Gold collapsing. Bitcoin UP.

[_mr_e]

the economically naive may ask, "well, why wouldn't miners like BF want to diversify their tx fee processing income away from POW mining to LN hubs? wouldn't that be smart?" answer: no.

you have to understand that this Valery Vavilov, aka BitFury, is one-of-a-kind. i remember reading/following his development thread when he was drawing his ASIC up in his garage as i was an active miner. this guy is a genius. the complexity of what he was doing was orders of magnitude above everybody else around BCT, which is why i mined with BF units soon after my Avalons. now again today they are leading the mining industry with chemical liquid cooling facilities that have no competition. and they're building 3 of these things in Georgia. how stupid would it be to let LN hub owners into the tx fee game who would number in the hundreds if not thousands by simply spinning a server up with code? no, BF is going to try and snuff them out and the only viable way i see that happening is by increasing blocksize to increase throughput on the mainchain to keep fees where they are. they'll help figure out transport improvements through pruning, IBLT, weak/thin blocks, etc. it will be worth it for them to try and maintain a monopoly for as long as possible.

So do you think BitFury is going to be capable of pulling China back below the 51% mark? So much hashing behind the GFC This has really been bothering me lately. Bigger blocks would actually help china and hurt the rest of us in the current condition since blocks take longer to get to the rest of the world, while > 51% of the hashpower can keep building off of each others blocks sooner, sort of like a natural selfish mining attack. They could even give themselves a subtle advantage by putting priority on building off eachothers blocks.

 

[Zangelbert Bingledack]

The Stolf is banging Core hard.

[doublepost=1453566475][/doublepost]Also, XT joins the 2MB party, will track consensus with Classic. @dgenr8 can you comment on the thin block support and how it differs from BU's proposed Xtreme Thinblocks?

So now: Classic, XT, and Unlimited all support the fork. Core doesn't.

 

[cypherdoc]

@_mr_e the

yes, i think it is inevitable that the Chinese mkt share gets eroded. the mkt will demand this. it knows that the HR is too concentrated and represents a potential risk from either collusion or gvt interference. this is why Bitfury is willing to take the risk on building out it's proprietary tech to the degree it has and will obviously leverage it's BW superiority. other Western mining pools will do the same. this is a highly competitive mkt all due to the fixed supply currency which has eliminated the unknown variable of a Fed debasing the money; in this case BTC. the incentives are huge as we are talking about a possible new world reserve currency. ppl laugh at this but i point at the USD which has NOT rallied despite a coupla trillion being shaved off global stock mkts, unlike 2008. in fact, the next move should be DOWN for the USD when it becomes clear they will have to QE all over again.
[doublepost=1453567386,1453566694][/doublepost]

Then why aren't they taking the wind out of those sails NOW (or last year, really) by getting together and deciding to mine bigger blocks?

they are but let's review the process over that time frame.

it started out with the XT proposal being backed by miners at 8MB. then TSHTF with miniblockists assasinating everyone in sight who supported XT. as a result, miners got quiet from confusion and for fear of retribution. this became their holding pattern pretty much until now while the entire community was hood winked by the two Stalling Conferences with promises from core dev that a scaling solution would arise from the Dec conference. the deception was finally revealed when Adam Back tweeted that a "surprise" solution would be presented in HK that would solve all our problems; that was SW. also, the miners stated quixotically that they would follow whatever core dev recommended during the mining panel. i think some were engaging in gamesmanship understandably, like maybe Bitfury who was biding time for them to build out their three immersion cooling centers. note this quote from Petrov:

"Right now, the Bitcoin Core team is slowly introducing a solution, and it’s a really complicated solution, and it's been a half a year and they are still in tests. Bitcoin Classic is a fast answer."

once they looked at SW they realized it is too complex and risks the system. not to mention it will take a long time to adopt safely.

so it is happening; in slow motion but accelerating. awareness and understanding take time. it's taken years to get to this point and i'll dare say i first brought it up here:


that was June/July 2014.
[doublepost=1453567632][/doublepost]actually, that post of mine was probably April 2014, as if you look at the Techcrunch article the OP links to was dated April 2014.

 

[_mr_e]

@cypherdoc You called so many things way early in Bitcoins life, I think many of these uprisings started right from your mouth! I sometimes wonder where we'd be now if you hadn't stayed so stubborn and gave up.

