Technical feasibility (doubts about ownership, doubts about computer technology)

[Skybuck]

Hello,

I've always wondered if it was possible to create some kind of (possibly p2p) digital coin system, therefore I think it's great that somebody who seems to deeply know how the financial system works give it some thought. Some of the predictions in the videos have already become true. Ofcourse it could be a self-forfilling-prohecy or it could be the real thing So far it's starting to seem to be the real thing

The digital coin concept which he presents is quite new it seems, at least to me, the video on youtube is from 11 october(10) 2011, I just found it today at 30-october(10)-2011.

I also took a quick look at the documents/specifications/pdf's for Digital Coin. So far the documents seem to be about the general idea but are faint/vague on technical details or implementation details which is probably because it's still in an early stage/just an idea. Perhaps the author/inventor(?) of digital coin idea is not a programmer, to me it seems the inventor(?) assumes that his idea has no technical hurdles, I am very much in doubt about that at this point in time. I do not expect this forum to solve any technical problems any time soon, but I do hope that perhaps some more light can be shine on the technical problems which I will discuss below. Perhaps I will also discuss these problems on other forums/usenet to see if technically-minded people can invent technical solutions to make digital coin a (safe/secure) reality.

But before I ask/consult with technical people it could be wise to first consult with this forum which is specifically ment for digital coin. Perhaps my technical concerns have already been solved, in either idea's or perhaps in implementations in the alternative systems which are also mentioned on this forum.

My concern is with the following vague ideas about digital coin:

1. Uniqueness.
2. Transferring of ownership.
3. Use of computers.

My main concern is with 1 "uniqueness" and 2 "transferring" this goes directly into what a computer basically is: a computer can be thought of as a copieing machine. So the whole idea of "transferring" information from one computer to the next worries me. This is not what happens in practice in practice information is "duplicated/copied" not transferred.

My second general concern is introduction of computers into something which should be pretty simply: "exchanging money from hand to hand". By introducing computers enormous complexity is introduced.

I am willing to put aside my second concern just for the sake of discussing the technical feasibility of the idea and to explore if perhaps some kind of new technology could be created for it. However software/hardware doesn't really matter since it's interexchangable. What matters is if the idea is theoretically secure.

I will add one more word of advice/caution: In all my life as a gamer I have yet to come across a video game which was not hackable/cheatable. The only game which was somewhat secure was stratego, because players are able to keep information secret. My diamond rule for warfare is: "never leak information to opponent" perhaps that diamond rule might somehow apply to conceiving a secure system. Having writing this weird advice I will now go on to my first question:

So my first question is with "transfer of ownership" of "credit coins".

Pretend for a moment that I am a hacker/cracker not interested any more in hacking video games but instead want to hack the system. Do not underestimate the extent to which hackers will go, they will stop at nothing to break the system. My first attack vector would be the ownership issue.

What stops me from "keeping ownership of the coin" and only "pretending to transfer ownership to the other party" ?

For example the hacker's computer makes the other computer believe that ownership was transferred but in reality the hacker still has ownership and can now trade the credit coin again and again and again, creating duplicates everywhere it goes ?!?

In theory perhaps it's impossible to create "uniqueness" without a "global view of all data". How can local systems know if something is "unique" if they have not a "global view of all data" ? This could be a mathematical question or perhaps even a mathematical/logically impossible.

My intention is not to demotivate anybody with persueing a digital coin idea, however I must be critical at current idea's/specification to point out any possible weaknesses, perhaps solutions can be found for those if not already existing.

So do not let my critique demotivate but please do try to explain or find solutions for it, that would be great. Also if it's agreed upon that these are potential weaknesses, then perhaps others can be asked to get involved to see and try if they can solve it somehow.

Bye,
  Skybuck.

[Skybuck]

From some guy usenet, which I will post here too:

"
This problem has been completely solved.
See, for example:

http://falkvinge.net/2011/05/29/why-im-putting-all-my-savings-into-bitcoin/
"

Skybuck/my reply:

Ok link not working, but I did find this link which is kinda interesting:

http://bitcoin.org/bitcoin.pdf

Apperently here bitcoin uses some kind of chain of calculations/transactions/hashes and such.

I am not yet sure if it's per coin, or just a chain of all kinds of transactions, I suspect the last.

I am kind of suprised that people actually take part in this complex specification. It's nice but complex

From reading the document it seems bitcoin does not yet use the concepts of "digital coin".

So perhaps bitcoin would still be vunerable to certain financial problems which are explained by that link I gave earlier !

