Topic: Forkless Towncoin Subnet proposal - final sketch - please tear it apart (Read 10138 times)

Do you literally want faster block generation than BTC's average of one block per ten minutes?

Ten minutes was chosen based on a planetary scale of operation: too fast to allow computers on Mars (maybe on Venus too) to partipate but slow enough that even old accoustic coupler modem connections might suffice if located on Earth, or something like that.

How fast do you want blocks?

How fast would allow even the slowest connections in town ample time to participate yet possibly effectively at least discourage distant towns, especially on distant continents, from participating?

I do not think trying to identify miners at their end is a good approach, I think better than trying to certify miners at the mining-software end is to simply have a set of approved addresses that minted (or even mined, meaning minted plus transaction fees) coins must go to in order for a block to be valid.

That way anyone can mine, even "angel GPU miners" (or angel ASIC miners etc etc etc), but either their only reward is the fees or their only reward is the warm fuzzies plus any gains they can get by trying to attack the network, unless they manage to log in to an approved pool.

Approved pools have means to cause the addresses all minted (or even all mined) coins go to to release funds. They can thus use these funds to reward pool participation.

Obviously we are talking alternate blockchain of course, since running at a high rate of block-creation will cause parts of Earth to experience problems similar to those experienced by all of Mars (maybe even Venus too) with the BTC network, depending on just how fast exactly you want your average rate of block creation to be.

-MarkM-

There will be Angel GPU miners producing coins. The rate of distribution to the town will be based on the relative hashrate of the town CPU pools.

Towns need to be able to cycle coins multiple times to get an amplified productivity from them. $10,000 cycled through a town 10 times can build an $80,000 house.

Cycles come from a resonance between participation and transactions. For this to work, block generation must increase. Based on certain variables the movement of money is preferred to the reward per block.

But I'm going to just go forward with Xenland and davout's work and tweak that.

Then could you once again explain your main point as specific and simple as possible?
I don't think anybody else has understood it.

Does your system make some coins special, e.g. have a different value from the rest? If yes, how does this happen? Do you want to piggyback the Bitcoin network for security but essentially create your own block-chain?

No trusted party is necessary. No movement of decimal places is made. These have to be determined by the subnet statistics not by top down decisions. I won't support solutions based on abritrary separation of participant and administrator. That is unacceptable.

Here's what I see as a workable system: a trusted party acts as a central bank. One or more blocks are used as money supply of arbitrary size, as you fix the exchange rate high enough and shift the decimal point.

The users can rely on the Bitcoin network for secure transactions.

You're giving up protection against government interference, though.

If what you're suggesting does not use any backing, I don't see how anything could be different from the current vanilla system.

I'm a bulldog. Everything I've seen people complain about can be solved through community action. With an angel GPU miner producing the acceptable coins, you can have whatever system you like behind that veil.

Saw your post about your initiative in 'my' topic today. Tried to read it but couldn't really get through. Love to read some more if you can put it together a little more coherent and preferably without insults.

I've got the means to do it.

This idea of the "towncoin" is good one, and if the technical issues
can be overcome, I see a huge potential. Small towns all over
American (and the world) could effective implement this as way
to keep the economies viable even with the collapse of the dollar.

If a town were to issue their own "bitbills" it could become a
local system of exchange among the merchants.

LETSystems often have all of the problems confronting any voluntary,
not-for-profit, non governmental, community based organisation. LETS
organizers often complain of being overworked, and may suffer burnout.
Many schemes have ceased operation as a result. Many of these problems
can be overcome through the use of Bitcoin. Moreover, the value of local
money is greatly enhanced if it can be converted and spent elsewhere.

I could see a day when various communities issue "Towncoins" in the
form of physical "bitbills" and then individuals can use this money in
the local economy or over the Internet. If the idea were to catch fire
we would see a rapid adoption the of the Bitcoin worldwide.

I, for one, would like to see this move forward.

Yes. Although most of the dynamics will be filtered by the fact it's a share based pool. So even the concerns TI3966 has disappear behind the angel GPU miner curtain. To be honest the small inconvenient shocks to mainline would actually be good in the long term.


I'm fairly certain it may require less monitoring than the mainline pools.


Initially, we are talking about a town ravaged by storms. There would be inspired individuals making the temporary connection, much like jump starting a car.


Once blockrate jumps inside difficulty outside jumps. GPUs will drift into town.


Mostly correct


Not quite. The townpool difficulty gets reduced, the reward is reduced for all. However, that special calculation is the attractive ratio aka soft gradient using the square root. I don't know if it will be needed.


Perhaps, but angel GPU miners should be able to kick that off.