 

[cypherdoc]

@_mr_e

well, take comfort from my wife calling me stubborn as well.

thing is, i'm lucky to not have ANY financial COI's in Bitcoin other than my coin and my career depends not on any geeks or tech related company employment. it's liberating to be able to speak freely and with intent only to hope for a better world thru Bitcoin, altho it still takes alot of fortitude to take the abuse i have. once this HF affair is settled, the truth can come out, and it'll get even worse for core dev
[doublepost=1453569259,1453568144][/doublepost]haha. Hill representing Core?:

 

[cliff]

@cypherdoc - lol, omfg. These "cyber boys" [h/t random banker], even if ever on the right side of something, should NEVER be left to speak in public, they do themselves no favors and seem to be clueless about the ubiquity of power of and power relations* (the lack of self-reflection is disturbing) Where is their chief of strategy, John Dillion - i mean John Dilley?

*M. Foucault, (1980), Power/Knowledge: Selected Interviews and Other Writings 1972–1977
*Martin Heidegger, "The Question Concerning Technology," Basic Writings Ed. David Farrell Krell (Harper & Row, 1977)

 

[YarkoL]

@dgenr8 can you comment on the thin block support and how it differs from BU's proposed Xtreme Thinblocks?

Even though the question was not addressed to me, I might contribute few words, based on info that I have received from @Peter Tschipper ...

Both XT's thinblocks and BUIP010 originate from some code written by Hearn just before he left, but there is a considerable divergence in the two approaches. Without going into tech details, the main difference is that XT's thinblocks are designed to be compatible with all the nodes, whereas Xtreme Thinblocks can be requested and received only by nodes who have implemented BUIP010.

In other words, BUIP010 defines a relay network for Xtreme Thinblocks, and XT's thinblocks can be seen as an optimization of the already existing block transfer mechanisms.

Aside from that, Xtreme Thinblocks really are smaller, faster and require less re-requesting at times of high traffic congestion. Also, XT Thinblocks rely on the assumption that the nodes that they connect with support bloom filters, which is optional now with Core nodes, and may be phased out at some point in future.

But it is good that XT (and possibly Classic) are implementing thin blocks, because it will increase the general awareness of this solution and prepare the public to appreciate BUIP010 much more better if and when it rolls out.

 

[Mengerian]

@cypherdoc @albin
I've been thinking about the meme that Joseph Poon raised of "no block-size limit = transaction fee race-to-the-bottom". This is a concern that gets raised repeatedly, and I agree that the concern is misplaced. I do have some sympathy for this view however, as I was worried about this for a little while (years ago), but have since realized that there are many ways to avoid this situation.

One solution we could imagine is miners tacitly agreeing to not produce blocks above a certain size, or not include transactions with a fee below a certain cut-off. If the majority of miners follow along, they could orphan blocks that violate the agreement. But then this idea gets pilloried as collusion, the formation of an evil "mining cartel".

But this got me thinking, what is the functional difference between a "mining cartel" and "consensus"? Fundamentally, there is no real difference.

So that leads me to view the current "consensus" on block size limit through the lens of cartel economics. A group is trying to restrict the supply of a good (on-blockchain transactions) to an artificially low level. But the more they restrict it, the higher the incentive becomes to break away from the "cartel". As more and more participants signal willingness to break from the cartel, it becomes weaker and vulnerable to defectors. This also makes me think that the Bitcoin Unlimited BUIP 005 User Agent String signalling will be a very effective tool to help bust the small-block cartel.

Regarding Lighting Network, I think it's a cool idea that could have some interesting use-cases. But it is not a panacea for scaling, and cannot suitably replace on-chain transactions for many/most Bitcoin uses.

 

[Peter R]

In case anyone missed it, Vitalik Buterin hit it out of the park in the "complicated soft-fork solution to avoiding a simple hard fork" batting competition.

In this Reddit post, he presents a soft-fork method of decreasing the block interval from ten minutes to two minutes, thereby increasing Bitcoin's transactional capacity from 3.5 to 17.5 transactions per second. The technique requires cooperating miners to fudge the blocks' timestamps to trick the old nodes "into retargeting difficulty in such a way that on average, a block comes every two minutes instead of once every ten minute."

I don't suspect we will actually deploy this solution, but the theoretical implications have had me thinking all morning. Over the last year, we've learned that Bitcoin can be changed in radical ways with only a soft fork. And Vitalik has just shown that even the block time can be changed this way too!

For the purposes of discussion, let's assume for the moment that the following statement is true [1]:

1. Any change to Bitcoin's nature that can be accomplished with a hard fork can also be accomplished with a soft fork at the expense of added complexity.

