According to Antonopoulos, this statement suggests that Bitcoin is not yet centralized to some extent.Without second-level technologies, there are two ways to solve the network capacity problems and the increased requirements for its operation.

1. Increasing the size of the base unit and placing the cost of scaling on the site operators, which will lead to some centralization of nodes and mining.
2. The removal of transactions beyond the network, where the level of security and trust is lower.This is already happening and leads to the fact that transactions are stored in private databases of major exchanges.

Thus, there is no question of the work of a fully decentralized or fully centralized system. This is the choice between off-network private transactions and off-network second-level transactions with a system that does not require trust.

Note. Ed .: On this point in the arguments of Antonopoulos there is a logical error. They can be reduced to the following: if the network is not already fully decentralized, then a little more centralization will not hurt. In fact, he confirms that the massive introduction of the Lightning Network will lead to the centralization of processing of transactions in the Bitcoin network in the hands of large players, but considers this to be less evil than the centralization already present, or the concentration of processing within large platforms (which Lightning implementation will not reduce, but rather increase). In addition, there is a third way to solve the problem: optimization of the protocol to reduce the size of the transaction transactions and stored data. Bitcoin Core also works in this direction – for example, aggregating signatures will reduce the size of complex and repetitive transactions.

Channels and Routing

Many believe that every transaction occurring on the Lightning Network, if it occurs through a chain of nodes, should open a new channel. The cost of financing and calculation of each transaction in this case will be more than in the blockroom. For a transaction, only the path is actually needed, and Lightning Network clients can use automatic processing of transaction routing.

Some also believe that the channel must be closed to calculate each transaction, but this is also incorrect. In an open channel, many transactions can be made, so the funds can remain in the network and do not need to close the channel. Most of these processes will be carried out by wallets, not manually.

Note. Ed .:  If such a myth does exist, then Andreas explained everything correctly. Routing to the Lightning Network in the presence of permanently open channels is carried out at no extra cost. But constantly keeping a lot of open channels is beneficial only for large payment processors, which leads to the centralization of processing described above.

Routes and Sites

In this model, each node transmits the transaction as a “hot potato” (note: principle, similar to the children’s game “Hot Potato”, when players throw a ball that can not be dropped) to the next node in order to bring the package closer to their destination. In this case, each intermediary knows the source and purpose of the package. However, Lightning Network uses a completely different model called routing by source.

The source node receives information about all available nodes for routing, such as throughput and commission fees. Then he creates an optimal path based on the minimum total cost of the route.Further, this route is encrypted at each step by a multilayered principle. Each intermediary knows only the node from which the transaction arrived. Then it deletes the layer available to it and detects the next node on which the packet should be sent.

The destination end node is the only node that receives the information that it is not necessary to transmit the packet. The source node is the only node that knows that the packet has not arrived from another node.

Note. Ed .:  The concept of multistage transactions so far exists mainly in theory, since the operating network is not yet sufficiently complex for multiple transitions.

Locking bitcoins on the network

Each channel needs sufficient funding to process the required amount of transactions. But the channels can be replenished only for making large payments, sending bitcoins to the Lightning network address, and then taking them from there.

In fact, this is a way to prevent centralization. A node connecting many channels with a large number of bitcoins will become a target for hackers. It is more profitable to have more channels for more nodes in a network with a grid structure than high-cost centralized nodes.

Note. Ed .:  Not the most successful explanation. As long as the channel exists, the amount of bitcoins necessary to keep the balance closed for the moment is blocked. This amount (with effective configuration of nodes of the sides of the channel) can be changed dynamically, and it will be the smallest if the exchange occurs in both directions, but a certain amount in the channel is always necessary. With a large number of channels, the blocked amounts will be significant. Moreover, if the channel transmits transactions exclusively in one direction, the blocked amount will be equal to the sum of all the transferred transactions. And the presence of a large handler of multiple nodes does not reduce the concentration of transactions “in one hand”, only creating the appearance of its absence.

Combating Money Laundering

Not necessary. In most cases, the Lightning network is designed to transfer small amounts. As a rule, regulation on combating money laundering applies only to cash flows of a much larger volume. It would be difficult to find and bring to justice those who work on their home computers. However, if the problem does arise, this will result in the network beginning to better mask its operations.

Note. Ed .:  Also an unfortunate explanation, actually confirming the original “myth”. For regulators, the transaction amount of a particular person is important, not the amount of each individual transaction. In general, the problem lies in the administrative, not the technical plane.

Insufficient support for Lightning wallets

It is true that the early implementations of the Lightning Network carried technical data to the surface and uncovered elements that should have been hidden behind the user interface.

Purses with Lightning support should look the same as multi-currency wallets. Sending and receiving will be implemented using QR codes, and the wallet will be able to determine whether the address is intended for Bitcoin’s or Lightning Network.

Note. Ed .:  It is still too early to talk about full support for LIghtning wallets in the graphical interface – full-fledged work with LN for an unskilled user is still a matter of the future.