Sonic的升级,是换皮还是换骨髓。
Sonic's upgrade: a change of skin or a change of core?
作为亚洲视线范围内,有一定发言权的团队(我们21-22年近一年占据了
@0xSonicLabs 链上超过50%的交易量。),我们对这次sonic的升级进行了一次全方位的审视。有趣的是,似乎每一个从业者都知道Sonic的升级,但是没有从业者认真研究过到底升级了什么,甚至英文资料都很少。关于Full node之争其实Sonic也有自己的答卷,它非常有讨论意义,我们不应该忽视房间里的大象。
As a team with a notable presence within the Asian perspective (we have accounted for over 50% of on-chain transactions on
@0xSonicLabs in the past year, 21-22), we have conducted a comprehensive review of Sonic's recent upgrade. Interestingly, it seems that every practitioner is aware of Sonic's upgrade, yet few have delved deep into what exactly has been upgraded. There is even a scarcity of English resources on the matter. Concerning the Full Node debate, Sonic has its own response, which is highly debatable. The elephant in the room should not be overlooked.
到底升级了什么?为什么它称自己是L1?
我们翻阅了Fantom Foundation的所有Doc,替他们总结了升级中值得列出来的创新。包括:
-内部(Fantom Virtual machine & Storage)和
-外部(Bridge)两个部分。
Sonic的新命名:
-Fantom virtual machine(FVM)
-Storage(DB)
-Bridge(Fantom Gateway)&Validator
What has been upgraded exactly? Why does it refer to itself as L1?
After going through all of Fantom Foundation's documentation, we have compiled a list of notable innovations from their upgrade. These include:
Internals (Fantom Virtual Machine & Storage) and
Externals (Bridge).
Sonic's new nomenclature:
Fantom Virtual Machine (FVM)
Storage (DB)
Bridge (Fantom Gateway) & Validator
内部升级: -Fantom virtual machine(FVM)&Storage(DB)
在Fantom Foundation的doc中先提出了DB和EVM分别占据了84%和13%的消耗时长。而其中重点解决DB的效率显然会更高
详情参考Doc:
blog.fantom.foundation/fanto…
而新升级的这两个部分互相成就,FVM合并和预设了多个指令,建立了很多index,希望用这样的方式来提高效率。某种程度上来说更像是树状结构让用户更快能通过路径执行。这个类型的优化在其他公链也能看得到。
Internal Upgrade: Fantom Virtual Machine (FVM) & Storage (DB)
In the documentation provided by Fantom Foundation, it was initially highlighted that the DB and EVM respectively accounted for 84% and 13% of the consumed runtime. However, focusing on enhancing the efficiency of the DB appears to be more advantageous.
For more details, refer to the documentation:
blog.fantom.foundation/fanto…
The new upgrades in these two components complement each other, with the FVM consolidating and predefining multiple instructions, establishing numerous indexes in an effort to boost efficiency. To some extent, it resembles a tree-like structure enabling users to execute tasks more swiftly along paths. Such optimizations can also be observed in other public blockchains.
重头戏是DB的部分,Sonic直接拆分了传统DB成为了LiveDB和ArchiveDB,见配图。大白话来形容就是LiveDB只管当下此刻的数据,文档中的描述是‘Current Blockchain Only’;而ArchiveDB更像是一个记录历史数据的数据库,所以本质是通过拆分DB功能来提升只有一个DB的运行效率。
详情参考Doc:
blog.fantom.foundation/fanto…
这也是23年1月Doc文档中描述Storage处理时间下降98%这句话的来源(毕竟只用处理Current Blockchain Only)。
这拆分之后,马上就要面临的问题是dispute问题,换句话说,如果产生了争议的交易如何处理呢?
The highlight is the part about the database (DB). Sonic directly splits the traditional DB into LiveDB and ArchiveDB, as shown in the diagram. In simple terms, LiveDB only deals with current moment data, described in the document as 'Current Blockchain Only'; whereas ArchiveDB is more like a database that records historical data, so essentially, it enhances the operational efficiency of having just one DB by splitting the DB functionality.
For more details, refer to the document:
blog.fantom.foundation/fanto…
This is also the source of the statement from the January 23 document that describes a 98% decrease in storage processing time (after all, it only deals with Current Blockchain Only). Following this split, the imminent issue to be faced is the dispute problem. In other words, how are disputed transactions handled if they arise?
