Quote from: gendal on June 30, 2014, 09:44:26 PM

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.