Quote from: nioccoin on August 07, 2014, 10:37:27 PM

Quote from: DarkhorseofNxt on August 07, 2014, 02:40:17 AM

Probably we can accomplish it step by step. First a wallet that is running with nodes. Stabilise that can do some basic transaction. Then ora 2.0 comes with all the fancy plugin ideas.

Edit : also please consider adding multisig for public accounts.

I will very likely implement multisig, or some similar solution. It falls under our previously-discussed MFA scheme, specifically under the "something you have" category.

Anyone hungry for a can of worms? Here is one I wrote earlier in 2014. It is an excerpt from one of my unpublished white papers. This explains why I am not fond of the whole multisig (escrow) solution for commerce, though it is still valid for theft prevention.

I have reread this dozens of times, hoping for a lightbulb to come on, in an attempt to solve this issue in a decentralized manner. (Proof of Trust is an elephant name for the as yet-to-be-determined solution to this puzzle):

(the markup in this is because it was written in TeX)

Quote

\section{Talk more about proof of trust}

Proof of Trust is not a new innovation, but a ubiquitous one. The inspiration comes from human observation. We are social creatures, and with that social imperative comes the issue of trust. We all come from parents who may or may not have been trustworthy. We all have had friends, spouses, and family, some of whom have betrayed the trust vested in them. Officials betray trust all the time. What is needed is not to cast trust out, but to embrace proof of trust, and then still prepare for its unwinding. Proof of trust, like an SSL cert or trusted algorithm can sometimes be found wanting. When you read Satoshi's seminal work - he discusses the problem of trust in regards to financial transactions. But his solution is inadequate because it is one-sided. His Bitcoin system is designed for trusted transactions, but in his Introduction he writes "Transactions that are computationally impractical to reverse would protect sellers from fraud, and routine escrow mechanisms could easily be implemented to protect buyers."

Wait, this sounds as if Bitcoin does not use cryptography to eliminate trust! Exactly. It is one hand clapping. It solves only the merchant's trust problem. The buyer must \textit{hire} a trusted partner to escrow the transaction or otherwise certify the goods are as described. What is an escrow system but a third party that provides the trust component? Would anyone trust an escrow whose main office is a travelling carnival? So even Bitcoin does not eliminate trust. And that is not what it is trying to do, but it is not accurate to say that Bitcoin fixes the trust issue. What Bitcoin does is fix the \textit{seller's} risk in the protocol, and the buyer's commodities are of unknown risk, pending a second financial transaction procured by the buyer to the escrow system!!! Alibaba at long last has an escrow system that works seemingly well, and it places the onus on the seller to deliver the goods prior to the escrow releasing the money. But it increases the cost of doing business for the buyer and the seller. The seller will raise his price to accommodate the new basis in cost to his business. One way or the other, the buyer will end up paying the difference.

Further, it is clear from this that the trust part of the equation is necessarily centralized, if not by one company by a dozen, but still far less than the decentralized peers that secure the blockchain. Supposing there exists one escrow service that sifts Bitcoins and goods, attempting to satisfy both sides of the transaction. That entity will necessarily need to sit at the crossroads of the physical and digital boundaries, unless the goods being purchased are themselves digital. Therefore, that escrow entity must be physically centralized, and how will they be trusted? Won't they have to post a bond of some sort to certify their business as legitimate, and purchase insurance on the goods they hold? Won't they need to satisfy not only the financial authorities as well as the commodity authorities? It seems that an escrow service is not nearly as trivial as was envisioned. The Alibaba system is a good example, but notice they provided it to satisfy the buyer for a very good reason. Fraud from the seller side was rampant, causing the reduction of trade. The frauds were hurting the legitimate buyers and sellers. The escrow service put the onus of trust on the sell side, because the fraud wasn't bad money per se, but bad or grossly misdescribed or fraudulent goods. What if the money is fraudulent? In Bitcoin, it doesn't matter. If the Bitcoin is stolen, it is spendable without recourse.

But if the money is held in abeyance, that completely defeats the elegance of the fast payments upon which Bitcoin is based. The point is that one of the best parts of Bitcoin is in reality not feasible, and not usable, because it defeats itself in practice. As has been discussed, where Bitcoin is most interesting (purchasing goods), it is least useful, and where it is most useful (transferring money), it is prosaic, offering little benefit to the user and, as of recently, incurring a capital gains tax that makes the baseline expense of using it at least 15\% more expensive than cash.

Even if it was trivial, being at the crossroads of the physical and digital worlds requires real expenses, for facilities, for employees, and the like. Therefore, it would seem that the purchaser will pay additional fees, probably in excess of what the transaction would have cost using a credit card or cash at the local brick and mortar warehouses.

In sum, trust cannot be decentralized when it must cross the physical boundary.

This was written some time ago, and some of my opinions have changed, so consider this nioccoin version 1.0. But I still am haunted by the problem of having to centralize trust. That hasn't worked out too well for the payment card industry, and even SSL companies have been robbed blind, causing heartache, or heartbleed, as the case may be Cheesy

As long as we are dealing in physical goods, I don't think we can solve this problem "online".

Here's my vision for "the" winning currency: In-person transactions using digital anonymous payment, where the buyer can inspect the physical goods, and the seller can confirm the payment on the spot. No escrow needed. No credit card with your name on it, and being asked for your photo id. 95% of all commerce still takes place locally anyway. Digital cash needs to behave almost exactly like real cash for the buyer, and be verifiable instantly by the seller, no less than if you pay the seller with a crisp new hundred-pound note and he has to use a special pen to mark it to check for counterfeit.

kind regards,
nio

Very interesting topic nio! I agree that decentralised 'proof-of-trust' might be impossible to guarantee, but I think decentralised 'probability-of-trust' is very attainable using reputation systems, especially if the reputation is itself attached to something valuable, like a 'reputation' coin.

I think Bitmark is trying something interesting with their 'Marking' proposal - https://github.com/project-bitmark/marking/wiki

Maybe an acceptable situation for both sellers & buyers of physical goods can be found using a reputation system where the costs associated with breaking a previously earnt good reputation is made prohibitively high, compared to the potential gain from defaulting on an agreed sale or exchange. We all know that a scammer can reboot a new identity on a market like ebay any time they choose, but they can't take any 'good reputation' with them into the next 'life'. If that good reputation was actually exchangeable into another asset, then an honest buyer/seller could literally bank their good reputation from time to time, and a dishonest act would literally cost the offender real value.

Of course honest people can fall on hard times and get tempted to act dishonestly if they're in desperate circumstances, and genuine scammers can 'imitate' an honest trader to 'set up' for a big scam later on (like zipllibrary who did the shareXchange scam), but if the loss in monetary reputation was close in value to the size of a particular trade then the other party could say with a higher 'probability' that the likelihood of getting scammed was a lot smaller than if the other party had just cashed in some of their 'reputation coins'.

I think loss of reputation capital *could* be used as an effective disincentive against rogue behaviour, and it could be monetized using a crypto currency

Can we have a reputation coin built into ORA somehow?