Okay, in the example of the poor community, the rich community can trade some of the coins back to the poor community for labor. Assuming that the farmers aren't as poor, the poor people have no flour or eggs to sell, but they can sell off some labor.
However, let's assume the third community, who are homeless and permanently unemployable for mental health reasons, and (for the sake of argument) will never ever earn a coin from anyone else. Since this system enforces a 1-to-1 exchange rate, can you name any practical difference whatsoever between Alice trusting one of those people for 1 coin (i.e. they'll be able to trade with someone in a scenario where that someone will get 1 AliceCoin and Alice will get 1 Homelesscoin) and simply gifting 1 AliceCoin as charity? IMHO there is none, except Alice would be able to "keep score" of how much she has given to charity.
This is a key issue here, that with this system trust isn't simply "Bob is who he claims to be", in this system trust means "I'm willing to guarantee that Bob's coins are as good as mine by agreeing to an automatic 1-to-1 trade", in essence you're willing to co-sign any purchase Bob makes since sellers will happily take your coins instead of Bob's; and in any environment where's an innate disbalance (i.e. your coins actually are more valuable than Bob's because more people will trust your coins) this arbitrage will be exploited up to your trust limit, resulting in you losing your coins and gaining Bobs coins instead.
The same applies with community coins; no matter what happens, the coins of some communities will be de facto more valuable than others, so any link of trust between these communities will be arbitraged by that built-in 1-to-1 trade option until that trust runs out and trade between those communities stops. If someone makes a 1-to-x trade reflecting the actual difference of market power, then again people can exploit the built in 1-to-1 trades - so there's a strong disincentive for anyone to make that trust between those communities, since you have to bet (and spend!) your resources in keeping the different currencies linked. The system doesn't provide for "partial trust" of say, 1-to-0.8; you have to be totally sure that Bob's coins are as good as yours, since by making the trust link you're willing to trade all your coins for Bob's coins automatically (with no confirmation or refusal possible) at a 1-to-1 rate. There's a lot of theory and experience for national currencies pegged to each other - and how it can get exploited for financial gain (e.g. Soros had some profitable deals).
I see what you're saying, but you haven't convinced me of this disbalance. What incentive do sellers have to accept AliceCoin over HomelessCoin if everyone adheres to this 1 to 1, even though de facto this is not the case if you consider economic resources?
> Since this system enforces a 1-to-1 exchange rate, can you name any practical difference whatsoever between Alice trusting one of those people for 1 coin (i.e. they'll be able to trade with someone in a scenario where that someone will get 1 AliceCoin and Alice will get 1 Homelesscoin) and simply gifting 1 AliceCoin as charity?
The key difference is that she has lost no value if everyone holds to the 1 to 1 always. This is an important distinction, you're right. However, I don't see any incentive to break from 1 to 1 if everyone is holding to it. If the grocery store accepts HomelessCoin, I have no preference for AliceCoin. If everyone breaks from 1 to 1, then you suddenly have to calculate and consider the value of billions of currencies now. Who wants that? What value do I get out of that change?
This coin is inherently built upon the idea that we are not tying individual coin value to someone's economic value. The seller should just want more coins, and it doesn't matter who they come from.
The "... if everyone holds to the 1 to 1 always. " part is a lot of wishful thinking. As I said, look at the real world examples of currency pegging failures, many of them artificially engineered by large scale market activities.
1) There is an innate difference in practical value, initially even if some coins have just a bit more liquidity i.e. wider range of acceptors as others.
2) If everyone holds to the 1-to-1 agreement forever, then there's no difference between the coins and everyone can freely make the trust deals.
3) If the 1-to-1 does not hold forever, then the people who were betting on the link failing can earn a lot of free coins in this arbitrage, taking them off of people who have made the 1-to-1 trust deals - draining their resources until the assigned trust limits run out.
4) Thus, there's a strong incentive for skilled, resourced people to try to break any 1-to-1 peg between any arbitrary communities, since they can earn a lot of coins if this trust breaks. Especially since it's an automatically enforced 1-to-1 without fees, it's a safe bet - if you fail, then there's no effect, if you succeed, you win a lot. Someone will eventually do so, and the 1-to-1 will break. This system design contains rewards for would-be exploiters of trust, so this design will be exploited.
You have an incentive for 1 to many trades makes sense. What possible incentive does someone have for making that trade with you though when they can go anywhere else and get 1 to 1. This only becomes a problem if everyone ignores 1 to 1 entirely.
If 1 to 1 breaks and it is known, the system could collapse and the coins would now be worthless.
Even if you secretly broke 1 to 1 and managed to find someone willing to take the short end of the stick, you could also be identified and people could simply blacklist all new coins from you and avoid trading with you. There's a lot of risk in failure.
I still don't see the incentive/means for breaking 1 to 1 once implemented. And again, I'm not saying adoption is easy.
In a simplistic manner, the system for breaking 1 to 1 is as follows:
1) The attacker (either an individual or organization with substantial wealth to invest in this endeavor) identify a potential mismatch - i.e. a "subnetwork" that is weakly linked in terms of trust and have an substantial import/export imbalance; i.e. that there is ongoing demand for that subnetwork to buy stuff from elsewhere, and due to that historical imbalance that subnetwork has comparably few outside coins, and the main network has comparably a lot of the subnetwork's coins. Ideally, that subnetwork is one large "coin validator" as described in the original paper, but constitutes a small part (say, 0.1%) of the whole market. Lets call those coins PoorCoin.
