Topic: [ANNOUNCE] The Proposal for EnCoin - page 3. (Read 9455 times)

How does "smart" = "efficient"
and "stupid" = "inefficient"

in the case of achieving a stable currency based around a commodity? This is the exact same thought process that suggests "smart" = "early" in bitcoin.

A commodity is neither smart nor stupid, efficient nor inefficient. Megahashes are not the commodity, electricity is. I am not looking to reward who can get the most megahashes in a watt of electricity. I am not proposing the exchange and trade of megahashes. Megahashes are not a product. The goal is not how much profit can be made, the goal is a stable medium of exchange. When the goal becomes who can make the most money, the medium of exchange is lost. Instead you have an ever-continuing competition and the absolute clusterfuck waste of resources that accompanies it and absolutely no one benefits besides whoever comes out on top--for the time being. Is it so wrong that I want miners, sellers, traders, hoarders, savers, spenders, and aunt jemima all to benefit from my system? Not just the privileged, "smart" few?

Actually I wasn't going to bother since there is so little interest here. But I checked the survey results on Suggester's poll and 26% of the people support his proposal and 10% partially support him. That is 36% interest in stable money. I guess I will start a thread.

Funny? Why do you think all these people are not interested in you?


Yeah, it is going to come as a real shock to bitcoiners that a small group of people, make control and distribute their money supply! Most of the posts on this forum are from people who are in that group or want to be in that group.

P2P means peer to peer. Anyone in my system can be a peer. Anyone can be a trusted peer. They just can't be an "anonymous trusted peer". As you pointed out, that concept is just stupid. Anyone can however mint all the coins they want. Even anonymously! But only the ones who are both smart and efficient will be able to do so profitably.

Which is as it should be. Because only a complete moron thinks you should penalize the smart and efficient, to benefit the stupid and inefficient. That is the closest thing this site has to a core principle.

That is what I call "trusted central authority." That's great if that's what you want, but it's not p2p. Make your own proposal.

"HEY GUISE I got this idea where a small group of people make, control, and distribute the money supply! And if someone disagrees, they can post a message on a message board!"

Woot! One more!

That makes at least 5 people who have shown up and expressed support for stable money. I think if we get two or three more we can take a shot at occupying wall street!

But seriously, Suggester. Thanks for joining the discussion!

I was completely unaware of your thread, thanks for pointing it out. I don't believe the Bitcoin developers will be listening to the sound of reason any time soon. And the vocal majority around here is very pro-pyramid, so that could very well be used as the reasoning for not changing it.

I haven't read the whole thread, but it appears you want to stabilize the price of a bitcoin with the dollar, based on the estimated number of CPUs mining. This has 3 major problems, imo: 1) it doesn't protect against dollar inflation (which isn't terrible, but if the fed prints more dollars so does bitcoin), 2) estimating the number of CPUs is difficult if not impossible with the bitcoin proof-of-work design, 3) the security/continuity/dependability of the network still relies on massive amounts of hashing power, and always will. Bitcoin also has terrible scalability issues.


This opens up the network to attack, unfortunately. edit: Or would your "cooling down" settle up transactions too? Then that isn't bad. I'll have to read some more.

Although it's not in the proposal yet, I did mention somewhere in this thread I came up with a very reasonable solution to not require hashing power at all: let merchants put their money on the line to secure the network, and in return, refund most or all of their mandatory transaction fees. So if the economy is completely stable, no hashing power is necessary.

I'll peruse the rest of the thread when I have time as there might be some good discussion in there somewhere (I hope). I'm probably going to work on one final, major revision to the proposal, then that will be it from me for the rest of the year at least. If enough interest has been generated, I might be persuaded to even hire some coders if necessary. I simply don't have the skills to make this happen and I have no problem admitting that. And I don't have the time to invest in acquiring the skills for a non-profit venture of this scale. I love the concept of bitcoin, it's such a shame its vision is based on greed.

