Ultra Scalable Ethereum?

tl;dr: But does it really scale? That’s the big question facing Ethereum.

There is a whole heck of a lot that I don’t understand about Distributed Ledger Technology, of which blockchain is one type.

I’m relatively technical, but not a jedi master of it, so I’m sure there are things I am not seeing.

Plus, one of the most common lessons or marketing throughout history is that the best technology doesn’t always win.?First-mover advantage matters. So does ‘skin in the game,’ momentum, and the?Lindy effect.

All that being said, my doubts about Ethereum’s long-term viability have increased as of late.

I am a huge fan of the Bankless community and really enjoy their newsletter. They were probably the first place where I heard Ether referred to as “ultra-sound money,” following the passage of?EIP-1559, in which a portion of all transaction fees are “burnt” in order to increase the value of the token.

I know they are big-time Ether bulls, probably because they both believe in it and their Ether bags, as they say, are quite heavy.

So, I was excited to read their perspectives in the lengthy write up called?Ultra Scalable Ethereum.

They wanted to make the case the “modular” blockchains are the future and that “monolithic” blockchains are the past.

It’s a strong position to take because it both takes a swipe at Bitcoin (monolithic) and rides the wave of where many other chains (e.g. Near, Avalanche, Solana, Tezos, etc.) are going in terms of their design (modular).

Sharding and Roll-ups

The basic hypothesis behind Ethereum’s planned scalability is so-called “sharding” and “rollups,” which takes the singular/monolithic blockchain we all know and love, and splits it into shards, coordinated by a so-called “Beacon” chain and then accelerated by “roll ups” which allow for faster execution and lower fees.

At a high-level, it makes sense.

If you have ever used the Ethereum main chain and experienced its gas fees and then used a roll-up, such as?Optimism, Arbitrum, or a zk-rollup, the first time you do is like a revelation.

Compared to the main chain (‘Layer 1’), these Layer 2 (‘L2’) solutions are lightning fast and much, much cheaper. It’s great (until it gets too crowded and the fees start going up). Still, these were necessary?market responses.

The Underlying Assumptions of Ethereum Scaling

There were three underlying assumptions that jumped out at me while reading this essay that I am not quite ready to accept, which leads to my doubt.

DLT and Blockchain

The first?is that the only/best form of decentralized ledger is a blockchain.

Yes, blockchains are amazing in many ways, but we are still relatively early in this transformation and there are many other approaches out there to solving the problem of a centralized database that don’t necessarily do it through a specific blockchain. That’s a subtle, but important distinction.

Remember, we used to think that the best way to search for a website was by having humans curate them…then Google released Page Rank. BTW, Google was the 16th search engine to come on the market.

EVM

Second, is that the?Ethereum Virtual Machine?(EVM) is a given as the execution environment. While it is the most commonly used and perhaps the most widespread-at the moment- it is incredibly inefficient and makes fast decentralized applications on a Layer 1 near impossible.

This is why, for example, the team at?Dapper Labs, which basically unleashed the NFT craze in 2017 with?Crypto Kitties, decided to build and release their own blockchain,?Flow. They didn’t have confidence in Ethereum’s ability to scale.

Composability

Third, and perhaps most importantly, was that sharding and rollups break, or at least seriously inhibit, so-called “composability.”

Composability is crucial for the dApp space. It’s essentially the ability for any developer to use/re-use any other piece of code from anywhere on the network to “compose” a new application, without having to rebuild it from scratch.

This is, for example, what allows?Yearn?to run an optimized interest bearing vault service because it can pull in?Compound, Aave, and others without having to get API permission or, more importanly, re-write the code.

However, when these dApps start living on different shards (and within different rollups), it’s much more difficult to make that work.

There are ways, one of them is known as “bridging,” but, like in the real world, if you build a bridge, you need to maintain it. Plus, with each shard (and Ethereum wants to have 1024 of them, plus rollups spreading out from there- as I understand it), I suspect you would need to build bridges from each of those. That’s a lot of bridges to build and maintain.

I may not be understanding exactly how that part works exactly, but I found it really fascinating that nowhere in the article at all did it mention the word “composability.” This is really one of the killer features of the DLT/blockchain revolution and, for me at least, it’s a big miss.

The Path Forward

I did think that the presentation of Ethereum’s token economic design was really strong in the article, explaining the difference (and importance) between issuance and transaction fees. In that, I thought it was spot on. There are way too many chains out there where the issuance is through the roof, but nobody looks at it…kind of like how people look at the Dow Jones index, but not the?M2 money supply. You need to understand both.

However, the part that lit me up was in the conclusion, which I just felt was total hubris:

The point is if we went back in time, or hopped to different parallel universes and re-rolled the dice 10,000 times, the crypto industry would find itself at a modular design conclusion 99.9% of the time.

I just find this non-sensical as he’s calling the marathon (wishful thinking?) in after the first mile.

Ethereum’s innovation in the form of the concept of decentralized smart contracts is one of the most significant of my lifetime. That I believe. Vitalik Buterin’s contribution is second only to Satoshi him/herself.

And Ethereum may definitely emerge as the standard. VHS beat Betamax after all. They have the money and they have the developers.

But just because you are up by 2 goals in the first 20 minutes doesn’t mean you have won the game.

With most of the world’s developers (and people) not yet in a decentralized network, there’s still a lot more to learn and to unfold.