Glad to see this is being critically looked into. Kamino started rolling out multi-feed oracles with max deviation, max age checks and reference checks recently. For example an oracle can be the earliest of two feeds (Pyth and Chainlink) as long as both are within 5% of each other and not 5% further than another reference feed (a fixed price or a more centralized feed which hasn't yet been battle tested). This should work 99.9% of the times, which is better than 99% of the times when one of the two feeds can go down, but not as good as 100% of the times since it requires some intervention if both prices are bad (but in that case it's really exceptional and requires judgement). This is on top of price bands (where it makes sense), non-zero checks and twap checks.
The decentralization review of @aave by @defiscan_info went live last week and was discussed in @DeFi_FR @lemiscate chipped in 4 times, never tackling the report, but instead attacked its publishers: a classic deflection attempt Why? Let me dive into what he wants to bury Aave's review is simply the biggest one ever produced for @DeFiscan_info. Because of Aave's codebase's size and relative complexity, we are talking about months of effort. Every finding was verified several times, and experts on the codebase, like @The3D_, were consulted before publication. The decentralization analysis concluded that Aave is a Stage 0 protocol, just like Compound, and for similar reasons. The main lack here lies in three dimensions of the analysis: autonomy, exit window, and upgradability. The autonomy dimension assesses the protocol's critical dependencies, which could seriously affect its users if they were to fail. For a lending product like Aave, the star culprit is the oracle(s) used—ChainLink. "The protocol does currently have limited validation on asset prices provided by Chainlink. These checks include upper caps for stablecoins and LSTs and a sanity check that the price is above 0 for all assets. If the reported price by the price feed was below 0, a fallback oracle would be queried. Aave has currently no fallback oracle price feeds instantiated. As a consequence if the price was equal to or below 0, user actions on the Pool contract that require a price would revert." ==> Aave uses ChainLink data with barely any validation and no fallback mechanism ===> It inherits all Chainlink-related risk (an upgradeable system without decentralized ownership over those permissions.) For the two other dimensions, I'll keep it simple and let you read the report, because today, I'd like to highlight @lemiscate's manipulation tactics as well. Aave gets a High Risk on the Upgradability and Exit Window dimensions because some emergency features are without a timelock (such as pausing markets), enabling potential manipulations, and several multisigs do not adhere to the security council criteria: The report, of course, does not fail to highlight what is worthy of praise in Aave's current model, such as their redundant and fault-tolerant implementation of cross-chain governance with a.DI. A precise list of what needs to be done to reach the next stage is also provided for each item at fault. --- Now, let's take a minute to analyse @lemiscate reactions to this report: In order: 1. Marc: "The two people at the end of the room who still care about this kind of stuff are shocked" ↪ Plenty of actors care about contextualized information, including savvy degens, institutionals, etc. A classical fallacy: you are not popular, and thus you are wrong. 2. Marc: "@bluechip_org gave A rating to LUSD and D to GHO, and still look at their market cap." ↪ @bluechip_org evaluates the economic model of stablecoins, not their potential for growth. This is a classic case of pretended confusion (Marc knows this) to discredit the report altogether without even discussing it. Another classic: don't like something? Discredit it by misinterpreting its goals on purpose. 3. Marc: "BOLD was supposed to kill GHO? It was so superior as a concept" ↪ It is. GHO is a caveman stablecoin with no redemption, a manually managed interest rate, and uber-centralized peg maintenance. No matter how many millions Aave splurges on GHO, BOLD will eventually prevail. However, growth/market cap is not the topic here; we discussed decentralization. Seeing Marc dance on the grave of BOLD's first iteration as the relaunch was announced was surprising, to say the least. I assumed he was pro builders and DeFi. The reality is that he is pro-Aave, and if he has to damage the DeFi ecosystem to grow Aave further, I am now convinced that he would do it without hesitation. 4. I mentioned that "Emilio from BGD double-checked the review", Marc attacked again on the fact that it is now Avara and not BGD. ↪ You're starting to see the pattern? Argue, discuss, and attack everything around so that the core (the review) is ignored, discredited, and most importantly, never addressed. Marc has commented 4 times about the @defiscan_info review in @DeFi_FR, but still hasn't said anything about the content itself, apart from calling it a "brain fart". It's a usual tactic of Marc that when faced with a topic he dislikes, he will tackle everything around it and resort to personal attacks on the poster and attacks on the reputation of his projects. It was exactly the same when @bluechip_org published the $GHO review, again with the exact same twisted argument that the Bluechip rating (which assesses economic safety) is not correlated with the stablecoin market cap. I'm simply at a loss that such behavior can be tolerated and even supported in our industry, and I wanted to share it with you. Those are classic bully tactics seen in many profiles in our space, and the only way to eventually improve is to call them out. I have the luxury of not caring a second about my "employability", allowing me to stand my ground and my values even if that could be deemed risky for my career. Indeed, I realize this is an exception rather than the norm, and that's one reason I wrote this post and led the charge today. Indeed, Marc is an investor in many projects, which could explain why many are shy to denounce his abusive behaviour. Marc and I obviously have a history, as I am one of the few people in the space willing to tell him when he's out of place, which happens often. Since I started doing this, he's relentlessly attacking my character on random occasions, as if he were spreading poison as an insurance policy to protect his past, present, and future abuses. I remain baffled by his behavior, which often exhibits adolescent characteristics. During one of our last altercations, he literally flexed that "he has a girlfriend, and thus is not available to answer, unlike me, the ugly, lonely nerd" (and as often with Marc, he assessed this without any knowledge of my situation). This behavior is net damage for the whole DeFi space, as Marc is not a random person but someone with influence that many still look up to. It's up to us to hold our champions to the highest standards, so here I am, doing exactly that.
8,933
0
本页面内容由第三方提供。除非另有说明,欧易不是所引用文章的作者,也不对此类材料主张任何版权。该内容仅供参考,并不代表欧易观点,不作为任何形式的认可,也不应被视为投资建议或购买或出售数字资产的招揽。在使用生成式人工智能提供摘要或其他信息的情况下,此类人工智能生成的内容可能不准确或不一致。请阅读链接文章,了解更多详情和信息。欧易不对第三方网站上的内容负责。包含稳定币、NFTs 等在内的数字资产涉及较高程度的风险,其价值可能会产生较大波动。请根据自身财务状况,仔细考虑交易或持有数字资产是否适合您。