[ANNOUNCE] Zero Reserve - A distributed Bitcoin Exchange

[herzmeister]

Maybe I overread the part where you explain how you will manage to implement the whole idea without a gateway to the traditional banking system, as you promise in OP?
I think your idea is great and deserves attention but how exactly are you trying to skirt the need for any bank accounts tied to the system?

Because the bank is YOU. Or ME. If we're friends, you'll see in the system that the balance between you and me is currently maybe EUR +2000,-. And tomorrow maybe EUR -600,-. Depending on all the action between you and me or our friends and friends' friends.

We may choose to settle the balance between us eventually from time to time with some fiat cash, or not. Either way, we'd all continue to use our regular bank accounts as usual, without our respective banks suspecting anything.

[anu]

Maybe I overread the part where you explain how you will manage to implement the whole idea without a gateway to the traditional banking system, as you promise in OP?
I think your idea is great and deserves attention but how exactly are you trying to skirt the need for any bank accounts tied to the system?

Because the bank is YOU. Or ME.

We need to understand that the banking industry is very much unlike, say, the semiconductor or the aerospace industry in that we all can trivally create their products. There is nothing banks can do we can't also do. What we all (the Bitcoin community) are doing boils down to making the entire banking system redundant.*

Ripple** and hence Zero Reserve are ancient ideas: The Hawala Money Transfer System is around since well over 1000 years. Ripple / Hawala works - it it the oldest money transfer system currently in existence. The difference to Zero Reserve: Modern networked computers make it possible that everybody can be a Hawaladar - the computers are taking care of the complexity to find a line of trust through the social graph. That was the challenge all along, not the fact that we can grant each other credit***.

As Herzmeister said, we will be using the banking system for a while, because many accounts are bound to get unbalanced, for the simple reason that all our income all goes to the bank. The difference to a centralized Bitcoin exchange is that when settling that, it is a random transfer between private accounts, from the bank's POV. They can't easily block that as they can with MtGox.



* Making the world a better place as everybody gets the dividends. It will be also more difficult to crowd fund wars than to (central)bank fund them.

** Using the word Ripple to make explaining easier. But ZR is not associated in any way with Ripple Labs or any other company.

*** In fact that is also a challenge - we have been conditioned since 100 years to keep friendship and money apart: Friends are people you party with.

[joesmoe2012]

Here's my key to retroshare, if anybody wants to join in, PM me your key and i'll add you as well.

-----BEGIN PGP PUBLIC KEY BLOCK-----
Version: OpenPGP:SDK v0.9
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=Oq0s
-----END PGP PUBLIC KEY BLOCK-----
--SSLID--25a6aee526312320868964d10bff257d;--LOCATION--Vienna;
--LOCAL--169.254.219.79:7540;--EXT--70.194.1.14:7540;

[O_Shovah]

Adding my key for testing purposes:


-----BEGIN PGP PUBLIC KEY BLOCK-----
Version: OpenPGP:SDK v0.9
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=mLAe
-----END PGP PUBLIC KEY BLOCK-----

[phelix]

A potential possibility for paying nodes:

https://bitcointalk.org/index.php?topic=62558 - Sustainable nanopayment idea: Probabilistic Payments

[phelix]

Just yesterday I realized a big difference between ZeroReserve and Ripple.com: ZeroReserve's orderbook and normal transactions can be anonymous. You just pay to a hash you can transport otr.

Also the current state of development is impressive already. It seems there is only little missing towards a working implementation.

[anu]

Just yesterday I realized a big difference between ZeroReserve and Ripple.com: ZeroReserve's orderbook and normal transactions can be anonymous. You just pay to a hash you can transport otr.

Also the current state of development is impressive already. It seems there is only little missing towards a working implementation.

Indeed, a remote payment is a payment to a hash, not something that makes any sense to anyone except the payer and the payee, like a RetroShare ID. Just like Bitcoin, it is not perfectly anonymous - every hop can tell the previous and next hop. Since the number of hops is small (less or equal to 6, most likely), the chance is significant that the previous hop is the payer and the next is the payee. But AFAIK there is no way to be sure of that and hops don't know what the payment is for.

