"Testnet Rewards Program" is coming! Analyzing Filecoin usability from the perspective of TPS

Driven by events such as "cloud computing power, the launch of the Filecoin mainnet, and the computing power war among major miners", Filecoin has become the third hot spot in the blockchain industry in 2020 after DeFi and ETH2.0.

At the same time, Filecoin's grand vision of "providing peer-to-peer storage services for billions of files around the world" has also given many adventurers huge room for imagination.

On July 31, the official announcement stated that in any case, the test network reward plan was decided to be officially launched on August 12. The 4 million reward plan worth 460 million is finally here, which once again attracted huge attention.

However, there are many different opinions about Filecoin, including doubts, such as the "small movie theory" proposed by Jiang Zhuoer, the founder of the Litecoin mining pool, and the "storage efficiency theory" proposed by Yang Haipo, the CEO of the ViaBTC mining pool. There are also affirmations, such as those from early Filecoin investors such as Xie Dapao. Who should we listen to? As an observer, how can we sort out the chaos and see the essence?

Reset 5 times, updated 6 times, modified 5 times,

The testnet rewards program is finally here!

On July 31, Filecoin officially announced that although the testnet reward plan may be delayed, the reason is that it hopes that more capable miners can meet the corresponding standards, that is, most large miners can achieve a transaction success rate of more than 75%. However, if most large miners fail to meet the above standards, the official will still start the testnet reward plan for qualified miners on August 12.

In short, don’t wait for those unprepared miners, the real testnet reward plan must be launched on August 12!

The testnet reward program invites miners from all over the world to compete for global and regional rewards totaling 4 million Filecoin tokens. At the current price of FIL tokens, 4 million tokens are worth 640 million - this is probably the largest "candy distribution plan" in history.

Looking back at the history of Filecoin testnet, we entered the first phase of the testnet on November 3, 2019, and the second phase of the testnet on May 15, 2020. Since the start of the first phase of the testnet, there have been 5 resets, 6 updates to the proof scheme, and 5 modifications to the economic model.

Just on July 24, the test network experienced a crash. According to analysis by industry professionals, the crash was caused by a "computing tipset state failed" failure, which means that the state of the current highest "block set" could not be calculated.

It is worth noting that this problem had been discovered and a patch submitted by the official team on July 20, four days before the failure occurred, but users were not notified to update the patch - which was really careless.

So why so much testing and adjustment?

The fundamental purpose is to simulate the real environment of the real mainnet as much as possible, discover various problems in advance, such as code vulnerabilities, rationality of economic models, etc., and make corrections - if there is a problem on the testnet, it can be restarted and repaired. If a serious code vulnerability appears after the mainnet is launched, it will be hard forked at the least, and directly "failed". If all problems are solved on the testnet, the mainnet can be officially launched.

In any case, after going through so many trials and tribulations, the testnet rewards are finally here!

However, in response to numerous doubts, especially Jiang Zhuoer’s “small movie theory” and Yang Haipo’s “storage efficiency theory”, on the one hand there is the hot Filecoin mainnet launch and the skyrocketing FIL tokens, and on the other hand there are doubts, how to choose?

"Small Movie Theory" and "Storage Efficiency Theory"

Let’s put aside the complicated cloud computing power advertisements, nor talk about the dazzling mining equipment, let alone how miners should arrange mining machines and mining farms. Today we will only talk about one issue: from a data perspective, look at the availability of Filecoin, as well as the "small movie theory" of Jiang Zhuoer, the founder of Litecoin Mining Pool, and the "storage efficiency theory" of Yang Haipo, CEO of ViaBTC Mining Pool.

On May 27 this year, Jiang Zhuoer said on his personal Weibo:

"Many people have asked about Filecoin, let me tell you about it. According to our evaluation, Filecoin's performance is very poor at present, not enough to store small movies, and the economic model is wrong (miners rely on some GPU computing, not storage to get rewards), so we have no plans to participate in Filecoin mining. Currently, Filecoin mining machines are basically leeks, so everyone should participate with caution. If you want to mine, just build your own machine."

What is a short film?

In fact, it is a short video, such as the short videos on platforms such as Tik Tok and Youtube.

On June 22, Yang Haipo also expressed his personal views through his personal Weibo. He believes that decentralized storage is a false demand. The fundamental reason is that the capacity and efficiency of blockchain are inherently low, and not everything can be stored.

So, are the "short film theory" of Jiang Zhuoer, the founder of Litecoin Mining Pool, and the "storage efficiency theory" of Yang Haipo, the CEO of ViaBTC Mining Pool, correct?

