*Coin Complaint List

[Bicknellski]

I'll add

*setting it up so that you get everything and everyone else gets nothing (freicoin)

That is lie. It is 80% to a foundation that will redistribute coin agreed to by the USERS not one person or small group it will be a consensus.

What really happens:

20% goes to the miners now...

5% demurrage will go to the miners as Freicoin loses value.

And if we can't find an equitable way to distribute Freicoin then that 80% will be destroyed and demurrage will recoup 100% of that for the miners.

*Hate it when people BULLSHIT and LIE about altcoins.

[LaggedOnUser]

Here is a coin that seems to be addressing one of those criticisms.  A new MC2 currency plans to incorporate democratic elements of inflation adjustment in its design (https://bitcointalk.org/index.php?topic=169204.0).  That seems to address the complaint that "Hard-coded parameters don't allow for non-procedural feedback mechanism" to some extent.

I thought of some more criticisms:
* Complex in design, opaque in operation, hard to understand, hard for novices and non-techies to use.
* Lack of regular mining payback for people without advanced rigs discourages them from signing up and participating.  I.e., they mine for 6 months and get one coin... why should they feel involved?
* Too many alternative coins with very little feature differentiation confuses the potential user.

That gives me an idea.  Imagine a coin that gives 1/2 the reward to those directly mining it and splits the rest of the reward evenly to everyone else on the network in proportion to their stake.  Thus their account functions like a bank account and returns a regular rate of interest.  It would also give regular payouts of interest... say once a month... to everybody just for participating.  If coin age can be measured, this can be done without actually transferring coins just by valuing coins with a mathematical formula based on their age.

[markm]

Hmm sounds like that could turn out to be bread-and-circus coin.

(The masses voting themselves more and more goodies...)

-MarkM-

[LaggedOnUser]

Hmm sounds like that could turn out to be bread-and-circus coin.

(The masses voting themselves more and more goodies...)

-MarkM-

I think the design limits their choices to +/-1% inflation.  Although I agree that historical examples of crowd-controlled currency, such as Revolutionary France, ended badly with hyperinflation.

[LaggedOnUser]

I thought of one more "criticism", actually feature request.  The cost and certainty of a transaction don't scale with the transaction size.

It occurred to me previously that we could spend less energy confirming transactions if they scaled with the size of the transaction.  This would use random sampling like they do in quality control.

How it would work is that the person receiving the currency could pick a random number of servers depending on the size of the transaction, i.e., a transaction for a $2 cup of coffee should be quick to confirm, lightweight on computer time, and doesn't require much certainty.  A $200,000 house transaction can be slow to confirm, take a lot of computer time, and requires much certainty.

Imagine that you start off with a 50% chance of having a good or bad server, and each new randomly chosen server cuts the odds of a bad transaction in half (a simplifying assumption intended only to demonstrate the concept).

10 servers = About a 1/1,000 chance of a bad transaction
100 servers = About a 1/1,000,000 chance of a bad transaction
1000 servers = About a 1/1,000,000,000 chance of a bad transaction
10,000 servers = About a 1/1,000,000,000 chance of a bad transaction

So for a $2 cup of coffee, you pick 10 servers and have an expected loss per transaction of 1/1,000 * $2 = $0.002 (i.e., 2 tenths of a cent).

For a $200,000 house, you pick 10,000 servers, and also have an expected loss per transaction of 1/1,000,000,000 * $200,000 = $0.002 (i.e. also 2 tenths of a cent).

If you have to pay each server that you pick for each transaction, this enables scaled transactions all the way from microtransactions to arbitrarily large purchases.  Paying each server 1/10 of a cent for each transaction would result in a 1 cent server charge for the coffee transaction and a $10.00 charge for the house transaction.

This is similar to the trusted server approach used by Ripple coin, except the size of your server pool would scale up with the transaction (they always use the same number of servers) so it would scale down all the way to small transactions with larger volume, i.e. microtransactions.

Rather than or in addition to adjusting the number of servers, you could also adjust the difficulty of the problem that they have to solve.

