Subscribe to finestone
>100 subscribers
Subscribe to finestone
>100 subscribers
Share Dialog
Share Dialog
There are levels to Ethereum-equivalence: beyond EVM-equivalence and beyond equivalence with the entire Ethereum architecture. Equivalence extends to network attributes like permissionless participation (ultimately leading to Ethereum-validator inclusion and decentralization), and to philosophical considerations. This post attempts to define a “hierarchy of Ethereum-equivalence”, and make the case for why its apex (Taiko’s path) is a compelling path for rollups, and for why having an ‘Ethereum on Ethereum’ outcome is important for users (and Ethereum). We’ll look at the following topics:
The conventional definition of Ethereum-equivalence
The extended definition of Ethereum-equivalence
What each level of Ethereum-equivalence gets you
Why care about full Ethereum-equivalence
In the world of rollups, many teams have aimed to differentiate themselves on their level of EVM-compatibility, or one level deeper, on Ethereum-compatibility.
For the sake of clarity, relative to the target in question (EVM, Ethereum): perfect compatibility = equivalence.
This is especially true for ZK-EVMs, as proving EVM or exact-Ethereum-block execution with ZKPs is not trivial and presents tradeoffs vs generating ZKPs efficiently. The same is not true for optimistic rollups, where most of the leading implementations all started with Ethereum-equivalence because you can do that “for free” (without a corresponding steep cost or lengthened ZK engineering effort). This ZK-EVM dynamic is captured in Vitalik’s diagram on “The different types of ZK-EVMs”.
The battle (see: ZK-EVM wars 2021-) for narrative supremacy, developers, and users (but mostly narrative) wages primarily within that tradeoff space of Ethereum-compatibility vs proof-generation performance.
Taiko takes an extreme approach: Ethereum-equivalence — in both the conventional definition from Vitalik, and what I put forth as the extended definition. In the conventional definition, an Ethereum-equivalent (type-1) ZK-EVM is able to prove Ethereum blocks exactly as they are on L1. Meaning:
it implements the EVM exactly as is on Ethereum L1, including the gas schedule. This feat can be called EVM-equivalence, or type-2 ZK-EVM
it implements all other Ethereum architectural aspects (state trees, transaction trees, hash functions, precompiles) exactly as is on Ethereum L1.
To my knowledge, Taiko is the only ZK-Rollup aiming directly and initially to be a type-1 ZK-EVM (though notable to mention Risc Zero’s path for proving Ethereum blocks as well). To learn more on where the differences between Type-1s and other Types manifest themselves, especially on a smart contract compatibility level, please refer to this article.
For a gross simplification so that less technical readers may gain an intuition on the differences, it looks something like this:
However, in my mind, the conventional definition of Ethereum-equivalence is only part of a more comprehensive definition of Ethereum-equivalence (hereafter “EE”). I like to extend EE to mean more, specifically along two further dimensions:
Permissionless, Decentralized and Based
Philosophically-aligned and ✨Based✨ (can perhaps be called “Ethereum Enshrinement”)
These two additions form the 3rd and 4th levels of my interpretation of an EE hierarchy pictured at the top of this post. Let’s explore what these new additions mean, why they matter, while also reviewing why the conventional EE components matter.
I’d like to state in advance that this is not for the sake of starting another rollup semantics war. This is about the hard properties that such a network/rollup can provide to users and digital economies. By defining it, I hope many current and future rollup builders, dapp deployers and end-users can see its merits. It is also about the hard and soft (social) properties that can feed back into Ethereum and further enshrine it as the value layer of the internet, and ultimate social coordination tool.
Finally, implicit in these EE definitions is that we are actually talking about Ethereum rollups through and through: choosing Ethereum as your data availability, consensus, and settlement layer (and execution - but that is implicit in the conventional definition already).
Most smart contracts and developer tooling work out-of-the-box
Security in a battle-tested VM
Developers
Basically, a treasure trove of tools, learned patterns, people and projects amassed over 8 years so developers and dapps can hit your rollup running.
Same as EVM-equivalent above, except ALL contracts and tooling works (no edge cases)
Reuse battle-tested Ethereum infrastructure with minimal diff, such as execution clients and PBS infrastructure
Security in a battle-tested VM (as above), but also from the overall system, including clients
Simple and robust cross-chain communication with Ethereum L1/equivalent-L2s via Merkle proofs
Emotional benefits and fulfilment of your work being able to prove Ethereum blocks as they are on L1, and thus being a stepping stone to a SNARKified and scalable L1
Future-proofness by staying in step with the Ethereum brain trust, and allowing the massive community to help solve problems with you
Basically, everything from EVM-equivalent, but even smoother developer experience because everything works with no edge cases nor need for re-audits. Beyond dapp developer benefits, the rollup builders themselves and participants in the network (node runners, proposers, builders) can reuse and run battle-tested infrastructure for max simplicity and security. Finally, the rollup itself inherits some superpowers from “speaking Ethereum” perfectly, like cross-chain communication, and paths to extend those powers.