Etlase, the EnCoin concept seems very promising! It's what I've been hoping for since I started my infamous thread almost two years ago. You're welcome to read the initial post and kick some skeptic butts if you feel like as I revived it again today after the hacking of Bitcoin7. I believe the instability of prices is making users vulnerable to exchangers as they need to keep their "bitcoiny" wealth in fiat form to avoid price fluctuations, which defies the whole point as puts us back to square one.

I don't address any security issues or drastic changes like you do here though. The whole premise is about the economic necessity and the practical plan to tie bitcoin's price to electricity.

It'd be very sad if you had to go through publicity and advertising from scratch for EnCoin though. I was hoping the Bitcoin's developers would listen to the sound of reason and introduce these changes to the original client instead. If no steps are taken to ensure the coin's price stability it will never go mainstream. It'll be regarded as a very risky investment almost like gambling instead of being a friendly medium of exchange which everybody can use to escape the control of big corporations and big brothers.

This is a nasty sounding, but pretty fair summary of what I am suggesting.

Arbitragers will compete, not on how fast they can hash, but on how cheaply they can hash. The arbitrager with the lowest overhead wins. The mathematics of the minting algorithms prevent even the last rogue minter from driving the system away from stability.

If the latter holds. The former is always perfectly sensible. Otherwise, you are arguing to allow others to do exactly the same job, but at a higher cost. As all electrical costs are directly born by clients transacting in the *coins, minimizing this overhead expense benefits everyone.

Anyone is allowed to mint. However, the vast majority of people using this system won't mint, and won't want to mint. The arbitrage competition requires too much *human* mental effort, and it requires gambling with ones personal dollars.

However, as with bitcoin, anyone can run a client which keep tabs on the honesty of the rest of the network. These clients require trivial hardware to run and very little electricity. Absent all the needless hashing, monitoring honesty is a trivial problem.



This however is not a correct summary for any reasonable definition of "secure the network". I claim that security will be provided (as it is with bitcoin) by cryptography. This prevents all theft, fraud, forgery and transaction related mischief. Denial of service (preventing valid transactions from being acknowledged) and history substitution (chain swapping) is not subject to the minters at all.

Network continuity, and denial of service prevention become the responsibility of those non-anonymous parties that profit in the actual *coin marketplace. However, it is never compromised by 51% attacks. Denial of service/history modification requires 100% consensus among well known non-anonymous parties that can be held legally responsible for their actions.

So, I claim if even 99% of non-anonymous parties attempt to block valid transactions or change history, the last 1% has 100% of the evidence necessary to file charges against (or publicly disgrace) the others. Even if 100% of non-anonymous parties are compromised, every anonymous monitor holds the same evidence.

That is what I call "security".

And if you are curious about Red's proposal, this is the gist of it:


The issue he keeps trying to bring up with Encoin is precisely the same. Except that with his system, you have a technological elite with a monopoly on the money supply (sure Bill will keep going at a .009% ROI for 720 computer-hours a month of work) who can simply just hoard and drive up the prices at will without ever a whit of a chance for anyone else to make coins because he can simply make them when the difficulty falls back in his profitability range but no one else's. Drive up the price/let difficulty fall, sell, mine. Hey look we've got a more convoluted version of Bitcoin. And no incentive for anyone to secure the network but arbitragers and those who they've got by the balls.

Why would I care if a botnet is trying to earn coins? It doesn't undermine the system. And with the payout structure, it will be a much more complicated endeavor (nigh impossible with randomly assigned peers as I am now describing) trying to maximize the payout. With Bitcoin, all an entire botnet needs to do is send data to one peer and that peer will earn coins as though he were the entire botnet. Now instead the controller must often force his botnet to compete against itself and try to be average to avoid penalties of being too fast or too slow. If we assume these botnets are CPU mining (as the bitcoin community does), they need to be split into 40-50 computer computer sub-botnets, and each sub-net must actually participate in the network, unlike bitcoin. This means over 20 times the bandwidth required for a 1k computer botnet. Admittedly, this wouldn't mean much for a long time. In the mean time, anti-malware or firewall software should have no problem detecting this activity.


The payout structure has nothing to do with the sybil attack scenario I described. I wanted to make sure that for an agency trying to subvert the reputation system that it would cost dearly and accomplish little.


