please delete

[markm]

People exchange actual commodities for fiat regularly/constantly, it happens all the time.

So obviously the problem of how to deal with the reversibility of fiat has already been solved.

But, the solutions might not be very scale-able down to individuals who only occassionally sell something in return for fiat.

The suggestion users of Kijiji are given is only accept cash in person.

Basically being a seller in today's world is not very scale-able. If you aren't going to do enough trades to package your statistical losses from chargebacks into your price without thereby driving your price up above what people are willing to pay, you are likely to lose customers to a larger / higher volume merchant. So maybe commerce itself is not really very susceptible to decentralisation nowadays? Maybe if it ever was it was only because the technology had not arrived to allow really large operations?

On the other hand, there have been worldwide networks of money-changers and money-transmitters operating without regulation (other then by themselves / their network and their customers and how many troops / thugs their customers can raise up against them) for centuries. How did they do it? Maybe friend to friend, basically? Trust networks?

-MarkM-

[Rassah]

People exchange actual commodities for fiat regularly/constantly, it happens all the time.

So obviously the problem of how to deal with the reversibility of fiat has already been solved.

Yes. Centralized trusted exchanges, similar to MtGox, Tradehill, or Bitcoin7, all susceptible to laws, regulations, errors, seizures, shutdowns, and being central targets for attack. The unsolved problem is p2p exchange. I fear that problem may not be solvable.

On the other hand, there have been worldwide networks of money-changes and money-transmitters operating without regulation (other then by themselves / their network and their customers and how many troops / thugs their customers can raise up against them) for centuries. How did they do it? Maybe friend to friend, basically? Trust networks?

Friend to friend and trust (with a sword as an incentive). I do believe that system (like Ripple) will work, but my problem (and I'm sure it's one I share with others) is that I don't have any friends who use Ripple, and maybe one who uses Bitcoin. There's no Ripple network for me to hook into
Maybe eventually the network will be that of stores you can walk into who can exchange your fiat for BTC for you in person, with a proof of I'd and other things.

[markm]

You do know about the #bitcoin-otc channel on Freenode IRC network, don't you?

Right now people even mention exchange of bitcoins for fiat out in public there.

Maybe someday we'll need to know some secure, encrypted way to contact the people we learn to trust there, and knowing their GPG codes from back in these good old days we'll be able to know it is really them when we contact them through some crypted channel.

Thanks to #bitcoin-otc I do know a few people who use bitcoin and are willing to trade them.

I don't know anyone locally in meatspace but so far #bitcoin-otc has sufficed for my needs.

-MarkM-

[jtimon]

The unsolved problem is p2p exchange. I fear that problem may not be solvable.

You can pay for the bitcoins with ripple and then settle the debt with your friend using cash.

[markm]

Except that ripple doesn't seem to be working or something, at least not working well enough to bother actually implementing some kind of p2p protocol for it. Maybe a lot more people have to actually use it before anyone will consider it worthwhile to do some work on a p2p implementation?

-MarkM-

[Rassah]

I've used  #bitcoin-otc to get a few grand last summer, with escrow service run by Gavin. Good point about that being a p2p exchange.

[jtimon]

Except that ripple doesn't seem to be working or something, at least not working well enough to bother actually implementing some kind of p2p protocol for it. Maybe a lot more people have to actually use it before anyone will consider it worthwhile to do some work on a p2p implementation?

Maybe it doesn't work for you because you don't have connections. It's like saying that facebook doesn't work while you don't have any friends. I can buy bitcoins for villages.cc hours.
I think Ryan is doing some work with the p2p protocol, a design is done at least. I'm impatient for helping him out, but I cannot right now. Hopefully on summer. It's probably too much work for just one person.

[Mike3574]

Just giving a little update here. For the moment I'm only working on crypto-crypto trades because it is not possible to get fiat into a p2p environment at the moment. I've made some updates to the GUI. You can see picture of it here: http://litebit.co/project-item/p2px/ I'll upload some pics here later.

Also, in order to encourage developers to volunteer their time, I am releasing the source to my GUI. It can be downloaded here: http://pro2host.com/cryptocoin/p2px/

DISCLAIMER: This is a non-functional GUI! Only design and layout!

I've also been brainstorming a lot of ideas on this project. I'm a bit strapped for time now but I'll be writing about them later. I really want to get this project off the ground. It's been sitting for too long. I have zero developers!

[cryptoanarchist]

I've been thinking about this and trying to conceptualize how this could be done, and there are a lot of ideas on this thread to go through.

