Recent Critique of Ethereum’s Scalability Solutions

Hi everyone, I came across this discussion on Twitter around Ethereum’s scalability solutions. Adrian Brink from Namada has challenged its effectiveness and has expressed concerns about decentralization and scalability. Thought I’d summarize his thoughts here but if you wish to join the discussion, here’s a link of the exchange on CT.

Ethereum has progressed from ‘Payment Channels’ to ‘State Channels’ followed by ‘Plasma’ and ‘Plasma Cash’ and now to ‘Rollups’. So far, none of these proposed solutions have been effective in achieving real decentralization and scalability as there is a tendency within the blockchain community to ignore potential shortcomings.

Brink asserts that Rollups are not fully optimized in their current form and warns that within the next 2 years the first Rollup operator will probably crash at scale and cause the loss of 20bn USD similar to that of Terra’s one with Luna as its main asset and UST as its algorithmic stable coin. (Not sure it’s fair to make such a comparison but I believe Adrian has a point.)

For example, EIP-1559 does not account for off-chain collusion as fraud-proofs are either permissioned or not implemented. In addition, there is a bounded cost on purchasing all ETH gas. This means that there is a maximum value that a Rollup can safely custody and looks like we are all ignoring it.

Miners can also artificially keep the base fee low to make a DoS attack possible and that there is a real possibility that even if fraud-proofs for Optimistic Rollups work, they are still never going to be secure due to Ethereum’s gas constraints. Supposedly if an attack can only be sustained for 14 minutes, and a large coordinated attack took place, billions could be lost. This is only a hypothetical but it's better to prepare than to be sorry.

Having summarized these highlights, I’m curious to know what everyone else thinks. Do you think that the Ethereum community has a culture of over-optimism when it comes to newer progression?