The difficulty isn't fixed, and inflation is created by demand. As of yet there are no supermachines that are less costly, more efficient, and capable of keeping up with everyday GPUs in both speed AND versatility. Until this is the case, GPUs will remain king of setting the difficulty. The best I can do is not adopt a scheme similar to *brix as it is aiming to be a CPU miner; rather, the two could be simply combined, or any other various algorithms could be used over straight SHA or whatever the hash is at the time.


Darn, EnCoin can't account for a major energy crisis. I knew there was something I was missing in trying to solve every single existing or future problem of the world.


So it takes a world crisis for an early adopter scenario to emerge with ENC, but with BTC it takes 1 person and a pre-mine. Hmm.

Awesome post BubbleBoy! I'm going to jumble the order of some of your quotes so I can clearly delineate where I am in agreement and disagreement with the EnCoin concept.



I totally agree with you. Which is odd, because it clearly I'm saying something that seems very close to the opposite!

The part I'm agreeing with is 1 *coin should *not* be pegged at 10 KWH as EnCoin suggests. Specifically, by requiring each coin to represent burning an average of 10 KWH. I find that impossible for the same reasons you do. What I'm saying is, electricity is the only commodity I've noticed that has even the potential to become a tool for stabilizing *coin prices.

Electricity has one unique characteristic. As Etlase2 points out, we can require someone to purchase at least some energy prior to creating a new *coin. This means the coiner has "skin in the game" prior to creating/selling coins. I don't see any other basket of goods that has these characteristics. I can't force someone to burn wheat in the process of manufacturing a *coin.

The question then shifts to how much electricity do we need to require at this moment? While it is impossible to determine exactly how much any particular person must purchase, fortunately, it is possilble to dictate that they purchase more relatively speaking or less than was required in the previous period. (Changing difficulty)



I totally agree with these statements. Both would be bad and should be avoided.

What I'm proposing is the (coin/kwh) requirement is constantly changing algorithmically based upon market conditions and the demonstrated need for new coins. The price of electicity serves as a limit constraining two non-linear functions. The (coin/$) exchange rate and the (kwh/$) exchange rate.

Again, even that sounds impossible. But, here is the edge.

I'm not suggesting the algorithm dictates when coins must be made. That remains a human decision. Human's make coins when Arbitrage between (coin/$) and (kwh/$) makes it profitable to do so.

In a stable economy with a *coin price at a stable equilibrium, there is ZERO need to create new coins. Therefore, the proposed algorithm must give zero incentive to creating new coins. At this point 1 Coin = X kwh and there is no arbitrage profit. An algorithm CAN monitor the lack of coin creation.

If the economy is increasing (more $ of goods to trade), then the (coin/$) relationship will vary from the (kwh/$) relationship. This creates arbitrage incentive to create new coins. An algorithm CAN monitor demonstrated coin creation.

If the economy is decreasing (less $ of goods to trade), then the (coin/$) relationship will vary from the (kwh/$) relationship. This creates an incentive destroy coins and/or discourage selling coins. I propose a transaction TAX burning coins for this. Again, an algorithm CAN monitor coin destruction.

I proposed the beginnings of an algorithm in a prior thread. But the gist goes.

If people are minting excess coins, then make it more expensive to mint coins. (increasing economy)
If people are being taxed, then make it cheaper to mint coins. (decreasing economy)
If people want to mint just enough coins to rebate their tax, let them. (stable economy)



I totally agree with you. I'm suggesting competition between arbitragers to monitor the market and keep prices stable. This is the opposite of "Everyone can run minting in the background without thinking about it."



Totally agreed. In this model I want to minimize the number of people who have to burn needless electricity. That is just unnecessary added overhead to the *coin economy. Only arbitragers would need to solve proof-of-work problems. And then, only when they see it as profitable to do so.

--- Note ---

If electricity prices vary dramatically, or even if the economy varies dramatically (Walmart accepts coins), then coin/kwh relationship we move from the previous equilibrium.

