These transactions are not occurring on the bitcoin network - they are occurring on the lightning network which is a second layer on top of the base bitcoin layer. Transactions here are virtually free and use virtually no energy. This is how bitcoin scales for many small transactions.
The only knowledge you have of the lightning network is a YouTube video from 3 years ago posted on a subreddit with a financial interest in badmouthing BTC?
I'm not really sure if this is the proof I'm looking for.
The author of the video was explaining how the LN works and it seems to be an exponential routing calculation problem ( similar to traveling salesman )
So while it could work for 75 k channels, it's not going to work for 1 million.
If the explanation of that video is correct (since you didn't disagree that), then how many channels it has now is irrelevant from my POV.
Confirmed by who? Coinbase currently requires only 3 confirmations[0]. Binance requires only 1 confirmation for deposits and 2 for withdraw[1].
Actually. Was there a successful double spend in Bitcoin ever?
I mean, this kind of attack can't be done in secret, right? Such attack wold immediately make the news, because the blockchain is public. And to pull it off you need a substantial % of the mining capacity, so only a handful entities in world can coordinate such high hashrate attack.
It just doesn't seem very likely, and specially not if you're just reversing a low value retail purchase. I'd wager the rate of counterfeit money you receive when dealing with cash is way higher.
It's not about double spend. It's entirely possible two miners successfully mine a block simultaneously but one of the miners didn't include your transaction in the block they mined. Now there's two competing blockchains and every miner chooses which one is the truth... longest blockchain wins, so whichever gets the most participants is likely to be the final blockchain
Now your transaction may not be included until a future block or be lost forever, though I'm not certain what it takes to be dropped from the mempool.
The wait for confirmation is exactly to avoid double spend (but not only), as the Bitcoin wiki says[0].
> It's entirely possible two miners successfully mine a block simultaneously but one of the miners didn't include your transaction in the block they mined.
The example you make would work with just 1 block of confirmation, just keep try to include the transaction into the next block.
> longest blockchain wins, so whichever gets the most participants is likely to be the final blockchain
It's more about hashrate power rather than number of participants, however the longest blockchain of confirmed blocks wins, unconfirmed blocks may be invalidated and transactions pushed to the next ones or dropped.
> I'm not certain what it takes to be dropped from the mempool.
After 2 weeks[1] the transaction is considered invalid and should get dropped from the mempool (not all nodes use the same software version or they may run a fork with different settings)
> To be clear, the respective blockchains will confirm blocks as normal and are not controlled by Binance. For the faster service, the exchange is accepting a reduced number of confirmations as sufficient to alter wallet balances within its own systems.
Binance and Coinbase on their own nodes can choose a lower number of blocks for confirmation but the network may reject it later, if that's the case they will have to put the transaction in another block to be mined.
