A fair and strong Bitcoin fork design exploration, "Faircoin"

[deepceleron]

Why? Bitcoin isn't fair; unknown early adopters (not you) mined with little competition for the first half of it's existence. They can crash the currency value if they want or cash out for hundreds of thousands of dollars. Deflation: A 10,000 BTC pizza is now a 10,000 BTC Mercedes. Other alternate coins have had launch problems: selfish premining, monopolizing landrush on low difficulty launch, easy denial attacks.

How do we launch a new currency that otherwise is identical to Bitcoin? Delayed mining!:
-The first N blocks have no reward,
-The genesis difficulty is high (~100,000 vs Bitcoin's current 3.3M),
-Insertion of full coinbase block reward into generated block only becomes valid if the previous difficulty period was > 1 million
-"pre-launch" blocks may either have only this 0 generate value, or optionally a maximum testing value of 0.01 if deemed necessary.

Design:
Except for these fair bootstrap rules and putting it on a different P2P network with different addresses, it IS Bitcoin, and stays a minimal patch against the current Bitcoin client. This gives a long period for setup and adoption, as at least the hashrate of 2-3 large pools is required before anybody starts earning coins (plus you get around 2 weeks warning of the block that begins coin launch) The currency must be protected with hashrate before any coin generation, discouraging attack or exploitation by a single entity.

Outcome:
-A long period for adoption, the hashrate must first scale near the size of Namecoin's current merge mining to be "launched".
-Installed base of client software, users with addresses, and miners possible before any coins exist.
-Merged mining this Faircoin is a no-cost bonus to pools and miners, now your pool pays Faircoins too.
-Currency starts rare, and continues to be rare. Miner earnings likely to be similar to Bitcoin earning rate (depending on pool merged mining adoption) due to high difficulty threshold.

Use: Will it be worth anything? Namecoin is worth 1/100 of Bitcoin, and it has little current use and a sub-par command-line-only client. This starts hard to mine and also is almost "forced" on everyone by the necessity to merge mine, so it is widely distributed from the first reward block.

Tell me (besides it being another alt-currency) why this is bad. The only flaw is that it might not "launch" due to pools never putting enough merge mining hashrate towards it.

[grich]

-The first N blocks have no reward,

How can I use Faircoin without coins?

[deepceleron]

-The first N blocks have no reward,

How can I use Faircoin without coins?

-A long period for adoption, the hashrate must first scale near the size of Namecoin's current merge mining to be "launched".

You would be able to download the client, synchronize with the blockchain, generate addresses, put your "Faircoin" (example name only) address into your pool setup, add other's addresses to your address book, etc. Coins can only start being mined after the hashrate is high though, making them hard to get from the very first "earning" block.

It may take weeks or months for enough pools to adopt it and start hashing enough to "turn on" the coin generation. That's the point - lots of time for everyone who is interested to get involved, adopt and implement software, no unfair "rush". The miners of the first pool to find a 50 Faircoin block then have them all, until the next block 10 minutes later. A typical one-rig miner might get 0.5 faircoins a day, so they start rare, encouraging immediate value and even speculation.

600 MHash/s rig at perhaps 1,100,000 first-coins block difficulty = .549 coins/day

[Syke]

unknown early adopters (not you) mined with little competition for the first half of it's existence.

So early adapters that helped make the coin successful were rewarded. Sounds fair to me.

Tell me (besides it being another alt-currency) why this is bad. The only flaw is that it might not "launch" due to pools never putting enough merge mining hashrate towards it.

So your incentive to start mining is that you get the same reward as someone who comes along later, after the coin is successful?

Your design isn't fair. It's anti early adapter.

[ElectricMucus]

The idea has been around.

What you'd have to do is make mining a sigmoid function. That way the block reward would follow a probability distribution curve.
So the bock reward would be low, although not zero at the beginning, high during the median period and again low at the end. One way to do that with discrete values is to calculate a long row of the pascal triangle and make consecutive block reward follow that row.

It's an interesting concept, and trying it out shouldn't be that hard. The thing is however it solves neither the lost coins problem nor the need for pooled mining.