[Cheshyr]

Sorry for the delay in updating; real world has kept me too busy to be too involved during the day.  I've appending summaries of the new points... let me know if I misconstrued anything.

Personally I'm enjoying the MC2 thread.  It's got some good points and I think it'll grow into something useful.  I think its vision is a little too narrow, but incremental improvements are sorely needed right now, and TacoTime certainly knows his way around cryptocurrency, so I trust he'll reach a reasonable approximation of his vision.

Lotta good suggestions and complaints.  You could almost design a new coin from this list, and actually have something innovative.

[LaggedOnUser]

Now that we have defined the problems, does anyone have any solutions?

I have a solution to propose for the following:
* Excessive Early Adopter Rewards
* Inflation/Deflation enormous compared to normal currencies
* Erratic income despite regular contribution

My suggestion is to make the rate of coin production constant.  Imagine an altcoin that produces exactly 1 billion new coins a year.  In that case, it doesn't matter when you sign up... you have the same chance to mine a coin at any time (solves "no excessive early adopter rewards").

The "inflation" starts out high, but that's necessary, since coins usually start out with a bang of popularity.  The inflation steadily decreases.  After 1 year, it is 50% inflation.  After 10 years, it is 10% inflation.  After 30 years, it is 3% inflation.  After 100 years, it is 1% inflation (solves "Inflation/Deflation enormous compared to normal currencies").

Plus, as long as you are mining constantly, the inflation alone doesn't take away your purchasing power.  If I mine 500 coins in the first year, nothing (except increased competition) prevents me from mining 500 coins in the 2nd year, 3rd year, etc.  Your share of the total coin base, thus your purchasing power, could remain roughly constant.  This also avoids the "deflation" and "hoarding" problems, since constant purchasing power makes the coin neutral between saving and spending.

Also, unlike national currencies, the inflation is mathematically predictable.  (Of course, nothing guarantees that the currency would be adopted or continue to be used, like the other altcoins, so that is still an uncertainty that you would have).

Finally, if you set the coin frequency high enough, that can give more steady mining income (solves "Erratic income despite regular contribution").  If 100,000 people mine the coin, and there are 1,000,000,000 coins mined per year, then there are 10,000 coins mined per person per year.  That is enough for them to earn 30 coins a day in steady mining income.  In 30 years of mining, they could earn 30 x 10,000 = 300,000 coins, which is about a typical level of lifetime savings in dollars as well.

The steady supply of new coins should also solve the "lost coin" problem.

Hm... anyone feel like starting a new coin? 

[Malawi]

Now that we have defined the problems, does anyone have any solutions?

I have a solution to propose for the following:
* Excessive Early Adopter Rewards
* Inflation/Deflation enormous compared to normal currencies
* Erratic income despite regular contribution

Hm... anyone feel like starting a new coin? 

+8395,4

Been thinking about the same. Also-the blocks should be generated at short intervals, to make transactions faster.

BTC is doomed in the long run, simply because at some point all the coins will have been lost in random events.
Not sure what raste of losses one are to expect from a cryptocurrency, but I think people look better after their alt-currency now than when BTC was really just an encryption experiment.

Only downside I can see, is that the blockchain will grow pretty fast.

[LaggedOnUser]

Now that we have defined the problems, does anyone have any solutions?

I have a solution to propose for the following:
* Excessive Early Adopter Rewards
* Inflation/Deflation enormous compared to normal currencies
* Erratic income despite regular contribution

Hm... anyone feel like starting a new coin? 

+8395,4

Been thinking about the same. Also-the blocks should be generated at short intervals, to make transactions faster.

BTC is doomed in the long run, simply because at some point all the coins will have been lost in random events.
Not sure what raste of losses one are to expect from a cryptocurrency, but I think people look better after their alt-currency now than when BTC was really just an encryption experiment.

Only downside I can see, is that the blockchain will grow pretty fast.

I think a majority of the early coins were probably abandoned or deleted.  It is a funny sort of design that requires an ever-growing amount of space to keep track of an ever-smaller pile of coins.