2) You borrow PoorCoin (assuming that the currency is in wide practical use, that shouldn't be an issue - none of the cryptocurrencies eliminate the need for credit, and if it becomes popular, credit will be made in it). A lot of it. If collateral is an issue, you provide it in other coins or US dollars. A lot of it - cornering the market is the goal. In a sensible market, actually, that will be very visible and that by itself will be enough to break the 1-to-1 peg; that's what happens with real pegged currencies - pegged currencies require sufficiently huge reserves to clearly demonstrate that breaking the peg takes more resources than anyone can afford, and the peg is lifted as soon as it's obvious that a credible attack is possible. If it happens now, then you buy PoorCoin at the new (lower) exchange rate, pay back your loans, and profit from the change.
3) Otherwise, for every single Alice in the 'rich' part of the network who trusts poorcoin, you "trust Alice", and immediately send through as much funds as you can to yourself, converting the PoorCoin to AliceCoin. This is the key flaw of the system - you can force an arbitrary large number of 1-to-1 trades without the possibility to get refused, unless Alice stops trusting PoorCoin (in which case you win anyway). Then, use AliceCoin to borrow more Poorcoin. You have to control more wealth than all the Alices combined (that's the goal of #1, to identify a market where you can do that), so that you can drain all or at least most of them. This is the key part of the attack - you remove the possibility for 1-to-1 trades by taking all such opportunities and saturating the market.
4) People on the PoorCoin subnetwork want (need!) to buy stuff from the main network, and find out that they can't. The pool of people who were willing to take PoorCoins at 1-to-1 is exhausted or highly depleted because of your actions; some people are willing to do so but not enough to sustain the ongoing PoorCoin trade imbalance. In the main network, people become wary of PoorCoin, since it has become obviously different from the rest of the coins, it's harder/impossible to spend with the usual, automatic 1-to-1 trade. So..... eventually some of those people offer a trade at less than 1-to-1, because they want those imports even if the price has suddenly raised. And that's it; you've accumulated an enormous short position of PoorCoin, you gradually buy it back with normal coins and due to the lower exchange rate in the end you have a lot of normal coins left over.
This profit happens at the expense of all the Alices who trusted PoorCoin (and thus would be rightly fearful and less able to trust any other such subnetworks in the future), and at the future expense of everyone earning PoorCoin, whose income just fell.
Blacklisting all new coins from me is a non-issue - this attack doesn't work on small scale, it fails with amounts that are too low to break the peg and succeeds when the amount is large enough; one person's lifetime UBI doesn't make a difference on a large scale exploit. It takes some effort to make, but when implemented, it brings great rewards, so it'll attract organizations that are able to do it in an organized manner as sneakily as possible (i.e., it's worth for them to create sham organizations/buyers to mask their intentions at some point). In real world currencies, this is futile with hundreds of millions, but feasible with many billions for a smaller currency (IIRC some $30 billion for the Thailand baht back in 1997).
1. A centralized trust/identity system (as I have said is need with this) means that there could be no such thing as PoorCoin as defined here. The idea is that because it was an XCoin, created on this network and tied to an identity endorsed as a single user by Centralized Thing X (this could be registered to SSN for example, though of course that has problems right now. You get the idea.), anyone will accept any trusted coin.
2. You're applying concepts like pegged currencies that don't really seem to apply. This isn't really a currency market - to everyone, it's just Coin. It doesn't matter what name is on it, they all have the same value. A low/high availability of XCoin doesn't drive the price anywhere because it's equivalent to literally any YCoin. There is no market to corner because no one cares. To put it in economic terms, buying up a market and hoarding doesn't matter if there are literally billions of equal value substitutes.
3. It will be impossible to run out of accepted 1 to 1 trades for PoorCoin, even if it were ever to exist.
4. Anyone on PoorCoin, if it were ever to exist, can then still make trades.
However, let's assume the third community, who are homeless and permanently unemployable for mental health reasons, and (for the sake of argument) will never ever earn a coin from anyone else. Since this system enforces a 1-to-1 exchange rate, can you name any practical difference whatsoever between Alice trusting one of those people for 1 coin (i.e. they'll be able to trade with someone in a scenario where that someone will get 1 AliceCoin and Alice will get 1 Homelesscoin) and simply gifting 1 AliceCoin as charity? IMHO there is none, except Alice would be able to "keep score" of how much she has given to charity.
This is a key issue here, that with this system trust isn't simply "Bob is who he claims to be", in this system trust means "I'm willing to guarantee that Bob's coins are as good as mine by agreeing to an automatic 1-to-1 trade", in essence you're willing to co-sign any purchase Bob makes since sellers will happily take your coins instead of Bob's; and in any environment where's an innate disbalance (i.e. your coins actually are more valuable than Bob's because more people will trust your coins) this arbitrage will be exploited up to your trust limit, resulting in you losing your coins and gaining Bobs coins instead.
The same applies with community coins; no matter what happens, the coins of some communities will be de facto more valuable than others, so any link of trust between these communities will be arbitraged by that built-in 1-to-1 trade option until that trust runs out and trade between those communities stops. If someone makes a 1-to-x trade reflecting the actual difference of market power, then again people can exploit the built in 1-to-1 trades - so there's a strong disincentive for anyone to make that trust between those communities, since you have to bet (and spend!) your resources in keeping the different currencies linked. The system doesn't provide for "partial trust" of say, 1-to-0.8; you have to be totally sure that Bob's coins are as good as yours, since by making the trust link you're willing to trade all your coins for Bob's coins automatically (with no confirmation or refusal possible) at a 1-to-1 rate. There's a lot of theory and experience for national currencies pegged to each other - and how it can get exploited for financial gain (e.g. Soros had some profitable deals).