I'm sorry. It so obvious to me that you do not have computer who can handle 24 GB block every 10 minutes.  And if you think you have or you WILL HAVE (in next decade) then it is like debating with 5 years old child.

That's just plain wrong and we've been through that before.

"b) current technology cannot handle 24 GB blocks (and will not handle any time soon)"
Your answer was that it is my feeling. So show me computer what can handle 24 GB blocks b/c I have never seen any. (maybe NASA has some supercomputer)

this sounds exactly like something i'd say:

The Blockstream Business Plan (self.Bitcoin)

submitted an hour ago * by Celean

Note: This was previous posted and (self-)deleted, but has been revised to address some factual inaccuracies.

A lot people seem to be confused about exactly why the developers that are getting a paycheck from Blockstream - most of which you can find on this page - are all so vehemently opposed to any and all discussions about increasing the block size, even by a moderate amount, much less in a way that scales naturally over time in a way miners can influence.

As most regular readers will know, Blockstream received 21 million US of venture capital funding less than a year ago in order to develop sidechain/payment channel concepts for Bitcoin. Among other things, they have joined development on the Lightning Network - for example, Rusty Russel is a Blockstream employee who is a confirmed prototype LN developer.

Now, obviously it would be hard to attract $21M of funding unless you have a plan to make a profit on the development, and while they haven't published any business plan that I'm aware of, it is by now increasingly obvious how they are planning on obtaining this profit.

How the Lightning Network works

The paper presented for the Lightning Network is a whooping 59 pages, and as such, I expect that the actual number of people who have read it numbers in the dozens. There is a more succinct explanation here, written by Rusty Russel himself, but essentially (and highly simplified):

    The system is trustless, and no node can run away with funds that haven't been agreed by both the sending and receiving parties, but in case one party misbehaves, funds will be locked down for a period of time until a set timeout occurs.
    It is conceptually based on a hub-and-spokes model with large centralized "payment nodes" that numerous people and companies open payment channels with. Payment nodes can be interconnected, thus forming a chain of payment channels from the sender to the recipient.
    To open a payment channel, a leaf node (end user) has to commit an "opening transaction" with a specific payment node (or any other leaf node) to the blockchain. The funds committed at this point is the largest amount that can be spend during the life of this payment channel, and every payment channel you open requires one such transaction.
    When a payment channel has been opened, multiple transactions can be created and signed on the channel without being published to the blockchain, up to the amount of funds committed.
    The funds in the opening transaction are locked to that specific payment channel. To make funds available again for either party, all the final transactions have to be committed to the blockchain, thus finalizing the BTC transfer (if any).

Centralization drivers

The Lightning Network, by design, consists of what is effectively one-way payment channels between two nodes. In order to avoid the need for end users having to open a large number of payment channels (and thus having to commit a large amount of funds for these), it is conceptually based around centralized "payment nodes". If a sender already has a payment channel open to such a payment node, and that payment node has direct payment channel open to the recipient, or can route a chain of payment channels through other payment nodes, the payment is essentially instant. If it's not, a new payment channel has to be created by committing (and waiting for) a blockchain transaction, which is not faster than making a direct transaction on the Bitcoin network.

As a number of blockchain transactions are required to create and subsequently close out a payment channel, and you have to lock down funds for each separate payment channel, most people would only want to have one or a handful of such channels open at any given time.

In other words, payment nodes will be subject to a massive network effect. The more people use it, the higher chance that an existing chain of payment channel can be found, which means that you get a low-fee, almost-instant transfer of coins, instead of an awkward wait for the blockchain to confirm the transaction.

Worse yet, as the signing keys need to be Internet-accessible for payment channels to work near-instantaneously, the payment hubs will require having the full balance that is committed to a payment channel in what is effectively a hot wallet. This will be a huge security risk for most people, further cementing the centralization of that network to those that can manage a highly secure infrastructure.

How Blockstream plans to profit

The essential question of "how can anyone profit from the Lightning Network" is easy: payment nodes will have the ability to charge fees for the payment channels that connect to them. Note that there will be very real costs in running a Lightning Node, both in terms of hardware and in the cost of having funds being locked down in payment channels (and subject to theft), so that by itself is fair enough.

Less connected nodes will have a significant handicap and have to charge higher fees for two reasons: first, for the blockchain transactions required to establish their own payment channels to the better connected nodes, and second, because the better connected nodes will presumably charge fees for the less connected nodes to use their payment channels. This assumes that well-connected nodes will allow less-connected nodes to open payment channels at all, which they may opt not to do.

