Why is the long-awaited Filecoin so exciting? | An article on IPFS

Have you ever had this experience: a friend shared a file link on a network disk with you, but you were too busy to check it in time. When you remembered to check it, you found that "Oops, the page you visited no longer exists"?

The shared file does not exist. It may be that the sharing link is invalid, the file may be deleted, or the network disk service provider has gone bankrupt (for example, there was a large-scale wave of network disk closures in 2016). If it is the latter, it means you may have lost the file.

It is not just the data on the network disk. When we visit some websites, we will also encounter "404 Page Not Found". Website 404 may mean that the file on the server has been permanently deleted, or the server may be shut down. In either case, it means that the webpage you are looking for has disappeared from the Internet.

Content can be permanently deleted, which is an obvious drawback of the current centralized storage on the Internet.

The two major online storage services, iCloud and Dropbox, charge more than 800 yuan per year for 2TB of storage space. Dropbox's 2018 financial report shows that each user paid an average of $117.64. High storage costs are another drawback of centralized storage on the Internet.

On August 23, Amazon's cloud service AWS failed, causing a brief service interruption on many trading platforms in the blockchain industry. As more and more companies host their data and information on cloud servers such as Amazon AWS and Alibaba Cloud, once these centralized cloud service providers have problems or are attacked by hackers, the impact may be greater than many people imagine. Centralized storage will inevitably face security issues.

In addition, the centralized storage of the Internet is not good enough in terms of privacy protection. This is why we need to back up a lot of privacy-sensitive data with mobile hard drives instead of storing them directly on cloud drives.

So, is it possible to come up with a storage method that allows valuable content to be "stored for a long time, at low cost, with good security and strong privacy protection"?

This is what IPFS (InterPlanetary File System) does.

IPFS is called "a new direction for Web 3.0". If successful, it will not only subvert the existing Internet storage model, but also become an important infrastructure for blockchain.

Today, let’s take a deeper look at IPFS and its integration with blockchain technology.

01

What is IPFS?

The full name of IPFS is The InterPlanetary File System, which is translated into "interstellar file system". It is a peer-to-peer distributed file system that aims to create a more open, fast, and secure Internet, supplementing or even replacing the existing Internet HTTP protocol.

IPFS is not a blockchain, nor a project, but an underlying network transmission protocol. As mentioned above, if it succeeds, it will become an important infrastructure of the blockchain.

The emergence of IPFS is inseparable from the HTTP protocol, so let’s first understand the HTTP protocol.

Tim Berners-Lee, the father of the World Wide Web

The HTTP protocol was proposed by Tim Berners-Lee, the father of the World Wide Web, in 1991. It mainly specifies the communication format between the client and the server.

For example, you are reading this article on WeChat on your phone or on your computer (both are clients), but the data of this article is actually stored on Tencent's server. HTTP is a network transmission protocol. As long as you follow this protocol specification, you can read this article stored on Tencent's server on any mobile phone or computer.

It can be said that the development of the Internet is inseparable from the hypertext transfer protocol HTTP, which is one of the reasons why Tim Berners-Lee won the 2016 "Nobel Prize in Computer Science" - the Turing Award.

But with the explosive growth of the Internet, many drawbacks of the HTTP protocol have gradually become apparent.

The HTTP protocol relies on a centralized service network, and the server must be powered on 24 hours a day. If the server is shut down and the webpage is 404, users will not be able to access it. In addition, it will also have the defects mentioned at the beginning of this article, such as high cost, vulnerability to attack, poor security, and weak privacy protection.

In May 2014, Juan Benet, who graduated from Stanford University, proposed the idea of ​​IPFS and established the Protocol Lab in 2015 to develop IPFS. In 2017, the Protocol Lab announced the establishment of Filecoin, a blockchain project based on IPFS.

Unlike HTTP, which relies on a centralized service network, IPFS is a decentralized peer-to-peer hypermedia transmission protocol that integrates peer-to-peer network technology, BT transmission technology, Git version control and self-certifying file system technology.

After reading the above description, you may be confused. Don’t worry, let’s use a simple example to introduce how IPFS works.

02

An example to understand the operation process of IPFS

Suppose Dabai wants to upload a PDF document to the IPFS system. The first step is to add the PDF document to the IPFS client. The IPFS client will perform a hash calculation on the PDF and give a hash value starting with Qm. One of the features of IPFS is that content-based addresses (hash values ​​starting with Qm) replace the domain-based addresses of the traditional Internet.

Then, the IPFS system will make multiple copies of the PDF document and split each copy. Each split part will be stored in a decentralized IPFS network node.

