EIP-4844 would effectively reduce rollup fees by up to 100x- even before danksharding.
Let's explore how it works under the hood ๐งต ๐
Calldata is read-only data used by smart contracts for external calls. Rollups currently utilize L1 calldata for data storage, which is limited to ~10KB per block.
EIP-4844 is a stop-gap solution until danksharding is fully implemented. It introduces the transaction format that danksharding will eventually use, giving up to ~1MB average per block for data storage!
Instead of using calldata, it introduces new blob-carrying transactions, which are used for data storage by rollups. These are independent from EVM execution, and pruned after a month or so to not burden nodes with additional storage overhead requirements.
EIP-4844 is one of the first steps needed to turn Ethereum into a unified settlement + DA layer, and is forward compatible with all future sharding plans. This being said, it is still a work-in-progress EIP, but it's very promising!
crList is a key to a censorship-resistant modular blockchain, but what exactly is it?
Breaking it down ๐งต ๐
In the future, when PBS separates block proposals from a smaller subset of builders, how do we make sure the proposed transactions aren't censored or blocked by builders?
There are many ways to solve for censorship, but it's important to use a method that doesn't use a lot of bandwidth to keep requirements low.
What's Data Availability Sampling (DAS), and how does it factor into a modular blockchain? Let's break it down.
๐งต ๐
Data availability is an important area of active research. In order to scale a blockchain, the data needs to be both stored efficiently and retrievable for nodes (which may not be able to store the entire blockchain state).
Data is stored in blobs, which are made up of the original data, extended data, and proofs.
Why are Layer 2s a superior solution vs low-fee alt-L1s? In order to understand this, we must abstract what a blockchain actually is from a macro level.
A ๐งต ๐
While Bitcoin pioneered the blockchain for currency, it is clear they can do much more, and the rise of smart contracts proves this. Modern blockchains now provide a service: secure, decentralized data publishing and distribution.
Fees are dynamic according to how much users want to pay to add/change data at any given time.
Saying "ETH's fees are too high, nobody will use it!" is equivalent to saying "Too many people are driving on the highway, nobody will use it!"
You've heard of EIP-1559, but what if it was applied to more than just gas? Enter Multidimensional EIP-1559, a proposal from @VitalikButerin.
Let's break it down in a ๐งต ๐
EIP-1559 smooths out gas costs, reducing gas spikes due to dynamic blocksize and pricing. This reduces usage bursts in the network, which makes it easier to use at any given time instead of pricing out users at random intervals.
This has been a success in reducing waiting periods for users among other improvements, as shown in this research paper: arxiv.org/abs/2201.05574
What's Danksharding and crList, and why are they so groundbreaking?
A ๐งต๐
Danksharding gets the name from the core dev who proposed it, @dankrad. Instead of the previous design which was simply parallel data shards, Danksharding flips Ethereum upside down with a new role: the block builder.
The builder grabs all of the existing L1 data as well as rollup data and puts them together into blocks which are then broadcasted to the network. Since we're giving builders this massive power, there needs to be a check involved to avoid censorship, which is crList.