This means that the first mover advantage is incredibly significant in the establishment of this network. And Blockstream, as a significant developer, will obviously be perfectly situated to be the primary provider of this service, and collect all the fees this entails. Depending on the openness of the codebase and timeliness of its distribution, other players may or may not be able to compete, but this isn't known at this point.

How this relates to the block size

The reasons laid out above perfectly explain why these developers completely reject any notion of increasing the capacity of the base bitcoin network. They want a fee market to be established so that when the Lightning Network is ready to operate, there is a significant cost in placing a transaction on the blockchain. This, in turn, will encourage people to shift their transactions over to Lightning, which will allow the payment node operators rather than the miners to collect the fees in question.

Furthermore, the more expensive it is to place a transaction on the blockchain, the more advantageous payment channels will be, and the higher fee can be charged by the payment node operators. It also makes it more expensive to sustain multiple payment channels, which will further boost growth for already well-connected payment nodes.

The Lightning Network is a genuinely revolutionary invention that will allow Bitcoin to scale to a much higher degree than before for micro-transactions and frequent small purchases. However, it is important to keep the bias in mind when you read debates about the block size. It is essentially pointless to discuss it with many of the involved developers, as they have too great a stake seeing the block size remain where it is. The only way the block size will ever be increased is to outvote them and ignore their frequent demands for "consensus" (which will never be reached).

Blockstream developers frequently use the argument that a larger block size will increase centralization of the bitcoin network. This is somewhat hypocritical and disingenuous, as the Lightning Network by its very nature will be far more centralized than the core network with a larger block size will ever be.

tl;dr: Blockstream wants to choke transactions on the blockchain in order to spur adoption of sidechannels and the Lightning Network, where they will be perfectly situated to collect fees for providing that service.

You're saying it's wrong that they've been trying to argue that?

<shrug> I was just explaining the context of my making the above points. If you don't think they've been denying that their CoI is an issue at all, then good, feel free to ignore my arguments for why it is at least a concern, since you think they already agree as well.

If they can't handle the pressure from the investors, let others try to implement lightning or something like it. Satoshi did not have 21 MUSD. Lots of people did not have 21MUSD.

You must not understand what proof means.

The problem with your statement is that you've expressed it in terms of an unsolvable problem, which is begging the question.

Processing 24 GB of data in a 10 minute period is possible with technology that exists today. It does not require faster-than-light communication, or violating mass/energy conservation, or solving the halting problem, or anything else that's actually impossible.

Perhaps what you actually mean is that handing a 24 GB every 10 minutes would be expensive using existing technology. Perhaps even so expensive as to cost more than Bitcoin users would be willing to pay.

Expressed in those terms, you have a statement that can actually be rationally evaluated.

You'd need to establish how much it would cost to pay for the hardware and operating expenses to operate a 24 GB/10 minute network, and then estimate how much the users would be willing to pay per-transaction. If the estimated costs exceed the estimated budged, then you'd be correct to say that the network would probably be too expensive to operate.

Of course, once you've phrased the problem in a manner that allows for potential solutions, then it becomes possible to talk about productive things like, "what steps could we take to reduce the cost of operating the network?"

On the other hand, if you phrased your "obvious truth" in a form of an unsolvable problem deliberately because you have a preference for the problem remaining unsolved, then keep doing what you're doing.

The odds of miners building accidentally on a false chain seem low. By definition they have to be able to download a block in less than 10 min (otherwise they couldn't keep up with the transactions themselves).

Yes they might build on a false chain for a short period of time if blocks are randomly found fast, but the statistics works out that this won't last long and waiting x blocks solves the issue. You're better than me at the math but I'd bet that 6ish confirmations works out to good enough.

If a majority of miners are intentionally doing this, then we have a 51% attack underway and lots of assumptions break down.

I question the need for new messages to help SPV clients find the right path because 1) they seem exploitable to me and 2) all they need is the header chain to find the right path and headers are already short and fast to transmit. Waiting for x blocks again seems to protect them if you assume a majority of well connected miners.

Well you have 16 years to come up with better scaling solutions and prevent 24gb blocks from happening.     Unless better solutions are ready to go, then we will have no choice but to allow the blocksize limit to increase every 2 years.