Perhaps the new ideas of digital coin could be integrated into bitcoin or a new system created based on bitcoin !

Bye,
  Skybuck.

[Paul Grignon]

One has to separate BitCoin the technology from BitCoin the money system.
For their money system, the inventors chose the primitive concept of money as "a single uniform commodity in limited supply".

This is a money concept made necessary in the past because we had no way to transfer value over long distances except by means of physical "money", portable (small) items of high value, such as gold and silver coins.   We overcame that technological limitation long ago.   

In Money as Debt III, I argue that this is the ROOT MISTAKE of our current money system, money the value of which is based on its scarcity.

BitCoin also makes the next fundamental design mistake of money, making it an unredeemable token of exchange subject to speculation.

It is designed for pump-and-dump volatility.

BitCoin might be adaptable to being used as credits for specific goods and/or services from specific suppliers, and thus DEFINED in value by what it buys.  But I suspect not.

As for the technological premise of DIGITAL COIN, I am not a programmer.  The assumption is that it is possible to create a completely secure file as claimed by maidsafe.net  This file is the serial number of a time-limited contract for delivery of advertised goods and/or services at advertised prices from a specified supplier.

Nothing is "transferred".  The file is a single unique serial number that is rewritten to designate a new owner as the only one that can rewrite (spend) it.

[Jordan]

One has to separate BitCoin the technology from BitCoin the money system.
For their money system, the inventors chose the primitive concept of money as "a single uniform commodity in limited supply".

BitCoin might be adaptable to being used as credits for specific goods and/or services from specific suppliers, and thus DEFINED in value by what it buys.  But I suspect not.

I see no reason why Bitcoin cannot be used in this manner in the near future.  It's one of my longer term goals to implement a variant of Bitcoin that can be used to issue a commodity credit, or even a stock of a company. I'm hoping some else will do it before I get around to it though     'cause I might be a while with all my other projects.

I do see the need to create organizations that offer insurance against fraud for such issuances. It's one thing to purchase a few 'loaf of bread' credits from someone you know, it's another thing to invest in a company on the other side of the globe.

I can imagine most individuals will use an investment firm or manager who will utilize independent insurance firms that investigate each issuer.

[Paul Grignon]

I suggest that all who are interested look into http://www.getsglobal.com/.

[Jordan]

RE http://www.getsglobal.com/   I'm unsure about their products/services yet, but their website is stellar!   

[jtimon]

I have to read the digital coin proposal more carefully but from what I've understood about it here are my two first cents.
The "credit coin" is already implemented under the name of ripple, but only in a centralized server.
There's a draft for a decentralized protocol. It would in most cases need registries to timestamp the transactions but that doesn't makes the protocol centralized. Think of them as something like trackers in the eDonkey network.
I proposed to use a blockchain (the technology under bitcoin) instead of the registries, but although that would have some advantages, that would not allow instant payments in most cases (you would have to wait an hour or so to be sure there's no double spending).
The Ripple concept is the more flexible mutual credit system that exists so I think it should be functional for your "credit coin".
It can be used with any unit: USD, EUR, hours, bitcoins, ounces of gold or silver, terras, valuns, perpetual coins...

http://ripple-project.org/

The part I really doubt I understood is the "perpetual coin". To achieve a stable unit my answer is a reference currency defined as a great basket of commodities and services but not being ever issued. Something like Lietaer's Terra but without backing at all: only for contracts and credit.
I called that referenceCoin once:
https://bitcointalk.org/index.php?topic=11614

So, if I've understood digital coin right, it is feasible and through several technical methods.

[segovro]

The Ripple concept is the more flexible mutual credit system that exists so I think it should be functional for your "credit coin".
It can be used with any unit: USD, EUR, hours, bitcoins, ounces of gold or silver, terras, valuns, perpetual coins...

http://ripple-project.org/

Yes, indeed. I have been following a bit the Bitcoin movement, and it seems to be able to play the role (at the Digital Coin theory) of the Perpetual Coin, able to cope with the current mass of money in circulation when (if) it reaches the 21 million (if we use the mBC and microBC). It could play the role of the "digital gold" reserve.

I was missing the other piece, the Debt Coin. Ripple theory resembles much what I read at Digital Coin.

Ripple is currently based on servers. But a P2P version seems to be under development. From what I read is a flat debt coin, without the time and value components explained at Digital Coin. I imagine the developers are currently strugling with encryption. It does not seem to be a substantial investment to improve the result with these parameters.

Bitcoin and Ripple look promising to me. Worth following.

