Hacker News Reader

Comment by jimrandomh | Hacker News Reader

The key thing here is that this is strictly about what happens to transactions which don't have confirmations yet. The standard recommendation has always been to wait an hour (6 confirmations) before crediting someone as having sent you Bitcoins, and doing so is an absolute defense against all the shenanigans being talked about.

View on HN

I have not heard Bitcoiners routinely advise each other "If you buy something with Bitcoins, you probably should not attempt to buy something else for an hour." That's a necessary component of any defense against these shenanigans, at least until the One True Bitcoin Client starts to differentiate between change which is 6+ confirmed and change which isn't yet.

> The standard recommendation has always been to wait an hour (6 confirmations) before crediting someone as having sent you Bitcoin

Seriously, how should Bitcoin replace real money if that remains true? There's a book store round the corner accepting bitcoin - so I go, select a book, pay and hang around an hour until I get to leave with the book? At least I'd have enough reading material...

The Bitcoin community has two answers to this:

#1: "Well if you buy a book with a credit card, the bookstore has no guarantee that that charge actually settles successfully."

#2: "Most people won't use Bitcoin directly. They'll keep their Bitcoin with a trusted provider. The trusted provider can verify capability to pay as quickly as a SQL transaction, via any mechanism mutually acceptable to it and the bookstore, which need not involve the blockchain. Providers which routinely allow their customers to steal from other members of the ecosystem won't be trusted for very long, and be forced to transact through the blockchain versus being being extended instantaneous credit."

The response to #2 is typically "We could call those providers 'banks'" and the Bitcoin community is, how shall I put this, divided by how much they like that outcome.

I tend to buy my books in cash. That kind of settles immediately with the transfer of the note. Disputing a credit card charge for a card present charge with either a signature or a pin is much much harder than the typical internet "card not present" transaction, so #1 is sort of moot. Actually, in germany (and other parts of europe) it would be uncommon to pay a book with credit card, we all have debit cards that clear immediately.

#2 just replaces the current system with the same system but another name and different actor, so I still don't get where the advantage lies :)

I can totally see how bitcoin can work in online transactions where credit cards are currently used, especially if fulfillment happens after some delay, but I don't see it replacing cash unless we mint little bitcoins with an embedded chip that contains a non-removable amount of satoshis. But that's very much not viable.

WRT #2, it's very different to need a wallet for small spending amounts (say, $100) and a bank.

You don't get to wield much power when your costumers are using you for buying coffee at Starbucks, but not when buying a car or another big purchase, or something shipped from somewhere (since it's reasonably to wait one hour for those to process).

So I need two wallets? One that I keep with a <bitcoin-bank> for small purchases that need to clear quickly and one that I keep for larger purchases. That doesn't sound practical.

Also, if I understood patio11s first post correctly, that large-purchase wallet would still need an hour after each purchase to clear properly [1]. Now, I don't run around and buy a couple of cars one after the other from different provides but it quite often happens that I need to book flights, hotel accommodations and rental cars for a business trip and I can't wait an hour after each step to go forward to the next one. More wallets here? Am I missing something?

[1] The ideal case seems to be to wait until a transaction is in the blockchain before sending the next one from the same wallet, but that could in theory take even longer.

Many implementation details can be hidden by the user interface. A wallet might keeps a set of addresses with small balances and ones with larger balances.

I don't think book flights and hotel reservations would require you to wait until previous transactions clear, these kinds of businesses always have the option to cancel your booking if they notice a double-spend on your side.

Edit: in general, businesses that receive your payments days in advance of delivering the product, Bitcoin is a better alternative than Credit Cards which can have chargebacks weeks after or charge a high fee

You can avoid the issue by not sending a large quantity of BTC all at once to a single address. For example, after buying from Coinbase, fan out your BTC in small increments to many addresses.

In Electrum, you can right-click on a receiving address with 0.1 coins in it, and choose to spend coins only from that address ("Send from").

Same principle as not carrying around a wallet full of $1000 bills.

#2 seems reasonable but #1 isn't, if the card payment is accepted I'm sure the bookstore has some kind of a guarantee? Otherwise I expect the fraud would be overwhelming. Although that might just be a difference between the USA and Europe, I never quite understood this concept of credit cards in the USA.

And regarding the creation of banks for bitcoin I find it interesting that we keep coming back towards old and tried solutions. As a bystander it's pretty educative, I feel I've learned quite a lot about how money works (or doesn't work) in all those bitcoin threads we've had lately.

#1: well, the bookshop can dispute the chargeback (I don't know what the success rate for this is) and if the cardholders makes too many chargebacks their card will probably be cancelled. But yes, I can go to a bookshop, buy a book, and then call my bank and say the charge was fraudulent or the product I bought was never delivered, and they'll give me the money back (and not pay the bookshop). Well, I could if bookshops in my country accepted credit cards :)

This is why credit cards are, I kid you not, one of the crowning achievements of the 20th century: there is no guarantee and they still work. Even before we had constant-on connections which could phone home and verify with a database, possession of plastic which looked mostly cardlike was Good Enough (TM) to suggest that a bank trusted a bank trusted a bank trusted you to make good on your debts, and accordingly the hotel/restaurant/bookstore/etc should grant you services immediately on the assumption that you'll settle up as promised and not attempt to cheat them.

The best guarantee you get when you run a card in the US is "As of the current instant in time, that card is valid and has sufficient credit on it to cover the amount you just charged", but that is very far from a guarantee that the charge will actually be settled without incident. But it's still good enough.

(And if you're scandalized by this, ask me about checks!)

To be specific, credit card companies put the bulk of the risk on the merchant. If someone steals your card and buys a book, the credit card company will reimburse you and (most likely) not pay the bookstore. They set up the incentives this way because customers are more likely to use their card when they are protected, but stores can't afford not to accept credit cards.

The risk might be on the merchant, but the premium will of course be paid by the consumers.

Don't forget, merchants also pay a premium in their fees for every transaction.

Which is paid by everyone, even people that settle their transaction in case. I use a cash-back credit-card and effectively my fellow shoppers are subsidising my purchases, because we're all collectively paying the card fees which I'm getting a rev-share on. CCs are a deeply weird model.

Mr. Frank Abagnale might also have a word or two to say about checks and trust ;)

Asking about checks.

(P.S. Your thoughts and response are always very thoughtful and appreciated).

Really, cash is no different. I remember when I was living in London it was really hard to pay for things with 50 pound notes because they had been subject to a lot of counterfieting.

I see a major problem with both of those.

You covered #2 - yep, those are banks. And you'd be giving them a lot of control over the currency if this became the standard means of transacting.

#1 ... if bitcoin is to replace fiat, then it must replace fiat. I can go ahead and walk into a bookstore with a $20, bill and walk out with a book and $10 in change.

It's not a replacement for fiat if there's an hour before I can leave the bookstore, and during the same hour can't buy anything else with my remaining $10...

I'm not sure why you think you speak for "the bitcoin community" here.

My answer is much simpler:

#3 It will be fixed.

And I know I'm not alone with that answer.

There is no easy way to fix it without changing BTC protocol dramatically towards something more like LTC or Doge(ie shorter confirmation times). So #1 and/or #2 it is.

Reply on news.ycombinator.com