What is Filecoin pre-mining?

Recently, I read an article by Hu Feitong, "Filecoin Attack 2: Pre-mining Attack". The article explains in detail an "attack" in the Filecoin system: pre-mining. After reading this article, I have a new understanding and cognition of some of the problems that appeared in the Filecoin test network at the beginning, and I have also made some predictions about some possible situations after the launch of the Filecoin main network.

Today, I would like to share this rare and good article with you.

The so-called "pre-mining attack" is explained in the original text as "preparing a large amount of computing power before the network is launched, and then directly loading the computing power when the network is online, running ahead. By the time you go online and initialize, others have already prepared for several days."

The term "pre-mining" is actually not new in the public chain field. Many digital currencies are pre-mined, but in most cases, the project party mines a portion of the digital currency before the main network goes online, and then the miners start to participate in mining after the main network goes online.

Since "pre-mining" has always existed, why does "pre-mining" become an attack in Filecoin? The fundamental reason is that certain features in Filecoin will cause some miners who participate in mining to have an unequal inherent advantage and mine more coins than other miners.

So what factors in the Filecoin system may lead to possible inequality? This depends on the key points of Filecoin mining.

In previous sharing, I have introduced that there is a key data to calculate in Filecoin mining, which is "Proof of Replication". There are three parameters required to calculate "Proof of Replication":

1. Miner actor address

2. The data itself

3. sectorID

Among these three data, the first and third ones can be created in advance by miners before the test network goes online, while the second one can be prepared in advance by miners. So in the end, these three parameters can be created by miners themselves.

Therefore, miners can create all the data and wait for the test network to go online. Once the starting gun is fired, they can pile computing power into the system and start mining.

The Filecoin project believes that this is a system problem and has proposed a solution to this problem: introducing random numbers. This random number is generated after the mainnet is launched and cannot be predicted by miners. Therefore, in this case, miners cannot pre-mine.

But in my opinion, the article does not show that “pre-mining” is unfair to the nodes participating in mining.

The article quoted the words of the Protocol Labs project team to describe "pre-mining". Judging from the date of the citation, the project team publicly talked about the issue of "pre-mining" as early as August 2019, 3 to 4 months earlier than the test network launched on December 12.

So I believe that all miners who are interested in participating in the Filecoin testnet mining should know this "pre-mining" trick. Since everyone knows this trick, they will definitely make preparations in this regard.

If everyone is ready for pre-mining, it means that everyone is at the same starting line. In the end, who can grab the pre-mining will depend on who has better technology.

A sentence at the end of the article says: "Will there be pre-mining when the mainnet goes online? I don’t know, but it seems so. The network always needs computing power to maintain, just don’t make it too high."

What this means is that it is very likely that the mainnet still has the problem of pre-mining. So in order to prepare for pre-mining before the mainnet goes online, various mining service providers will definitely vigorously attract investment, expand equipment and systems, and actively prepare for the battle!