How The Blob Saves Ethereum — EIP-4844’s Rollup Rodeo

My wife was just starting to take Ethereum seriously when I told her about scaling.

She mistook me saying “sharding” for “sharting,” and I lost a year’s worth of concept legitimization.

Needless to say, “proto-dank sharding” had her rolling with laughter.

Anyways, EIP-4484. Let’s start with the TL;DR version.

TL;DR

EIP-4484 creates “blobs” that attach to Ethereum’s existing blocks. These blobs are purpose-built for Layer 2 rollups and give them some guaranteed, predetermined place to store their data that eventually ends up on Ethereum. These little tag-along blobs are automatically deleted after 30–90 days, making rollups much less expensive. We end up with a bottom-up, market-driven approach to scaling Ethereum. Cool!

The Problem

Ethereum doesn’t scale well. It goes back to the ol’ Blockchain Trilemma, in which blockchain is posited to have three features: speed, security, and decentralization. You get to pick two out of three to scale.

This trilemma is why blockchain technology hasn’t seen wider adoption.

Secure and fast? You just reinvented Visa.

Secure and decentralized? Takes forever — ain’t nobody got time for that.

Fast and decentralized? That’s where all the exploits and hacks come from.

The (Proposed) Solution

Ethereum has outgrown its clothes due to the endless feast of new users and builders seeking to leverage the emerging world computer as their ecosystem of choice. It’s suffering from success!

To scale, Ethereum must do the unthinkable: Shard its pants off.

Was that one worth the setup? Doesn’t matter. I am a dad. There’s this ongoing joke/observation in the US about how once you become a father, your puns and jokes become increasingly corny and stale.

So far that’s proven true. Apologies, dear reader, but there will be more.

Sharding entails breaking up the main Ethereum blockchain into a bunch of little chains that run in parallel. These shards stay aligned by leveraging the Beacon Chain for coordination.

Here’s a picture from Big Daddy Vitalik on how that works:

The (Defunct) Solution — Shard Yourself

It turns out that breaking the world computer into over a thousand pieces while maintaining coordination is really hard. Who knew!?

In the meantime, people began creating quasi-trusted environments that operated off-chain to keep the party going, eventually posting compressed versions of their data back to Ethereum’s main chain — clever… and cheap NFTs for the time being.

And it turns out this approach sort of… worked. Really well.

From Polygon to Optimism to ZKSync, an entirely new ecosystem exploded on the scene, boasting naturally aligned economic incentives to scale the Ethereum Mainnet.

We ended up with an even better way to scale Ethereum, which is all that really matters at the end of the day (or decade. Take your pick.)

The (Actual) Solution — The Rising Rollups Regime

You’ve probably heard of rollups by now. I intentionally avoided the term until now. If they’re totally new to you, don’t fret. Rob over at CoinDesk does a fantastic job breaking down rollups from a high level.

Basically, they’re self-contained systems that allow transactions to happen outside the main blockchain. After a set period, they bundle, or roll up (clever, right?) these transactions and submit them to the main blockchain, saving tons of money in transaction fees.

It’s a clever solution but not a silver bullet.

Let’s use dropshipping as an example to explain why. Those who get to the party early have a great time — there’s low competition, and selling cheap Chinese blenders on Amazon can make you rich. But then the masses get involved. Before long, you have margins so compressed they make a 240p YouTube video look sleek. If you were strategic in the early days, you can corner the market. If not, you’re fighting for air with the rest of them.

Same thing with rollups. Polygon took the limelight for being an Ethereum-centric, independent blockchain. Now they’re deploying rollups left and right, and their zkEVM is in the limelight. But these rollups have to fight for block space on Ethereum, which gets very expensive very quickly. Simply being a rollup just isn’t sexy enough anymore — you have to fight for your margins (or, in this case, users) like the rest of the crowd.

To maximize market share, we end up with general-purpose rather than application-specific rollups. General purpose rollups that can act as centralized gatekeepers to the main chain.

Same problems, new layer. So… rollups aren’t good after all? 🤔

NOT TRUE. That’s why I’m writing this article.

And Whose Data is it, Anyway?

Before revealing the blockbuster solution that you’ve already seen in the TL;DR section, let’s touch on another challenge facing today’s rollup regime: data availability.

Whether they’re ZK or Optimistic rollups, they have a serious problem. They don’t talk to one another and are constantly fighting over block space. This drives up transaction costs and clogs the network, transforming rollups into ouroboroses eating their own tails.

But it’s not their fault. In fact, it’s actually best practice.

You see, rollups call data from Ethereum to determine the canonical state of other rollups. It’s simply the safest way to minimize attack vectors while ensuring that Ethereum serves as the single source of truth between network participants.

That makes sense. One of Ethereum’s biggest draws is its security. Why trust a third-party rollup when you can rely on the world computer? All is right with the world.

Buuut this means that data availability is now the number one problem on Ethereum.

What we need is a way for rollups to queue their bundled transactions in an orderly fashion while simultaneously ensuring that these attestations are optimistically canonical.

If only we had some sort of — GOOD NEWS, WE DO!

>EIP-4844 KICKS IN THE DOOR

Have you ever looked at SegWit and thought, “Man, I wish that were me…” Well, Ethereum definitely has. EIP-4844 uses a similar conceptual approach to segregating bundled transactions from rollups, relegating them to a rather chunky tag-along piece of data they call a blob.

Blobs are basically like side-carts you see on a motorcycle, whereas the actual Ethereum blocks are the motorcycle itself. No more rollup transactions holding your hips while you ride the open road!

The blob is incredibly sexy because it solves several multi-dimensional problems in a single network upgrade, creating not just a stopgap for Ethereum’s scaling but introducing a configuration with significant potential for underpinning the entire thesis of scaling.

Here’s a highly professional diagram we can talk about. It took me way too long to make in Miro:

Wrapping it Up

EIP-4844 is a thing that does a few really big things for the Ethereum ecosystem. Big things like:

• Scalability: Before the EIP-4844 upgrade, all rollups had to fight for block space, just like the rest of us. Post-EIP-4844, these rollups have a dedicated, partitioned area for their data, allowing for greater predictability and much cheaper transaction costs for rollups.
• Communication: Before EIP-4844, rollups were implored to use the Ethereum chain as their single source of truth regarding the canonical state of their rollups. Post-EIP-4844, rollups can query data from blobs to make optimistic assumptions regarding the state of other rollups.
• Diversity: Before EIP-4844, rollups were economically incentivized to bring general-purpose solutions to market to maximize their market share. Post-EIP-4844, application-specific rollups can provide maximally applicable solutions to market niches due to deprecated network administration costs.

It’s a nice EIP. I don’t think it deserves a name like “proto-dank sharding,” but given everything it’s poised to do for the ecosystem, we’ll let it slide.

Catch ya in the next one!

Oh, and follow me on Twitter, LinkedIn, or add me on Discord or Old Scool Runescape.