Permissionlessness for network participants — node runners, proposers (sequencers) and provers — can lead to a decentralized network and provide the properties of censorship resistance and liveness
Permissionless proposers (especially in an Ethereum-equivalent-architecture rollup) leads to an outcome of a based rollup, where Ethereum L1 validators become the possible set of L2 proposers (with the reuse of existing PBS infrastructure, involving searchers and builders)
One of the most decentralized network topologies possible, with the same strong guarantees of censorship resistance and liveness as L1
MEV flowing to L1: enriching L1 validators, adding to the economic security of L1, and in the case of eventual MEV burn, benefiting ETH holders
A component of deeper alignment with Ethereum (see next section)
Prover schemes should also be permissionless, though the degree of decentralization is less important compared to proposers. What’s required is that the protocol (or proposers) can attract some set of efficient provers to generate proofs on a reasonable timeframe at reasonable cost. You can learn more about different proving designs Taiko has implemented here and here.
Basically, permissionlessness and decentralization emulates the “rest of Ethereum”, and in some regards, what many of us are here for: a credibly neutral, censorship-resistant environment that can be the basis of a multi-trillion dollar economy, provide freedom to individuals, and withstand maximally adversarial conditions.
It’s popular to postulate that the L1 (DA and consensus layer) should be maximally decentralized, and that the rollup (execution environment) and its sequencers can (and maybe should) be centralized. It’s certainly true that the rollup design space allows for a flourishing of configurations with different tradeoffs while maintaining user asset safety by allowing forced inclusion via the L1.
But I believe it’s also true that in that flourishing, certain rollups should exist that emulate Ethereum across the board, including on what it offers, not just how it offers it: decentralization, credible neutrality, censorship resistance, and liveness. After all, when Ethereum L1 block/blobspace is primarily being bought by rollups (or super high-value apps), applications and users in that rollup-first world will still demand those important Ethereum-like properties for certain applications spanning financial to social. We cannot crowd them out. If rollups will host the activity and applications of the future, it seems reasonable that Ethereum on Ethereum is an attractive proposition, and a critical one.
Support and legitimacy from the Ethereum community
Technical resources and collaboration from core devs and other world-class teams
Awareness and interest from talented developers with a long-term view
Social support, coordination, and user base
Emotional benefits expanded, from contributing to something you love and that is changing the world
This invites talent to build with and alongside you, and to help educate and evangelize
Long-termism by default, building technology that seeks to be critical for centuries to come
Competitive edge. Philosophy/Ethos/Vibes may be the main competitive dimension going forward. The (rollup) tech will be commoditized and converge
Because this is part of the extended EE definition and a bit hand-wavy by nature, let’s define some of the things that actually make a rollup philosophically-aligned with Ethereum (not just talk about what it gets you as above).
Open source and permissively licensed. All of it
Upstream contributions to Ethereum
As mentioned, by proving Ethereum L1 blocks as they are, a stepping stone to SNARKified L1. Besides proving its feasibility and implementing it, it also gets provers, chip makers, and hardware/algorithm optimizers ready to prove Ethereum L1 blocks.
Work on community codebases, like PSE ZK-EVM. This benefits others, but also your rollup, as other great teams and individuals contribute.
Based rollup: let L1 validators and the rest of the Ethereum block-building pipeline easily propose for your rollup
Enshrine ETH as the gas token on your rollup
Empower your community (the Ethereum community)
Invite and enable contributors. Set yourself up to incentivize and empower multiple core contributing teams (client implementations, etc).
Elevate and hire from the community; paths for contributors to become full-time.
Reward Ethereum core devs and other shoulders you stand on via mechanisms like Protocol Guild, token distribution, etc.
I do realize there’s a nuance between having a philosophy that aligns with Ethereum’s (what Ethereum cares about), and a philosophy that enshrines Ethereum. This may be a good subject for a future post, but for now I blend them.
An important takeaway from this hand-wavy section is a practical one: even if you do not care about Ethereum philosophical-alignment for Ethereum or altruistic reasons, you should care about it for competitive reasons. I believe the bottom 3 levels of the stack/hierarchy will be commoditized, and will converge on a few variations (with the demonstrated flavours being a popular combination). The top tier will be the competitive dimension. Enshrine Ethereum because you love it, but also for your own good.
