Counterparty and Coloured Coins

[No_2]

Could someone explain to me how the system of counterparty and coloured coins are different? From my limited understanding of both they seem pretty similar.

[justusranvier]

In order to use Counterparty, you've got to buy special tokens.

Anyone who has bitcoins can create and use colored coins without that restriction.

Other than that, they are identical.

[gendal]

Could someone explain to me how the system of counterparty and coloured coins are different? From my limited understanding of both they seem pretty similar.

From what I can tell, they're similar from an asset issuance/transfer perspective but differ when it comes to exchange.

In particular, it seems that counterparty (and mastercoin) support issuance/tracking of tokens issued by third parties ("colored coins" if you like) but also have a concept of decentralised exchange...  i.e. bids and offers are also processed (and persisted) in the platform.

This has many downsides (blockchain bloat, etc) but does open the possibility that offers to buy/sell can be executed with certainty...  valid, matching orders result, inevitably, in an exchange of the assets.

Contrast with colored coins, where buy/sell orders are processed external to the platform - and so there is the risk that one or other party could renege.

I don't have a strong view on which I approach I prefer but they *are* different.

I wrote about it here: http://gendal.wordpress.com/2014/06/10/a-decentralized-securities-trading-and-settlement-system-is-being-built-hidden-in-plain-sight/

[Peter R]

Contrast with colored coins, where buy/sell orders are processed external to the platform - and so there is the risk that one or other party could renege.

Colored coins can be traded for bitcoin using a single coinjoin TX.  

If Bob wants to buy Alice's colored coins, they can write a single TX that spends Alice's colored coins to Bob and spends Bob's bitcoins to Alice.  The transaction is only valid after they both sign it.  If either one of them successfully "double spends" one of the inputs, then the entire transaction would be invalid.  So either the trade happens or it doesn't.

[gendal]

Contrast with colored coins, where buy/sell orders are processed external to the platform - and so there is the risk that one or other party could renege.

Colored coins can be traded for bitcoin using a single coinjoin TX.  

If Bob wants to buy Alice's colored coins, they can write a single TX that spends Alice's colored coins to Bob and spends Bob's bitcoins to Alice.  The transaction is only valid after they both sign it.  If either one of them successfully "double spends" one of the inputs, then the entire transaction would be invalid.  So either the trade happens or it doesn't.

Agree - but how do Bob and Alice find each other?

In the traditional securities exchange world, a valid, current bid/offer that is matched with a corresponding valid, current offer/bid is binding: the counterparties are mandated to exchange and there are rules and institutions (clearing houses, etc)., to enforce it.

From what I can tell, counterparty (and mastercoin and, I *think*, NXT) have implemented a scheme that does something broadly equivalent:  if you enter bids/offers on those platforms that are matched, then the underlying assets will be automatically transferred by the platform.

On the colored coins systems, there is a decoupling between exchange and settlement.  There is no in-built mechanism that will inevitably lead to an exchange of assets as a result of a successful match on whatever exchange platform is used.

You're right that one can create atomic settlement transactions in the way you outline, but that doesn't address the point I was making.  You need Alice and Bob after they have agreed to an exchange to then subsequently create and sign the transaction. What's to stop me making an offer to buy a colored asset for BTC on some exchange and then subsequently simply deciding that I'd really rather not go through with the exchange?  Answer: nothing.  Hence the section in my article where I outline ways of remedying the situation (performance bonds, escrow, etc, etc)

This isn't a criticism of the colored coin projects - I know many of the guys behind some of the projects and I have a lot of respect for them and the "embed bids/offers in the blockchain" architecture of counterparty et al is distasteful to me...  but this is, I think, the essential difference between the approaches.

[justusranvier]

On the colored coins systems, there is a decoupling between exchange and settlement.  There is no in-built mechanism that will inevitably lead to an exchange of assets as a result of a successful match on whatever exchange platform is used.

The word you're looking for is "modular".

Colored coins are just smart property tokens, that could be used in a wide variety of systems. There's no reason to weld them to a particular settlement system. The best results come from teaching an existing settlement platform to benefit from the features colored coins provides.

Note that this solution in no way requires anyone to invest in appcoins, which evaporates the entire value proposition for appcoins.

[gendal]

Thanks Justus - good article.

What's your take on the problem I discuss in my piece about ensuring inevitability of settlement once two orders have been crossed?

i.e. regardless of how assets are represented or issued, my take is that anybody wanting to trade two assets will want certainty that a validly matched order will lead to the assets actually being exchanged on whatever platform they happen to be represented on.

Ultimately, this means those making bids/offers need to relinquish some control at the point they make their bid/offer....  in other words, there needs to be a way for them to be "compelled" to exchange if their order is validly matched. And you can't do that if they have the ability simply to send the asset/coin somewhere else... that freedom has to be relinquished.

Counterparty et al do it through the protocol (and my distaste stems from the pollution of the blockchain but I also accept having app-specific coins is troublesome) - but I don't see any way of achieving something similar with a colored coin system...  unless you're prepared to put your assets into escrow at the time you post your bid/offer.  And that might be fine - but it does require you to trust an identifiable entity rather than "the protocol".

Is this an issue you recognise?  Is it only a problem in theory or are you perhaps arguing there is something in OT that can help here?  (I don't know enough about OT but it's on my to-read list)

Richard

[justusranvier]

What's your take on the problem I discuss in my piece about ensuring inevitability of settlement once two orders have been crossed?

i.e. regardless of how assets are represented or issued, my take is that anybody wanting to trade two assets will want certainty that a validly matched order will lead to the assets actually being exchanged on whatever platform they happen to be represented on.

Ultimately, this means those making bids/offers need to relinquish some control at the point they make their bid/offer....  in other words, there needs to be a way for them to be "compelled" to exchange if their order is validly matched. And you can't do that if they have the ability simply to send the asset/coin somewhere else... that freedom has to be relinquished.

Counterparty et al do it through the protocol (and my distaste stems from the pollution of the blockchain but I also accept having app-specific coins is troublesome) - but I don't see any way of achieving something similar with a colored coin system...  unless you're prepared to put your assets into escrow at the time you post your bid/offer.  And that might be fine - but it does require you to trust an identifiable entity rather than "the protocol".

Is this an issue you recognise?  Is it only a problem in theory or are you perhaps arguing there is something in OT that can help here?  (I don't know enough about OT but it's on my to-read list)

I'll answer this in very general terms.

Centralization is not a problem, and decentralization is not a virtue.

Thefts are problems.  Fraud is a problem. Counterfeiting is a problem.

Some people are proposing to solve those problem by blindly throwing "decentralization" at them like it was some kind of holy water. A more rational approach is to:

1) Develop a threat model.
2) Collect a variety of potential solutions
3) Examine the trade offs present with each potential solution
4) Select the most solution that best set of tradeoffs for the given problem.

Given that, the first question to ask is, "What are the benefits and downsides of performing order matching and settlement in a blockchain rather than in a traditional matching engine?"

Anyone who answers, "Downsides to using a blockchain? What are those?" doesn't have anything useful to contribute to the discussion.

So far my experience has been that includes most of the "Bitcoin 2.0" people and projects.

[gendal]

Thanks - actually, answering in general terms was very helpful...  highlighted exactly where I think we agree and where we were talking at cross purposes