First of all, if you want to actively trade with this client, and not just make single exchanges, there HAS to be some kind of substitute fiat anchored cryptocurrency, and barring some genius solution, there has to be a web of trust system to make it work.

The fiat-anchored cryptocurrency would be for the p2p client wallet, so that you could make fast trades without waiting on government institutions. Each 'crypto-dollar' would be backed by a user on the network and the blockchain would keep track of reputation/feedback for each user. The network would also track each cryptodollar as it was traded around so that anyone can see who is backing each trade in the order book and what their rating is.

Transactions in fiat world would only happen when a user wanted to withdraw from his wallet. The network would know who was backing the crypto-dollars and initiate the exchange and record feedback. I'm sure people would get scammed at first, but as time goes on people will build reputations for having sound currency, just like certain banks in the old west.

I think it should work in physical exchange too while you're at it. Give users the option of setting a physical exchange point like bitcoinmap and what they exchange for. It'd be cool if you can just click and see where you can pick up physical silver coins for bitcoin and vice versa.

[Mike3574]

Thank you once again to everyone that has posted their interesting ideas. However, I've canceled all work on this project a while ago. That's why I posted my GUI source. So if anyone wants to work on it then go right ahead! I'm only one man who has absolutely no programming experience and can barely manage to create the non-functional GUI. There was one coder involved at one time but he bailed out before he had even done any substantial work. Realistically, due to the scope of this project, a fairly large team would be needed to develop the many features and above all, ensure that everything is secure. I haven't even been able to develop a secure method in theory. It may not even be possible. Then again, nothing is impossible really. But this is just something that is beyond my own capabilities of bringing to life.

Please continue brainstorming! Maybe one idea that somebody puts forth here will actually be workable. Brainstorming never does any harm. Good luck to whoever decides to give a go at this project!

All Best,

Michael

[spndr7]

Order opening,closing and modification
A block-chain is maintained which keeps all valid open orders.A node requesting to open an order sends order quote to the network.The network verifies the order by periodically querying the corresponding blockchain and check that order is sent by the node is less than its
balance or not.In this way fake orders could be weeded out.A node can modify or cancel the order by querying the network with message signed with its private keys.

Orders open could be accessed by all nodes.Any node interested could send request for accepting an order.After similar verification about buyer's balance the order status is changed to accepted.

Money Allocation to intermediate nodes
Once order is accepted,the network divides the money to the other nodes such that,the money send to any node is less than the trade amount,that node wants to execute.This trade assignment is sent to both trading parties.They are required to send money to assigned intermediate nodes.These transactions are verified by the network.

Intermediate transaction
After all intermediate transactions are over,the network sends the address of buyer and seller to the nodes which received money from seller and buyer receptively. Until they successfully execute their routing assignment,their trade money is not released from their intermediates.

Assumptions
1. To start the trading, first two nodes should be assumed that they conducted their routing operation.
 
Advantages
1.Buyer and seller don't know each other until trade is successfully conducted.
2.Using a block-chain,all trading activities can be recorded securely so that trade is not tampered until its over. 

[jtimon]

I think the solution to trade across chain should part from the contract defined in the wiki which enables an atomic trade.

[fellowtraveler]

Yes. Centralized trusted exchanges, similar to MtGox, Tradehill, or Bitcoin7, all susceptible to laws, regulations, errors, seizures, shutdowns, and being central targets for attack. The unsolved problem is p2p exchange. I fear that problem may not be solvable.

OT is not a p2p exchange, but I believe that it solves this problem (or that it will, once all the pieces are in place.)

The first piece is a transaction server which cannot forge transactions and which cannot escape an audit.
OT fits this in theory, but the auditing protocol has not yet been coded.

The second piece is the ability to store Bitcoins safely on an OT server, in such a way that the OT server itself cannot steal them.
The solution here is to use multisig to store BTC in voting pools on the blockchain. See my other posts on this forum where I discuss this.
Now that multisig is actually available, this functionality could be added to OT.

The third piece is separation of powers. Meaning, while issuers who hold and wire dollars are susceptible to laws, seizures, shutdowns, etc, they are separate entities from the OT server(s) themselves, which could operate anonymously on Tor, at a profit. (This is made possible by Bitcoin. We see Silk Road doing this already -- operating anonymously yet at a profit -- again made possible by Bitcoin.)

You were right to fear this problem may not be solvable -- I have put a lot of thought into it myself. However, I believe the above pieces do solve this problem effectively. (Though p2p technologies are necessary in several ways for the above scenario to work, it's still not a "p2p exchange" but you still get all the benefits of one, i.e. it still solves the problem.)