The Blockstream Core Devs often make the following statement:

2. Soft-forking changes require only the support of the majority of the hash power. [FALSE?]

If we also assume that #2 is true, then #1 + #2 imply that any change to Bitcoin's nature can be made with only support from the majority of the hash power. And if that is true, then the other statement that Blockstream Core Devs often make would be powerless:

3. Hard-forking changes require overwhelming consensus. [FALSE?]

In fact, it would be counterproductive to fight hard-forking change that the miners want to make, because they could always make those same changes with a more-complicated/costly soft fork instead.

What I'm thinking this morning is that in fact both #2 and #3 are false [2],[3]. Soft-forks require more than just the majority hash power and hard-forks require less than "overwhelming consensus" (whatever that means). In fact, I suspect they both require the exact same thing:

4. Changes to Bitcoin's nature require Nakamoto consensus only.

[1] I'm not claiming that it necessarily is true; but as far as I know no one has shown that it is untrue.

[2] As an example of #2 being false, assume that the miners implemented Vitalik's proposal but everyone else was violently opposed. Perhaps the community would change the PoW to render the miners' hardware useless. The miners' chain would eventually die (no one wanted those coins anyways), and the new "economic majority" chain would persist and eventually become the longest.

[3] We've already discussed many examples of #3 being false (e.g., the sinking ship analogy from @Roger_Murdock).

 

[Mengerian]

@Peter R
Agreed that Vitalik is currently leading the "ridiculously complicated soft-fork thought experiment" contest

I'm really looking forward to real-life empirical evidence on your point 3 in the next few months. A hard fork, contentious or otherwise, will reveal information about Bitcoin's nature. This will provide invaluable learning to all participants in the network, and increase the value of the network. I also anticipate this turmoil will yield some good trading opportunities.

 

[Zangelbert Bingledack]

Seems like we could float it as a puzzle: "1, 2, and 3 can't all be true. Which one is false?"

 

[albin]

I've been thinking about the meme that Joseph Poon raised of "no block-size limit = transaction fee race-to-the-bottom".

My take on it is that the issue creates concern simply by virtue of word choice. A transaction fee race-to-the-bottom is exactly what a market is supposed to do if prices are fixed above profit-maximizing conditions.

I think there are two reasons underlying this concern --

1) Over-extrapolation to slippery-slope condition of price being low to price actually being zero, in essence violating underlying economic assumptions to create "singularity" condition.

2) Corollary to #1, fictitious infinite supply brought to market equated to infinite quantity demanded @ market price (which conflates infinite demand due to monotonicity of preferences with actual quantity demanded at market equilibrium), used as a theoretical fiction to support portraying the correlation that increasing transaction volume as a function of time and decreasing reachable nodes as a function of time have a strong causative relationship, to the point where transaction volume is portrayed as the main driver of dropping node count. Where in reality no realistic blocksize proposal seems to be in any way threatening to the function of nodes, and we don't even have any way of reliably understanding the trends in node count, since we're anecdotally comparing current node count with an idealized state of perfection with possibly >100,000 extremely crappy, potentially largely inbound-unreachable, intermittently-connected full nodes because that was the only way to run a wallet.

 

[Peter R]

@Graphics:

I was pretty surprised to receive this text this morning (I guess she saw it on her Facebook):

 

[albin]

I can't believe it took Stolfi to rapidly dismantle the FUD that Peter Todd has been spewing for years:

 

[awemany]

So do you think BitFury is going to be capable of pulling China back below the 51% mark? So much hashing behind the GFC This has really been bothering me lately. Bigger blocks would actually help china and hurt the rest of us in the current condition since blocks take longer to get to the rest of the world, while > 51% of the hashpower can keep building off of each others blocks sooner, sort of like a natural selfish mining attack. They could even give themselves a subtle advantage by putting priority on building off eachothers blocks.

The transactions need to come from somewhere, though...

 

[Mengerian]

@albin
Agreed, regarding the flaws with the "transaction fee race to the bottom" worries.

I also think that it takes very little market pressure to incentivise small transaction fees (think fractions of a penny). Even with unlimited block sizes, it should be pretty easy for miners to encourage small transaction fees that would be inconsequential for all but the smallest micropayments (where Lightning Network could find a niche).

The most plausible way to get to the "Bitcoin achieves world domination" scenario, I think, is to have a high on-chain transaction volume. Low fees per transactions, high cumulative fees per block. Luckily the network incentives seem aligned to nudge us in that direction, so I'm hopeful that is how events will unfold over the next several years.

 

[Roger_Murdock]

What I'm thinking this morning is that in fact both #2 and #3 are false [2],[3]. Soft-forks require more than just the majority hash power and hard-forks require less than "overwhelming consensus" (whatever that means).

