Original text: "Several Major Misunderstandings of the Public about Blockchain"

Author:ChainLinkGod

Blockchain is an innovative computing model. The security of data and applications is not guaranteed by a centralized third party, but by a decentralized computer network. Because the blockchain itself is permissionless, anyone can join the network and independently verify the authenticity of the calculations. Ultimately, blockchain technology establishes a power checks and balances mechanism between users and network nodes.

Blockchain technology has brought us innovative digital currencies (such as Bitcoin) that can write monetary policies into code; it has also enabled programmable trust-minimizing applications (such as Ethereum). Currently, blockchain is rapidly entering the public eye as a new digital infrastructure, and this development momentum is also called Web3. Although many people recognize the value proposition of blockchain, there are still some misunderstandings about the underlying trust model of blockchain.

This article will explore the role of full nodes with autonomous verification capabilities, especially how full nodes supervise block creators (i.e. miners or validators), and the key impact of reducing verification costs on the scalability of blockchain. After reading this, you will have a deeper understanding of the differences between blockchain and traditional computing in terms of trust model.

Bottlenecks of traditional computing models

Most applications today use a client-server model that sends requests to a centralized database server. There are several types of participants in the traditional computing model:

Client: The end user usually creates a request through a personal device, and the request is sent to a centralized data server through the TCP/IP protocol. The client trusts that the database server will faithfully respond to the request. Database server: This is a remote computer, usually a centralized cloud service provider, that provides data or specific services. The back-end database server is usually a group of computers managed by a traditional organization. In the traditional computing model, although the centralized server lacks transparency, users can only trust it.

This computing model is also known as Web2, where developers can quickly publish and expand applications according to their needs. Since data centers can ensure low latency and high throughput of computing, a very smooth user experience can be achieved. Applications deployed on centralized servers provide huge value to billions of users around the world, but they also have some bottlenecks.

The biggest bottleneck is that end users cannot verify whether the calculation process of the response request is reliable, nor can they confirm whether the data entered into the database has been manipulated. In fact, such applications are operated in a dark box, and users must fully trust that the third party will not harm their interests.

This trust assumption of centralized applications may have a negative impact on users. Users may encounter censorship, have their accounts deleted by social platforms, have their personal data leaked, have content recommended by algorithms, have higher reconciliation costs, have their data tampered with or lost, and even lack accountability in the entire system. Ultimately, this will lead to a collapse of social trust and increase economic coordination costs.

Blockchain computing model

Blockchain breaks away from blind trust in centralized entities and adopts the following technical mechanisms to achieve trust minimization and credible neutrality.

Encryption technology is used to authenticate the ownership of data/assets and verify the authenticity of transactions; a decentralized consensus mechanism is used to sort transactions and execute protocol rules; an economic incentive mechanism is used to ensure the immutability and robustness of network ledgers.

It is worth mentioning that blockchain is an open network that anyone in the world can join, and can actively contribute or passively monitor messages. End users can personally verify whether the results of the network output are valid and whether the data in the ledger has been tampered with. Because blockchain is transparent, it also reduces the risk of information opacity, eliminates information asymmetry, and ensures that all participants are fully aware of the risks.

In the blockchain computing model, end users can choose to join the blockchain network and verify the calculation results.

In order to ensure the normal operation of blockchain, different types of network participants are required. One participant often plays multiple roles, and each role plays a key role.

