Quote from: Stedsm on September 28, 2014, 12:35:45 PM

Quote from: Pkofet on September 28, 2014, 11:26:07 AM

-snip-
Or, the supermarket will accept your 0-confirmation bitcoin transaction as long as it has minimum fee attached.
Double-spending is extremely hard unless you have a significant proportion of network hashrate, and so it is very very unlikely a casual supermarket customer has the ability to double-spend successfully.

As far as I know, there hasn't been a double spend in over 2 years. I doubt there will be one ever again.

This all depends on what you understand as "double spend". Its especially hard if you want to revoke a confirmation (">50%-attack"), which seems not to be the case here. Why be afraid of a 1 confirmation revoke if you are talking about 0 confirmation TX?

Anyway a 0 confirmation TX can be "double spend"* if the TX is badly propagated through the network and has a low or no fee. In this case it would be possible to generate another TX that spends the same inputs but pays a (higher) fee and uses a full node with plenty connections to broadcast it to the network.
This is indeed nothing a casual supermarket customer would or could pull off. On the other hand the supermarket could demand a fee on the TX and set up several nodes to see how good the TX has been propageted to make a possible doublespend even harder.

* its not actually spend twice, but for the sake of the argument lets call it that.

Quote from: Pkofet on September 28, 2014, 01:05:17 PM

-snip-
While double-spending is rare, I don't think it is that rare.
You can take a look at https://blockchain.info/double-spends to check the latest 20 double-spending attempts.

People generally have different ideas of what a double spend actually is. I have no idea what this list by bc.i want to tell me, most of the TXs look pretty normal.