[deepceleron]

Your design isn't fair. It's anti early adapter.

You still get a two-year mining head start on the person that discovers digital currencies in 2015. This also doesn't take away anybody's Bitcoins, which is likely to always be the dominant Satoshi-based coin. However, this gives you a coin that doesn't have that mystery element of "who has the 4 million difficulty<1,000 bitcoins?"

[ElectricMucus]

Just had an idea: If the pascal triangle is used anyway it could be used to derive a sort of "soft difficulty" where it is possible to mine at several difficulties at once and the reward for the blocks would be proportional to how near the row is to the top of the pyramid. Also the block reward would adjust more quickly for low difficulty mining and slower for high difficulty mining.
So that would solve the instant transaction problem as well as pooled mining. The only shortcoming it cannot solve on it's own is the lost coins problem.

[deepceleron]

The idea has been around.

What you'd have to do is make mining a sigmoid function. That way the block reward would follow a probability distribution curve.
So the bock reward would be low, although not zero at the beginning, high during the median period and again low at the end. One way to do that with discrete values is to calculate a long row of the pascal triangle and make consecutive block reward follow that row.

It's an interesting concept, and trying it out shouldn't be that hard. The thing is however it solves neither the lost coins problem nor the need for pooled mining.

Another (not-at-all-recommended) wild idea completely different from anything I've seen put out there yet - an inflationary coin, where the number of satoshis mined in a block = block difficulty, with no limit to number of coins that will exist. This lets the currency base grow with the adoption rate (measured by mining), however a divergent series as naively implemented as this could quickly "inflation" the value to zero over time.

Probably a good median is a permanent 50 BTC reward (to counter "lost coin" and also address the "fees must replace reward" problem), but that doesn't address this fair launch of a new coin, which is my primary focus here.

[ElectricMucus]

The idea has been around.

What you'd have to do is make mining a sigmoid function. That way the block reward would follow a probability distribution curve.
So the bock reward would be low, although not zero at the beginning, high during the median period and again low at the end. One way to do that with discrete values is to calculate a long row of the pascal triangle and make consecutive block reward follow that row.

It's an interesting concept, and trying it out shouldn't be that hard. The thing is however it solves neither the lost coins problem nor the need for pooled mining.

Another (not-at-all-recommended) wild idea completely different from anything I've seen put out there yet - an inflationary coin, where the number of satoshis mined in a block = block difficulty, with no limit to number of coins that will exist. This lets the currency base grow with the adoption rate (measured by mining), however a divergent series as naively implemented as this could quickly "inflation" the value to zero over time.

Probably a good median is a permanent 50 BTC reward (to counter "lost coin"), but that doesn't address this fair launch of a new coin, which is my primary focus here.

I don't think that would matter if the coin would be inflationary anyway the block reward could remain fixed for all practical purposes. At some point there would be an equilibrium where the number of lost coins is as high as the newly generated coins. Geistgeld used such a model iirc.

If a non-inflationary model is desired there could also be a "storage fee" which would have to be tuned for a Nash-equilibrium with the transaction fees. This way lost coins would eventually be mined back into the pool. This would also limit the hoarding problem and would prevent a deflationary spiral even with a fixed supply. I also think a cryptocurrency could have two representations at the same time: "Coins" where there is a fixed supply and "Cash" where there is an inflationary supply. The limits are only rounding errors.. The only thing which would need to be hard coded, everything else could be voted on by the poof of work mechanism.

[crazy_rabbit]

I think you still have failed to describe why the early adopters of Bitcoin making it rich is a bad thing. They made money, got motivated to make more money, got the word out, built services, etc....

The idea of waiting until a certain point before giving miners coins just creates and incentive for people to NOT mine the coin until it reaches some theoretical point. Because of course- our power is limited so why give it over to a coin were we get nothing back for it? And why wouldn't just EVERYONE wait to mine hoping someone else will spend all the time/energy and then you can jump in and reap the rewards?

No point in making something anti-early adopter unless you plan on doing and the early adopter work yourself.