This might be a nice feature especially for people to get accustomed to trading decoupled from fiat, as the value of a towncoin is 1 BTC. What towncoin will do is set a lower bound for prices for household items in town. This will have a feedback effect which causes mainline BTC to go way up. Gresham's Law will be hacked. Prices in towncoin will be more stable because of higher transaction rates so people will want BTC but towns will have a strong effect on price.


Think hammers not dollars. Trading for dollars will only have local effects. Trading .005 BTC for candy on one side of town and a hammer on the other side will force mainline to shuffle a bit. Whichever price wins out (hammers make things which bring BTC which make more things so probably hammers win) will force mainline BTC prices to shift upward. And nobody will be able to call it pure speculation. Physical speculation will be in the back of the minds of traders and then hold onto your hat.

Ok now we're getting somewhere. This would be a flaw except that towncoins would not be produced by a person. Only townpools would be able to produce them. Ultimately this would be a possibility only within the first few days of slow start pool.

And there's a feedback loop.

But it would disappear if many people connected quickly such as when the town establishes the townpool system. Bingo 36,000 people on the network. Also mainline difficulty would jump after 2016 blocks were generated. So yes it is an issue, but only a short term one. As mainline difficulty jumps more mining drifts toward the townpools and the gap shrinks and local difficulty goes up and mainline difficulty drops.

Paraphrasing from the first page: Every rule is based on a bargain.
Rule A. Reduced difficulty = Reduced reward. Reason: fast block generation. Bargain: effective reward is the same. Amount of reduction is equal to Global hash rate / pool hash rate.
As more CPUs connect the difficulty rises again.

Rule B. Increased mining per pool = Reduced reward. Reason: mining means more participation. Bargain: mining brings an abstract benefit, reward goes down.
As more people mine the smaller the reward per block which drives local prices down.
Towncoin food prices don't go up because some developer builds a hotel.

Rule C. Increased transactions reverse Rule A so difficulty goes up. Reason: transactions mean more participation. Bargain: transactions = abstract benefit, difficulty up.
As more people make transactions the likelihood of superconducting cycles (1000 BTC cycled 6 times build 5000BTC house) increases.

The reason for Rule C counter balancing Rule A is to return the system quickly back to mainline rhythm once the health and wellness of the town is established.
Also the purpose of Rule B isn't to generate more mining, but to react to it. Town people will mine to get back to working order. Same with Rule C.
Towncoins don't provide the incentive, survival provides the inspiration.

I'm gonna wait a bit longer before I go forward as that was a good find. It's easy to get sucked into a good idea with flaws. Please guys keep tearing it apart.

That's why he only wants to do it in his own "pond" within the real Bitcoin network. Transactions would still be done via the vanilla client, only the money generation would be different. There just needs to be a trusted authority (as in every pool) that watches over money distribution. Additionally he needs to have some incentives for people to mine - especially GPU mining must be more profitable than real bitcoin while returning far less coins. If there are 0 GPU miners in the pool, the few CPU miners won't find blocks themselves and the pond dries out because no money is generated at all.

If I get it correctly, he'd need:

- A pool that somehow detects CPU miners (Hashrate could be cheated by GPU miners launching multiple instances, so there must be another way - probably a quick additional calculation or something that is easy/inexpensive on CPUs but hard on GPUs but should not cause too much loss of hashrate...?!) and hands out getworks of different difficulty but rewards the submitted shares the same. Probably using a PPS system.
- A few Bitcoins to start the system until the first block(s) have been found
- A way to track Bitcoins entering the system and a modified client that gives people the chance to see if a coin or fractions of it is a special "Towncoin" or a regular "Bitcoin" (this might be the hardest part - not the detection, but the distinction so everyone gets the menaning of it!)
- An exchange where you can trade your Dollars for Towncoins and vice versa, with guaranteed exchange rates (that would be quite high, something around a few hundred USD for a full Towncoin probably)

In "Fiat money" this would mean, he'd collect 500 1$ notes, write down the serial numbers and guarantee that he will trade each of these notes for 5 $/piece. Whoever believes him, can now treat these as well like 5$ notes.

AntiVigilante, I read your post a few days ago and thought A was a pretty good idea, ie, decreasing the difficulty and lowering the reward proportionally. For one thing, it would practically eliminate the need for mining pools. However upon starting to read the bitcoin spec I realized that the entire point of having the difficulty scale with the CPU power of the network is to keep the block generation rate (near) constant. If your idea is implemented, it would allow any attacker to reverse transactions with very little processing power. All they would have to do would be to reduce their local difficulty to 1, effectively netting them 0 reward, but allowing them to overwrite large sections of the block chain, since they could get many low difficulty blocks accepted in the time it ordinarily takes to generate a full-difficulty block.