Of course, there are many perspectives to examine this point. Let us temporarily stand from the perspective of TPS (system throughput) and analyze it through data - the data speaks for itself.

From the perspective of TPS,

The rationality of the theory of watching short films

As early as 2018, Protocol Labs defined Filecoin as "a distributed network where data is not restricted by Internet companies."

It’s a multi-trillion dollar business. More than 2.5 quintillion bytes of data are generated every day, including 294 billion emails, 500 million tweets and 64 billion WhatsApp messages. That doesn’t include climate science reports, disease tracking maps, connected car coordinates and more.

Behind these data is extremely high TPS support.

What would happen if platforms like Youtube and FaceBook were moved to Filecoin?

For Youtube, users upload an average of 500 hours of video per minute and watch 1 billion hours of video every day. The average time a person spends on Youtube is 11 minutes and 24 seconds per day. These huge amounts of data can be quantified as TPS, or visits per second.

How high will the TPS requirement be?

We know that Bitcoin has a peak of 7 confirmed transactions per second. The calculation method is: each Bitcoin block is 1MB, and the average transaction size is 250 bytes, so only 1024*1024/250=4194 transactions can be recorded. It takes an average of 10 minutes to generate a block, so the number of transactions per second is 4194/10/60=6.99 transactions, which is why Bitcoin TPS is 7.

For the TPS of Youtube and FaceBook, we use the simplest calculation method. According to the data from similarweb.com, the total page and app visits of the two websites in June were 31.05B and 24.48B respectively. So, with simple calculation,

Youtube's TPS is: 31.05B/30 days/24 hours/60 minutes/60 seconds=11979;

Facebook's TPS is: 24.48B/30 days/24 hours/60 minutes/60 seconds = 9444;

That is to say, Youtube had 11,979 visits per second in the past June, and FaceBook had 9,444 visits per second in the past June.

So, what will the TPS be after the Filecoin mainnet goes online?

The official deliberately avoided this data. Based on a block rate of 25 seconds, an average of 5M allowed for each block, and an average of 250 bytes for each record, the maximum TPS value is 5*1024*1024/250/25=839, and about 1/4 to 1/2 of this needs to be used by the storage proof itself, so the actual supported TPS value may not exceed 600.

This value is less than 1/10 of the current demand of Youtube and FaceBook respectively. This is the peak value of Filecoin and the average value of Youtube and FaceBook. If an emergency occurs, the traffic of Youtube and FaceBook is likely to increase tenfold.

Therefore, from the perspective of TPS, Jiang Zhuoer and Yang Haipo’s “Small Movie Theory” and “Storage Efficiency Theory” are very correct!

However, we are living in a world that is changing rapidly. Will Filecoin, like ETH, gradually transform from the "world computer" to the foundation of decentralized finance under the blessing of DeFi? Will Filecoin have such a transformation? After all, everything is just beginning.

Inspiration from ETH development path:

Maybe we are all wrong!

When interviewing a Filecoin industry leader, he gave the following explanation for the theory of comparing the TPS with Youtube and FaceBook:

“Filecoin’s market positioning is different, just like Ethereum wanted it to be the world’s computer, and now it has found its position as a DeFi platform.

In fact, for Ethereum, its "world computer vision" was proposed at the very beginning.

On July 31, 2018, when Ethereum was three years old, Vitalik Buterin elaborated on his concept of "world computer" in an interview.

Vitalik Buterin said:

I continue with the analogy of a "world computer". The idea is that the blockchain as a whole functions like a computer. It has a hard drive, and on the hard drive, all the accounts are stored.

It stores the code for all smart contracts, accepts incoming instructions — which are signed by many different users — and processes them according to a fixed set of rules.

To put it simply, it is similar to a global operating system that allows DApp to run on it.

However, in the subsequent development, DApps based on Ethereum could not escape fraud, Ponzi schemes, gambling, etc., so investors were disappointed. As a result, Ethereum was shorted and sold by a large number of institutions at the end of 2018, causing the price of Ethereum to fall to as low as US$80 per coin, and market sentiment was extremely low.

After more than three years of wandering in the wilderness, the Ethereum community has finally found the promised land - DeFi. The current prosperity and popularity of the DeFi ecosystem is self-evident. Whether from the perspective of business logic or reality, DeFi has enabled Ethereum to form an on-chain closed loop, and even allowed the entire blockchain industry to finally see the "dawn" after years of wandering.

Even Ethereum’s two co-founders, Vitalik Buterin and Lubin, admit that Ethereum is becoming the Global Decentralized Finance Ledger, which is a good thing for Ethereum.

So, can Filecoin find its accurate position in the process of development? Perhaps only time can answer this question.