Topic: The deflationary problem - page 6. (Read 32501 times)

I think the fact that someone had a transaction with six confirmation that was then undone was extremely significant.

True, but it happened at a time when Bitcoin's rise was big news and there was a high rate of influx of new users. These are the conditions under which you would never expect such a drop. And it happened due to a failure at a choke point whose significant was not, I think, generally recognized.

But I agree with your point. You can see these events as Earth shattering or meaningless. The important thing is that bad things happened and Bitcoin barely noticed. So if you're worried about bad things happening, the evidence suggests that Bitcoin will survive.

From my own perspectives, the block chain split was of zero consequence.  If it were not for the activity on this forum at the time, and my interaction with this forum, I (personally) wouldn't have even noticed.  The same would have been true for most users who either don't spend bitcoins on a daily basis or don't fret over the time to confirm.  Once users are competing for blockspace, a 24 hour long time to confirm will become more commonplace for much more mundane reasons.  And the price crash wasn't even the worst crash that I've seen since I've been here.  On a percentage basis, dropping from about $250 to about $50 over the course of a week or two isn't as bad as dropping from $32 to $2 in 2011 (2012?), although it happened much faster.

I just wanted to wrap this thread up by pointing out that moores law will break down at some point once we can't make smaller silicon... or at least change to a different time frame. Once bitcoin asics are at the practical limits of silicon's molecular size (is it 4 or 5 nano meters?) we're dead in the water on this issue until quantum computing... at which point we'll have to change the proof of work anyway.

Moore said it himself: the following an extracted from this transcript  http://ftp://download.intel.com/museum/Moores_Law/Video-Transcripts/Excepts_A_Conversation_with_Gordon_Moore.pdf  in 2005.

Interviewer: How long can it continue?

Gordon Moore: I think Moore’s Law will continue as long as Moore does anyhow! Ha ha ha... I’m periodically amazed at how we’re able to make progress. Several times along the way, I thought we reached the end of the line, things taper off, and our creative engineers come up with ways around them. I can think of at least 3-4 things that seemed like formidable barriers that we just blew past without any hesitation and in fact, the board meeting we came up I saw pictures that go a couple of generations beyond anything I had seen previously in regards to the highest transistors we can build... so I think we’ve got quite a bit of life yet. I’ve  ever been able to see beyond the next three generations of technology. Three generations of technology is about 6-8 years and I can see that far now, things haven’t really changed. Eventually they’re going to have to. Materials are made of atoms, and we’re getting suspiciously close to some of the atomic dimensions with these new structures, but I’m sure we’ll find ways to squeeze even further than we think we presently can.

I hope so. I think it's worth working on ways to reduce the probability and severity of this attack. So, in an academic sense, it's well worth worrying about. But it would be really odd to argue that people should avoid Bitcoin because of the very small risk of this kind of attack.

In addition, though the known solutions (at least known to me) to a 51% attack (from a motivated attacker with significant resources) are pretty awful, it's not necessarily the total death of Bitcoin. So while it would be a severe blow (as the block chain split was, as the recent price crash was), I would expect Bitcoin would survive somehow. Bitcoin's death has been predicted so many times and Bitcoin has weathered crises.

ASIC's are tremendously beneficial to the future of cryptocurrencies.  The fact that bitcoin lasted long enough to herald in ASICs is a great accomplishment for all future cryptocurrencies that will compete on the basis of which economic protocol is most efficient.

That said, once the hashrate matures we cannot simply rely on the fact that ASIC miners will not stop mining.  The hashrate must double every two years, otherwise the monetary barrier to a double spend become  smaller and smaller.

I can't refute those points, but the next one makes them all sort of moot.

They could, but would they? If they feared the 51% would continue, they'd have no assurance of any return on their investment. If the price falls, some existing miners might be discouraged and stop mining, possibly faster than new mining could be added, especially when it wasn't profitable in the short term.

That assumes that their profit comes from Bitcoin rather than a competitor (which could even be conventional fiat currency). Consider, for example, a future time when Bitcoin prices are high but the block reward is low enough that mining isn't all that profitable. If someone shorts Bitcoins (perhaps even by investing in a competitor, which could even be something like Western Union, PayPal, or Visa), they might actually find it profitable to launch a 51% attack.

One thing that helps Bitcoin resist this kind of attack is that it can't be mined efficiently with general purpose hardware. This means that miners probably won't stop mining at the drop of a hat because they can't sell their mining hardware to someone who is going to do something else with it anyway or use it for something else -- they will have to try to keep Bitcoin healthy event if that means short term losses otherwise their expensive mining hardware is just scrap metal. And it also means that an attacker has to invest in hardware that they can *only* use to mine Bitcoins. They can't rent general purpose hardware to attack the network at a low cost. Coins that use ASIC-unfriendly mining schemes are much more vulnerable to this kind of attack than Bitcoin is.