There is some related discussion to "Coin Complaints" on the Hardfork Wishlist for Bitcoin (https://en.bitcoin.it/wiki/Hardfork_Wishlist).  One of the things that has been repeatedly proposed is to "Flip the chain, instead of committing to new transactions, commit to the summaries of open transactions", which has several forum discussion references there.  If possible, that seems like it would reduce the size of the blockchain.

[Cheshyr]

Interesting.  I'm not certain I understand the entirety of your suggestions.  "make the rate of coin production constant".  Constant compared to what?  Hash rate?  Time?  Existing Coin Supply?

In your example, you say 1 billion coins a year, which sounds like a fixed rate in relation to time.  I've been working with this concept as a foundation as well, but this seems to exasperate the inflation problems.  It makes the inflation constant and predictable, but I'm not certain we can sell a userbase on "Wait 30 years, and the inflation will settle to something about the same as the USD."

I believe your second paragraph about a constant 500 coins a year in a static market is misleading.  While academically correct, you still have to pay to generate these coins at a variable cost to yourself, and I don't believe it's healthy to assume no change in competition or advancement in technology.

The erratic income bit is really a function of the randomness of the hash.  That same randomness is what enables distributed competition, as opposed to the person with the highest hardware investment always getting the answer first.  A larger number of coins, produced faster, would allows us to reach the statistical expected values quickly and more often, so that seems like it should be a goal.  But then we need to address network latency issues and blockchain size...

...which leads into your link to that other thread.  I'll have to check it out.  The blockchain problem really is one of the lynchpins of all of this.

[Malawi]

I also understood the constant production as constant in time. X coins per Y time(average).

With BTC the inflation rate has been 100% 50% and now 12.5%, still people buy and use it based on future decrease in inflation.

With a fixed output, the inflation rate would be the same for the first two years, and then a bit higher for the rest of the duration. Still, it would mean that you can always get new coin into circulation to replace lost coins. Since the future inflation rate is known, that will be taken into considderation just as with BTC.

Since inflation is high at the start, and difficulty will be low due to fewer adopters, there would still be a great incentive to start mining. And what a coin really need in the beginning, is miners who think the coin has a future. As the coin gets a price, the future of it is more certain, and you might get more buyers/miners/users.

If a new block was created about every minute, it should mean that a fully confirmed transaction should take up to 6 minutes instead of 1 hrs with BTC.

About the blockchain itself - maybee there could be some kind of "purging events" where old transaction details are put in a seperate blockchain mainly to be used at places like blockchain.info and the blockchain in use becomes a lot shorter since it only needs to keep the static wallets and more recent transactions.

[LaggedOnUser]

Interesting.  I'm not certain I understand the entirety of your suggestions.  "make the rate of coin production constant".  Constant compared to what?  Hash rate?  Time?  Existing Coin Supply?

In your example, you say 1 billion coins a year, which sounds like a fixed rate in relation to time.  I've been working with this concept as a foundation as well, but this seems to exasperate the inflation problems.  It makes the inflation constant and predictable, but I'm not certain we can sell a userbase on "Wait 30 years, and the inflation will settle to something about the same as the USD."

I believe your second paragraph about a constant 500 coins a year in a static market is misleading.  While academically correct, you still have to pay to generate these coins at a variable cost to yourself, and I don't believe it's healthy to assume no change in competition or advancement in technology.

The erratic income bit is really a function of the randomness of the hash.  That same randomness is what enables distributed competition, as opposed to the person with the highest hardware investment always getting the answer first.  A larger number of coins, produced faster, would allows us to reach the statistical expected values quickly and more often, so that seems like it should be a goal.  But then we need to address network latency issues and blockchain size...

...which leads into your link to that other thread.  I'll have to check it out.  The blockchain problem really is one of the lynchpins of all of this.