Alignment is of course easy to say and hard to do. And that’s exactly why some teams will outcompete here. You can’t copy-pasta care and vibes. Vibes inform technical decisions. If you care about Ethereum and what it provides for the world, you don’t take shortcuts, and you provide an environment with compelling guarantees. In a period where a new rollup will launch every week and it’s convenient for even L1s to snuggle up next to Ethereum for a narrative boost, a credible commitment to Ethereum may differentiate you and be rewarded. Projects that actually care about Ethereum and scaling it securely will continue to rise to the top.
So let’s see what we’ve built by extending the EE definition:
Like Maslow’s hierarchy of needs, if you wish to reach the apex of Ethereum-equivalence as a rollup, the lower tiers must be satisfied before the higher ones can be reached for. Indeed, EVM-compatibility or equivalence is table-stakes for many projects trying to harness Ethereum’s developer community, infrastructure, and reach. I hope the higher levels of the hierarchy may also soon be deemed table-stakes, certainly so for rollups trying to produce blocks that truly emulate Ethereum — where censorship resistance and liveness assurances are solid, where new digital nations may spring up, and where Ethereum (and its community, validator set, and native asset ETH) is enshrined.
A bit counterintuitively: a rollup that implements a different VM, but satisfies the 2 new pieces of the definition would also be a great path for the outcome/properties I describe here - even if not “EE”; but with a different VM comes greater difficulty to reuse the infra, and importantly for the L1 validator/PBS pipeline to build & include blocks like they are used to, so it may be impractical or impossible to reach the same level of decentralization.
Full Ethereum-equivalence is of course not the only configuration, but it’s the one I’m most interested in, and one that I believe Taiko follows to a strong degree. I do not yearn for a monoculture where all rollups implement the “perfect” Ethereum-equivalence I described (I am all for the outsourced innovation and options for users), but I do hope that within the rollup free market there will be some rollups that give as much as they take from Ethereum, and realize that its properties flowing from decentralization are important for their users — not just for them as rollups in relation to L1. I believe Ethereum deserves rollups that offer people similar environments and strong guarantees, and that offer Ethereum a fertile ground for scaling itself, a means to perpetuating its mission of openness and freedom, and an enshrined position.
There are levels to Ethereum-equivalence: beyond EVM-equivalence and beyond equivalence with the entire Ethereum architecture. Equivalence extends to network attributes like permissionless participation (ultimately leading to Ethereum-validator inclusion and decentralization), and to philosophical considerations. This post attempts to define a “hierarchy of Ethereum-equivalence”, and make the case for why its apex (Taiko’s path) is a compelling path for rollups, and for why having an ‘Ethereum on Ethereum’ outcome is important for users (and Ethereum). We’ll look at the following topics:
The conventional definition of Ethereum-equivalence
The extended definition of Ethereum-equivalence
What each level of Ethereum-equivalence gets you
Why care about full Ethereum-equivalence
In the world of rollups, many teams have aimed to differentiate themselves on their level of EVM-compatibility, or one level deeper, on Ethereum-compatibility.
For the sake of clarity, relative to the target in question (EVM, Ethereum): perfect compatibility = equivalence.
This is especially true for ZK-EVMs, as proving EVM or exact-Ethereum-block execution with ZKPs is not trivial and presents tradeoffs vs generating ZKPs efficiently. The same is not true for optimistic rollups, where most of the leading implementations all started with Ethereum-equivalence because you can do that “for free” (without a corresponding steep cost or lengthened ZK engineering effort). This ZK-EVM dynamic is captured in Vitalik’s diagram on “The different types of ZK-EVMs”.
The battle (see: ZK-EVM wars 2021-) for narrative supremacy, developers, and users (but mostly narrative) wages primarily within that tradeoff space of Ethereum-compatibility vs proof-generation performance.
Taiko takes an extreme approach: Ethereum-equivalence — in both the conventional definition from Vitalik, and what I put forth as the extended definition. In the conventional definition, an Ethereum-equivalent (type-1) ZK-EVM is able to prove Ethereum blocks exactly as they are on L1. Meaning:
it implements the EVM exactly as is on Ethereum L1, including the gas schedule. This feat can be called EVM-equivalence, or type-2 ZK-EVM
it implements all other Ethereum architectural aspects (state trees, transaction trees, hash functions, precompiles) exactly as is on Ethereum L1.