Bump.

Nope, it must not. In order for the network to be secure, 51% of hashing power must remain in "white-hat" hands.

Even if a 51% attack happened, it wouldn't do much, nor last long (as the other miners could simply buy cheap hardware and knock him back under 51%).

It's more profitable to play by the rules, than to try and subvert them.
See: https://en.bitcoin.it/wiki/Weaknesses#Attacker_has_a_lot_of_computing_power

https://bitcointalksearch.org/topic/criticism-of-bitcoin-174378

I've posted an equation that relies on declining rates of return on mining, allowing increased mining to produce more hashes, but also for more coins to be produced.

I am not defending Bitcoin I see flaws too, to your point, gold is a finite resource, and it inflates very slowly over thousands of years, energy and environmental cost of gold inflation dwarfs that of Bitcoin. Bitcoin is finite and inflates (if only it would inflate like gold) the cost of Bitcoin inflation is insignificant in when compared to gold.

I think you are one step closer to realizing the problem.

Hash must always grow.  

The inclination of early Bitcoin theorists was that Hash would fall if miners did not profit, and the mechanism would regulate itself until miners profited.  This is true for mining profit and always will be.

But if miners do not profit, and the difficulty must come down for them to profit, then the whole network can potentially become compromised.

Thus, as i originally stated hash must increase in accordance with Moore's Law, as well as difficulty, and the protocol should reward mining profit through inflation rather than punish it with deflation.

Because if Network hash rate doesn't at least keep up with computing power, it can be outcompeted, and lose the 51% battle?

You seem to have forgotten the second half of Moore's law. Hardware also gets cheaper as it gets more powerful. Fees will more than keep miners in the latest hardware.

Gold has been inflating for thousands of years.  Your arguments are nonsense.  It takes energy to find gold and it takes energy to mine bitcoins.  Neither is a cause of mal-malvestment because it is not free to obtain.  Mal-investment only occurs when inflating is cheap relative to what is gained by inflating.  Inflation of fiat is bad, of gold, bitcoins, copper, uranium, etc not.  It is necessary and good.

Please spare me the mundane economic discussion.  If you did your due diligence i already addressed the inflationary aspect earlier, and how it relates to other known stores of value such as gold.

I will have to address this in the first post so people don't keep repeating the same thing.

The inflation in the economy is cause for Malinvestment, it is a cost industry pays at the expense of efficiency and environment, the net energy in mineral  / energy mining and environmental damage and wasted recourses as a result of the inflation caused Malinvestment, will be much higher than the cost of a fixed currency infrastructure through the Hash rate. 

To make your argument you will have to prove why the hash rate will scale relative to number of transactions, the to functions do not scale symmetrically, your assumption that they do is the error that makes your argument.   

Network hash must grow at least at the rate of Moore's Law.  This is also known as Sweft's Law.  If Bitcoin fails to satisfy this condition, then it will be prone to attack.

I'm not sure what you're talking about, honestly, and neither do i think that you, yourself, know what you are talking about.

The other issue is that there must exist a base hash which would be proportional to world hash computational output on which then Moore's Law is applied to satisfy Sweft's Law.  Growing at the rate of Moore's Law is insignificant if the baseline threshold for attack is low.

Let me make it more clear.

If we take 70,000ghash and divide by 60 and multiply by 2000 we can arrive at the cost to replicate the bitcoin network security.  This is assuming 60ghash can be purchased for 2000$.

So, currently, the network hash is worth about 2.3m and valued at 600m.  So bitcoins are worth about 230x network hash.

To have a secure network from which we can then apply Sweft's Law the ghash should be about 100,000 times larger.  This would cost 230,000,000,000$ to replicate the network or 230b.   From there we apply Moore's Law and the network must grow in accordance to the Law to ensure minimum security

I just don't think that deflation will satisfy Sweft's Law and I believe thst a competing alt 2-3% inflationary cryptocurrency will eventually outcompete bitcoin for the various reasons i have outlined.

If bitcoin is still experiencing exponential growth in 2140, the world will have much greater problems than securing the bitcoin network, because that will require many times the current world population and that the population also be growing exponentially.

I already addressed this issue, please read the thread.

Fees are capped at 100% so increasing fees cannot support exponential growth.

You have three means to secure the network once bitcoin becomes deflationary.

1) exponential growth of velocity which is almost impossible given the deflationary nature.
2) exponential growth of price which would lead to hoarding and less coin velocity
3) mining at a loss of profit to protect the network and your coins

All of these are situations are irrelevant given moderate inflation.  Inflation supports network hash by supporting mining profits

computers won't vibrate anymore
natural vibrations is the key

What the fuck does this mean?