[Jordan]

I am generally encouraged, as more and more people become aware of monetary system problems (thanks to people like Paul Grignon), more and more brainpower and effort is being expended to work out solutions.  The future is bright, but we've a hell of a price to pay for the failures of the existing paradigm.

[segovro]

It looks to me that the Credit Coin proposal is nothing else that a coin that expreses certain family of Ricardian Contracts., but each of them would express a different one, depending on the issuer.

Ricardian Contract
http://iang.org/papers/ricardian_contract.html
http://www.systemics.com/docs/ricardo/issuer/contract.html
http://www.webfunds.org/guide/ricardian.html

The most finished cryptographyc development I have found to sopport the issuing and circulation of such coins on a P2P basis is Open Transactions

https://github.com/FellowTraveler/Open-Transactions/wiki

Video Walkthru: Part 1: http://vimeo.com/28141679 Part 2: http://vimeo.com/28142096

Hello,

I've always wondered if it was possible to create some kind of (possibly p2p) digital coin system, therefore I think it's great that somebody who seems to deeply know how the financial system works give it some thought. Some of the predictions in the videos have already become true. Ofcourse it could be a self-forfilling-prohecy or it could be the real thing So far it's starting to seem to be the real thing

[jtimon]

The most finished cryptographyc development I have found to sopport the issuing and circulation of such coins on a P2P basis is Open Transactions

While Open Transactions is a great project it's not p2p. It relies on servers on which the transactions take place, but the great thing is they cannot run away with the digital vouchers so you don't really have to trust them. The problem is that to implement ripple on them, you need every participant on a transaction to have an account on the same server. You cannot make atomic transactions with the "cash" instrument.
I've made a proposal for a distributed protocol that would enable ripple and more using cash-like tokens within atomic transactions.
https://bitcointalk.org/index.php?topic=60591.0
Maybe it's worse at the "untraceable" part. In fact I have some questions open to fellowtraveler about this:
https://bitcointalk.org/index.php?topic=53329.msg711924#msg711924

Anyway, I should put more work into that draft (a lot of things come from the instructive discussions with Ryan Fugger about the distributed ripple protocol, one could say that this is my preferred design that protocol).
And I undesrtand that fellow traveler doesn't have the time to discuss my proposal, he and mark are doing a great job testing, correcting and improving OT, using it for a bitcoin exchange and bond, shares, etc market (read the previous thread if you're interested).

[jtimon]

The part I really doubt I understood is the "perpetual coin". To achieve a stable unit my answer is a reference currency defined as a great basket of commodities and services but not being ever issued. Something like Lietaer's Terra but without backing at all: only for contracts and credit.
I called that referenceCoin once:
https://bitcointalk.org/index.php?topic=11614

Reading more about perpetual coin. It seems that this idea has been already explored. The RICI index (which I didn't know of) was chosen. I prefer this index unit over a currency backed by a basket of national currencies.
But in any case, people can use Ripple (Credit Coin) with the units they like more, not necessarily PCs.

The point of the specification I don't like is this one:

9. Issuers such as governments, corporations and societies would issue Credit Coins traded on a
global automated public exchange. Parity of any Issuer’s Credit Coin with Perpetual Coin could
only be achieved by that Issuer achieving perfect balance-of-trade. Thus the system enforces
balanced budgets on everyone by automatic re-valuation of credit in real time such that Issuer
credit issued = real demand for the products and/or services of that Issuer at all times.

Ripple doesn't need anything like this. If someone issues IOUs denominated in PC then he owes that quantity of PCs. As long as he makes good on his debts, there's no need for this system. The parity of CC to PC is in the promise (CC) itself.

Well, another thing I don't like is:

3. A “digital object” is a secure unique serial number

This should be out of the analysis, of the specification. What you want to say is that it has the properties of a serial number such as uniqueness, but you don't really care if they're implemented through serial numbers or not. That's part of the design and not the analysis.
The digital objects will be likely implemented with cryptographic certificates, using public key cryptography.

All we need is p2p Ripple and you have your digital coin.
An index for the PC is interesting but the PC is really optional. People can use Ripple with usd, eur, hours (of unskilled labor) or whatever unit they like. It would not surprise me that people prefer hours to RICI units. Just let the people decide the denomination of their credit coins. Someone may prefer denominate his IOUs in the actual good or service he sells (5 massages, 10 loafs of bread, 2 hours of math classes, etc). People should be also free to choose the exchange rates between different credit coins (even if they're denominated in the same unit). We don't need more rules for voluntary exchanges.