Yes, I was suggesting a constant rate of coin production per year, that is, a fixed constant amount of coin production per unit time.  A more realistic assumption is that the currency steadily gains in popularity (rather than a fixed user base), and that it gets steadily easier to mine the coins (due to hardware advancement).  While the first issue means you will be mining fewer coins, I think it will be compensated for by the law of supply and demand, meaning that more users demanding the coins will make them worth more.  The second issue, technological advance, doesn't seem like much of a problem, since you can upgrade your hardware from time to time and keep up with the hardware advancements.

As for the inflation issue, even 10% inflation is manageable.  A lot of currencies in the past have experienced inflation at that level, including even the US dollar during the 1970s.  Some world currencies are experiencing 20-30% inflation even today, such as the Argentine peso.  It could actually help the currency by keeping the money flowing, rather than stagnating and unavailable because the savers don't want to let go of it.  Unlike the peso, it would have the advantage of being predictable.  If usage of the currency were growing at near the rate at which the coin base is growing, which isn't unreasonable, that would seem to mitigate the cost of inflation, because they would bid up the currency in USD, which would compensate for the extra coin creation.  Furthermore, the USD may not be looking so healthy during the next 30 year period, given the way our politicians and central bankers treat it, which would make altcoins seem like a more viable alternative.

A problem that I thought of with coins that favor early adopters is that just gives you incentive to create your own *coin alternative, so you can actually get a chance to be one of the favored few.  That would lead to the endless multiplication of such currencies, since the rewards are so unbalanced.  A currency like I suggest could therefore have more staying power because people can learn about it and sign up any time without being especially disadvantaged.

[scrybe]

Now that we have defined the problems, does anyone have any solutions?

I have a solution to propose for the following:
* Excessive Early Adopter Rewards
* Inflation/Deflation enormous compared to normal currencies
* Erratic income despite regular contribution

My suggestion is to make the rate of coin production constant.  Imagine an altcoin that produces exactly 1 billion new coins a year.  In that case, it doesn't matter when you sign up... you have the same chance to mine a coin at any time (solves "no excessive early adopter rewards").

The "inflation" starts out high, but that's necessary, since coins usually start out with a bang of popularity.  The inflation steadily decreases.  After 1 year, it is 50% inflation.  After 10 years, it is 10% inflation.  After 30 years, it is 3% inflation.  After 100 years, it is 1% inflation (solves "Inflation/Deflation enormous compared to normal currencies").

Plus, as long as you are mining constantly, the inflation alone doesn't take away your purchasing power.  If I mine 500 coins in the first year, nothing (except increased competition) prevents me from mining 500 coins in the 2nd year, 3rd year, etc.  Your share of the total coin base, thus your purchasing power, could remain roughly constant.  This also avoids the "deflation" and "hoarding" problems, since constant purchasing power makes the coin neutral between saving and spending.

Also, unlike national currencies, the inflation is mathematically predictable.  (Of course, nothing guarantees that the currency would be adopted or continue to be used, like the other altcoins, so that is still an uncertainty that you would have).

Finally, if you set the coin frequency high enough, that can give more steady mining income (solves "Erratic income despite regular contribution").  If 100,000 people mine the coin, and there are 1,000,000,000 coins mined per year, then there are 10,000 coins mined per person per year.  That is enough for them to earn 30 coins a day in steady mining income.  In 30 years of mining, they could earn 30 x 10,000 = 300,000 coins, which is about a typical level of lifetime savings in dollars as well.

The steady supply of new coins should also solve the "lost coin" problem.

Hm... anyone feel like starting a new coin? 

BBQCoin or Fairbrix both have constant creation rates, pick one of them or a host of other early clones...

[LaggedOnUser]

BBQCoin or Fairbrix both have constant creation rates, pick one of them or a host of other early clones...

Fairbrix is no longer recommended according to its creator, who says, "As the creator of Fairbrix, I should chime in. First of all, Fairbrix is dead and I have abandoned it since launching Litecoin." (https://bitcointalk.org/index.php?topic=46528.120)  BBQCoin is a fun coin not for serious use, according to its creator, who says, "This new *coin is not made for be traded, sell/buy. It's just for 'fun'. I have spent too much time to share it only with my friends, so i'm saying here his existance." (https://bitcointalk.org/index.php?topic=93437.0).

