Litecoin founder talks about Bitcoin expansion: security and decentralization should be prioritized

The author of this article is Charlie Lee, the founder of Litecoin and currently the director of engineering at Coinbase.

I first discovered Bitcoin in early 2011. Like many people, I thought it was awesome. Bitcoin transactions are: decentralized, secure, fast, cheap, and unlimited.

Let's take a closer look at what this means:

Decentralization: Thousands of nodes verify and relay all Bitcoin transactions.

Security: Millions of dollars of hardware ensure the security of the Bitcoin network.

Fast: Verification is fast. Since it is impractical to launch a double-spending attack, zero-confirmation transactions are very secure.

Cheap: You pay almost nothing. Your transaction will be mined and included in the next block.

Unlimited: At that time, Bitcoin blocks were very small compared to the block size limit.

The delays in transactions are due to the fact that there is no limit on the block size. So there is actually no limit on the transaction flow.

This is Satoshi Nakamoto’s unique creation. Bitcoin is run like a startup: employees are paid equity at first, until the company makes enough money to pay higher salaries and reduce the equity paid to employees.

In the case of Bitcoin, the block mining reward initially offset the decentralized network costs, but gradually shifted to Bitcoin users paying for the fees. Just like startups, Bitcoin miners also serve as an investment in Bitcoin's success. In essence, early Bitcoin adopters had the best of both worlds.

Unfortunately, this doesn't last long.

By design, the block reward is halved every four years, and its trend is towards zero.

If this happens, we will have to choose between Bitcoin’s five core features (decentralization, security, speed, cheapness, and unlimitedness) and figure out which one we are most willing to sacrifice.

I'll explain why this is the case using an analogy.

Satoshi Castle

King Satoshi built a castle for his peasants. He decided to pay 50 gold coins a day to hire soldiers to protect his castle and peasants. Since his wealth was limited, he planned to cut the subsidy in half every four years.

So starting from the fifth year, he only needs to pay 25 gold coins every day.

After about 100 years, he no longer had to pay his soldiers any allowances. His idea was that the peasants would come to value the safety and security of their castles, and they would eventually be willing to pay their own soldiers.

The security of Satoshi Castle was so good that many peasants from other castles began to migrate to Satoshi Castle.

Satoshi was a very welcoming king and he let everyone in. All was well until one day the castle became too crowded and could no longer accommodate all the peasants who wanted to come in.

What can Satoshi do? He can:

1. Expand the castle to welcome everyone.

2. Start collecting taxes from the peasants to pay for the soldiers. Collect taxes so that the poor peasants cannot afford to continue living in Satoshi’s castle.

Of course, no peasant wants to pay taxes. They want Satoshi Castle to remain the same utopia it was before. So, the vast majority advocates expanding the boundaries of the castle.

Although Satoshi himself felt uneasy, he decided to listen to the peasants and double the size of the castle every two years.

As time went on, the problem became clearer. As castles grew larger, more powerful soldiers were needed to protect them if they wanted to maintain the same level of security.

As Satoshi's subsidies became less and less, some taxes were inevitable. But the poor farmers could only pay a little tax, and the rich farmers were unwilling to pay more than the poor farmers. Because the soldiers did not get enough money, they started to strike.

This made the castle vulnerable to attack. One day, soldiers from Yellen Castle came, sacked Satoshi Castle and destroyed it.

From this analogy you can see that it is not the increasing taxes (transaction fees) that cause the soldiers (miners) to strike and security to erode, but the increasing block size limit (castle size).

BIP 101

Bitcoin's early adopters are a spoiled generation

After several years of using a decentralized, secure, fast, cheap, and unlimited network, they were unwilling to give up any of these five characteristics. The vast majority of these early Bitcoin adopters did not understand how Bitcoin worked or why the Bitcoin network worked so effectively.

The most secure network cannot be highly decentralized, have huge transaction flow, be fast, and have low fees at the same time.

Something has to give.

Take credit cards, for example. You give up decentralization and low fees, but you gain speed, security, and no limits. Not a bad tradeoff, but that's not Bitcoin.

Compared to the overall cost of a decentralized system, operating a centralized system is more efficient and less expensive. If Bitcoin has 5,000 nodes, the cost of a Bitcoin transaction is 5,000 times that of a credit card transaction.

This is the essence of decentralization.

Because every node has to do the same work as every other node, the fees increase exponentially. The block mining reward pays for today's fees. As the block mining reward approaches zero, transaction fees will need to be paid to offset that fee.

People think BIP 101 can solve the problem, but I think BIP 101 is too dangerous.