As an aside, this article may interest you:

http://www.webisteme.com/blog/?p=1032

[segovro]

While Open Transactions is a great project it's not p2p. It relies on servers on which the transactions take place, but the great thing is they cannot run away with the digital vouchers so you don't really have to trust them. The problem is that to implement ripple on them, you need every participant on a transaction to have an account on the same server. You cannot make atomic transactions with the "cash" instrument.

I am not that familiar with technologies to understand your propositions at the BitCoin forum. You are right, Open Transactions is not p2p. But I wonder if for the Credit Coin we really need it to be p2p.

During its (more or less) yearly cycle from production to sales, the issuer of the Credit Coin has to be a well known company, only the intermediate transactions have to remain anoymous. We can take advantage of this "weakness" to facilitate transactions, and avoiding the multiple node validation that comes along with solutions like BitCoin.

I think there is a great feature of Digital Transaction, which is that all contracts by all issuers (could) follow the same XSD schema for Ricardian contracts.

http://iang.org/papers/ricardian_contract.html

http://www.systemics.com/docs/ricardo/issuer/contract.html

In this way, each issuer could write in the contract whatever he is ready to pay at redemption, but users could have universal WALLETS that would be able to store, manage, pay or get paid with any of these Credit Coins or currencies. These WALLETS consist of sofware able to connect to any Digital Transaction based server, and display the information written at all currencies Ricardian XSD schema.

I visualize it always as an Android wallet, cause all this has to work on mobile or it will not work. A wallet able to deal with many different Credit Coins has to look like the Financisto wallet.

https://market.android.com/details?id=ru.orangesoftware.financisto&feature=search_result#?t=W251bGwsMSwxLDEsInJ1Lm9yYW5nZXNvZnR3YXJlLmZpbmFuY2lzdG8iXQ..

Now, what if every transaction has to be validated at the "Bristol Pound" server, the "Occupy Berlin solidarity network server", or the "Food fair trade server"? I cannot see any problem with this.

The Credit Coin is by definition vulnerable to legal attacks as the issuer is a known entity.

Furthermore, a wallet should allow for a transaction in the absence of Internet connection, waiting for validation. Two persons trusting each other could do this kind of transactions.

In my view, only the Perpetual Coin has to be truly p2p, to become the world, non vulnerable "gold" of the crypto coins.

[jtimon]

You are right, Open Transactions is not p2p. But I wonder if for the Credit Coin we really need it to be p2p.

No it doesn't have to be p2p, but the servers need to interoperate so that I (in server A) can pay you (in server B). Like mail servers, for example.
As far as I know, this is not possible with OT. And when OT implements ripple (credit coin), it won't be possible to make ripple transactions across servers.

During its (more or less) yearly cycle from production to sales, the issuer of the Credit Coin has to be a well known company, only the intermediate transactions have to remain anoymous.

Wait, wait. Every user issues its own credit coins that are only accepted for the people who trust him. Those credit coins are traded transitively for other CC until you reach CCs that the seller accepts.

I think there is a great feature of Digital Transaction, which is that all contracts by all issuers (could) follow the same XSD schema for Ricardian contracts.

In my proposal only digital tokens are moved and traded. They cannot be counterfeited and are atomically traded, but the protocol doesn't say anything about what those tokens represent.
It is possible to bind those tokens to legal contracts, but the protocol doesn't need to take that into account. 

I visualize it always as an Android wallet, cause all this has to work on mobile or it will not work. A wallet able to deal with many different Credit Coins has to look like the Financisto wallet.

Yes, this needs to be scalable to smartphones and need to manage different credit coins.

The Credit Coin is by definition vulnerable to legal attacks as the issuer is a known entity.

The tokens can be linked to a person or legal entity, but it's not mandatory. Of course, no one will accept CCs issued by an unknown entity, but you can tell your friends "these are my CCs" without making that information public or official certificates.

Furthermore, a wallet should allow for a transaction in the absence of Internet connection, waiting for validation. Two persons trusting each other could do this kind of transactions.

That would be possible only when the two parties trust each other because:
1) One have to accept the CCs of the other directly.
2) They don't have a timestamper to ensure atomicity. 

In my view, only the Perpetual Coin has to be truly p2p, to become the world, non vulnerable "gold" of the crypto coins.

I prefer the perpetual coin that doesn't exists in the real world. A unit of value that depends on an index or something similar.
Anyway, I don't think that a uniform unit is feasible or even desirable. Different units (linked to different index) should be able to compete so that people can chose what is the more stable and desirable one.
So I'm not in love with the Perpetual coin part of digital coin.