DEFIDEVELOPMENT

Translation: Dragonfly founder Haseeb Qureshi: "I'm very concerned that no one will care about Rollup."

share
Translation: Dragonfly founder Haseeb Qureshi: "I

The following content is translated from an article by Dragonfly founder Haseeb Qureshi. If in doubt, please refer to the original article.

Over the past 6 months, the biggest story on Ethereum has been the explosive growth in transaction demand. Transaction fees have hit record highs, making it unaffordable for many regular users to use Ethereum.

Of course, I'm talking about Rollups. Vitalik has laid out the future of Ethereum scaling with Rollups. Leading Rollup development teams have now secured over $100 million in venture capital. Every major DeFi protocol has made the decision to choose one of the upcoming Rollups as their future home.

If you're not sure what Rollups are or need a refresher, here's a brief summary. Rollup is a mini blockchain that inherits the security properties of the blockchain it's based on. Even if Rollup validators or operators are untrusted, they can't steal your funds assuming Rollup is correctly implemented.

There are two basic types of Rollups: optimistic rollup and zero-knowledge rollup (ZK Rollup).

Optimistic rollup is protected by fraud proof—anyone can freely prove if a computation is incorrect; whereas ZK Rollup is protected by cryptography—mathematical proofs that computations are correct. A more comprehensive explanation of Rollups goes beyond here, but if you want to delve deeper, check out Vitalik's article.

The beauty of Rollups is that they are completely trustless. If you trust Ethereum, you should trust Rollups. It's an impeccable scaling solution and has garnered favor from Ethereum's key leaders.

Now, let's look at the second biggest story on Ethereum in the past 6 months: the rise of Binance Smart Chain and Polygon (formerly Matic). These are not Rollups; they are more like side chains—independent replicas of Ethereum with proof of stake. They essentially take off-the-shelf Geth, copy the consensus code, increase gas limits, and bridge back to Ethereum using multisig.

Both chains have skyrocketed and now conduct more transactions individually than Ethereum itself. Other blockchains like Avalanche and NEAR are ramping up their own EVM-compatible systems that also connect to Ethereum.

The world now looks like this:

It's almost like sharding, but with Ethereum 1.0's "beacon chain." Cross-chain transfers and messages are facilitated by multisig, temporary bridges, and some trusted parties. I call this architecture "sharding for the poor." This is how DeFi is actually scaling right now.

But! This sorry state of affairs will soon end. Rollups will be ready soon.

I'm genuinely excited about Rollups. They are disruptive, elegant, and designed so well.

But I worry that users won't care about Rollups.

That's what I'm worried about.

The Savior Complex of Layer 2 Solutions

Let me tell you an ancient story.

Once upon a time, there was a blockchain that could not scale. But then, some very clever individuals invented: trustless Layer 2 technology that could scale the blockchain. Users were excited, and after years of effort, the clever folks finally built the Layer 2. When users could use it, no one cared because they were already using some simpler jank solutions.

Note: Jank refers to bootleg or subpar quality

Did This Story Ring a Bell?

Remember the Lightning Network? People went straight to WBTC. Remember Plasma? People went straight to xDai. And now here we are, not waiting for Rollup, people are going straight to Polygon and BSC.

A good story can solve everyone's problems. This story — the Rollup scaling story — fits everyone. Decentralization extremists tell a grand tale about scaling Ethereum without compromises. Traders can wax poetic about how Rollup will make ETH reach $10,000, while many nod in confusion as MATIC-based AAVE goes wild with yield farming and people are already playing digital horse racing.

I'm not joking! This is actually impactful. Rollups have overwhelming support from the Ethereum intelligentsia, Twitter mobs like to debate things that aren't decentralized enough, I get it, because I am one of them.

But what cannot be ignored is that the masses have already adopted Polygon and Binance Smart Chain. No venture capitalist thought about this. Developing countries like India, Indonesia, Thailand, and the Philippines are adopting these platforms, many of whom have seemingly never used Ethereum before (they couldn't afford it). Remember the whole "bank the unbanked" thing? These platforms, note: referring to MATIC/BSC, actually have global impact and attract users with what they truly care about.

I often say there are three motivations driving cryptocurrency users today:

  • Make money
  • Have fun
  • Ideology

Of these three, ideology is the weakest. And I fear ideology will eventually become the main driver for adopting Rollup.

That's the problem with Layer 2: they sound better in theory than in practice.

The Bitter Pill of Adopting Rollup

On Polygon, a simple Uniswap-like transaction costs $0.0001. On Binance Smart Chain, it costs $0.20. On Ethereum, it's around $7. And on Optimism, it's around $0.68.

Why are Rollups more expensive than these sidechains? It's because every Rollup ultimately has to publish call data to Ethereum, tying their costs to Ethereum fees. Each Rollup can only scale Ethereum by a constant factor. Therefore, compared to the fees many users are already accustomed to, the costs are not low.

