Home Old docs

Web3 beyond blockchains

Off-chain composability: our vision for a Web3 far beyond blockchains

Web3 is trapped on-chain

Web3 looks like it's heading into a bright future. By combining incentives and cryptography, blockchains have achieved unprecedented decentralization and security, and general-purpose L1s like Ethereum now host growing ecosystems of interoperable decentralized protocols. There's everything from DeFi staples like the Compound lending platform to fascinatingly novel experiments like the cryptography game Dark Forest. Key technical problems like execution scalability now have breakthrough solutions like rollups and zero-knowledge proofs. And even the UX is steadily improving, allowing more and more regular users to join the "crypto" world.

But on a closer look, there's a big problem: web3's revolution is trapped inside blockchains. Smart contracts running on blockchains may have the strong neutrality, decentralization, and censorship-resistance that Web3 promises, but everywhere else, Web2's problems seem to crop up with a vengeance.

Web3 gateways

Nowhere is the lack of "Web3 superpowers" clearer than in dApp frontends: off-chain programs used to interact with on-chain systems by inherently off-chain humans and devices. The typical dApp frontend is nakedly a Web2 service, with a centralized server hosting a webpage (like app.uniswap.org) calling a centralized SaaS service to interact with on-chain smart contracts on the user's behalf.

Unsurprisingly, users are at the mercy of the same sort of unaccountable centralized power we see in Web2 — just witness how a single regulatory agency (OFAC) easily cut off access to a widely-used protocol (Tornado Cash) simply by putting pressure on a few service providers (Infura, etc).

All of the smart contracts involved in not only Tornado Cash, but also Wormhole and Merit Circle, ran safely and correctly. But the apps—which include inherently off-chain frontends, infrastructure, governance, etc—did not.

Non-blockchain Web3

Moreover, there's a whole world of entirely non-blockchain projects that are arguably Web3 protocols — "alt-tech" protocols sharing the same goal of fixing Web2's problems with decentralized cryptography: Tor, I2P, Matrix, and the like.

Compared to blockchain apps, nearly all non-DeFi dApps struggle to gain adoption due to the difficulty of successfully combining decentralization, usability, and security, a need that blockchains seem ideally suited to satisfy. Yet puzzlingly, we don't see much blockchain adoption in these projects that truly need decentralized security.

Mainstream Web2

And of course, mainstream internet users at the mercy of abusive Web2 desperately need Web3 censorship resistance, decentralization, and user-aligned incentives! But most of them don't know much more about Web3 beyond "crypto trading"...

Mel frees web3

We believe that the way to free Web3 is a new paradigm for Web3: off-chain composability. Current blockchains host complex on-chain ecosystems of smart contracts with "frontends" that struggle to interface with them. Instead, we envision full-stack off-chain dApps leveraging Mel as the web3 superpower component, not as a platform.

This will be a largely non-blockchain world with two key properties:

  1. Off-chain programs trustlessly compose with on-chain programs. Off-chain programs, whether dApp frontends or whole alt-tech stacks, must integrate on-chain functionality without sacrificing L1 security.

For instance, a frontend for a coin mixer can't be censorable by any SaaS provider, and no platform provider should be able to tamper with queries to an on-chain PKI used by an end-to-end encrypted chat program. The "web3 superpowers" end up successfully crossing over the boundaries of the blockchain.

  1. Decentralized protocols compose into dApps off-chain. Decentralized protocols with blockchain-backed security should not primarily compose using on-chain constructs such as standardized smart-contract interfaces. Instead, they are composed in off-chain programs like mobile apps and frontends to create a "full-stack" decentralized system.

For example, clients for a decentralized encrypted chat platform with cryptocurrency payments can be built from a Sybil-resistant DHT, a blockchain-backed naming system, a micropayment network, etc, all of which are protocols trustlessly compose with on-chain logic to provide robust decentralized security. But none of these systems need to have interoperating on-chain logic.

In Mel's ecosystem, trustless light clients take Web3 superpowers out of the on-chain box, then decentralized secure protocols pass the superpowers to user-facing apps. Web3 is freed from the blockchain box, so much so that users and developers of the end applications never even need to think about the blockchain.

A great example of a member of this sort of protocol ecosystem is Earendil (https://earendil.network), a decentralized, strongly censorship-resistant communication protocol we are building into core communication system for the Mel ecosystem.

Why do we need a new blockchain?

But why do we need a new blockchain to implement off-chain Web3? Can't we build this world on an established blockchain like Ethereum?

It turns out that ease of off-chain composability requires many design trade-offs in all the aspects of a blockchain, and features for on-chain ecosystems hamper both exporting and producing Web3 superpowers. As examples:

VM improvements: These provide easy opcodes for cool stuff like ZK and make smart contracts happy by saving gas costs. However, the governance involved both breaks thin client compatibility and threatens neutrality

Standard contract APIs: Decoupling contract interface from implementation, these APIs power many on-chain ecosystems (ERC-20 DeFi etc). However, the disparate implementations complicate off-chain verification, and buggy/centralized contracts introduce systemic security risks.

"Moar TPS!!": This makes on-chain things fast and cheap, essentially making a bigger box for bigger contracts. However, thin clients are way harder to run, and the much bigger nodes lead to centralization.

Unfortunately, current blockchains, especially fancy new ones, double down on one or more features, trading away off-chain composability for an on-chain ecosystem. That's why we made a new L1 from scratch, optimizing every part of Mel for being the decentralized security keystone that enables an off-chain composable Web3.

An overview of the docs

Throughout this website, you'll find

NextData model

Last updated