传统EVM和DB下,比如有
@VitalikButerin 拍脑门定的7天挑战期(是的,我指的是乐观挑战),通过二分法的方式,大家统一称为默克尔树的解决方案,来溯源定位到精准的那一笔争议交易,而毫无疑问的是,这个做法听起来就极大的消耗了DB的处理时间,实际上也是如此。
而Sonic的解决方案看起来是放弃默克尔树这一套,并且将传统的Full node拆分成Validator和Archive node。大白话翻译一下就是Validator负责EVM(即FVM)和LiveDB(即当下数据),所以Validator的要求也大幅度降低了,很好理解,因为处理的命令量和数据量都降低了。(这个地方需要格外留意,因为关联到后面的Bridge的创新)。
而Archive node就承担了查询调用ArchiveDB(即历史数据)等复杂一些的功能,基本应该都是项目方和类似我们这样做链上生意的人需要准备的,但目前并没有看到有对Archive node的激励措施,这十分的原教旨主义。(因为Vitalik当时提出类似观点的时候也没解释)
In the traditional EVM and DB setup, for example, the 7-day challenge period set by
@VitalikButerin (yes, I'm referring to optimistic challenges) is resolved through a binary method commonly known as the Merkle tree solution. This method traces back to pinpoint the exact disputed transaction. Undoubtedly, this approach appears to significantly consume the processing time of the database, which is indeed the case.
Sonic's solution seems to abandon the Merkle tree approach and instead divides the traditional full node into Validator and Archive node. In simpler terms, Validators are responsible for the EVM (i.e., FVM) and LiveDB (i.e., current data), thus reducing the requirements for Validators significantly, which is easily understood as the workload and data volume decrease. (This point requires special attention as it relates to the innovative Bridge later on).
The Archive node takes on more complex tasks such as querying ArchiveDB (i.e., historical data), which are likely requirements for projects and businesses conducting transactions on the chain, like us. However, there currently seems to be no incentive measures for Archive nodes, which is quite purist in nature. (This lack of incentives is reminiscent of Vitalik's earlier views on a similar topic, which were left unexplained at the time.)
外部升级-Bridge(Fantom Gateway)&Validator:
外部升级和内部升级高度关联,甚至我们认为很可能是外部升级(Bridge)倒推出了如今内部升级的结构。否则我们的确很难解释
@AndreCronjeTech 回归之后为什么主要负责的就是Bridge(Fantom Gateway)
如果了解Fantom的历史,就好理解Bridge对Fantom生态的重要性。早几年
@MultichainOrg 官宣成为了Fantom指定合作的跨链桥,因此有了Print的稳定币。而后的事情就是Multichain突发被中国警方控制,而跨链桥的倒塌让Fantom的TVL一蹶不振。八卦一句,当年我们追踪过Multichain关联地址在Fantom上印钱然后转去其他链的举动,我不清楚 Foundation 到底有没有参与,但是这个举动当年获得了巨量的利润。
External Upgrade - Bridge (Fantom Gateway) & Validator:
External upgrades are closely intertwined with internal upgrades, to the extent that we consider it highly probable that the external upgrade (Bridge) retroactively influenced the current structure of internal upgrades. Otherwise, it would be quite challenging to explain why
@AndreCronjeTech, upon his return, primarily focuses on the Bridge (Fantom Gateway).
Understanding the history of Fantom makes the importance of the Bridge to the Fantom ecosystem clear. A few years ago,
@MultichainOrg was announced as Fantom's designated cross-chain bridge partner, leading to the creation of the stablecoin Print. Subsequently, Multichain faced unexpected control by Chinese authorities, causing the collapse of the cross-chain bridge and a significant decline in Fantom's Total Value Locked (TVL). On a side note, back then, we tracked Multichain's associated addresses minting money on Fantom and then transferring it to other chains. Although I am unsure if the Foundation was involved in this activity, it did generate substantial profits at the time.
所以摆在Fantom升级面前的问题是:1)跨链桥是否安全 2)有没有稳定币。
2)可以通过运营获取,目前USDC已经要发行在Fantom上,不知道代价如何,传言当年
@opensea 来Fantom的话开了一个天价,最后没谈拢。1)的问题是技术问题,也是这次升级的重心,而且它必须是原生的,否则无法向市场解释如何避免Multichain的事故。
但我们说是桥,不如我们说是另一种的node创新,为了拿到一个原生的桥,
@AndreCronjeTech 进行了一个非常精妙的举动,即让运行一个Sonic节点的人,同时再运行一个节点在ETH上,通过这两个节点的同步来加快原生桥的跨链速度。这也是为什么Sonic十分骄傲的说他们的Bridge是Fast&Decentralized。
详情可以参考Doc:
blog.soniclabs.com/sonic-gat…
The issues facing Fantom's upgrade are: 1) the security of cross-chain bridges and 2) the availability of stablecoins.
Stablecoins can be acquired through operations, with USDC already in the process of being issued on Fantom. The cost implications remain uncertain; it is rumored that when
@opensea considered moving to Fantom, negotiations broke down over exorbitant fees. The first issue is a technical one and a focal point of this upgrade. It must be native to the platform to effectively explain how it prevents incidents like those seen with Multichain.
Referring to it as a bridge might not be entirely accurate; rather, it's an innovative form of node. To achieve a native bridge,
@AndreCronjeTech made a clever move: requiring individuals running a Sonic node to also operate a node on Ethereum. Synchronizing these two nodes accelerates the cross-chain speed of the native bridge. This is why Sonic proudly declares that their Bridge is Fast & Decentralized.
