Nah, channel factories solve that problem.

https://bitcoin.stackexchange.com/questions/67158/what-are-c...

A new layer between the layer 1 and layer 2, sure...

Why don't you think it's possible?

It is possible, but absurd. Why more layers if layer 1 works fine just increasing the maximum block size? Bitcoin Cash did it and works flawlessly. Just try it.

It's called the blockchain trilemma.

Bitcoin cash pays the price through centralisation. It can still only manage 100 transactions/second so certainly isn't flawless.

Far better to make the base layer slow but very robust, and then add layers to scale.

> Bitcoin cash pays the price through centralisation

What centralization? If you are referring to the node implementations, it has several [0]: Bitcoin Node, Bitcoin Unlimited, Bitcoin Verde,... Bitcoin BTC has basically one: Bitcoin Core [1]. If you mean mining, it's the same as BTC, same mining algorithm, same miners.

> It can still only manage 100 transactions/second

True for now, but it's way more than it needs right now now [2]. But don't worry, this month upgrades to ABLA [3], and will be able to scale to as many tx/s as needed.

[0] https://bitcoincash.org/#nodes

[1] https://en.wikipedia.org/wiki/Bitcoin

[2] https://blockchair.com/bitcoin-cash

[3] https://gitlab.com/0353F40E/ebaa/-/blob/main/README.md

The increased block size increases the amount of storage required to run a full node (and also the bandwidth required to sync nodes). Even at just 100 tx/sec, you need 14x the storage of a bitcoin node (so around 10TB vs 700GB for bitcoin). That puts it out of reach for many people to run and so making it more centralised.

To compete with a payment rails like Visa and Mastercard, you'd need to increase that to 10000+ tx/sec, making running a node complete unfeasible for most people to run.

There's simply no such thing as a decentralised blockchain network with unlimited block-size. At some point if you want to remain decentralised, you have to create layer 2+ networks that can handle the smaller, high-bandwidth transactions, and so you may as well commit to that model now and focus on making the base layer as robust as possible.

Ultimately for a money to attract the most value, it needs to optimise for the very largest transactions, and offer the most robustness (i.e. most decentralisation). This is precisely what bitcoin has done.

If you mean a full non-mining nodes (which only a few people really need, merchants, exchanges, ...), read points 7 and 8 of the whitepaper. For full mining nodes, you need expensive specialized hardware anyway.

Non-mining nodes are just observers that do nothing good to the network. Is like going to war with popcorn as a weapon. Mining nodes are the only ones than can include transactions to a block.

https://www.bitcoin.com/bitcoin.pdf

Yes full non-mining nodes. I understand about pruned nodes, but that is a bonus for both networks - it means people can run a node on their mobile phone for example. The nodes have the real power in the network - they're are the people’s vote on what bitcoin really is. You can mine what you like but if my node says it's not bitcoin, it's not bitcoin. And there's many more who think like me. This has already been proven in the block-size wars - both the consensus on block-size and proof of the power of nodes vs miners.

The HW to run bitcoin is extremely cheap - less than $100. All you need is an old laptop or Raspberry Pi with a 700GB of storage, and a ham radio/dial-up internet connection to another node.

So you're telling me that you can have a say in the blockchain with no proof-of-work whatsoever on your part

Yes. It's network effect. Even if all of the miners simultaneously decided to go big block or some other controversial change that most people don't want, the majority of the population's nodes would then not accept the mined blocks. They'd see the rate of new blocks drop to 0, and people would instantly start mining on their PCs to get the block reward. Some of the miners would start to change their mind and start mining with ASICs for the easy block-reward and eventually the hash-rate would build back up again, with the previous miners or without them.

The non-mining nodes in fact can completely destroy the ASIC miners by upgrading to a new POW algorithm. Hopefully it will never come to that, but ultimately the people, with their non-mining nodes have the power, not the miners. The only way miners can successfully change bitcoin is if the vast majority of people agree with the change.

Why “non-mining full nodes” are a terrible idea:

https://medium.com/@olivierjanss/why-non-mining-full-nodes-a...

The comments to that article (especially the one from dooglus) make a decent rebuttal to the idea that full nodes are a "terrible idea".

If you're convinced a block chain can scale to 50k transactions / sec and also remain decentralised and offer enough privacy, then keep at it, but I'm afraid I'm unconvinced.

Even if it were theoretically possible, the solution would likely be so complex, or based on a long chain of "hopefully unlikely" events that I wouldn't have enough confidence in it to store large amounts of value.

The slowness and simplicity of the Bitcoin network is what gives it robustness, and if I'm storing large amounts of money, that's what I want even if it means I can't buy a cup of coffee on the same network.

There's also a rebuttal to that comment: https://medium.com/@olivierjanss/first-of-all-i-want-to-appl...

> The slowness and simplicity of the Bitcoin network is what gives it robustness, and if I'm storing large amounts of money, that's what I want even if it means I can't buy a cup of coffee on the same network.

Then why not just use gold? It is slow, simple, and you can't use it to buy coffee.

Because it's not as liquid, it's not as easily verified, the supply doubles every 40 years, it's not as easily protected (compared to multi-sig bitcoin), I can't send it around the world in 10 minutes, and I can't take it across borders in my head. It doesn't form the basis for the entire world's future reserve asset and monetary unit, and the market is mature/saturated whereas bitcoin is completely misunderstood by almost everyone and so is wildly undervalued.

Bitcoin is demonetising gold.

You are describing Bitcoin Cash too, with the difference that with BCH in that 10-minute window the transaction is irreversible (like the original Bitcoin), so a 0-conf tx. And extremely low fees, of course (also like the original Bitcoin).

back to the start we go :-)

Bitcoin transactions are also "irreversible" providing the fee paid is reasonable, but ultimately the clearing time for both networks is down to the likelihood of a 51% attack for which bitcoin is far more protected against due to the far larger amount of hash power. It's not comparable.

If I'm moving/storing $1 billion, do I care if the transaction costs $1 instead of $20? No - I care that I'm storing the money in the safest place to store value I can find.

If we're talking about amounts much bigger than the transaction fee, then I don't see any problem with BTC's layer 1, for now.

If we look simply at network growth in economic terms, Bitcoin is only going to attract larger and larger sums of money. This halving epoch we now have institutional investors (ETFs) and once it has proven itself at that level, the next stage will be nations openly buying it.

So even if fees increase in fiat terms, so will the size of the transactions. Enough to take the market cap far, far beyond that of gold. Once bitcoin has grown to the point that its value has stabilised, there will be substantial demand for transacting in it directly and there will naturally be a lot more work invested into higher-level payment networks/solutions. Right now though, Gresham's law sees to it that most people are happier to hoard than spend and that's likely to be case until bitcoin is at least $10M/BTC