To my knowledge, Taiko is the only ZK-Rollup aiming directly and initially to be a type-1 ZK-EVM (though notable to mention Risc Zero’s path for proving Ethereum blocks as well). To learn more on where the differences between Type-1s and other Types manifest themselves, especially on a smart contract compatibility level, please refer to this article.
For a gross simplification so that less technical readers may gain an intuition on the differences, it looks something like this:
However, in my mind, the conventional definition of Ethereum-equivalence is only part of a more comprehensive definition of Ethereum-equivalence (hereafter “EE”). I like to extend EE to mean more, specifically along two further dimensions:
Permissionless, Decentralized and Based
Philosophically-aligned and ✨Based✨ (can perhaps be called “Ethereum Enshrinement”)
These two additions form the 3rd and 4th levels of my interpretation of an EE hierarchy pictured at the top of this post. Let’s explore what these new additions mean, why they matter, while also reviewing why the conventional EE components matter.
I’d like to state in advance that this is not for the sake of starting another rollup semantics war. This is about the hard properties that such a network/rollup can provide to users and digital economies. By defining it, I hope many current and future rollup builders, dapp deployers and end-users can see its merits. It is also about the hard and soft (social) properties that can feed back into Ethereum and further enshrine it as the value layer of the internet, and ultimate social coordination tool.
Finally, implicit in these EE definitions is that we are actually talking about Ethereum rollups through and through: choosing Ethereum as your data availability, consensus, and settlement layer (and execution - but that is implicit in the conventional definition already).
Most smart contracts and developer tooling work out-of-the-box
Security in a battle-tested VM
Developers
Basically, a treasure trove of tools, learned patterns, people and projects amassed over 8 years so developers and dapps can hit your rollup running.
Same as EVM-equivalent above, except ALL contracts and tooling works (no edge cases)
Reuse battle-tested Ethereum infrastructure with minimal diff, such as execution clients and PBS infrastructure
Security in a battle-tested VM (as above), but also from the overall system, including clients
Simple and robust cross-chain communication with Ethereum L1/equivalent-L2s via Merkle proofs
Emotional benefits and fulfilment of your work being able to prove Ethereum blocks as they are on L1, and thus being a stepping stone to a SNARKified and scalable L1
Future-proofness by staying in step with the Ethereum brain trust, and allowing the massive community to help solve problems with you
Basically, everything from EVM-equivalent, but even smoother developer experience because everything works with no edge cases nor need for re-audits. Beyond dapp developer benefits, the rollup builders themselves and participants in the network (node runners, proposers, builders) can reuse and run battle-tested infrastructure for max simplicity and security. Finally, the rollup itself inherits some superpowers from “speaking Ethereum” perfectly, like cross-chain communication, and paths to extend those powers.
Permissionlessness for network participants — node runners, proposers (sequencers) and provers — can lead to a decentralized network and provide the properties of censorship resistance and liveness
Permissionless proposers (especially in an Ethereum-equivalent-architecture rollup) leads to an outcome of a based rollup, where Ethereum L1 validators become the possible set of L2 proposers (with the reuse of existing PBS infrastructure, involving searchers and builders)
One of the most decentralized network topologies possible, with the same strong guarantees of censorship resistance and liveness as L1
MEV flowing to L1: enriching L1 validators, adding to the economic security of L1, and in the case of eventual MEV burn, benefiting ETH holders
A component of deeper alignment with Ethereum (see next section)
Prover schemes should also be permissionless, though the degree of decentralization is less important compared to proposers. What’s required is that the protocol (or proposers) can attract some set of efficient provers to generate proofs on a reasonable timeframe at reasonable cost. You can learn more about different proving designs Taiko has implemented here and here.
Basically, permissionlessness and decentralization emulates the “rest of Ethereum”, and in some regards, what many of us are here for: a credibly neutral, censorship-resistant environment that can be the basis of a multi-trillion dollar economy, provide freedom to individuals, and withstand maximally adversarial conditions.
It’s popular to postulate that the L1 (DA and consensus layer) should be maximally decentralized, and that the rollup (execution environment) and its sequencers can (and maybe should) be centralized. It’s certainly true that the rollup design space allows for a flourishing of configurations with different tradeoffs while maintaining user asset safety by allowing forced inclusion via the L1.
But I believe it’s also true that in that flourishing, certain rollups should exist that emulate Ethereum across the board, including on what it offers, not just how it offers it: decentralization, credible neutrality, censorship resistance, and liveness. After all, when Ethereum L1 block/blobspace is primarily being bought by rollups (or super high-value apps), applications and users in that rollup-first world will still demand those important Ethereum-like properties for certain applications spanning financial to social. We cannot crowd them out. If rollups will host the activity and applications of the future, it seems reasonable that Ethereum on Ethereum is an attractive proposition, and a critical one.