For more details, please refer to the documentation:
blog.soniclabs.com/sonic-gat…
要实现这个逻辑,有个现实问题就是同时运行2个Validator实际运行成本是翻倍的,这也是为什么内部创新时,那么强调要降低validator的成本,因为本质是两个node。这也是为什么Sonic一直强调自己是L1,不是L2。但是目前我们对这个有趣的方案定义为L1.5。
至于为什么是ETH?大概是因为如果对其他的链也用这个方案,同时运行Validator的数量还需要继续增加,而显然ETH上的TVL迁移过来性价比最高。我们猜测Sonic一直就觊觎ETH上的合成资产并未得到比较大的激活。希望通过这个举动来盘活类似stETH和wBTC这类资产。在
@michaelfkong 和
@AndreCronjeTech 的只言片语中也能略知一二。
To implement this logic, a practical issue arises: the actual operating cost doubles when running two Validators simultaneously. This is why emphasizing the need to reduce validator costs is crucial during internal innovations, as essentially, it involves managing two nodes. This is also why Sonic consistently stresses that it is L1, not L2. However, currently, we label this intriguing solution as L1.5.
As for why Ethereum is chosen, it is likely because if this approach were applied to other chains, the need to increase the number of validators running simultaneously would persist. Clearly, migrating TVL from Ethereum offers the best cost-effectiveness. We speculate that Sonic has long coveted synthetic assets on Ethereum but has yet to see significant activation in this regard. It is hoped that this move will revitalize assets like stETH and wBTC. Insights from
@michaelfkong and
@AndreCronjeTech provide some clues in this regard.
Sonic到底想做什么? 对于一个尚未list的项目,通常情况下都会按照list的要求去反推运营和增长的路径,这才诞生了Airdrop目前的阶段性业态(交互&贿赂TVL)。这种以果为因的做法对一个已经list在各大交易所的标的,显然是不需要的。这个部分我已经放在关于Sonic如何发钱的视频里面进行了讨论。梗概来说,
#SonicArcade 为了DAU;
#SonicBOOM 为了ECO;
#MemeSeason 为了拉新。
从FDV的角度看,从L2到L1的定义无疑会潜在推高估值水平,但是现在除了技术创新外,共识都会倾向Killer dapp是什么的问题,或者说差异化在哪里。这里就涉及了AC第二个工作重点,
@TheBlockBureau 。
@TheBlockBureau 目前还看不到产品,但是从几个月前我们关注开始,它似乎就是一个关于【信用贷款】相关的基建。看上去像是基于地址活动抽象出了一个模型来进行评估和授信,像是Alipay的信用分。
这个产品是符合Defi基因创新的,也符合整个市场对Fantom升级的理解和估值水平。配合上Bridge的创新,我们很容易有一个较大的想象空间,比如 信贷 合成资产;或者信贷 合成资产 稳定币or算法稳定币。
What does Sonic really aim to achieve? For a project that has yet to be listed, the typical approach involves deducing operational and growth paths based on listing requirements, leading to the current phase of Airdrops (interaction & incentivizing TVL). Such a practice, which starts with outcomes rather than causes, is evidently unnecessary for assets already listed on major exchanges. I have discussed this aspect in a video on how Sonic distributes funds. In essence,
#SonicArcade for DAU;
#SonicBOOM for ECO;
#MemeSeason for user acquisition.
From the perspective of FDV, the shift from L2 to L1 undoubtedly has the potential to elevate valuation levels. However, beyond technical innovation, current consensus leans towards questions about what constitutes a Killer dapp or where the differentiation lies. This is where the second focus of AC,
@TheBlockBureau, comes into play.
@TheBlockBureau has yet to unveil a product, but since our attention was drawn to it a few months ago, it appears to be infrastructure related to "credit loans." It seems to have abstracted a model based on address activity for evaluation and credit granting, akin to Alipay's credit scoring.
This product aligns with DeFi's genetic innovation and fits the market's understanding and valuation levels regarding Fantom's upgrade. Coupled with the Bridge's innovation, there is ample room for imagination, such as credit synthetic assets; or credit synthetic assets stablecoins or algorithmic stablecoins.
我们对Defi 2.0的时代充满热情,无论是高性能公链还是链上的高频交易、信贷都是符合下一个Defi版本的故事。非常欢迎 Fantom Foundation联系我们,每次讨论Fantom,华人用户似乎都不太关心,这让我们对此感到失落,它应该被关注。我们很乐意在这次升级中参与让链上更繁荣。
回到开头的那句话,不要忽视房间里的大象。
We are enthusiastic about the era of Defi 2.0, where high-performance public chains and on-chain high-frequency trading, as well as credit, all contribute to the narrative of the next version of Defi. We would be thrilled for the Fantom Foundation to reach out to us. Whenever Fantom is discussed, it seems that Chinese users are not particularly interested, which leaves us feeling somewhat disheartened; it deserves attention. We are more than willing to participate in making the blockchain more prosperous through this upgrade.
Returning to the initial statement, let's not overlook the elephant in the room.