As the castle analogy illustrates, BIP 101 values ​​cheapness and unlimitedness over security and decentralization. (BIP 101 has no effect on fastness, but as we’ve seen with replacement-by-fee protocols, fastness doesn’t last forever).

BIP 101 makes the block size grow too quickly. This destroys any chance of fee pressure due to supply constraints.

In order to keep throughput high and fees low, security and decentralization will eventually be sacrificed. This will happen because miners will create large blocks that are not economical for the network.

This will force other miners to leave because it will be unprofitable for them. So, as the block reward decreases, security and decentralization will no longer exist. What happened to Dogecoin will also happen to Bitcoin. In this case, merged mining cannot guarantee the security of Bitcoin.

This is not the Bitcoin I signed up with.

Marginal cost of transaction

Okay, so if we generate a lot of transactions, and each transaction requires paying some fee, will those fees add up to enough to pay the miners to ensure the security of the network?

Unfortunately, the answer is not enough.

The Bitcoin network is a zero-sum system in equilibrium: transaction fees plus block mining rewards must be sufficient to cover the total cost of operating the Bitcoin network.

When the block reward is zero, transaction fees must be sufficient to cover all costs. Bitcoin is designed so that a miner receives a transaction fee, and any miner must pay a fee to relay, verify, and store the transaction.

It is this unbalanced cost-reward relationship that prevents Bitcoin from achieving equal security and decentralization without transaction supply constraints.

For example, if there are 5,000 miners and nodes, the marginal cost for a miner to process a transaction (i.e., verify it, add it to a block and store it) is $0.0001.

The total cost for the entire network is about 5,000 * $0.0001 or $0.50. The average transaction fee needs to be $0.50 for the system to be sustainable.

If the fee is lower than this value, less efficient miners will earn less than their costs, so they will have to stop mining and give up. This will lead to lower security and more centralization.

The only way to ensure that transaction fees are sufficient to pay miners to ensure security and decentralization is to set transaction supply constraints based on the block size limit.

Without a small enough block size limit, miners will mine and include transactions into blocks if the transaction fee is greater than its marginal cost but less than the total cost to the network.

For example, let’s analyze a transaction that incurs a fee of $0.01.

Because it only costs the miner $0.0001 to add the transaction, he will mine it and include it in the block. But the transaction is not cost-effective for the entire network. When mined, it will cause harm to the network.

On the contrary, if there is a block size limit, miners will only mine and include transactions that pay the most transaction fees into blocks. This ensures a healthy balance between transaction flow and transaction fees.

Ultimately, we are just trying to figure out what adjustments should be made to help Bitcoin grow. We need to be especially careful about adjusting the block size limit, as it has the potential to destroy Bitcoin’s security and decentralization.

Bitcoin is not a one-size-fits-all electronic currency

My argument is that we should design Bitcoin with security and decentralization as the top priorities.

Some transactions require the highest level of security and decentralization, which requires paying higher transaction fees to use the Bitcoin network. Not all transactions can afford this fee, so they may not need security and decentralization as much as they should.

That was just perfect.

They can use Litecoin and altcoins, sidechains, payment channels, the Lightning Network, off-blockchain networks, and other networks yet to be created to conduct that transaction.

Heck, they can still use credit cards, if the merchant is willing to pay the fee.

You can use Bitcoin to buy a house or a car. Wait 60 minutes, pay a $1 fee and enjoy a very secure, decentralized and irreversible transaction. How perfect!

If you are buying coffee, need a cheap and fast transaction, and don’t care about security or decentralization, you can use Litecoin, the Lightning Network, a sidechain, or even transact off the Starbucks blockchain.

As long as everything is seamless, users don’t care. Payment networks will triage the best transactions based on what the transaction type requires.

Technologies like on-chain transactions, the Lightning Network, payment channels, and sidechains also allow Bitcoin to be seamlessly, cheaply, and freely converted to and from anything else.

The wallet will hide all the complexity from the user. We are not there yet, but the future is very exciting. Not all transactions will use Bitcoin, but everyone will use Bitcoin.

PS: I currently support scaling Bitcoin via 2-4-8, Segregated Witness, LN, and verification cost metrics.

Original article: http://www.coindesk.com/eating-the-bitcoin-cake/
By Charlie Lee
Translator: yyy
Reward address: 1AZ8NqLgxnhxZSH44V1RhxuUp6Z6a4nhF6
Editor: printemps
Source (translation): Babbitt Information (http://www.8btc.com/eating-the-bitcoin-cake)