So indeed, you can pay for furry porn without letting your friends know of your preferences

BTW: I used the word "Ripple", but I think this may give too much credit. Ripple is not doing anything substantial that is not practiced by the banking system, say, in the SWIFT network. In particular, it is pretty much the same as Hawala. And Hawala goes back more than 1000 years. If we can believe certain sources, it goes back 5000 years and was practised (albeit by another name) for payments between the ancient cities of Ur, Uruk, Askalon, Ninive, and Babylon. I'll try to research that a bit deeper and possibly remove all references to "Ripple". Any pointers to that is highly appreciated.

[phelix]

In case you are still opting to start out with bitcoind:
http://dot-bit.org/forum/viewtopic.php?p=5196#p5196   [ANN] Namecoin RPC interface in C++

[phelix]

In case you are still opting to start out with bitcoind:
http://dot-bit.org/forum/viewtopic.php?p=5196#p5196   [ANN] Namecoin RPC interface in C++

also: https://github.com/libcoin/libcoin/wiki

Should compile with MinGW 4.6

[bluemeanie1]

A tech paper is here:
https://mega.co.nz/#!vZ80yQJS!ccrCBREYZrOPr8oK7C9StVGuDmYENNYwrFiPXZVQldM

And the code is here:
https://github.com/zeroreserve/ZeroReserve

sorry, this looks pretty weak.

lot of attempts these days by the community at large to use 'Distributed Exchange' to attract attention to themselves.

I am working on a distributed exchange and it has a rich background of theory:

https://docs.google.com/document/d/1cKlN55wX7n0SLvxidLoFVrJnNMJO-Iefr8bVyeHBseg/edit

the link is not my best writing but it explains how to do it with Confidence Chains.

the problem is far more complex than just posting an order book to some shared file somewhere.

you already have something like this with Bitcoin-OTC, http://bitcoin-otc.com/

[anu]

sorry, this looks pretty weak.....

.....it has a rich background of theory:

Sorry, mine doesn't boast any theory. Reason is that I didn't actually invent anything - I just combine known inventions. And the main idea is, as I said several times already, more than 1000 years old - possibly as old as 5000 years. I find it preposterous to put theory into a paper that is not the theory you invented. Look up the Hawala/Ripple theory on the original papers, or rather, clay tablets. The more I get into the topic, the more I find it preposterous that Ryan Fugger wrote a paper claiming stuff.

I believe I referenced all the relevant inventions I am using: E3PC, Turtle Routing along Hawala trust chains, signatures.

the problem is far more complex than just posting an order book to some shared file somewhere.

I agree. That is why I put all the stuff mentioned above together to make this work. You may have noticed it if you had actually read the paper or played with the code.

No idea what you mean by "shared file". Are you trolling?

you already have something like this with Bitcoin-OTC, http://bitcoin-otc.com/

Bitcoin-OTC has about as much in common with ZR as Liberty Reserve had with Bitcoin. They do something vaguely similar, but that is it. Again, are you trolling?

[phelix]

Vote for ZeroReserve as your favorite Bitcoin Project of Autumn 2013

[ning]

I have a question regarding the implementation of the matching engine (exchange) used in Zero Reserve. There isn't a reliable global clock in a distributed system. How can the distributed nodes collectively determine which order comes first and which order comes sometime later?

[anu]

I have a question regarding the implementation of the matching engine (exchange) used in Zero Reserve. There isn't a reliable global clock in a distributed system. How can the distributed nodes collectively determine which order comes first and which order comes sometime later?

Nodes don't do anything collectively - the philosophy of ZR is to be strictly local (no reliable information beyond friends). Each node only matches own orders with other orders. Nodes may only have a subset of all orders floating around in the network and that subset may differ from node to node.