Support and legitimacy from the Ethereum community
Technical resources and collaboration from core devs and other world-class teams
Awareness and interest from talented developers with a long-term view
Social support, coordination, and user base
Emotional benefits expanded, from contributing to something you love and that is changing the world
This invites talent to build with and alongside you, and to help educate and evangelize
Long-termism by default, building technology that seeks to be critical for centuries to come
Competitive edge. Philosophy/Ethos/Vibes may be the main competitive dimension going forward. The (rollup) tech will be commoditized and converge
Because this is part of the extended EE definition and a bit hand-wavy by nature, let’s define some of the things that actually make a rollup philosophically-aligned with Ethereum (not just talk about what it gets you as above).
Open source and permissively licensed. All of it
Upstream contributions to Ethereum
As mentioned, by proving Ethereum L1 blocks as they are, a stepping stone to SNARKified L1. Besides proving its feasibility and implementing it, it also gets provers, chip makers, and hardware/algorithm optimizers ready to prove Ethereum L1 blocks.
Work on community codebases, like PSE ZK-EVM. This benefits others, but also your rollup, as other great teams and individuals contribute.
Based rollup: let L1 validators and the rest of the Ethereum block-building pipeline easily propose for your rollup
Enshrine ETH as the gas token on your rollup
Empower your community (the Ethereum community)
Invite and enable contributors. Set yourself up to incentivize and empower multiple core contributing teams (client implementations, etc).
Elevate and hire from the community; paths for contributors to become full-time.
Reward Ethereum core devs and other shoulders you stand on via mechanisms like Protocol Guild, token distribution, etc.
I do realize there’s a nuance between having a philosophy that aligns with Ethereum’s (what Ethereum cares about), and a philosophy that enshrines Ethereum. This may be a good subject for a future post, but for now I blend them.
An important takeaway from this hand-wavy section is a practical one: even if you do not care about Ethereum philosophical-alignment for Ethereum or altruistic reasons, you should care about it for competitive reasons. I believe the bottom 3 levels of the stack/hierarchy will be commoditized, and will converge on a few variations (with the demonstrated flavours being a popular combination). The top tier will be the competitive dimension. Enshrine Ethereum because you love it, but also for your own good.
Alignment is of course easy to say and hard to do. And that’s exactly why some teams will outcompete here. You can’t copy-pasta care and vibes. Vibes inform technical decisions. If you care about Ethereum and what it provides for the world, you don’t take shortcuts, and you provide an environment with compelling guarantees. In a period where a new rollup will launch every week and it’s convenient for even L1s to snuggle up next to Ethereum for a narrative boost, a credible commitment to Ethereum may differentiate you and be rewarded. Projects that actually care about Ethereum and scaling it securely will continue to rise to the top.
So let’s see what we’ve built by extending the EE definition:
Like Maslow’s hierarchy of needs, if you wish to reach the apex of Ethereum-equivalence as a rollup, the lower tiers must be satisfied before the higher ones can be reached for. Indeed, EVM-compatibility or equivalence is table-stakes for many projects trying to harness Ethereum’s developer community, infrastructure, and reach. I hope the higher levels of the hierarchy may also soon be deemed table-stakes, certainly so for rollups trying to produce blocks that truly emulate Ethereum — where censorship resistance and liveness assurances are solid, where new digital nations may spring up, and where Ethereum (and its community, validator set, and native asset ETH) is enshrined.
A bit counterintuitively: a rollup that implements a different VM, but satisfies the 2 new pieces of the definition would also be a great path for the outcome/properties I describe here - even if not “EE”; but with a different VM comes greater difficulty to reuse the infra, and importantly for the L1 validator/PBS pipeline to build & include blocks like they are used to, so it may be impractical or impossible to reach the same level of decentralization.
Full Ethereum-equivalence is of course not the only configuration, but it’s the one I’m most interested in, and one that I believe Taiko follows to a strong degree. I do not yearn for a monoculture where all rollups implement the “perfect” Ethereum-equivalence I described (I am all for the outsourced innovation and options for users), but I do hope that within the rollup free market there will be some rollups that give as much as they take from Ethereum, and realize that its properties flowing from decentralization are important for their users — not just for them as rollups in relation to L1. I believe Ethereum deserves rollups that offer people similar environments and strong guarantees, and that offer Ethereum a fertile ground for scaling itself, a means to perpetuating its mission of openness and freedom, and an enshrined position.
No comments yet