The new wave of Bitcoin L2s are sidechains

I heard a lot of people saying that no one could define an L2 at Bitcoin Asia. The problem is that we have a definition, and most people just want to ignore it. Marketing, eh.

“Bitcoin L2s” are the hottest thing on the street. People are using a lot of jargon to distract users from trust assumptions and ruin Bitcoin Season 2.

Why all the sudden energy? See, about a year ago some teams figured out how to use Bitcoin as a data availability layer for rollups. Others have worked to improve trust assumptions regarding bridges (also called two-way linking). Research is making great progress and many projects believe we will have rollup-style blockchains in production by 2025.

2025? Some projects now claim to be on Mainnet?

Teams have picked up on this energy and are prematurely promoting the modular thesis for scaling Bitcoin. Projects are launching with bridge contracts on non-Bitcoin blockchains, marketing themselves as Bitcoin L2s. Infrastructure providers are amplifying their message and boasting that Bitcoin is back.

But these solutions don’t scale Bitcoin. They are completely independent, centralized side chains.

Layers they say? More like layers of trust assumptions.

Definitions

Many of these projects attempt to adopt the modular theorem for Bitcoin scaling. This essentially means that each aspect of the transaction lifecycle can be its own specialized system. Execution, transaction ordering, and data availability can all be managed by independent actors. Bitcoin will be the settlement layer underlying all of this.

It’s not a terrible thesis when you dive into it. But the current implementation on Bitcoin is a bit worse for wear and tear.

Many of the new projects claim to be ‘rollups’. Rollups would use Bitcoin for data availability, and post their latest state root, and enough transactions to recalculate the blockchain’s state from inception to Bitcoin. If they want to scale Bitcoin’s transaction throughput, they will also need a trust minimizedbridge contract where users can deposit money to earn on the rollup.

Dive into a few documentation sites and you’ll see that none of these new projects (in production) use Bitcoin for data availability. They want to use an alternative DA solution for performance reasons. This means that they want to be “validiums” or “optimiums”.

These constructions are similar to rollups. They are blockchains that similarly have a bridging contract with the parent chain, but use a different system for DA. This improves performance, reduces costs, but comes with some security tradeoffs.

In the validium design, the L1 contract would be responsible for verifying the validity certificate associated with a specific state transition for settlement. After completing a specific state transition, the Validium Bridge Contract can process withdrawals for users wishing to leave the chain, including unilateral exits that users can self-submit if the state data is available. Optimiums are similar, but rely on a tamper-proof mechanism instead of proofs of validity.

But none of the production implementations use a mechanism, on Bitcoin, that supports verifying SNARKs or fraud proofs…

Everything is verified on a completely different Layer 1 or their own authorized sidechain network!

Most of these chains use an Ethereum L2 SDK. They settle on Ethereum or on a fully centralized fork of geth that they have scraped together.

So there is no relationship with Bitcoin. Maybe it settles on Ethereum, uses the most popular DA layer and has a great execution layer.

But it’s not Bitcoin.

So side chains?

All new Bitcoin L2s are just modular sidechains. And when I say “modular sidechain,” I mean they use an alternate blockchain outside of their parent blockchain for performance purposes. They also make security tradeoffs by using an alternate DA layer for better performance.

Their bridge to Bitcoin? Run by multi-sigs.

So the general trust assumptions that users adopt are:

  • I hope multi-sig operating the Bitcoin bridge won’t hinder them
  • I hope the centralized sequencer will record and execute their transactions
  • Rely on the alternate DA layer to ensure data is readily available
  • I hope the centralized prover will post state transitions to the L1 contract OR hope centralized challengers will challenge malicious state transitions
  • Rely on the parent chain of the side chain to validate state transitions (finality)
  • Rely on an admin key not to upgrade the chain and steal user funds

Using a modular Bitcoin sidechain is fine if users know they are relying on a fully centralized chain and bridge program to use their BTC. A few projects are completely honest about this approach, and I’ve said publicly that I’m not completely against it from a go-to-market perspective.

The problem is that the majority of teams abstract away security details and try to give the impression that their designs are even remotely similar to modular constructs in Ethereum or other ecosystems.

Not all hope is gone

You may be reading this post and thinking the whole situation has gone to hell and isn’t worth investigating. It may feel like it some days, but there’s a lot of cool R&D work happening around improved sidechain designs.

Teams like it Citrea And Alpine Labs want to develop rollups on top of Bitcoin. There’s a lot of great work being done from the BitVM community and the ZeroSync team to improve bi-directional pin designs and develop a SNARK verifier that works today. This work also inspires a number of bridging proposals of various rollup and sidechain projects.

In these situations, you can’t throw away the good with the bad. It is not entirely hopeless. But all the nonsense we see in other ecosystems around complicated scale proposals, symbolic incentives and the roadmaps for ‘progressive decentralization’?

That comes down to Bitcoin a hundredfold.

So yes. These new chains are not L2s.

Leave a Comment