Block creator (BP): responsible for sorting transactions, packaging them into blocks (note: this is a special data structure), and then submitting the blocks to the network for verification. If two valid blocks are created at the same block height, the block creator will decide which version of the blockchain is "orthodox" (i.e., follows the longest chain principle). A consensus mechanism is used to decide which block creator will create the next block. The two most popular consensus mechanisms are PoW (miner) and PoS (validator). Full node: Full nodes are the core of the blockchain network. This type of node is responsible for downloading and verifying each block submitted by the block creator. If the block is verified to be valid (i.e., it complies with the protocol rules), it will be added to the personal ledger of the full node and the state change will be executed. Invalid blocks that do not comply with the protocol rules will be ignored and discarded, and no state changes will be executed. Archive node: Archive nodes not only store all the information in the full node, but also calculate and store the previous state of the blockchain. Archive nodes can be used to request any historical data, such as the balance of a user account at a certain block height in the past. One thing to note here is that a full node can be converted to an archive node at any time without downloading any additional data from the network. Archive nodes are usually very demanding hardware and are usually run by service providers (such as block explorers). Light Clients: Light clients are full nodes with limited functionality that only download block headers (i.e., unique cryptographic fingerprints). Light clients can verify that a transaction is included in a block, but because they do not download or execute all transactions in the block, they essentially need to trust that most block creators are honest. Light clients are also called "Simplified Payment Verification" (SPV) clients and were originally proposed in the Bitcoin White Paper. RPC Providers: RPC Providers are also full nodes that can connect other network participants to the blockchain and read and write data. Users of RPC (Remote Procedure Call) services often do not have the ability or willingness to run full nodes or light clients themselves, so this service can greatly reduce the barrier to accessing the blockchain.Users must trust that the RPC service provider can provide services honestly, because users cannot verify the results themselves. End users: End users are ordinary users who want to conduct transactions in the blockchain network. Such participants may run full nodes or light nodes, or they may access RPC service providers. The only purpose of the blockchain is to provide services to end users.

Although each type of participant plays a key role in the smooth operation of the blockchain ecosystem, full nodes are undoubtedly the core of the blockchain network because they are responsible for maintaining a complete copy of the network ledger. Other participants, such as block creators and light clients, connect to full nodes to keep abreast of the status of the network. Therefore, the topology of the blockchain network is shown in the figure below.

Schematic diagram of the topology of a blockchain network, where light clients and block creators are connected to full nodes. Block creators have limited power.

A common misconception about blockchain is that since block creators are responsible for adding transactions to new blocks, they have unilateral control over the blockchain. In other words, these people believe that if a majority of block creators are bribed or collude (note: this is often also called a "51% attack"), the network will collapse and everything of value will be completely destroyed. In reality, the power of block creators is limited, so only a few types of malicious attacks can occur.

Block creators have the following powers:

Censoring transactions: Block creators can decide which transactions to include in their blocks, and can also selectively ignore certain transactions. If block creators collude, it is indeed possible to exclude certain transactions, but in fact, only one honest block creator is needed to include transactions in a valid block. Rewriting the blockchain: If a majority of block creators collude (such as a 51% attack), then the previously created valid blocks may be forked due to block reorganization (reorg), and a new blockchain will emerge as the main chain. It should be noted here that some blockchains will establish a checkpoint mechanism to implement the concept of "economic finality". Once the blocks before the checkpoint are reorganized, the blockchain will be suspended, which requires social consensus off-chain to resolve. Block reorganization will roll back the previously accepted valid blocks and replace them with a new set of valid blocks.

Block creators do not have the following abilities:

Changing protocol rules: Block creators cannot modify protocol rules at will, such as adding/removing functions (such as opcodes); increasing/decreasing the upper limit of block space or the frequency of block generation; and increasing/decreasing block generation rewards. Stealing user funds: Block creators who do not hold the wallet private key cannot spend the tokens in the wallet because the cryptographic signature of the transaction will be invalid. It should be noted here that blockchain reorganization can lead to double-spending attacks, which delete transactions signed by the private key holder in the blockchain ledger instead of forging transaction signatures.

To understand why block creators are limited in their capabilities, we must first understand the relationship between block creators and the full nodes responsible for verification.

How do full nodes supervise block creators?

The responsibility of the block creator is to submit new blocks to ensure the smooth operation of the blockchain. However, the block creator does not have the right to decide whether the block is valid on behalf of the entire blockchain network. The full node is responsible for determining the validity of the block. The full node stores the blockchain ledger, is responsible for independently verifying each block, and executes all transactions in the block to confirm whether the block complies with the protocol rules.

