2 weeks seems like an outlier. A recent thread [1] described someone getting cash from an ATM via bitcoin in 20 minutes. One of the comments theorized that thi

2 weeks seems like an outlier. A recent thread [1] described someone getting cash from an ATM via bitcoin in 20 minutes. One of the comments theorized that this was because the ATM service used only 2 verifications instead of 6, taking a slightly higher risk due to the small transaction amount (which implies a typical wait time for a large amount of an hour, not 2 weeks). Of course, even 20 minutes is too long to wait for a vending machine transaction, so your point still stands. I also found an article [2] that incorporates the problem of delay in even determining what the fee would be. What was also interesting is that it [2] detailed a transaction cost 8x what was quoted above, and seems to be backed up by https://bitinfocharts.com/comparison/bitcoin-transactionfees.html https://bitinfocharts.com/comparison/bitcoin-transactionfees... . It brings a new concern, volatility, to someone like a vending machine operator whose transactions aren't "micro" but are always going to be on the small end of "every day". Presumably, it's possible to have a cryptocurrency without these issues. Lightning may sufficiently solve this for Bitcoin (and/or other cryptocurrencies), but that has yet to be seen in practice, at scale. If a DDoS attack forces all (or even just some) Lightning transactions on-blockchain, what does that do to transaction times and costs? [1] https://news.ycombinator.com/item?id=17898721 https://news.ycombinator.com/item?id=17898721 [2] http://moneyandstate.com/the-true-cost-of-bitcoin-transactions/ http://moneyandstate.com/the-true-cost-of-bitcoin-transactio...