Overview of the new initiative

[Oren Gampel]

(Yes, we'll have to name it soon...)

Here is a rough view of the network:

  Software architecture:

Basically, users connect to the ColoredCoins Network with a wallet. 

The ColoredCoins Network is a P2P network that connects all it’s users among themselves and to the Bitcoin Network.

eToro and JPM are displayed here as theoretical examples of two companies that provide market making for issued assets. (I'm not trying to imply ever talking to JPM about it!). All o the participants will be able to provide, issue and Market Make over issued assets. They will compete over conversion rates, and reputation.

Pale blue components are Open Source Components:

The color Lib is the common definition of a Color - how it's defined, issued etc. We're working on both a super naive mockup implementation, as well as using Alex's NGCCC color implementation, so we can have basic implementation in place, and show the "pluginability" of the solution.

The P2P Network is the transport layer for the users to exchange assets, and announce trading intentions.

The Color Trading Layer provides a clean interface that enable developers to implement applications in this space, such as (standard app) Wallet, Web based Wallet, Lightweight Wallet, or Market Making Engines.

In light green you can see the propreitary part of the solution. Here based on eToro as liquidity provider.

The Market Making Engine has a plugin-able MM Logic. This will be included in the open source project with extremly naive implementation, and each market maker would be able to tailor it as he sees fit. We plan to introduce our market maker to provide initial liquidity to the market, and we'll be hedged by aquiring the base assets through eToro. For example, if we'll issue an "eToro Gold Coin" we will buy Gold through eToro when we sell this coin to any user.

Note, Of course, that all wallet implementations use the same Color Trading Layer to make color related actions.

The Bitcoin network is of course an underlying technology that we use, but I don't draw for brevity.

I'll write separately about each component, the color definition, how trading is and our progress in coming days. How KYC issues are handled as well as many other aspects will also be presented separately. 

[danny....@gmail.com]

So, basically the P2P network is used for price and counterparty discovery, but the exchange of value actually occurs on the Bitcoin blockchain?

[Joshua Zeidner]

Hi Danny,

That's the basis of the problem.  Whoever has the most hash power will determine who buys what for what price.  In traditional Bitcoin, people are(usually) not competing on the order in which TX records appear.  The offer may be *discovered* through some p2p network(like Bitmessage) but the COMMITMENT happens in another way.  It's not even close to a true market, and absolutely cannot compete with any 'real' exchange.

One might be tempted to argue that people will have access to trades, but they will all be at sub-optimal prices.  Thus it could be used to create a limited sort of economy, but certainly people would rely on traditional exchange frameworks to get a good price.

Confidence Chains does address these problems.  There are also other theoretical problems with using gossip-messaging for market making.

-jmz

On Sat, Apr 12, 2014 at 12:13 PM, <danny....@gmail.com> wrote:

So, basically the P2P network is used for price and counterparty discovery, but the exchange of value actually occurs on the Bitcoin blockchain?

--
You received this message because you are subscribed to the Google Groups "Colored Coins" group.

To unsubscribe from this group and stop receiving emails from it, send an email to bitcoinX+u...@googlegroups.com.
To post to this group, send email to bitc...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Alex Mizrahi]

So, basically the P2P network is used for price and counterparty discovery, but the exchange of value actually occurs on the Bitcoin blockchain?

Yes. This is the idea behind p2ptrade:

 * any kind of a communication channel can be used for price and counterparty discover (p2p, centralized, whatever)

 * exchange of value actually occurs on the Bitcoin blockchain

It was first implemented in ArmoryX more than a year ago, later in WebcoinX, and in ChromaWallet.

The devil is in the detail, though: ArmoryX p2ptrade could work fine if all participants are well-behaving, but it was pretty much trivial to disrupt it.

As I understand, Oren's team is building a more robust version of it, something which won't have a single point of failure and will be able to deal with various kinds of abuse.

Note that many extensions to a basic trading layer are possible; say, parties might use some kind of decentralized reputation system and will be reluctant to trade with entities which do not have proven reputation.

[Oren Gampel]

I'll add that this rather simple scheme works well, since the P2P communication uses the public and trusted data of the blockchain, anyone can validate whether the exchanger (seller who sells for another asset) of an asset actually owns it. 

In my 2nd post on the system I'll elaborate on our solution (and get into some of the devilish details). I'll also show few different approaches to ensure that the exchanger is committed to his offer (until he cancels it). A problem which I currently plagues several decentralized exchanges I've seen.

jmz, I find your comparisons to what you call "true market" somewhat wanting:

First, it's not clear what you're comparing to it. Is it centralized exchanges? Is it our solution, which we haven't presented yet? 

Second, saying something is "not even close to a true market", doesn't reveal what bothers you. Is it that the commitment is done "differently"? That the book order is decentralized?