Therefore, a power check and balance mechanism is formed: full nodes only accept valid blocks and ignore blocks that do not comply with the protocol rules, so as to supervise and hold block creators accountable. Block creators are economically incentivized to only create valid blocks. Because if they create invalid blocks, they will waste a lot of economic resources and give up all block rewards and transaction fee income. Even if other block creators continue to create valid blocks on an invalid block, these blocks will be ignored by full nodes, and these block creators will eventually fork off the network.

Full nodes are run by key economic entities in the blockchain ecosystem, including cryptocurrency trading platforms, stablecoin issuers, oracles, L2 validators, RPC service providers, custodial wallet service providers, whales, and ordinary users. These key economic entities independently verify the blockchain ledger, so they will not be deceived into accepting invalid blocks. Full nodes do not need to coordinate and communicate with each other when confirming the validity of blocks. These nodes only need to follow the protocol rules when verifying blocks to reach a consensus on the validity of the blocks.

Invalid blocks will be rejected by full nodes, and full nodes will continue to track valid blockchains

The protocol rules for many blockchains are decided off-chain by the community, a process also known as "social consensus." Reaching social consensus usually requires a formal process (such as Ethereum Improvement Proposals, or EIPs), where the community discusses the benefits of protocol rule changes on various social platforms. Once an initial consensus is reached off-chain, the blockchain's core developers will develop a new version of the blockchain client to implement these changes. Full nodes and block creators can choose whether to run a new version of the software and upgrade the network through a hard fork (non-backwards-compatible update) or a soft fork (backwards-compatible update).

If a hard fork upgrade is controversial, the blockchain network may split into two different versions, and the two versions will run separately. Once this happens, each blockchain will share all the history from the genesis block to a certain block height before the fork. This is what happened with the hard fork between Bitcoin and Bitcoin Bash in 2017, when people disagreed on how to increase transaction throughput. This was also the case with the hard fork between Ethereum and Ethereum Classic in 2016, when the community disagreed on how to deal with the DAO hack.

Bitcoin Cash is a fork of the Bitcoin blockchain that emerged because of a disagreement over how to increase transaction throughput.

Some blockchains do not adopt an off-chain consensus mechanism, but choose to handle upgrades through on-chain governance, where token holders govern a decentralized autonomous organization (DAO) and decide on protocol rule changes. Each protocol upgrade method has its own unique advantages and challenges, and which solution to adopt depends on the specific needs of the community.

Reduce the hardware requirements for full nodes

If the hardware requirements for full nodes are too high (for example, industrial-grade hardware must be used), then not many entities will be able or willing to run full nodes, and they may choose centralized solutions to read and write data to the blockchain. This will make the blockchain network more susceptible to manipulation by block creators.

In order to build a healthy full-node network, many blockchains will deliberately limit the network's transaction throughput to be compatible with the lowest-configuration verification node hardware in the community. Lowering the hardware requirements not only allows full nodes to operate normally under normal circumstances, but also allows new verification full nodes to be quickly built when the blockchain is attacked by a group of colluding block creators. The specific hardware requirements of each blockchain are different, but they are basically compatible with ordinary consumer-grade hardware.

A full node of the Bitcoin blockchain running on a Raspberry Pi, a low-powered microcomputer.

The following logic illustrates how reducing hardware requirements will enhance blockchain’s core value proposition:

Lower hardware requirements → Lower verification costs for blocks → More entities can run full nodes → Better protection against attacks by block creators → Improved network security → Better reliable neutrality → Increased usage

Although the blockchain protocol does not explicitly specify the hardware requirements for full nodes (because the operating efficiency of hardware and software is constantly improving), it does specify transaction throughput (Note: transaction throughput usually depends on block size and block frequency) to specifically target a certain type of hardware configuration.

