11
1
On the bitcointalk forum, evoorhees wrote:
For a long time people said the "waiting for six confirmations" was a significant drawback, but most people understand that problem has been almost entirely eradicated.
and Revalin replied:
as you mentioned we have an elegant potential solution to the confirmation delays, though I haven't seen any moves to actually implement it yet
How has the problem of waiting for confirmations been eradicated? What is this elegant potential solution? The first quote seems to imply the problem is already solved, whereas the 2nd says the solution hasn't been implemented yet. Which is it?
It should be noted that green addresses can (in the future) be used with a split-key eWallet, with a default transaction sending back to the client in the future, so there's no need to trust an eWallet with your coins to use his green address. – Meni Rosenfeld – 2012-03-20T06:29:27.480
1
The consensus on multi-node listening is that it you have to do it to protect against a multi-node double spend attack, but since that's not the most serious double spend attack known, it doesn't change the number of confirmations you need to wait for. That needs to be based on the most serious known attack, the Finney attack, which doesn't involve submitting different transactions to different nodes.
– David Schwartz – 2012-03-20T17:00:15.587But without multi-node listening, the chances that the scammer is successful doing a race attack increase ... maybe to more than 50% even. That would create such an incentive that the merchant would be forced to discontinue accepting bitcoins. At least with multi-node listening, the risk can be marginalized down to acceptable levels. – Stephen Gornick – 2012-03-20T23:51:05.150