Although it is evident that the currency market is crowded and hard to break into.  Bitcoin has early adoption and is clearly dominating the rest, regardless of whatever its technical defects may be.

There still is room for experimental currency, particularly if we can solve any of the other issues above.  I think the constant inflation approach is valid.  I also think that combination proof-of-work/proof-of-stake systems like PPCoin have promise of a more efficient, attack-resistant system than Bitcoin.  It would take a proof-of-stake system (or something like it) to implement my previous suggestion of scaling transactions by size for more efficiency.  Without proof-of-stake, someone could sign up for so many nodes on the network that you are randomly picking them to do your transactions, so they can successfully perpetrate an attack.

So perhaps the following features:
* PPCoin-derived
* Scales transactions to higher certainty based on size
* Constant, high coin creation per unit time (1,000,000,000 coins per year)
* Reduced bitchain storage requirements

Scaling transactions based on size also has the potential to reduce network traffic.  It might also be useful to structure the blockchain into subchains somehow so that every transaction doesn't have to be broadcast to everyone.

[LaggedOnUser]

I have another proposal to address some of those issues:
* Confirmation delay
* Little or No Usage outside collection and exchange
* Coin Dust
* Lack of Decentralized Exchange Mechanism
* Difficult for non-technical users to adopt

The solution isn't a modification of the Bitcoin software itself, but rather an additional layer of software to wrap around Bitcoin: the Bitbank.

A bitbank would be a separate open source project from bitcoin to implement a standardized banking system for bitcoins.  The idea is that a trusted 3rd party, your local bank, stores your bitcoins for you.  They also cooperate with local businesses that accept bitcoins, who have an account there.  They keep a separate internal ledger from bitcoins that maintains everyone's balance.  Whenever a local transfer occurs, they instantly make a balance adjustment without actually transferring any bitcoins.

Whenever an accumulation of interbank transfers of sufficient size occurs, they can actually broadcast the transaction to other bitbanks or users, thus settling it.  If there were a trusted network of bitbanks, they could speed this process up as well.  Two bitbanks within the network could transfer funds on a ledger between them throughout the day, and only settle accounts when the difference exceeded a certain threshold ($100,000) or a certain time period (1 day).  Thus purchases could be made throughout the network continuously and only be settled intermittently.

This would greatly reduce the amount of actual bitcoin transactions needed, thus taking pressure off the network.  The bitcoin network would resemble more an "interbank clearing network" than an "all to all transaction network".  Since local transactions, such as buying a cup of coffee, only result in an internal ledger entry in the bank, they can be done instantly.  Furthermore, having less transactions on the network makes it run faster overall (solves "confirmation delay").

Given the existence of such software, it can be promoted to local banks and businesses, so they can start offering bitcoin services with less effort to their regular customers (solves "Little or No Usage outside collection and exchange").

Banks on an existing network trust each other, thus they can collaboratively defragment the bitcoin base.  A bank can also do this internally.  The reduction of actual bitcoin transactions also helps reduce fragmentation and loss (solves "bit dust").

Your local bank or business can convert cash or bank balance to bitcoins on the spot (solves "Lack of Decentralized Exchange Mechanism").

Finally, the bank can offer services at their existing branches and ATMs denominated in bitcoins, with support for the general public ("Difficult for non-technical users to adopt").

This proposal has the additional benefits:
* Works with existing Bitcoin network and does not require new currency creation
* Users don't have to act as their own bank and worry about accidental deletion, loss or theft of bitcoins
* Users can "bank with bitcoin" without even owning a computer
* Participation by existing financial institutions gives bitcoin an air of legitimacy
* Does not prevent existing anonymous, decentralized use of Bitcoin

I notice there is already at least one place calling itself a "bitcoin bank" (http://www.flexcoin.com/) that is trying to solve bitcoin problems by wrapping its software in a banking layer.  I think that is a good idea.

In addition, Bitcoin is already evolving toward a banking model given the high cost of storing a full transaction chain and performing mining operations.  This would help it along in the right direction.

Thoughts, anyone?