually. no one rushed to fill blocks up to 1MB starting back in 2010 when the 1MB limit was imposed. and look at the orphan rate; it's stayed the same during the whole time period, irrespective of the bigger blocksizes. which brings up another point the small blockists like to make; this concept of infinite demand. there's no such thing. demand during the last 7.5 yrs has respected the generalized Bitcoin economic growth. this demand was well below 1MB for most of our history. and yet, users paid fees! and predictably, miners have begun setting minfees which is perfectly fine and predictable. now that we've slammed up against the 1MB for the first time in our history, you are seeing all sorts of distortions, ie, delays, high fees, unconfirms, terrible user experiences. lift the limit higher or eliminate it all together and demand will continue to rise smoothly as it has in the past in this graph.

The other thing is that despite suggestions that miners will fill blocks with spam transactions and that if we lift the limit blocks will suddenly become gigantic, what that chart demonstrates is that the opposite is true. Blocks follow the transaction demand for the network over time and only began to hit the 1mb limit when the network became saturated.

 

[cypherdoc]

@Inca

yes, and i've argued that if a spammer wishes to pay minfees to attempt to lock up the network in a no limit world, let him. he'll fail. just like the spammers have over the last year or so. but w/o a limit, the spammer won't have a defined ceiling (like the ridiculous 1MB) to attack and he'll likely flail about wasting BTC only to financially strengthen the miners who eat up his free BTC within a short period of time.

 

[jonny1000]

The variance you base your argument on is so small it's counteracted by variables in operating costs and interns speeds on any given day.

It's such an insignificant centralization factor it's less than a rounding effort.

@Jihan has already told us what it is:
https://bitco.in/forum/threads/gold-collapsing-bitcoin-up.16/page-693#post-24285

You exaggerated the situation by manipulating the data.

Right now the orphan rate risk is generally very very low, even for a small pool. if orphan rate is below 0.2%, a 76% is only 0.15% advantage, which is very minimal. 0.15% advantage compared with a 2.5%~4% PPS fee rate, or 2%-4% PPLNS tx fee revenue, is also negligible.

This point keeps being made. You keep mentioning that the absolute level of orphan risk is low or insignificant. I keep mentioning this is not at all what I am talking about. When you keep mentioning this irrelevant point it gets quite frustrating. I will try to remind you again what I am actually talking about.

I am concerned about the following scenario:

• The block reward is insignificant
• There is no blocksize limit
• Orphan risk costs are used to drive mining fee revenue, in the sense that:
  • Miners keep including more and more transactions in a block up to the point where the fee equals the marginal orphan risk cost, for the most marginal transaction in the block
    
  • Advanced wallets lower the fee to the point where they pay a slight premium over the expected orphan risk cost

In this scenario I am concerned about the following ratio:

• Orphan risk cost / miner fee revenue

In what I consider to be a flawed model, where orphan risk drives the mining revenue, this ratio should be well over 50%, which I consider catastrophic for the network.

 

[cypherdoc]

and i keep saying, "quit trying to predict what the situation will be in 2140. we've have bigger things to deal with today". like onboarding new users, improving user experience by eliminating long delays, high fees, etc.

as an example, you didn't address my above game theoretic. what if the only miners out there in 2140 were few dozen gvts? to me, that could very well be a stable system. and, they don't have to worry about marginal costs. this scenario may seem far fetched. but you can't tell me this might not be a possibility and demonstrates the difficulty of projecting out so far.
[doublepost=1468860458][/doublepost]in the meantime, you're holding the entire Bitcoin economy hostage to 1MB worrying about a 2140 scenario based on a hunch.
[doublepost=1468860870][/doublepost]actually, i find my game theoretic highly likely come 2140. gvts will want and have to get into mining for national security reasons. look at the China situation today. depending on how you look at it, they are protecting their miners . they're already in the business so to speak. look at Georgia and Bitfury. gvt has their tentacles involved in that one i guarantee. this is going to spread with time. and as i said, i don't think it's necessarily a bad thing at all assuming that it evolves similarly to what we have in the nuclear arms race; mutually assured destruction if one gvt tries to destroy the mining decentralization.

what makes this all possible is the sound money function of Bitcoin. there's too much money involved for any one actor or gvt to screw it up. if you try, get left behind in an evolving, increasingly prosperous world based on a new revolutionary digital currency.

 

[cypherdoc]

@jonny1000

you also didn't address my pointing out your contradiction in logic concerning the alternating short/long term strategies of miners when it comes to big/small block construction. you think the solution to this is imposing a centrally planned limit by some all seeing kore dev as a way to mitigate that situation. i pointed out that you couldn't be more wrong. eliminate the limit and let the economic actors work it out btwn themselves (users and miners) b/c there is no way you could know what limit would be best.

 

[pekatete]

I am concerned about the following scenario:

• The block reward is insignificant
• There is no blocksize limit
• ......

I've maintained that @jonny1000 is our very own village troll and got short shrift, but here is proof if any were needed. As @cypherdoc says, block rewards becoming insignificant cannot be on our radar at this point in Bitcoin space and time. The time we know is 2140, as for the space, no one can know as the universe (and bitcoin) are expanding .... infinitely!

 

[cypherdoc]

it took me a while to conceptualize/distillate @jonny1000's arguments about marginal this, marginal that down to the same 'ol tired argument that has existed since 2009; a Tragedy of the Commons. except as applied to miners instead of users.

all the naysayers back then who bought into that theory have already been proved devastatingly wrong.

 

[freetrader]

My question is: which miners' opinions do @jonny1000 's views represent?

Could they be the expression of sincerely held beliefs that some mining parties are just unable to voice themselves (presumably through lack of other eloquent spokespeople) ?

Or are these just personal fears of a catastrophical end to Bitcoin, shared by a subset of developers?
Are they indeed even shared by a sizeable fraction of miners?

Is it pragmatic investment practice to declare one's investment exit should things go south in a matter of ideological difference, when both alternatives (indeed, more than 2) are possible simultaneously in a cryptocurrency?

 

[pekatete]

I doubt miners, and here I mean the real people who invest substantial fiat to make ASICs and mining gear, can hold any such views, completely disconnected from reality as they are.

 

[cypherdoc]

oh heavens, what a waste of time. and which of the kore devs has the time & money to host a wine & dine event for miners? why Blockstream, i'd guess! and why does BTCDrak get to remain anonymous with so much money on the line?:

A late July gathering to feature members of bitcoin's development and mining communities, and organized by contributors to Bitcoin Core, will not focus specifically on the issue of scaling, according to materials provided to CoinDesk.

Of focus, organizers said, will be reducing the "language and cultural gap" dividing attendees on larger issues through activities such as sightseeing, community meals and informal discussions.

Bitcoin Core developer BTC Drak said that organizers are looking to the event as a way to ease what he described as the sometimes "tense" relations between miners and developers.

http://www.coindesk.com/no-scaling-agreements-industry-bitcoin-meetup
[doublepost=1468866654][/doublepost]@Jihan try not to be tricked by this.

 

[Inca]

They are desperate.

When you are highly antisocial and deeply unlikeable, don't be surprised if noone trusts, likes or respects you. Gavin is deeply missed in this regard.

As far as I am aware drak hasn't contributed a single line of code to Core.

Let them carry on with their Rube Goldberg machine.

Personally I think Classic's time has been and gone. We should as a community concentrate on Unlimited and increasing the node count. Perhaps Roger Ver could help - his views seem almost perfectly aligned with our own.