Orders are matched as they come in - first come, first serve. After the match, the rest of the order (if any) is placed into the order book. In practise, many nodes won't try to match, because they have no own-orders in the book.

So strictly speaking, ZR is not an exchange, its OTC with the look and feel and the anonymity of an exchange.

Collisions are handled by the seller node - an incoming buy request is denied if it can't be served because another buyer won the race.

[anu]

Nodes may only have a subset of all orders floating around in the network and that subset may differ from node to node.

That sounds bad, but its not so bad: An order simply needs a few seconds to propagate through the entire network. During that time, nodes closer to the originating node get to see the order sooner.

First come - first serve may lead to sub-optimal deals for the originating node. We'll have to see how bad that is in practise. If its bad, it's possible to implement some counter measures. I expect it to be largely a non-issue.

[ning]

Thank you for the clarification. It's just that the non-deterministic behavior of an exchange/platform always bugs me:)

Aside from lacking a global clock (and thus a global orderbook), I still have some concerns about this project (all are very subjective):

1) The misleading concept of a friend

A friend I fight a war with is very much different from a friend who happen to attend the same class as I do. Take Facebook friends for example, one can easily make a dozen of new friends on a social network every day. If all my Facebook friends want me to trust them and buy things from me using the credit they issued, shall I comply? A "Yes" or a "No" is too simple to answer this question, but at least we can assume that it is hard.

Another issue is that people regret and people move a lot. If someone buys a bike from a friend, it's highly unlikely that he/she would regret afterwards. But if someone buys thousands of stocks from a friend, the market fluctuations may cause this someone to want the transaction to be reversed, regardless of whether it is right. This may cause conflicts and even end their friendship. Also, developing a friendship that one can trust their money with can take a long time. My personal observation is that as we grow older, even having a few true best friends is a blessing. And in a modern society, we get to meet a lot more people because the mobility of the population has dramatically increased. Establishing and maintaining a long lasting friendship, is (can be) hard.

All in all, people might not enjoy trading frequently with their friends.

2) An established Hawala system doesn't necessarily require modern technologies

I found some video introductions about Hawala system on web. And I got to know that this kind of system has a very long history. And there are currently many Hawala systems functioning in a global scale without a high-tech infrastructure -- cell phones and emails are adequate given that there is someone somewhere that we all trust. I have an assumption (possibly wrong) about these established Hawala systems, take a Hawala system that enables people to move large amount of money in and out of a country for example, my assumption is that there must be someone who has predominant control over both divisions in and out of the country. Thus, a trustworthy Hawala system is, in a sense, centralized. So, my question is, will people embrace a decentralized one?

(Please note that I am not trying to be harsh. A friend of mine pointed me this thread and I need to get a feel of it. My way of trying to understand something is to throw as many questions as possible at it. This project is more revolutionary than all the projects I have thought about developing by now. Keep up the good work!)

[anu]

Thank you for the clarification. It's just that the non-deterministic behavior of an exchange/platform always bugs me:)

Me too

Aside from lacking a global clock (and thus a global orderbook), I still have some concerns about this project (all are very subjective):

1) The misleading concept of a friend

In ZR, a friend is defined as someone who you grant a credit. So it might be someone who you don't like, but trust.

All in all, people might not enjoy trading frequently with their friends.

You don't need to. Friends only route payment. Should it so happen that the counterparty is your friend you ( and he ) would not know.

2) An established Hawala system doesn't necessarily require modern technologies

Indeed not. But AFAIK Hawaladar do not usually route over more than one hop and Hawaladar are professionals. ZR takes care of the complexities of routing over far more hops and it doesn't require you to work to participate in the network - ZR takes care of that.

Keep up the good work!)

 

[hieroglyph]

This sounds like something I'd like to try in the future so I'm will watch this thread.  Good luck with the project I hope you will have success.

[kdrop22]

We need the distributed exchanges now, more than ever.

[phelix]

We need the distributed exchanges now, more than ever.

this. is there a donation address btw?