⭐ Merited by pooya87 (1)
Ah, the good ol' discussion. And I see arguments haven't changed in the past year 
I believe the Lightning Network will surely have some centralizing incentives. But that does not mean that the whole system will be centralized. For me, it will be a mix of hubs (commercial ones and altruist ones) and decentralized routing happily coexisting. After all, LN (like every cryptocurrency!) is a social network, it's a network of individuals, and they decide how to build it up.
That has two advantages:
- first, you have to pay only one transaction fee for all the 15 individuals to open or close the channel. The transaction is bigger, so the transaction fee will also be bigger, but the fee per individual node would be >5 times lower.
- second, you can integrate a well-connected node (a little hub, for example) into the group, to be able to recharge channels that get "out of balance". That would mean that there will be very few necessary on-chain transactions.
So CFs could be an interesting part of the puzzle to avoid the scenario that "only wealthy individuals and businesses" could afford to open channels. Instead, you would open a channel factory with your friends, some merchant(s) you regularly visit and perhaps one exchange to re-charge the channel in an off-chain manner if necessary.
I believe the Lightning Network will surely have some centralizing incentives. But that does not mean that the whole system will be centralized. For me, it will be a mix of hubs (commercial ones and altruist ones) and decentralized routing happily coexisting. After all, LN (like every cryptocurrency!) is a social network, it's a network of individuals, and they decide how to build it up.
Therefore, I would consider the Lightning Network a semi-decentralized solution, since it doesn't affect Bitcoin's decentralization (as its separate from the main chain), but it has some level of centralization in the way channels are opened and closed, mostly by wealthy individuals.
Have you seen the Channels Factories proposal (Whitepaper)? In short, CFs are single channels opened between multiple parties (up to 15 with the current limitations of Bitcoin Core, but afaik many more if Schnorr signatures are implemented). Members of such a group can open sub-channels at no cost to the other members of the group, and transact freely.That has two advantages:
- first, you have to pay only one transaction fee for all the 15 individuals to open or close the channel. The transaction is bigger, so the transaction fee will also be bigger, but the fee per individual node would be >5 times lower.
- second, you can integrate a well-connected node (a little hub, for example) into the group, to be able to recharge channels that get "out of balance". That would mean that there will be very few necessary on-chain transactions.
So CFs could be an interesting part of the puzzle to avoid the scenario that "only wealthy individuals and businesses" could afford to open channels. Instead, you would open a channel factory with your friends, some merchant(s) you regularly visit and perhaps one exchange to re-charge the channel in an off-chain manner if necessary.
Quote
There could be some point where banks, and big exchanges will be capable of opening/closing channels, making the Lightning Network a system of trust. This might change though, if people choose other channels and/or LN nodes that are operated by individuals instead of big corporations, or financial institutions. Despite this, it will also be possible to transact on Bitcoin's main chain, for further security.
Exactly! There will be no mandatory route. Everything is possible. Let's wait and see. I think LN will be an excellent micropayment network, but not necessarily will replace on-chain transactions completely.Quote
One big question that I have, is how would the Lightning Network be as secure as Bitcoin's main chain, considering it's an off-chain solution? I think that it could be easy for a hacker to easily perform double-spending on LN, right? This has brought up a lot of confusion to me, and I'm still trying to figure it out when reading the whitepaper.
As LN transactions are "private", but regular Bitcoin transactions, they cannot be double-spent easily. I don't know if a large LN channel, however, could incentive a Bitcoin double spend attack.