[ptshamrock]

Yes. Centralized trusted exchanges, similar to MtGox, Tradehill, or Bitcoin7, all susceptible to laws, regulations, errors, seizures, shutdowns, and being central targets for attack. The unsolved problem is p2p exchange. I fear that problem may not be solvable.

OT is not a p2p exchange, but I believe that it solves this problem (or that it will, once all the pieces are in place.)

The first piece is a transaction server which cannot forge transactions and which cannot escape an audit.
OT fits this in theory, but the auditing protocol has not yet been coded.

The second piece is the ability to store Bitcoins safely on an OT server, in such a way that the OT server itself cannot steal them.
The solution here is to use multisig to store BTC in voting pools on the blockchain. See my other posts on this forum where I discuss this.
Now that multisig is actually available, this functionality could be added to OT.

The third piece is separation of powers. Meaning, while issuers who hold and wire dollars are susceptible to laws, seizures, shutdowns, etc, they are separate entities from the OT server(s) themselves, which could operate anonymously on Tor, at a profit. (This is made possible by Bitcoin. We see Silk Road doing this already -- operating anonymously yet at a profit -- again made possible by Bitcoin.)

You were right to fear this problem may not be solvable -- I have put a lot of thought into it myself. However, I believe the above pieces do solve this problem effectively. (Though p2p technologies are necessary in several ways for the above scenario to work, it's still not a "p2p exchange" but you still get all the benefits of one, i.e. it still solves the problem.)

Everytime i read such statements i am in awe at the possibilities and thankful to the devs making this possible!

[jtimon]

Yes. Centralized trusted exchanges, similar to MtGox, Tradehill, or Bitcoin7, all susceptible to laws, regulations, errors, seizures, shutdowns, and being central targets for attack. The unsolved problem is p2p exchange. I fear that problem may not be solvable.

You were right to fear this problem may not be solvable -- I have put a lot of thought into it myself. However, I believe the above pieces do solve this problem effectively. (Though p2p technologies are necessary in several ways for the above scenario to work, it's still not a "p2p exchange" but you still get all the benefits of one, i.e. it still solves the problem.)

The p2p exchange between chains will be possible when the contract I just mentioned is possible in bitcoin (enable lock time and tx substitution).

The p2p exchange with chaincurrencies/nacional currencies is also possible. But the minter (the entity that stores the actual national currency) will sell the cryptotokens (say mtgoxUSD) in exchange for deposits and will always need to be trusted. This is true for OT as well.
In some other thread I can't find I said that you can achieve that with the ripple protocol (atomic trades of bitcoins for crypto-IOUs).
Ripplecoin could also do that.
The mtgoxUSD could be issued as colored satoshis too.
There's some people trying to implement an atomic trade of "smart bonds" for bitcoins right now.
Here are some related threads:

https://bitcointalk.org/index.php?topic=92421.0
https://bitcointalk.org/index.php?topic=106449.0
https://bitcointalk.org/index.php?topic=112007.0

If you have a p2p bond market, you have a p2p exchange. What you will never have is p2p USD or EUR.
Ripple of course helps with that.
If you don't trust mtgoxUSD you can sell btc for mtgoxUSD and them for intersangoUSD (or whatever). And you can make both trades atomically!! So you're holding mtgoxUSD for 0 seconds.
And this can be extended to n trades.

So I disagree, the p2p exchange is completely possible and may be a reality soon.

[cunicula]

You could have a p2p USD or EUR derivatives. Suppose someone colors a satoshi and assigns a one dollar face value to it, promising to exchange this Satoshi for one USD of bitcoins in the future at the current bitcoin/USD market price. Suppose colored bitcoins and regular bitcoins can be traded via a p2p exchange.

Why would a company do this? Well, they can invest a percentage of the bitcoin they borrow by coloring satoshis (fractional reserve).

Why would anyone buy the colored Satoshi (i.e. lend to the company)? Well it might be convenient to acquire USD units of value. In addition to this, the company might pay some interest on these colored satoshis in exchange for the ability to raise debt.

If you have multiple companies doing this, then they could discount each others' colored satoshis which would increase liquidity. It would be like free banking and thus unstable. It would be nice to have a more stable solution as well (for example an overfunded reserve with company profit coming from txn fees rather than interest), but this seems more complicated to implement at this point.

I think of these as possible long-term goals for this project. A BTC/LTC exchange would be a good starting point.