Third, does a "true market" is the "best" solution that we should strive for? I can tell you after many years of trading on some of the largest "true market" in the world, that there is nothing fair, orderly, transparent or otherwise wonderful about them. Don't get me wrong - these are great arenas but we can do so much better... I don't think that coming up with a different approach means you have to be bothered by the old one. It means using new technology to rethink the problem completely and maybe coming up with a better solution than the old one.

Soon we'll publish each idea and component of our system, and then anyone could raise specific concern to them. I did it so on purpose, as I believe it would be easier to criticise each idea and part, while an academic whitepaper that tries to cover everything is usually either too high level to allow a reasonable debate, or so low level and technical that it makes it hard to figure out.

[Joshua Zeidner]

Hi Oren,

  Those are interesting but I'm afraid inaccurate characterizations of what I presented prior.

  You haven't addressed the basic problem I presented, which presumably is one your system faces.  How can you ensure a fair and impartial access to offers?

  Failing to do so doesn't prevent one from making trades, but it does seriously impair ones ability to create efficient marketplaces.  To make matters worse, this bias is determined by the owners of the hashing power- introducing a potentially fatal bias to the market.  You MUST be able to offer something that approximates perfect pricing otherwise people will not use it at all(they will use a traditional exchange), UNLESS there are no other options.

-jmz

--

[Flip]

Not coming down on a particular side of this specific protocol issue (because I don't yet grasp it) but I can say with certainty that these market structure issues will be not only important, but genuine make or break issues. Ignore them at your peril (and dismissing each established market protocol by saying we can do better than incumbent exchanges (without specific reasons why the new protocol is better and non-perverting) is indeed to ignore market structure issues).

There will be an implementation that employs the "right" market structure and it will swiftly swamp everyone else.

For timely pop culture backup, read about IEX in Flash Boys (or anywhere). They've *slightly* improved (on the fairness axis) market structure and they're devouring share from a centuries old, hugely entrenched industry. Far slighter optimizations in market structure among this set of solutions will surely separate the big fat winners from the merely interesting also-rans.

[Joshua Zeidner]

Hi Flip,

  You're absolutely right.  If you can get perfect prices(or very close), people won't use it.  I don't see any of these important concerns being addressed and that says a lot of other relevant things.

  "dismissing each established market protocol by saying we can do better than incumbent exchanges (without specific reasons why the new protocol is better and non-perverting) is indeed to ignore market structure issues"  very well put.  these aren't trivial problems.

  Anecdote:  would you rather buy a car from some person off of craiglist for a above Blue Book price or get it at a dealership for a known rate?  The answer is pretty simple, so why bother even creating scenario #1?  At best you'll waste your time, and even more importantly the money of whomever might have invested in your efforts.

  Bitmessage won't help you.  Actually this idea had been presented before with other exchange ideas as a way to turn what was NOT a p2p exchange into a p2p exchange.  The short story is that this doesn't give you a 'distributed exchange'.  Seems that most of the participants in this discussion were not privy to these prior debates.  Gossip-based messaging suggests incomplete price awareness and that is a show stopper for most serious applications.  What use cases remain is anyone's guess, the only one I can think of is creating markets for assets that are either to new or too obscure to put on a 'real exchange' - but the value of a technology that does only that is certainly negligible.

  TL DR; If you dont make the exchange right, you might as well not make it all.  If it's not fair pricing, it's not really an exchange.

best, -jmz

[Yoni Johnathan Assia]

Hi Joshua,

1. We believe the key to progress is shipping code that works. Once there will be version 0.1 of an opensource decentrelized exchange, better versions and forks will follow. Something is always better than nothing.
2. Most people 99.99% cannot access existing exchanges, so your comment of people using existing ones is not complete.
3. The market maker model in the p2p messaging, works better than gossip based messaging - the model assumes also we are creating an open source market maker module - which allows anyone to create spreads and add bids and offers to the market - since each participant can build his own model to add liquidity in the market, efficient market should promote price discovery.

Thanks,
Yoni

[Israel Levin]

Hi JMZ, I can't speak for anyone else, but I was not indeed not privy to the debates you mention. Can you please give me a synopsis? Even better, can you please describe a scenario in which the scheme fails?

We can't fix it unless you report it :)


Many thanks

[Oren Gampel]

I absolutely agree, Flip, these are critical issues. I'm actually very thrilled to see the debate bubbling even before we present our solution. It tells me that there is an anticipating crowd, eager to find a valid solution.

I'll try to post our market scheme today in a separate post, so it can be criticized specifically. I might even focus it solely on this issue, to keep it short and succinct, and allow it to be criticised separately.

[Oren Gampel]

Please note that I'm marking all posts related to this with the Iridis tag. You can see all of them using the Iridis tag link:

https://groups.google.com/forum/#!tags/bitcoinx/Iridis