The goal for the system, is to assure that it always moves to a new stable equilibrium rather than tending to zero or infinity.

Those goals are entirely different. Bitcoin achieves 300 coins per hour axiomatically, and it does not attempt to control who gets it. You are trying to guarantee an average computer gets close to one ENC, yet there is no way a computer algorithm can do that. You can only check for the solutions to a hard puzzle that takes about 50 hours to solve on a PC, 100 hours on a laptop and 10 seconds on a botnet. You will never be able to determine algorithmically that the results were submitted by a botnet.
The payout structure you have devised to counter this phenomenon seems arbitrary and stems from a lack of understanding of the sybil attack. If anything please (re)read the conclusions of JR Duoceur's paper on sybil attacks and try to apply it to your system. It seems you are achieving what he claims is impossible, a distributed system that can discriminate against powerful peers and exclude them without using a central identity issuer.

If you drop the whole egalitarian/community BS what you get is a bitcoin-like system with fixed difficulty, in which inflation is proportional to hashing efficiency. Various schemes have been proposed in this board that attempt to counteract the effects of Moore's law, but the general idea is the same, the most efficient miner wins. The best you can do is adopt a scheme similar to scrypt/*brix that should keep ASICs or FPGA at bay for a few years, but you open the door wide to botnets.


Stable prices are a worthy goal. Pegging currency to any one commodity is not a good idea. The greens are in kill-mode against nuclear energy. Germany is shutting down it's reactors. Renewables are too expensive even in this era of cheap oil. Assuming a major energy crisis hits the planet, the effect is high prices of energy, inflation and recession. The Enron-Coin algorithm counteracts by halting monetary expansion, hopelessly pro-cyclical. In such a situation inflation is necessary to enable higher prices to relocate the lost wealth.

If you are declaring that 1ENC=10KWh you are giving an unfair advantage to holders of ENCs, you treat them as if the energy crisis did not happen and they can buy the same amount of goods and services as before. Well, the crisis did happen, and this can only mean the rest of the populace needs to further constrain their consumption to fulfil the needs of the ENC aristocracy.

I also have a visceral dislike for proof of work currency (a.k.a waste of useful resources), but it's very hard to solve that problem with a distributed currency.

I'll admit the previous post was lame. I was a little defensive when I should have read a little further into your post.

Suppose that 6950 equilibrium comes at our long coveted 1 ENC = $1 = 10 kwh
Did you work through the math?

6950 sells coins at $1 with a cost of $1 for a net of 0.
Altera sells coins at $1 with a cost of $0.10 for a net of 90 cents per coin.

Even if the sell price goes up. And the 6950 begins minting 5 coins for each 1 Altera mints. Until the ENC price reaches $1.23, the Altera still generates more proft on its one coin, than the 6950 generates on its 5.


More than you might think!


I'm sure 90 cents on each generated coin will pay back even a $600 machine pretty quick. Much faster than the nickels you keep mentioning.

I was going to make that point but realized I didn't have to. The Altera is supported by every other GPU that is even a little more energy efficient than the 6950. The chart shows lots of possibilities.

And Altera and company don't have to keep driving the ENC market price down at the velocity the 6950's were driving it down at. They just have to hold it down. Its much easier to drown someone if they're already underwater.

And as you obviously know, all that fee money has to go somewhere. Now it is going to the most efficient where it belongs. The irony is, to bring the 6950s back to profitability, you have to increase the difficulty while the market is stable. This increase is only to drive out electrically efficiency and add overhead. Sounds totally sensible.

teehee, you forget that there need to be 4.5x the alteras as 6950s! Otherwise they can't support the demand. I wonder how many people are going to program circuit boards to make a few extra cents a month?


I believe you are making the wild claim that a $600 machine is free.

I have to admit, I don't even know. Too much cut and paste and bleary eyes. But too many little boxes for me to dig through and find a different example. But it doesn't matter! I don't have to match the 6950 in Mhash/s.

(I don't really remember any of the details for "minimum requirements" to maintain your reputation and assure your spot in the minting pool. But I think it was generate at least one FNBlock a day.)