[markm]

For a while the Brits and Canucks tried to peg their United Kingdom Britcoins (UKB) and Canadian Digital Notes (CDN) respectively to GBP and CAD, respectively. It maybe would have worked if they had not copied the bitcoin model for their blockchains, specifically the limiting of the number of coins to only 21 million coins. With the total number of coins limited it turned out that trying to prevent the value rising well above that of GBP and CAD was not feasible; in the end they gave up trying to push them down in value and just learned to live with the constant increase in value of their coins.

-MarkM-

[jtimon]

You could have a p2p USD or EUR derivatives. Suppose someone colors a satoshi and assigns a one dollar face value to it, promising to exchange this Satoshi for one USD of bitcoins in the future at the current bitcoin/USD market price. Suppose colored bitcoins and regular bitcoins can be traded via a p2p exchange.

Yes, but you still need trust. Existing exchange could issue USD deposits certificates by coloring satoshis and users will be happy to hold the proof themselves while still having to trust the company, not for the exchange anymore, which is conducted in the blockchain with colored coins, but for the deposits. Users should still be able to withdraw the paper by sending their usdCoins back to the issuer.

Why would a company do this?

Bonds, IOUs, future contracts, currencies, vouchers, tickets, ETFs, discounts, smart property, votes, certificate of deposit, academic certificates, property registry, beer tokens and poker tabs with friends.
The question is, why would a company want to use another technology for anti-counterfeiting?

You may need more privacy than what can the chain can offer, which I think will be enough for most if not for all with hardcore obfuscation. Maybe it's about the confirmation time.
The ripple protocol solves this same problem in a different way that is stronger in these two points, but there's many cases in which the public accounting is actually a feature. You don't need the issuer online to move the asset, the chain signs the movement for him. For the smart car property example, you want to be able to sell it even if the company that sold it to you goes broke.
I hope that there can be more coin colors than people (everybody issues, but only your neighbors accept yours, so you ripple your satoshi to pay). I don't think all of us will have our own server always online. So we will need to trust a ripple server for the assets we issue or just use the chain.

I think of these as possible long-term goals for this project. A BTC/LTC exchange would be a good starting point.

A BTC/LTC is feasible but requires changes in the current protocol (enable nLockTime and transaction replacement), but colored coins don't.
Maybe we have the p2p fiat/coin exchange than the p2p coin/coin after all.
I remember when I writed about exchangeCoin, which was a protocol which listened to other protocols to enable atomic exchanges with it. Other chains could do the same to be tradeable for others apart than from exchangeCoins. Now that I think about it...what a stupid idea I had then.

[cunicula]

The question is, why would a company want to use another technology for anti-counterfeiting?

There are several possible answers:

1) Anti counterfeiting is expensive. Colored Bitcoins can be a very cheap and reliable solution. Consider liberty reserve. Presumably they have overhead expenditures.
Turn liberty reserve dollars to colored bitcoin and you would greatly reduce overhead. Only redemptions with the company directly have to be handled by the company. The company would not even have to agree to process redemptions below a certain minimum value. There is very little that the company would actually have to do on a day to day basis. Transfers between users would have very small txn fees, i.e. whatever you need to pay to transfer some satoshi.

2) Colored bitcoins would be like bearer bonds, except that they can be exchanged digitially. Nothing like this really exists right now. Bearer bonds aren't even really legal anymore. The company could access demand for its notes by tapping this niche market.

3) A colored satoshi could be more easily traded than something on GBLSE. With GBLSE there are two counterparties. With colored bitcoins and P2P exchanges, you have the same functionality as GBLSE, but only one counterparty. No need to worry about attacks (hackers, gov't, etc.) on GBLSE anymore.

You also mention that more privacy would be good. That's true. Making bitcoin anonymous would help here. However, I doubt we could get to a state where a public investment in something like Silk Road would be a safe endeavor.

[cunicula]

For a while the Brits and Canucks tried to peg their United Kingdom Britcoins (UKB) and Canadian Digital Notes (CDN) respectively to GBP and CAD, respectively. It maybe would have worked if they had not copied the bitcoin model for their blockchains, specifically the limiting of the number of coins to only 21 million coins. With the total number of coins limited it turned out that trying to prevent the value rising well above that of GBP and CAD was not feasible; in the end they gave up trying to push them down in value and just learned to live with the constant increase in value of their coins.

-MarkM-

Not sure if serious. The company could very easily and happily release more satoshis if its debt appreciated.  Keeping the window open to redeem debt at face value would prevent any depreciation, except in the event of bankruptcy (closing the window). Again, not sure if serious. Yes, you need to trust the company.