This is the point that I don't get - how will this work together with the mainline chain?! No mainline client will accept any "fast block".

All in all what you want to do is set up a "pay per share" pool and artificially give GPU miners harder shares to solve while still paying them the same as CPU miners to make sure CPU mining is more profitable - correct?

What I then don't get - why would GPU miners even want to mine in this pool? How do you distinguish between "real" Bitcoin and "Towncoins" and what is the added benefit of owning "Towncoins"?
I guess a subsidized, stable exchange rate, until the bitcoin rate has caught up? Anything else?

Also it might be quite hard to distinguish between someone who mines at 10 MH/s on his 30 PCs and someone who mines at 10 MH/s on 30 instances on the same GPU.

I do not want to decouple from the mainline. It makes no sense. Millions of townpools would provide the physical trade features people keep talking about but not being able to start up.
Decoupling will cause all the hard work behind bitcoin to become a casino all because of a slight algorithm detail.


LOL. Nobody could understand the technical explanation. Now you want technical Ok.

The first implementation will be using a share based pool system with angel miners. Like slush's pool but with davout's base so I won't have to reinvent the wheel.

What people keep tripping over is the fact that towncoins are bitcoins period. They are just produced in smaller quantities at a faster block rate to match the network value. Towncoiners will be giving up a lot to stay in line. The benefit to the mainline will be the pumping up of the value to where stores can break away from dependence on fiat.

50 BTC at 500,000 difficulty is the same statistically as .05 BTC at 500 difficulty.

True decentralization. But instead of creating the plugin (I'll play with it), I'm gonna use a shared based pool as a bridge.

If I understand the underlying idea (between tornadoes, watermelons and whatnot?!) correctly, you want to set up a pool within a town, get some initial bitcoins (to not be dependent on waiting until the first block is solved) and distribute these amongst townsters for mining duties. The mining itself is rather to have "proof of work" than to actually mine bitcoins, as you will distribute only your own initial coins anyways.

How is this different from just starting a new blockchain with all earnings in the genesis block?!


Please get a bit more technical!

First, this is very similar to slush's pool. In slush's pool everyone gets difficulty one. The angel miners with GPUs who produce the lucky coins have to share the output with everyone else. This just makes it so no angel miners are needed. Towncoins are BTC that are harder to generate.

As for an exchange being $2000, in a town that's been wrecked by disasters? You mean like handing out $2000 debit cards to Katrina victims?

[qoute]
You're talking about changing the rules on over 6 million bitcoins currently in circulation.
[/quote]

No actually it's to be implemented as a plugin.

Options would be something like:
--accept_dynamic_difficulty
--enable_dynamic_difficulty
--enable_soft_gradient


Megaminers are the central banks. They drive the difficulty up which splits the network into players and spectators. I know miners will have cut down or go into physical production to maintain the profitability of mining. But towns struck by tornados can't wait.


It's not an exchange. A "towncoin" is a Bitcoin which is much harder to generate. A townpool is a fast block producing CPU pool. You can have a million townpools if you like. But these fast blocks only contain small amounts of BTC. This is the bargain made with the network.

The people in the town need liquidity and transactions. This system simply makes sure they never produce coins which are of less value than mainline BTC. The reduced value is not a problem. Their value will rise because people need to repair homes and sell milk and eggs not stare at a computer screen waiting for MtGox to update.

In return for the bargain, towncoin productivity will push mainline value up as well. But this will only be realized on the exchanges.


Why should a storm ravaged town starve while they wait for critical mass?

An exchange can be set up for a couple thousand.  You're talking about changing the rules on over 6 million bitcoins currently in circulation.  How are you not trying to be a central bank?  I propose there will be 1000 exchanges in small towns before this gets into mainline.  You might have better luck starting a new block chain, but how are you going to build critical mass?

The majority of the modifications are a bargain to be let on the Bitcoin network without forking. The change in the code would happen in three stages and the first is the smallest. Like maybe 20 lines at most. A readable if then else snippet is 5 lines.

"Towncoin" reduces its output in exchange for faster block generation so towns don't have to wait. The blocks are valid on the blockchain and even more secure than mainline. No town is going to accept being run by MMORPG players. And the start up cost is ridiculous.

Also forcing towns to trade physical assets with the network instead of exchanging coins is EXACTLY WHAT NAFTA, CAFTA, and GATT have being doing to countries.

That's unacceptable.

I've read this thread like 5 times in the thought that maybe it is useful.
unfortunately each time it makes less sense than the time before it.