Gnosis founder Martin Köppelmannn says implementing a web of 128 “native rollups” can save the Ethereum network from being captured by corporate interests and vanishing into obscurity.

Native rollups are layer-2 (L2) networks built to Ethereum’s native security standards, meaning they’re composable, uncensorable and economically aligned with the original ethos of Ethereum. 

Unlike “centralized” rollups — which include L2s like Coinbase-incubated Base or “based rollups” like Taiko — native rollups offer all of the inherent security and composability of Ethereum. 

Native rollups would be “built and governed by” Ethereum core developers, with synchronous reading to the L1 and all data availability, maximum extractable values (MEV), and congestion fees being paid back to the mainnet. 

Köppelmannn argues that implementing a swathe of these so-called native rollups could hold the key to solving the problem of liquidity and developer fragmentation that currently plague the network. It would also return the network to the original sharding vision abandoned in favor of rollups.

The proposal received a warm response from Bankless podcaster Ryan Adams, who praised the approach, saying, “They only add [to existing rollups], they rebalance power, they make ETH stronger.”

However, Uniswap CEO Hayden Adams said he preferred a “middle ground” approach of an enshrined proof system — which would look like a sequencer-agnostic L2 proving system.

Adrian Brink, co-founder of the “intent-centric” blockchain firm Anoma, told Cointelegraph that the difference between native and based rollups is largely just “made up” for marketing purposes.

“There is no defined computer science terminology for a ‘native’, ‘based’ or ‘vanilla’ rollup. Based just means the rollup is sequenced by the base layer. Native just means there is a native opcode to verify the rollup’s execution,” he said. 

The risks of “centralized” rollups

Köppelmannn is the founder of Gnosis, an Ethereum infrastructure firm that’s heavily prioritized on decentralization. Gnosis is also behind Safe Global and Gnosis Chain, which together form the bedrock of infrastructure that much of the Ethereum ecosystem rests on.

Speaking on stage at Devcon in Thailand last week, Köppelmann, clad in a Tornado Cash T-shirt, asserted that Ethereum needs to distance itself from centralized layer-2 rollups and deploy its own web of ZK-proven networks, all built to the high standards of Ethereum’s layer 1.

“I have absolutely the highest respect for Jesse [Pollak] and for what Base and Coinbase are doing, but I do think saying ‘we’re bringing the next billion people to Ethereum’ is wrong,” he said.

“You’re bringing the next billion people to Base.”

Köppelmannn warned that Base and any other centralized layer-2 network are in charge of how much profit they make, which gives them different incentives.

“It’s entirely in their hands to control how many fees they charge for transactions, and they’re perfectly capable of doing something where they take a 30% cut,” Köppelmannn said, referring to Apple’s controversial fee on all sales conducted through its App Store. 

Why does Ethereum need “native” L2s?

Koppelman said rollup developers are typically presented with two main choices: centralized rollups, which rely on a centralized sequencer, or “based” rollups.

Centralized sequencing is good for optimizing user experience and fast transactions, while “based” rollups are slower but optimize for deeper connectivity to the L1. 

“If most assets are not bridged from Ethereum, and the sequencing is also not done by Ethereum, the connection to Ethereum is reduced to occasional checkpointing,” he said. 

Gnosis, Layer2, Features

Köppelmann said rollups are presented with two main choices: centralized or based. Source: Ethereum Foundation

Unfortunately, Köppelmann said neither centralized nor based present enough in the way of major benefits to the Ethereum ecosystem in the long term. 

His solution is to create and deploy 128 equal and interoperable native rollups on Ethereum — no multisignature wallets, two independent proof systems, and rigorous testing of code. 

“These L2s would be highly interoperable with L1, fulfilling Ethereum’s early promise to provide sharding using L2 technology,” said Köppelmann. 

Brink said he agrees with Köppelmann, at least in part, saying that Ethereum should look to make a shift away from its reliance on privately owned ecosystems in the long term.

“In theory, you can still have native rollups with centralized sequencers, depending on the definition — but native rollups are definitely an improvement from the more centralized L2 status quo,” he said.

“Many rollups today are glorified multisigs. Native rollups would have much better censorship resistance and improved composability.”

Köppelmann argued that while Base and other L2s claim that their networks share the “inherent security” of Ethereum, this is very far from the truth, noting that funds can be stolen in several ways on L2s that aren’t possible on the mainnet or other sufficiently decentralized L2s.

He also made a nod to an earlier Devcon talk by Ethereum developer James Prestwich, who provided several examples of how centralized sequencers can censor and manipulate decentralized finance protocols like Aave. 

Related: Ethereum DApp volumes gain 38% in a month — Will ETH price follow?

Brink said that while native rollups would, theoretically, make it “much easier” for developers to deploy new rollups and add new programmability features, they still aren’t what he considers a true scaling solution.

“This is because the amount of data that needs to be posted onchain still increases with the amount of usage they get,” he said. “The only real scaling solutions today are Plasma-like constructions, where you can post a constant amount of data onchain regardless of how much activity is happening.”

“This is an old idea that fell by the wayside because we didn’t have advanced ZK tech back then, but now we do, and Plasma needs to be explored again.”

Ultimately, Brink said the most important goal for Ethereum’s scalability is to focus on “generalized intents” — a single intent standard that can be shared and understood by all users, apps and solvers across the ecosystem — and do away with the fragmentation that currently mars Ethereum. 

Magazine: AI agents trading crypto is a hot narrative, but beware of rookie mistakes

“We need to stop innovating at the margins. We need compelling user stories and a renewed wave of innovation in applications that people want to and love to use.”