Ethereum has blobs. Where do we go from here?

https://vitalik.eth.limo/general/2024/03/28/blobs.html

The Dencun hard fork has activated, and thanks to Blobscriptions the blob fee markets are now in "price discovery mode".

It has been well-understood for years that the future of Ethereum scaling depends on rollups backed by data space secured with data availability sampling. EIP-4844 is a key change that lays the groundwork for this future.

The next step is verifying blobs with PeerDAS, a conservative version of DAS that can offer some additional scaling. At first, we'll likely see some clients full-verifying and others doing DAS checks.

More and more advanced versions of DAS can be introduced over time.

Meanwhile rollups also need to improve, in four key ways.

First, data compression.

Second, optimistic data constructions, both Plasma and other "hybrid data availability" ideas that are designed to maximize security.

Third, improve on execution-related constraints, both parallelization and storage-related issues.

Fourth, continue improving on security.

We need to raise our standards. By the end of the year, I think we should only refer to rollups that are at least stage 1 as "rollups".

Moving to stage 2 (fully trusting code) is the next step and will require moving carefully.

What does this mean more broadly?

I argue that now that the merge and EIP-4844 are done, we are decidedly on the right side of the S-curve. Further L1 changes will be quite meaningful and significant, but relatively milder.

For a long time, the Ethereum ecosystem has been content to develop applications for a relatively small bubble, and high fees were a good reason why we could not go further. Today, no longer. And so, we no longer have an excuse.

Fortunately, the Ethereum ecosystem is already rising to the challenge.

Going beyond scaling, applications that are built today need to be built with "2020s Ethereum" in mind. Here's what this means: