"GPS, how long to the nearest Mc SubKings?"

"GPS, too soon, how long to the next on route?"

"GPS, order Mc Sub from aforementioned Mc SubKings..."



-MarkM-

I think for example BitPay already does this optionally. It does not wait for confirmations for small amounts for some merchants. This is not a problem.

Cypto will never be as fast as just handing cash to someone.  Even if you make an altcoin that is 100x faster than BTC in confirmations, those confirmations will be less secure than BTCs.

The best solution will be for trusted service providers that both parties can confirm.  For example, if someone made some sort of Bitcoin debit card...

Not really.  Merchants accept some risk of chargebacks from employing credit and debit card payment processing networks.  They monitor it and weigh the cost/benefit.

Here's why they will accept some risk.  Selling a Mc SubKing presumably provides profit.

A race attack is not easy to perform (when the merchant is properly configured -- no incoming connections, and outgoing explicitly to well connected nodes and/or subscription to double-spend monitoring service).  And there are no guarantees that the attacker will succeed.

So if I as an attacker some how were to get success 5% of the time, that means I pay for and eat 19 Mc SubKings to get the 20th "free" (on average).

I bet the restaurant would take that customer's business any day of the week, even with those odds.   Fraud is illegal so the attacker would probably be easy to identify -- as the only one ordering a Mc SubKing wearing a full facemask.

If the attacker attempts the Finney attack, success is more likely but it is very expensive.  In that attack, the payment is only made after a block (containing the double spend transaction) has already been mined but is held from being broadcast.  It costs the attacker about $2 per second (at BTC/USD of $50) to perform this attack.   This attacker would be easy to identify also.   He's the one waiting at the counter for hours until his miner gets a block -- then screams frantically about getting a few Mc SubKings ordered, fast, and sends payment as soon as the QR code is presented.  As soon as the merchant accepts payment the mined block is released and the funds double spent.   The restaurant can know of this right away, so again, the attacker would never order anything that can't be hauled out to the getaway car in a matter seconds.   So the full facemask would probably be helpful there as well.


It does however expose the merchant to the same risk of a double spend attack that the OP is suggesting is the reason to wait for confirmations -- BitPay doesn't insure or protect against it happening, they simply notify the merchant (via the same channel as the original payment notification is sent).

Now there is a scenario where a merchant would not want to accept on 0/unconfirmed.  

A coin change machine at an unattended laundromat.   The attacker could try over and over to get a successful race attack.  For the unsuccesful attempts, the atacker gets quarters at face value.     Essentially if the items is high value purchase (above $100), or low margin, and can be purchased without revealing identity -- that's not a good candidate for 0/unconfirmed.  

There's other solutions that could be used as well ... green addresses, BTC payments through an Open Transactions server, Ripple, redeemable codes, and much more.   The problem that Square just had with their gift e-codes shows that regulations make it expensive to pass around digital credits (as money transmitter licensing and consumer protection regulations would need to be considered).    So Bitcoin may have an opportunity to go where others can't.

Fearing a double spend with $10 products is well, totally unnecessary. Simply make sure the receiving Bitcoin node is connected to well connected nodes and after 5-10 seconds there is no easy double spend. Seeing one with a $10 purchase is super unlikely. Simply accept without confirmation, that is the solution.

For expensive products green addresses are one solution, or waiting for a confirmation.

What is Joe SIxpack's risk, and what his doubt, as he talks toward Mc SubKing's microphone from his car window to order a SubBurger?

That he might doubt that they accept bitcoins there is believable, also that he might doubt the microphone is equipped to process one-tap payments from a bitcoin one-tap card.

But how did he even come into possession of such things?

(I doubt he bought then a Mc SubKIng's...)

In short if its the consumer you are talking about, I am not sure how he ended up at Mc SubKing's at all in any context relating to bitcoins...

Lets maybe step back a bit, is here there to try to evengelise to them that they ought to accept his bitcoins?

Or is it assumed they already got talking into that yet their marketing department hasn't figure out how to motivate him to try this new currency they now espouse?

Or what?

(Is he wondering whether they are marking up the subs to cover the cost of finney attacks, or something?)

TL;DR Why is he even thinking about confirmation delay, is it a fast food joint or not?

-MarkM-

I certainly see this as something with legs.  People saying that the merchant needs to have a well-connected node, etc, are thinking like geeks, not like merchants.  A merchant doesn't want to deal with any of this - they want a POS system that will let them get paid, minimum fuss, minimum chance of being ripped off.

I see a "bitcoin insurance" company that would insure against double spends, for a small fee (same ballpark as the tx fee), as a real opportunity once bitcoin gets out of geek/cutting edge territory and closer to mass market.  A company with a network of geographically distributed and well-connected nodes could detect a double spend in a matter of seconds.  And if the merchant transaction were routed via this company (which it almost has to be), and this company were connected to all the major pools, it could almost guarantee that its transactions were the ones that the network accepts.  Add to that some mining capacity and you've potentially got a business.  Add to that bitcoin wallet storage and security, currency conversion and hedging, integration with existing POS vendors, and you've maybe got a real B2B bitcoin solution.

IMHO

Bitcoin allows a merchant to DIY or there are service providers as well.   The reason this reminder of "well-connected node" is described frequently is because a merchant that does it the DIY approach without following that little detail will get ripped off in a race attack -- a study determined 100% success by an attacker is possible if not configured properly.  The worst part about it, the transaction will appear, then disappear with no trace as if it had never been made.  So that is just a caveat for those DIYers.