While most blockchains are compatible with common consumer-grade hardware, some require the hardware of the verification nodes to be at least server-grade to increase transaction throughput. These blockchains realize that although this will reduce the number of self-built nodes participating in verification, there are still enough key economic entities with sufficient resources to run full nodes.

Reducing blockchain verification costs

An ideal scenario is that everyone in the world runs their own full node to verify blockchain transactions. However, in reality, building a full node takes a lot of time and resources, and may affect the user experience of many end users. As blockchain becomes more and more popular, users who are accustomed to using centralized service providers may find it too troublesome to build their own full nodes, and they prefer to interact with on-chain applications immediately.

Therefore, most of the end users on the mainstream blockchain do not build their own nodes, but connect to RPC service providers and trust these service providers to provide reliable services, send user transactions to the blockchain network, and synchronize the current status of the blockchain to users. These users may not care about the benefits of blockchain self-verification, and the security of the blockchain network has not reached the ideal state. However, compared with the traditional computing model, blockchain still has a huge advantage in trust minimization.

Nowadays, many end users do not run full nodes themselves, but instead connect to trusted RPC service providers.

There is a compromise solution to this problem, which is to integrate light clients into blockchain wallets. Although integrating light clients can reduce the dependence on RPC service providers and thus improve the current situation to a certain extent, the light client model still fundamentally requires trusting that the majority of block creators create valid blocks and that all transaction data in the blocks are public.

In order for a light client to verify the block creator's block without downloading and executing the entire block (i.e. running a full node), the light client needs to be technically optimized to improve the level of trust minimization and become a "trust-minimized light client". Technical optimization can be carried out in the following aspects:

Fraud Proof: The full node generates a cryptographic proof and sends it to the light client, proving that the block is invalid and violates the protocol rules. At the same time, the light client does not need to re-execute all transactions in the block. Validity Proof: In addition to the fraud proof, a zero-knowledge proof is also generated, and the block creator can directly prove to the light client that the block is valid and complies with the protocol rules. Data Availability Sampling (DAS): This is a probabilistic sampling technique that can verify that all data in the block has been submitted to the network without the light client downloading the entire block. Trust-minimized light clients can verify the block creator's blocks without downloading or executing the entire block.

As new trust-minimizing technologies are released, validators can more easily verify the blocks of block creators without downloading and executing the entire block, and the verification cost is lower. This will help improve the security of the entire blockchain network. It is worth mentioning that the creation of trust-minimizing light clients can further give birth to more scalable blockchain networks.

Scaling blockchain while maintaining self-verification

We usually equate blockchain expansion with increased transaction throughput. However, a more comprehensive understanding should be that blockchain expansion requires both increased transaction throughput and low verification costs. Fortunately, trust-minimized light clients have emerged, which can increase the throughput of blockchain networks (such as increasing block capacity or shortening block time) while keeping block verification costs at a low level.

In addition, L2 scaling solutions such as rollup also utilize technologies similar to trust-minimized light clients, using cryptographic proof technology to prove the validity of transactions, and the underlying blockchain network does not need to execute every transaction. The underlying blockchain is a recognized trust-minimized light client for L2 rollup. In addition, transactions are packaged and stored on the blockchain to ensure data availability, which can greatly increase transaction throughput and hardly increase any verification costs on the underlying blockchain.

L2 rollup uses the underlying blockchain as a recognized trust-minimized light client to improve transaction throughput Summary

Blockchain is an append-only ledger that is very secure, reliable and neutral, bringing us a new computing paradigm. Blockchain establishes a power check and balance mechanism between block creators and full nodes, which can provide services to the community, continuously upgrade and iterate, while maintaining its tamper-proof properties. Although we still need to continue to work hard to further reduce the verification cost of blockchain, blockchain currently has a huge advantage over traditional computing models in terms of trust minimization.

——

The opinions expressed in this article are solely those of the author and do not necessarily reflect the views of the Chainlink Foundation or Chainlink Labs.

END