As soon as I woke up today, it was very lively to watch. Then I saw Ethereum: 0x75231f58b43240c9718dd58b4967c5114342a86c surge by 15%. It should be because @ okxchinese has a new move. To be honest, I carefully read the content of Exchange OS and didn't quite understand it. Of course, I know it's an on chain exchange with spot and futures, including predictive markets. But what exactly is this? I have checked a lot, including the white paper, but I didn't understand it very clearly. I only know that the matching engine, margin management, clearing, risk control, settlement, etc. are directly placed in the blockchain protocol layer, and then divided into two layers, namely X Layer EVM and X Layer TradeZone. That is to say, a combination architecture of layer one and layer two networks is used, with the X Layer EVM underlying Ethereum's zkEVM Layer 2, and the technical framework based on Polygon CDK. This layer is responsible for on chain assets, contracts, deposits and withdrawals, pledges, governance, and final settlement. Users deposit assets into contracts, which record the balance, and then withdraw from the contracts after the transaction. The essence of asset custody, deposit and withdrawal, final settlement, and governance execution at this level is the same as that of most DEXs or PerPs today, which are all infrastructure. The complex one should be X Layer TradeZone, which is responsible for matchmaking, margin, liquidation, and liquidation PnL、 The white paper does not explicitly state what framework is used for fund rate settlement, but it provides four main cores, namely Margin Engine, Liquidity Engine, and Settlement Engine. There is nothing new about them, and the most difficult one should be the Matching Engine's matching engine. But the Matching Engine did not provide a very detailed introduction. If OKX were to establish its own exchange, the matching system would only serve OKX itself, with rules, assets, risk control, and liquidity all controlled by OKX. But the goal of Exchange OS is to serve the entire market, which means that the Matching Engine cannot only serve a single product, but must become the underlying matching capability shared by multiple markets. The difficulty here lies in the fact that different markets have different assets, different oracle machines, different leverage, different risk parameters, and different liquidity depths, but the underlying matching and account systems still need to be unified. This should be quite difficult, but if it can really be perfectly achieved, there is great potential. Then there is the pledge of OKB, which is understandable, but the actual purpose of OKB pledge is not very clear. However, it can be imagined that in the future, a large number of people will borrow "OKB" because the purchase cost is too high, but borrowing is relatively simple. If a "crowdfunding pool" for borrowing OKB can be realized, it will be more interesting. Is Spring Coming for Short Bird Brother @ Wangduanniao?