[Etlase2]

Another (not-at-all-recommended) wild idea completely different from anything I've seen put out there yet - an inflationary coin, where the number of satoshis mined in a block = block difficulty, with no limit to number of coins that will exist. This lets the currency base grow with the adoption rate (measured by mining), however a divergent series as naively implemented as this could quickly "inflation" the value to zero over time.

Oh this idea has been proposed in the past. It is not a good one though because tying coin reward directly with difficulty means that as computing hardware gets better, people make more coins. It is a very big disincentive to use the currency when inflation is a factor of computing power. That is why with the Encoin and Decrits proposals I've made, mining is separate from network security and is based around a per-user approach rather than the lottery approach of bitcoin.

Probably a good median is a permanent 50 BTC reward (to counter "lost coin" and also address the "fees must replace reward" problem), but that doesn't address this fair launch of a new coin, which is my primary focus here.

This is also addressed with Decrits. Early mining is rewarded in several multiples of the regular award, reducing over time until it reaches the regular award. Early adopters are still well rewarded, but they will not be able to purchase a mercedes from a couple dollars of electricity.

I think you still have failed to describe why the early adopters of Bitcoin making it rich is a bad thing. They made money, got motivated to make more money, got the word out, built services, etc....

I think he stated it somewhat clearly:

They can crash the currency value if they want

This is a significant detriment to business and trade.

[BlackBison]

Ok. Send me a PM when the first (N -10) blocks have been mined 

[ElectricMucus]

I think you still have failed to describe why the early adopters of Bitcoin making it rich is a bad thing.

Ideology. This is not debatable.
At least not in a rational way.

Ok. Send me a PM when the first (N -10) blocks have been mined  

You realize that difficulty would compensate so that there still would be a early adopter benefit? At least with my model.. It just wouldn't be as steep.

[deepceleron]

Ok. Send me a PM when the first (N -10) blocks have been mined  

The idea of waiting until a certain point before giving miners coins just creates and incentive for people to NOT mine the coin until it reaches some theoretical point. Because of course- our power is limited so why give it over to a coin were we get nothing back for it? And why wouldn't just EVERYONE wait to mine hoping someone else will spend all the time/energy and then you can jump in and reap the rewards?

The advent of merged-mining gives a zero-cost (except for the work of setting it up) ability to support and harden a second crypto-currency with mining. Any successful future coin must, in my opinion, be adopted through merged mining support. Your PPS pool may be merge-mining and not even be telling you.

As an example of merged-mining success, over 1/3 of all Bitcoin hashing is also currently mining Namecoins. However, this is not notable to most miners, due to Namecoin's difficulty of use (due to a sub-par client), it's predictable failure in it's primary purpose (an alternate DNS system), and it's low perceived value per-hash relative to Bitcoin (adding less than 1% to a miner's income after exchange).

Namecoin faltered primarily due to it's introduction before the advent of merge-mining. One had to choose to only mine one blockchain at a time. When Namecoin became more profitable than Bitcoin, everyone jumped on it, mining a two-week difficulty period in just a few days and multiplying the difficulty by four. After the new difficulty was set, making it unprofitable again, miners quit, paralyzing the transfer of money. The currency has been burdened by this since, and the value stays low due to the abundance of no-use-but-to-sell coin.

The introduction of a new coin that has a reward turn-on threshold like I describe above may have a similar "jump on and mine the shit out of it" contingent like the above posters, however setting the "turn-on" threshold required to over 1/4 of Bitcoin's present hashrate means that it cannot make a massive jump - every remaining pool turning on merge mining at once would be the max increase possible, and in reality even a doubling of hashrate at reward start is possibly more than can be expected. The perception of a profitable 2016 block rush may actually encourage and cement the wide and permanent adoption of such a new currency. The new currency won't be left behind to languish like Namecoin was, as there is no advantage to discontinue merge-mining of such a new blockchain once it is implemented, unless it completely falls into disuse. The only obstacle is getting it initially added "for free" and hashed by optimistic pools, which was my only concern expressed in the first post (however the pools that add it first are going to be ready for it's rewards first and attract the miners).