The reason for making multiple copies is to make redundant backups, so that even if some nodes are attacked or data is lost or offline, the files can still be found in other nodes, thus ensuring security. However, it should be noted that IPFS cannot 100% prevent the loss of some files.

If Dabai wants to share this PDF file with Xiaohei, Dabai only needs to tell Xiaohei the address (the hash value starting with Qm). Xiaohei can download this PDF document from the IPFS system through the address.

The download process is similar to BT download, which is to download different parts of the PDF from multiple nodes in the IPFS system at the same time, and finally "join" them back to the original file, because IPFS integrates BT's transmission technology. Therefore, compared with the traditional HTTP protocol, which requires the client to load and transmit data from a centralized server, IPFS has a faster transmission speed and is less likely to cause congestion. Traditional centralized servers are very likely to cause congestion due to too many visitors. For example, when buying train tickets to go home for the Spring Festival, the website will freeze from time to time.

The above is the workflow of IPFS.

It should be noted that IPFS is only a transmission protocol and does not encrypt files, which creates a privacy protection problem. As long as others know the PDF hash address above, they can download Dabai's file from the IPFS network, which is obviously not what Dabai wants to see.

So how do we solve this privacy protection problem? In addition, storing files has costs, so why should the nodes in the IPFS network help you store files?

At this time, the role of blockchain needs to be brought into play.

03

IPFS + Blockchain

IPFS is just an underlying transmission protocol. Blockchain technology can amplify the role of IPFS, and IPFS will become an important underlying infrastructure of blockchain.

This is why the Protocol Lab, which proposed IPFS, needs to establish Filecoin, a blockchain project based on IPFS. The relationship between IPFS and Filecoin is like the relationship between blockchain technology and Bitcoin. The former is the underlying infrastructure protocol, and the latter is an application.

So, how does blockchain technology make IPFS more powerful? Let’s illustrate this with an example.

Dabai first adds the PDF document to the IPFS client and encrypts the document using Xiaohei's public key. IPFS generates a hash value address for the encrypted PDF document, makes multiple copies at the same time, and then disassembles and distributes them in different IPFS network nodes.

Dabai tells Xiaohei the hash value address of the document. After Xiaohei downloads the encrypted PDF document, he can use his private key to unlock the PDF document. Other people on the network, even if they know the hash value address of the PDF, will not be able to open it after downloading it because they do not have the private key. The node storing the document also cannot see the content of the document because it does not have the private key, thus protecting privacy.

So, why are these nodes in the IPFS network willing to help Dabai store this PDF document?

This requires the incentive effect of the blockchain Token economic model.

Taking the Filecoin project, which is specifically designed for storing valuable and frequently referenced files, as an example, miners (nodes) can be divided into two categories: storage miners and retrieval miners.

Among them, storage miners are mainly responsible for storage. For example, Dabai's PDF will be divided into many parts and stored in the hard disks provided by several storage miners. Retrieval miners are mainly responsible for data retrieval. When Xiaohei provides the hash value address given by Dabai to the Filecoin network, the retrieval miner immediately finds out how many parts the PDF is divided into and which nodes they are stored in based on the address, so that Xiaohei can download it quickly.

When Dabai stores PDF in the Filecoin network, he needs to pay a certain number of tokens to the miners, and the system will also give these miners token rewards.

In addition to the reward mechanism, there is also a penalty mechanism. For example, storage miners need to stake tokens. If the storage task is completed, the staked tokens will be returned; if it is not completed, such as the file is accidentally deleted in the middle, or the storage hard disk fails, the system will deduct the miner’s staked tokens as a penalty.

This is the charm of "IPFS + blockchain technology".

05

summary

Jiang Zhuoer, founder of Litecoin Mining Pool, once said: “Bitcoin brings currency freedom, Ethereum brings equity freedom, and in the economic freedom that can be expected in the future, there is legal currency freedom brought by stablecoins and information freedom brought by decentralized storage.”

If IPFS succeeds, it will change the existing Internet storage model and provide storage support for many blockchain projects and DApp projects. This is why IPFS has the potential to become an important underlying infrastructure for blockchain.

However, Filecoin, a blockchain project established by Protocol Lab, which proposed the IPFS protocol, was originally scheduled to launch its mainnet in 2018, but has been delayed until recently when it announced that it will launch a testnet at the end of the year and a mainnet next year.

The vision depicted by IPFS is very beautiful, but whether it can be implemented, whether it can be applied on a large scale, and whether it can change the existing centralized storage model, all remains to be tested by time.