Limitations of Ethereum 2.0

I don't have complete knowledge about L2. However, let me tell you my impressions after reading various materials on Ethereum 2.0.

The two main purposes of Ethereum 2.0 are the environment and gas costs. As for the environment, the purpose was achieved by the POS. However, the gas cost issue does not seem to be very satisfactory.

In the announcement on Ethereum's official site, there is a word that always appears regarding gas costs. It's the words, "without sacrificing security." This means, "We will lower gas costs without sacrificing security." But this is an impossible story. Lowering gas costs will inevitably lead to loopholes in security. This is an inevitable result because gas costs were introduced as a security tool in the first place. The only way to lower gas costs without sacrificing security is not to use gas costs themselves as a tool for security.

Gas cost and security are concepts that have no relationship in their nature. Therefore, using gas costs as a security tool is quite expedient. The expedient can never last forever. Eventually, it causes problems. It is a bandage attached to the bullet wound.

Ethereum-related sites boast that Ethereum is now heading for 100,000 TPS. However, under the existing gas cost system, this is only a meaningless announcement. Imagine a case where A transfers 1 wei to B and executes such 100,000 TXs per second with a gas price of 1 wei. Even if you do this 100 million times, it only costs 20 cents in total.

With L1, the TPS was about 15 to 20. In this state, even if TX is executed by choosing a gas price at 1wei it will wander around only in memory (mempool) and eventually disappear. However, when the TPS reaches 100,000, TXs with a gas price of 1wei start to be contained in the block.

To prevent this situation, Ethereum Foundation has no choice but to come up with various methods. But in the end they are bound to be expedient. The reason why they can only be expedient is that it is just expedient to use gas costs as a security tool in the first place. For reference, Clayton, one of Korea's mainnets, set gas costs 30 times cheaper than other mainnets, but raised gas costs 30 times again when it was attacked. This case is very important lesson for Ethereum. Problems arise in any form. Whether direct or indirect, L2 will eventually have no choice but to artificially raise gas costs.

The above examples lead to the conclusion that gas costs should no longer be used as a security tool. In other words, gas costs should be used as gas costs, and separate security tools should be used for security. Fortunately, it is possible. It is to change the name of the existing gas costs to "security deposit". The reason why the name should be changed is that gas costs are criticized if they are set at a high price, but "deposit" is not criticized even if the amount is large. In other words, all you have to do is return the deposit you received for TX, which was found not to be an attack. This can reduce the gas costs even to zero. At the same time, the garbage disappears on the whole node.

Currently, gas costs are mixed with gas and security costs. They should be refined. Then all the problems are solved. Technically, it is possible. The technical difficulty is less than a tenth of the change from L1 to L2.