A reward ramp-up period may actually be counter productive. One of the caveats would be low perceived initial income, causing lackluster enthusiasm that may be insurmountable, but another consideration would be the difficulty pools may have in calculating and distributing payments with an inconsistent or unpredictable reward (plus the further departure from mainline code).

[FreeMoney]

You still get a two-year mining head start on the person that discovers digital currencies in 2015.

That doesn't seem fair.

[ElectricMucus]

The problem with merged mining is it restricts innovation and the experiential nature of cryptocurrencies. That is ok for something like namecoin where the economic aspect isn't that important and you are mainly concerned with security. And with the advent of Keccack there is even less reason to do it.

The main shortcomings bitcoin has cannot be solved with merged mining since from the technical aspect nothing can be changed except the contents of a block.

[deepceleron]

You still get a two-year mining head start on the person that discovers digital currencies in 2015.

That doesn't seem fair.

With Bitcoin, if I find it four years later, I get punished with half the reward and a difficulty 3.3 million times higher too...unfair x 6M?

What kind of approach to currency distribution might you suggest to make a coin fair for even late adopters?

I think the best one can aim for is as wide as possible adoption at introduction. Anyone mining now on a new alt-coin that demands network acceptance is likely to get some "newcoin" in their purse through merge mining. The mining of coins seems to find equilibrium where generation power costs approach the cost of just buying the coins, and if we have that at introduction along with relative scarcity, then the value should track money supply and not lead to the anti-late-adopter hyper-deflation that distracts from use as a currency.

The problem with merged mining is it restricts innovation and the experiential nature of cryptocurrencies. That is ok for something like namecoin where the economic aspect isn't that important and you are mainly concerned with security. And with the advent of Keccack there is even less reason to do it.

The main shortcomings bitcoin has cannot be solved with merged mining since from the technical aspect nothing can be changed except the contents of a block.

Several alt-coins have died from their old unmaintained completely forked client becoming obsolete, maintained only by one profit-seeker. Many ideas (constantly adapting difficulty, faster blocks, transaction message passing, different hashing, new rewards, etc) may find consensus as the ideal for a re-imagined Bitcoin v2.0, but would only succeed if they continue with the level of development that Bitcoin's client currently has, and would still require a launch technique that would be resilient against abuse from the 24THash/s of existing miners. Being able to patch the Electrum client or your web wallet software with just a few rules for "neucoin" instead of requiring a completely new (old-code) client may be more successful, unless several key Bitcoin developers or others with their skill level pledge support to the philosophy (and maintenance) of a complete alt-fork running alongside Bitcoin.

[ElectricMucus]

Several alt-coins have died from their old unmaintained completely forked client becoming obsolete, maintained only by one profit-seeker. Many ideas (constantly adapting difficulty, faster blocks, transaction message passing, different hashing, new rewards, etc) may find consensus as the ideal for a re-imagined Bitcoin v2.0, but would only succeed if they continue with the level of development that Bitcoin's client currently has. Being able to patch the Electrum client or your web wallet software with just a few rules for "neucoin" instead of requiring a completely new (old-code) client may be more successful, unless several key Bitcoin developers or others with their skill level pledge support to the philosophy (and maintenance) of a complete alt-fork running alongside Bitcoin.

Well, I have as an argument the relative success of litecoin which is a significant step away from the merged mining aspect.

I concur with the assessment that forking the bitcoin software is a wrong way to apporach alternate cryptocurrencies in general. I am not that interested in another bitcoin fork, if anything it should start from scratch. This is the only way real innovation can take place.

[SgtSpike]

I think this is an interesting idea, and though I won't attempt to predict the viability of it, I would like to see it happen.

Alternatives to Bitcoin are not a bad thing.  Bad alternatives are a bad thing, but I'd call this one a neutral alternative.  It has good intentions, it is just a matter of experimenting with the idea and see if it actually works in practice/

The most important aspect for success would be getting enough pool operators onboard.

[Akka]

I think this is an excellent idea. But it only delays the early adopter issue. A coin that would allow an equal distribution for adopters in the first x years, or maybe until a first 100.000 users are established would IMO have grate chance for being successful. But the how to enable this, is a problem.