Say when ENC hits "6950 equilibrium", the 6950 is generating 5 blocks a day and the Altera is generating 1 block a day. Now the (ENC_price - 6950_cost_to_produce) = 0. The 6950's ROI is zero. It has to stop minting or continue at a loss.

This is your intended behavior, nothing more. I'm not making any wild claims.

The Altera on the other hand is still profitable at 1 block a day. It can keep minting. So can all the other Altera equivalents. They are 10x more efficient than the 6950. From this moment on they continue earning a ROI.

If there are enough Altera equivalent peers to hold the price steady, just below the 6950's break even, the 6950 peers are for all intents an purposes dead. They don't have to be 10x more efficient. 2x is probably enough.

Oh did I mention that if the economy goes through a big expansion, the Altera will have 4.5x less coins to try to make a big profit with? A shame, really. (This is called Opportunity Cost - I'd give you a wiki link but I'm sure you can find it.)
Or that in the 3-4 years it takes to break even, the algorithm could change to something else and render it useless? Whether this is a new crypto hash or simply the same hash but with memory/cpu work. The EnCoin design is a shitload more forgiving of changing this. And merchants would be eager to accept any change that keeps the value of their ENC stable.

Or that in the 3-4 years it takes to break even, it might just break. LOL!

I'm honestly curious how the 6950's 360MHash/s turns into .336. I mean it's only off by a factor of 1000.
So if we use the real numbers, the Altera is 4.5x slower than the 6950.

I'll even give you a bonus and say the 6950 was specifically purchased for 50/50 gaming and mining. So the mining unsunk cost is $125.
$299 if you're a student, sure I'll give that to you, minus $125 = $174 unsunk cost difference.

Since the Altera is 4.5x slower, it will earn 4.5x less coins, on average. If 15 ENC is average and the average card is a 6950, that means it earns 3.33 ENC per month. Fuck it let's just say the Altera is completely free to run, and an ENC is worth $2. 26 months to break even ruffalo (and in the mean time the difficulty is going up as people gradually upgrade their computers, so really we're looking at more like 30-36 months. And since it will probably miss the minimum hash value on occasion, probably 36-38 months.)

Altera EP4CE115C7   80 Mhash/s 18.18 Mhash/J   $299 academic
6950      0.336062 Mhash/s   1.8 Mhash/J  $250

The Altera does 18.18 Mhash/(watt second) = 65,448,000 Mhash/kwh
The 6950   does  1.8 Mhash/(watt second) = 6,480,000 Mhash/kwh

The Altera is 10 times more energy efficient than the 6950. Both cost less than $300
The Altera is 240 times faster than the 6950.

The Altera doesn't have to produce 240 times as many coins as the 6950 thus burning more total dollars.
It can simply produce 10 times more for equal dollars and loaf.
Or it can an equal number of coins for 1/10 the price and loaf.

The Altera has the choice to decide what is the most profitable way to run the 6950 out of business. It just watches the ENC sell price fall until the 6950 quits. That means as you pointed out that:
ENC price = 6950 cost to product
ENC price = (10x) the Altera's cost to produce.

Now at this point, the Altera doesn't have to be a total dumbshit and mint until the coins fall to ENC/10. Or show off and invite everyone else into his game. He just holds the prices steady making maximum profit, while telling everyone else, he's losing money for the good of the network.

Still don't get it?

The whole point of the system is to reach price equilibrium. This is exactly what would happen if every node was a 6950. ENC would reach the exact same price. Lots of people would mint a little and sell them off. Sometimes they'd make money and the ENC price would fall. Sometimes they lose money and the ENC price falls.

The Altera just makes money every time. As a side effect everyone else makes less, without the ability to know they could be doing better. The Altera guy could just be a better market timer than them. He's certainly not doing anything crazy looking.

If they upgrade and mint faster, everyone will upgrade and what good would that be? Just more sunk costs for everyone and no more profit to offset it. Better to stick with Ryan's original plan.

No, you deliberately ignore them because they clash in the face of your argument. Your magical, 100% more efficient hardware is simply free! And nobody else knows about it! And 20 outdated PCs are somehow cheaper to run than 1 modern PC! I guess?!?!

