This is a ๐งต about Aurora Cloud and why it's superior to any L2 stack.
TL;DR: decentralised sequencer, built-in composability between chains.
Read on for more details ๐
L2 narrative has been out there for many months. The only substantial reason for this is the scaling of the L1 blockchain.
L2s are not enabling new use cases. Nothing new would happen on L2s compared to L1. It's a band-aid.
And the scaling approach is quite simple to understand: to scale, we need MORE EVMs. But we don't want to run more networks, so we need a way to connect these new EVMs to the L1 chain. And to do this, we would actually settle these transactions on L1.
I would agree with this direction in general: EVM is a standard, and all the tools are designed to work with it. However, it cannot be scaled directly due to synchronicity, so the only thing we can do is launch more of them.
But why on Earth did we decide to resolve running the L2 network with centralized sequencing (i.e. a single computer that is ordering the transactions)? This is a direct step back to SQL.
And I reject the argument that a user is always able to settle their transactions permissionlessly on L1.
A user can only do it slowly.
LUNA crashed in less than a week; USDC depegged and recovered (!) in less than a week. There's no way for an honest participant to be in time for a change on the network with a dishonest sequencer.
This theoretical escape hatch is a dangerous marketing illusion.
"Sequencers would be decentralized eventually" -- this is nonsense. We do have an approach to decentralization of sequencers. It's called blockchain consensus. Why are we not using it then?! Oh, because it's too difficult? Give me a break.
A better way to do EVM scaling is right in front of us.
Imagine an asynchronous settlement L1 blockchain, where transactions coming to different contracts can be executed in parallel. Now launch on top of it a contract that implements EVM, so it can run EVM contracts (a dream within a dream). Repeat as many times as you need.
Such an L1 may solve the scalability issue of blockchains: since the runtime is asynchronous by definition, one can parallelize the execution of transactions. This is often referred to as "sharding" (although it is not technically sharding, the term has already stuck).
Every EVM contract would have a consensus of the L1 as a sequencer. If the validator set of this L1 is decentralized enough, then each EVM would have a decentralized and secure setup from launch.
Even more: if we designed precompiles to asynchronously call contracts from other EVMs (even with callbacks!), then we could compose these previously separated ecosystems into a single one. Thankfully, that's just a cross-contract call on L1 -- no need to have bridges.
Oh, by the way, since the above-mentioned EVMs are just contracts, one can greatly update the basic principles of their operations: allow only whitelisted entities to transact or deploy contracts for compliance; introduce custom fee mechanics and base tokens...
Or even remove the fees altogether! EVM-level payment for transaction execution is now separate from L1 level: RPCs are becoming blockchain providers with a competitive market and a wide range of business models.
This is a window into mass adoption.
How great would it be to live in this better future world, right?
Wrong. We are already living in it.
A scalable, asynchronous L1 blockchain is @NEARProtocol with Nightshade consensus.
An EVM smart contract on top of it is @auroraisnear.
A solution that allows anyone to launch their own EVM-compatible blockchain is Aurora Cloud.
auroracloud.dev/
(Pssst: there are already four of them launched now, including Aurora MainNet.)
Every single Aurora EVM has a built-in cross-contract call precompiles:
dev.aurora.dev/posts/cross-ecosystem-communication
These, together with a decentralized approach for front-ends (BOS), allow some next-level stuff to happen:
dev.aurora.dev/posts/building-a-game-using-near-aurora-and-bos
And, by the way, the Aurora stack has been working for 2 years without any hacks and with a 100% chain uptime.
And you know what? Forget about Aurora and NEAR. Erase everything starting from the tweet #16 from your mind. Think for a minute about the current direction of L2s.
Let's cut the marketing BS. Every deployment of these separate ecosystems is weakening Ethereum by siloing liquidity, collectibles, and governance powers.
In the proposed "mythical" scenario, every single EVM deployment would strengthen the whole ecosystem through built-in composabilityโall while maintaining decentralization. A core principle for blockchain antifragility.
Got something to add? Comment and prove me wrong. ๐ช
Alex Shevchenko ๐บ๐ฆ
@AlexAuroraDev
Co-founder of @auroraisnear | CEO of Department of NEAR Efficiency | Building blockchain infra since 2015
