The Odyssey event is suspended, and Layer 2 is not a panacea
Original compilation: 0x711, Block Beats
This article is based on Cinneamhain Ventures partner Adam Cochran's personal social platform point of view, and BlockBeats organizes and translates it as follows:
With the Arbitrum Odyssey event driving up the gas price on the Arbitrum network, there has been a lot of talk about the effectiveness of L2. So it's time to revisit some of the basics behind the current L2.
The first thing to remember is that there are many types of L2, but the basic premise is the same: some elements are removed from the Ethereum mainnet, but The state is pinned to Ethereum while still inheriting the security of the mainnet. In simple terms, we separate state from computation to achieve this. Just like the CPU (computing), memory (activity) and hard disk (long-term storage) in a computer.
The current L2 technology, in a nutshell, mostly separates the computing process from the main network, but the storage remains on the main network. Therefore, this means that on L2, elements such as "calldata" and state storage in smart contracts may have relatively stable gas costs (or higher gas costs in some cases) than on mainnet, But the computational cost decreases exponentially.
Therefore, just migrating the contract may not save a lot of gas, but some optimization can bring great improvement . For example, Aave optimizes V3 for rollup, which optimizes the gas cost by 10 times. As a super simple example, if you have two variables A = 10 and B = 7, it is more efficient to compute (A - B) internally than to make any external calls to estimate the value or store a third state.
Now we can improve the contract not only by call data/storage transfer, when we use call data or storage we can also compress or byte it Pack. This is the same principle that ZIP files compress data on your computer.
But even after all these improvements, there is another bottleneck, we still need to write some state back to the Ethereum main chain, which itself is a fee of. Now rollup technology solves this problem by packaging these transactions. If we aggregate 100 transactions together, we only pay one fee to write the output state to the main chain, so each transaction on L2 pays a small fraction of the fee, saving the overall fee.
However, if there is only one transaction at a time, then they still need to pay the full fee to the mainnet. Currently, most L2s may abstract a way and pay for it out of pocket, but they won't be able to do so forever. So, as long as L2 doesn't hit another bottleneck or have an unoptimized contract doing a lot of transactions, the more transactions they have, the more transaction cost efficient they will be (as opposed to L1).
There are now other bottlenecks including sequencer or solver capacity, custom EVM implementations, and other improvements around data storage and compression, but the core elements already exists.
I think it might be disturbing to see L2 fees spike, especially when mainnet fees are very low compared to last year. But it's important to understand why.
L2s are not panacea, for Arbitrum, their Nitro (which made huge improvements in batching and compression) is already on testnet , it looks like they should do the Odyssey event after Nitro goes live.
To reiterate, Odyssey is running on Arbitrum unoptimized V1 and is running a lot of transactions calling data and storage (bridge) , and the transaction volume is still moderate. In the future, we can expect tx type rebalancing.
Recommended readingArticle by @epolynya, an expert on L2 models, for a deeper understanding of L2.
Welcome to join the official BlockBeats community:
Telegram Subscription Group: https://t.me/theblockbeats
Telegram Discussion Group: https://t.me/BlockBeats_App
Official Twitter Account: https://twitter.com/BlockBeatsAsia