Here's a link: https://en.bitcoin.it/wiki/Mining_hardware_comparison

Notice how the more efficient hardware gets significantly less mhash/$? :GASP: Who'd have thunk it? Oh, maybe me! And that today's 5970 is tomorrow's 5770 and what affect this has!


Difficulty - Known by everybody! For the sake of simplifying, we will say it means "amount of hashes on average it takes to find 1 ENC"!
ENC Sell Price - that would be market price in case this is confusing. Determined by competition (supply) and demand!

Then We Get:

(Difficulty / (MHash/s * 3600)) * (kWattage * kWhCost) = Cost To Produce!

Then We Get:

(ENC Sell Price - Cost To Produce!) / (Cost To Produce!) = RETURN ON INVESTMENT

If this number is positive that means it is profitable to mine! It does not matter if an ENC costs 10kwh or 8kwh or 5kwh or 15kwh to produce! The market will figure it OUT!


Computational Unpredictability LOL! How would a new user know that everyone else is only running at half mast? They would only see that they are making twice as many coins! You assume the same flaw in logic that you think I have! Why would they not want as many coins as possible when the price is profitable? Because half the network might be secretly pretending to run a less efficient system?

ROFL


See above!


LOL @ "claiming" & "minimum watt requirement"! "Hey Bravo, this is inefficient system ZQ-Alpha punching in at 220W. Please provide my subsidy!"

I deliberately ignore them because they have no bearing on any of my mathematical arguments. I could just have as easily said, talk 20 outdated old PC I got for free, and outsourced the hashing. Or created a botnet and outsourced the hashing.

You're answer is always, "Yeah, but everyone's going to notice. Then they'll all start doing it too! Inflation will start. Everyone will decide inflation is bad, so will will agree to increase the difficulty. So there!"


I clearly demonstrated how they are not REQUIRED to "benefit others as well." You did not refute my point.


Again, you just said, "Of course everyone will take the path that's least profitable to themselves. How could they not want to maximize their momentary gains while sabotaging their long term profitability?"


But what you fail to acknowledge is you've just run all the machines off the network that you designed the network for. Now you have a network of people who understand the benefit of minimizing hashes/watt. All secretly competing to minimize their hashes/watt cost.

That was what I have tried to explain from the beginning. You simply cannot algorithmically force the efficient to subsidize the inefficient. That has to be done at the point of a gun.



You took the word right out of my mouth! If you wouldn't mind, why don't you invite some of your IRC friends here to critique my density!

----

I know you addressed the FreeNet Sybil problem by adding a fee to create FNs. I'm assuming you've already solved the Peer Sybil problem.

I'm sure you already noticed that in the grand scheme of things, it is trivial for one computer to serve as a FN Peer to multiple FreeNets simultaneously. Most of the work is completely redundant. If I'm receiving transactions, validating them, and signing them with one FN key. Then it is no more bother to sign the blocks as a representative of 10 different FNs. The additional broadcasting is really not a barrier with today's bandwidth.

The only thing that can't be done simultaneously is hashing at 100% cpu power for each network. However, I've clearly shown that is not required.

I could take my one already efficient machine (or plug plus an outsourced hashing rig) and create 10 virtual machines. Each would hash at 2mh/s and claim to running at 220W (or whatever the minimum requirement). Then I would be the someone in your, "subsidizing someone getting 2mh/s for 220W" example. But Woot! I would be getting whatever subsidy you are proposing ten times over!

---

What if I did that with my 10 plug computers creating 1,000 virtual peers for $1,000 and then outsourced hashing to one of those big bitcoin rigs looking for a little steady revenue. I'd fake being 1,000 slow peers and take the subsidies you propose from the others greedy supercomputers.

Then if word got out, people would compete with me by faking having the most slow inefficient boxes! How perverse! We'd be 3 guys with 30 plugs, simulating 3000 crappy peers taking subsidies from 3000 slightly above average peers. All while paying off one high-end bitcoin GPU rig to do as little hashing for us as we could get away with.