To me, it's pretty obvious that once a simple majority of hash power actually "flips the switch" on a hard fork attempt (based on whatever support thresholds they're comfortable with), it will almost certainly set in motion a positive feedback loop that will end with the vast majority of economic actors converging on the larger chain. Obviously, that doesn't mean there won't be a few hardcore ("hard-Core"?) ideological holdouts continuing to spin their wheels on the losing chain, but the losing chain shouldn't be expected to have any serious economic significance compared to the winner. One exception is if the majority hash power (perhaps funded by a hostile government) is executing a clear attack on Bitcoin by trying to force an obviously value-destroying change.

Now as a practical matter, I do think that more than a simple majority of hash power will generally be required to trigger a hard fork attempt because of "hard fork conservatism" on the part of miners. (Most miners won't be comfortable helping to trigger a hard fork as part of a smallish majority.) That's why we've seen these proposals that are only triggered upon some supermajority target being met. I suspect that some degree of "hard fork conservatism" is healthy, but I think the current level is probably excessive and based on many people's failure to appreciate exactly how strong the incentives toward convergence are and thus how unlikely a persistent chain split is. (Unfortunately, that misunderstanding is only being fueled by the FUD from small-blockers claiming that "contentious" hard forks are somehow insanely dangerous.) BTW, I was kinda disappointed that no one seemed to like my proposal for a Bitcoin Unlimited-type approach to the problem of trying to guess the optimum place to set these triggers, i.e., a client that would allow miners to specify exactly the minimum percentage of hash power with which they'd be comfortable helping to trigger a hard fork. A fork attempt would then be triggered automatically whenever X percent of hash power has indicated that it's comfortable triggering a hard fork with not less than X percent of hash power.

 

[Zangelbert Bingledack]

[doublepost=1453581810,1453580685][/doublepost]

BTW, I was kinda disappointed that no one seemed to like my proposal for a Bitcoin Unlimited-type approach to the problem of trying to guess the optimum place to set these triggers, i.e., a client that would allow miners to specify exactly the minimum percentage of hash power with which they'd be comfortable helping to trigger a hard fork. A fork attempt would then be triggered automatically whenever X percent of hash power has indicated that it's comfortable triggering a hard fork with not less than X percent of hash power.

Actually that's a pretty good idea now that I think about it (initially I was imagining a kind of meta-KBC, which seemed too confusing to get traction). It should get the fork to happen as surely as possible because the number can be as near 51% as feasible, but no nearer (because 51+% of people would never agree to do anything risky like letting it get too close to 50%). The result might be something like 57%. And people could adjust it on the fly.

 

[AdrianX]

@Peter R
Agreed that Vitalik is currently leading the "ridiculously complicated soft-fork thought experiment" contest

I'm really looking forward to real-life empirical evidence on your point 3 in the next few months. A hard fork, contentious or otherwise, will reveal information about Bitcoin's nature. This will provide invaluable learning to all participants in the network, and increase the value of the network. I also anticipate this turmoil will yield some good trading opportunities.

I couldn't agree more. The level of behavior of the existing ruling developers have dropped to to hang on to power has made me very nervous about Bitcoins future.

I think we'll see the start of the next growth spurt should Bitcoin resist this attempt at centralized authoritative control. I can't imagine any serious capital buying up bitcoin given the uncertainty followed by wanting to invest in a system where a single corporation dictates the roadmap.

Bitcoin is too important to fail. So I can't see it not bouncing back with more strength.

 

[AdrianX]

@albin
Agreed, regarding the flaws with the "transaction fee race to the bottom" worries.

I also think that it takes very little market pressure to incentivise small transaction fees (think fractions of a penny). Even with unlimited block sizes, it should be pretty easy for miners to encourage small transaction fees that would be inconsequential for all but the smallest micropayments (where Lightning Network could find a niche).

The most plausible way to get to the "Bitcoin achieves world domination" scenario, I think, is to have a high on-chain transaction volume. Low fees per transactions, high cumulative fees per block. Luckily the network incentives seem aligned to nudge us in that direction, so I'm hopeful that is how events will unfold over the next several years.

It's not a race to the bottom it's typical free market competition. Miners are weaned off subsidies and then compete to find a market equilibrium for fees. The winners are the ones who out compete the competition and are marginally profitable. Each will have some competitive advantage and global variance in demand for transactions will keep it diverse.

Every 4 years they are forced to experiment with fees and block size as they have their subsidy reduced by 50%.

It's scales of economy that will bring in revenue and it's the size of the network that makes it valuable.

Bitcoin is either going to be worth a lot in 15 years (20 subtract 5) or very little.