Also, remember - LN is not designed to send millions - for this you will still have your conventional on-chain transactions.
Who is talking about millions ? I'm talking about monthly payments, energy bills, rent, things like that already are impossible in a decentralized version of LN. And sure, I can go onchain but look at the extreme costs, even with only VERY few people using it, it got at $60. Now imagine the whole world using BTC to pay for their monthly bills, transaction costs would run thousands of dollars per transaction !
It really changes when you trying to think what word "Centralization" really means, and what scares you in it. Bitcoin was made to fight control imposed by banking, not single point of interaction. If central hubs are trustless, incentivised and have no real control over my funds, I'd say it's not that same centralization Bitcoin was born to fight with. After all you may say Mining pools are "centralizing" bitcoin, when they don't, because even though they can have big share, they don't have real control.
It's just funny, bitcoin was supposed to be an alternate system and now it's really inevitable that it's going to be a banking system in the end.
It's a joke because you have ZERO guarantee you can pay the person you would like to pay, even if both have open channels and you're a billionaire.
What?
There has to be an actual route from person A to person B, consisting of people who are A) online, B) connected between you and the person you want to pay and C) EVERYBODY in that route has to have at least the amount that you want to pay, freely available. That's just not a very realistic situation. Only option then is going on chain but then you'll pay hundreds, maybe thousands of dollars, if bitcoin is ever to become big. Pretty much nobody is using it at the moment (compared to fiat) and costs already went over $60. That can easily go 10x, maybe 100x.
No. There a ways of routing on the "fly". Most efficient would be to use shared map, but blind routing is possible. How do you think TOR work?
Eh? How is that even supposed to work, 'on the fly' ? The network needs to know which user in a route can pay how much, otherwise the transaction can't go through.
Who is talking about millions ? I'm talking about monthly payments, energy bills, rent, things like that already are impossible in a decentralized version of LN. And sure, I can go onchain but look at the extreme costs, even with only VERY few people using it, it got at $60. Now imagine the whole world using BTC to pay for their monthly bills, transaction costs would run thousands of dollars per transaction !
So? That is the great use case for LN that whole world can use without loading blockchain. Problem solved. 7 tx/sec won't be enough, but if LN flies - it won't be a big deal to increase block size.
It's just funny, bitcoin was supposed to be an alternate system and now it's really inevitable that it's going to be a banking system in the end.
In worst case it will be both. In best case it will be what it is now but better. LN is optional that is what important. Like conventional TX? Go ahead!
It's a joke because you have ZERO guarantee you can pay the person you would like to pay, even if both have open channels and you're a billionaire.
What?
There has to be an actual route from person A to person B, consisting of people who are A) online, B) connected between you and the person you want to pay and C) EVERYBODY in that route has to have at least the amount that you want to pay, freely available. That's just not a very realistic situation. Only option then is going on chain but then you'll pay hundreds, maybe thousands of dollars, if bitcoin is ever to become big. Pretty much nobody is using it at the moment (compared to fiat) and costs already went over $60. That can easily go 10x, maybe 100x.
I think I already covered all the points, like transaction splitting etc.
Eh? How is that even supposed to work, 'on the fly' ? The network needs to know which user in a route can pay how much, otherwise the transaction can't go through.
Note that I'm theorizing here. From the top of my head each full node can have snapshot/map of nearest channels. If full map consists of N overlapping pieces those maps this will be maximum number of hops needed to reach destination. Also if you think about it how much nodes there will be really? Will this map be that large?