And no Rollup is fully compatible with the EVM Ethereum Virtual Machine—there are subtle differences between the virtual machines of each Rollup and the EVM. For Arbitrum, they use AVM; while Optimism uses OVM, each with subtle and crucial incompatibilities with Ethereum contracts and tools. For ZK-rollups, it's a whole different world—ZKR compiles Solidity into equivalent zero-knowledge circuits to run in the ZK virtual machine.

Now compare that to Polygon, where you simply copy and paste your contract, and all is well.

Next, let's look at how funds move in and out of Rollups.

For Optimistic Rollup, when you want to withdraw funds, there's about a 1-week challenge period during which your withdrawal is frozen. This is bad. So, to facilitate "fast withdrawals," market makers are ready with fees. The fees they charge you depend on their inventory and the liquidity of their assets. Transferring ETH might cost 0.2% or other fees, but if you try to move any dog coin, it might cost more, possibly 1% or higher. If there isn't enough liquidity, some assets might not be withdrawn quickly at all.

As a user, you need to consider these things when planning your Rollup DeFi portfolio. That is to say, if you use traditional multisig-based bridging on Rollup, you can avoid these withdrawal issues. But if you use multisig bridging to mitigate custodial risks, how much improvement is there compared to Polygon?

(Note that ZK Rollup is not affected by this issue, as their withdrawals are practically instantaneous.)

I'm concerned that with all these expenses, Rollup won't satisfy any type of user. If you're a super whale who cares deeply about security, paying mainnet fees is fine. If you're just an average Joe, then you can accept Polygon. Then who's left?

Well, When Will It Actually Happen?

Before we dive in, let's talk about how Layer 2 solutions will play out.

Every DeFi protocol on Ethereum will be committed to Layer 2 solutions—some will choose Optimism, some will choose Starkware, and the one with the most branding will eventually become the mainstream Rollup.

Now it's apparent this isn't the right idea. In most cases, DeFi protocols are multi-chain. AAVE, Sushi, and Curve have already launched on Polygon, pushing their TVL above $8 billion. Sushi operates on over 5 chains, while Curve has 4. For the longest time, Uniswap only wanted Optimism, but with Arbitrum's imminent launch, Uniswap changed course and will be multi-chain. Note: Uniswap will deploy the v3 protocol on Layer 2 solution Arbitrum through the latest governance proposal

Binance Smart Chain taught everyone: if you don't roll out a BSC version, we'll fork it and take the revenue you should have received. Looking ahead, I anticipate every major DeFi protocol will be ahead on every significant chain.

So, is it really the protocols deciding where users go? Or is it users deciding where protocols go?

Currently, the lesson from Polygon and Binance Smart Chain seems to be the latter: protocols follow users and, as a result, reap substantial rewards.

I can tell you, as an investor, the current consensus is that Rollup will win. Vitalik likes Rollup. Everyone likes Rollup. Rollup is important. Invest in Rollup.

But I'm worried. I'm worried no one will care. People already have what Rollup originally aimed to provide: a fast, cheap, EVM-compatible blockchain that seamlessly integrates with the Ethereum ecosystem.

So in the long run, how can Rollup win?

In my view, there are two ways: one is non-Rollup sidechains catastrophically failing, and we learn from industry history, as we did from the Mt.GOX incident. Catastrophic failure doesn't just mean "nodes can't sync," it means "money is gone" or "blockchain completely halts." It's possible but perhaps unlikely.

So, there's only one other way: Rollup must actually become significantly better than alternative solutions. The virtue of decentralization slogans isn't enough. For this, I personally see only one path forward, and that's the assurance of cryptography and zero-knowledge proofs.

The cryptography behind zero-knowledge proofs has followed a trajectory similar to Moore's Law in recent years, showing no signs of slowing down. What was once thought to be infeasible to execute EVM-like computations in ZK Rollup is now being done by zkSync and Starkware, proving arbitrary-length computation chains through recursive composition ZK-SNARKs. Over time, I expect we will see more than just the constant factor scaling of today's Rollup: massive computation compression, privacy-preserving smart contracts, MEV-resistant proofs, and more.

State growth isn't a big issue in ZK Rollup in the long run, as users can always verify the correctness of their state by simply verifying the SNARK at the chain's end.

Looking ahead, ZK technology will only get better.

But even in the short term, I'm excited about what Matter Labs is doing with zkSync 2.0 and their zkPorter architecture. zkPorter is a hybrid of Validium and ZK rollup, allowing seamless migration between the two. Data on the Validium side is off-chain, can charge fees comparable to Polygon, and those wanting higher security can still use more expensive ZK Rollup. This brings all users' choices under one roof and enables complete interoperability between them.

zkPorter architecture. Source: Matter Labs

In my view, this is the future direction. Simply saying, "There are no dumb users; you made the wrong choice, I don't care if the fees are lower there," is no longer enough. We shouldn't shut the door on further scalability innovations.

As for Rollup, that's where I'm placing my bets. But who knows! I'm wrong more than I'm right, and if users embrace Rollup from the get-go, it's good for them and good for Ethereum.

<This article was frantically written on a plane, so please forgive any oversights. Almost all content mentioned in this article is invested by Dragonfly. Thanks to Ivan and Celia for their brief comments.>