I’ll try to answer a few of your questions, but I’m commenting in a completely unofficial capacity so if someone who is working on B&C notices an error, please let me know and I’ll correct it.
Based on their responses to questions so far, I’m pretty confident that the person or persons posting as “BCExchange” on BitcoinTalk have been involved in the Nu project in some capacity. I do not know if this will be the situation in the future or if new contributor will be given the task of posting on the B&C “official” BitcoinTalk account.
They are both important, so I’m not sure I understand what root concern your question is trying to answer. Do these statements help?
A blockchain’s number of confirmations (such as six, for Bitcoin) is a widely-accepted suggestion, rather than something that is protocol-enforced. It follows that the deeper a specific transaction is buried in the blockchain the less likely it is to be included in a chain that ends up being orphaned because a competing chain reached a higher confirmed height.
For n-of-m multi-signature addresses there’s a relationship between the level of security provided versus the ease of use and cost, if you’re confident that the reputed signers are sufficiently decentralized and are not in collusion. The more required signers, the less likely the funds are to be stolen. On the other side, the higher the number of required signers the more “expensive” each transaction is to transmit (in terms of payout size in bytes and the associated fees incurred) and the more trouble it is to bring together the minimum required amount of signers when a transaction needs to be created.
(Personal Observation): I haven’t seen a use case yet that suggests that it will, natively. That is, until futures markets could develop, or until different forms of derivatives could be offered. The payment gateway projects that @mhps has proposed seem to be the best path to limiting exposure for LPCs.
If B&C is a major success, I imagine that there will be lots of smart people working on solving the problems associated with facilitating the exchange of fiat for cryptoassets.
Neither*. B&C will not charge commission fees on trades, only the transaction fees on the messages sent through the blockchain (charged by the byte, instead of per kilobyte like in Nu or Peercoin). The size of the fee per “trade” will have more to do with the number of signers used and the type of message being sent (“trade”, “confirmation”, etc.) than the amount of tokens being traded between the separate blockchains.
* Initially. It’s entirely within the B&C shareholders’ right to vote on a motion that requires a percentage-based or flat commission fee at some point in the future.
The speed of trading on B&C will be slower relative to other (centralized) exchanges, sure. I’d wager that there are a lot of different types of day traders – some may be following strategies that require sub-second APIs while others may not need that level of reaction.
B&C’s architecture will exclude the traders that are attempting to build high frequency trading agents, but then again, so do the rate limits and reaction times of a lot of centralized exchanges that exist today (including many of the “big” players).
It will definitely have some impact, but I believe it’s too early to tell how large it will be until test trades can be made using the production code and settlement timeframes can be benchmarked.
I posted a notice about the auction in the Bitcoin’s block chain using http://cryptograffiti.info
Perhaps this gives a little more exposure to the auction.
The announcement at bitcointalk.org is currently on page 4 of the project development threads if sorted by views in descending order.
Taking into account that the thread is very young it has received an impressive amount of attention!
Most threads that have that much or even more views are much older.
Only the mercury exchange thread can almost compete with the ratio of views per days of existence.
If we might derive one thing from my randomly picked and not in any way representative selection it is: people are interested in decentralized exchanges and we know why
I would call it good feedback instead of criticism. I share that the proposed reputation system is one of the weaker elements in the design. Not necessarily bad, but also not as robust as you would like to see. Definitely room for improvement, but not sure whether that can be achieved in the first release. Maybe an iterative approach is better.
Why is that new, almost on every page someone is bringing this up: It is impossible to make a decentralized random selection using a reputation system and it also cannot work in long term if user accounts are not redistributed to new signers if the old signers go offline. It was really pointed out since day one, by many different users, and the only answer is that it is expected not to happen.
I think B&C bets on the fact that the shareholders as a whole would make sure to act swiftly in a state of emergency where the system is getting dangerously closer to situation where there are not enough trustable signers as Nu has been betting on the fact that the shareholders would act swiftly in a state of emergency in which shareholders need to raise drastically the parking rates to keep the peg…
It is basically a bet on human inventiveness and rapid response by a decentralized organization.
BKS shareholders have no influence on that, i.e. they cannot force to move the credentials of a user account from one signer to another. Even if they increase the number of signers, it doesn’t solve the problem that your particular user account still was associated to the old signers and will never go to one of the new signers if not done on a voluntary basis by the old one. So you can not deposit something on B&C and more or less forget about it, you should frequently check about the current state of signers to be sure about your coins, and make a new account every time you don’t see a sufficient number of signers online.
I still have a hard time believing that the reputation system will be that easy to manage in practice.
There’s a huge amount of overhead required to maintain trusted signers that will fall (rightfully) on the shareholders. A large amount of automation will be required to make it something that is more than just a bunch of “rubber stamp” votes.
As I said in the bitcointalk it might be possible that all it takes to set a signer up is downloading a peerbox-like image, initialize it on a ras-pi or a droplet, inject funds to appropriate wallets that have already built-in, and get keys. The rest is getting enough shareholders’ vote.
Compensation for all the deposit for opportunity cost and security insurance (funds always in hot wallets) is not trivial. So I guess the operational cost of B&C is not much smaller than a centralized exchange.
This is the part that concerns me, not finding the signers themselves, but providing shareholders with enough information at the right times to make informed decisions about who to upvote/downvote.
Connecting actual performance metrics of signatories to that flow of information is going to be very challenging using the blockchain as a data transport mechanism. That is, at least until third-party data processing services start to arrive.
Exercise: Mentally scale trading traffic up to 10,000/per day and try to imagine ways that your average voting shareholder is going to be able to keep up.
I agree with you that a Peerbox-esque system is a viable solution for automating the signing process.
One way online shopping mall manage it is letting the user rate the transaction and use that data to generate metrics for improvement. B&C could reward the user for rating. If that can be done for B&C, voters just choose a data feed that somehow best uses the metrics. Yeah it takes work to implement.
That was one of the early comments I made when I first read the white paper. I personally believe it’s critical that there’s a direct connection between the actions of the signer (as experienced by the trader who utilizes their services) and the information provided to the shareholders who will be contributing their votes to the system.
It’s not an easy problem to solve and reputation systems with “open” ratings like this have a fairly wide number of attack surfaces, but the inverse is equally dangerous if the actions of the signers aren’t directly connected to their ratings.
Since the job is mostly rubberstamping, not much performance is need, I’d accept to have several signers who are well known to the Nu community initially, and gradually open the signer position to a broader community as responsive rating/voting mechanism is implemented.
