By
Dimitar Bogdanov
October 2, 2023
4 Min Read
Proto-dankshardigng takes us a step closer to a much more scalable blockchain technology. Danksharding is seen as the perfect complement to Layer 2, but it requires several upgrades to get to that stage. EIP-4844 and proto-danksharding represent the first steps toward realizing that vision.
In a nutshell, proto-danksharding seeks to introduce a way for rollups to add cheaper data to Ethereum. This is made possible by introducing a new type of transaction that will enable blobs of data to be added to blocks. And, what’s interesting about this ‘blob data’ is that it doesn’t need to be forever stored on the blockchain. Here are the details.
On these pages, we’ve discussed rollups extensively, examining how they work and why they are important. But since proto-danksharding is specifically aimed at making rollups even better, a quick refresher is in order.
So rollups are Layer 2 scaling solutions for Ethereum that are designed to process transactions off-chain, batch many transactions together and submit them via a single transaction to the Ethereum mainnet. This saves a lot of gas, but there’s more room for optimization. This is because rollups need to make that data available for review by independent actors. So the data is stored on the blockchain forever. However, the data is needed only for a time, so there’s really no need for it to be stored in perpetuity. This is what EIP-4844 and proto-danksharding aim to address.
At its core, EIP-4844 aims to establish a method that would allow rollup data to be deleted after a certain period of time without compromising the integrity of the ledger. Under the method described in the proposal, rollups post commitments on-chain while keeping the actual transaction data in blobs. Let’s elaborate further on this.
The aforementioned commitments are essentially cryptographically reduced expressions of the blob data. To produce a commitment, a rollup has to fit a polynomial function to the data and then evaluate that polynomial at certain points determined by randomly generated numbers. The result is then wrapped up in a cryptographic function and submitted to the chain as a small cryptographic commitment.
The validity of the commitment can then be easily checked by a prover. To do so, the prover evaluates the polynomial at the same points. If the results match, the commitment is valid, but if the values are different, that’s an indication that the data has been changed. The prover can challenge the data if they think it’s wrong.
The best part is that for proto-danksharding to work, the blob data only needs to be stored in nodes for a sufficient period of time. So, after a fixed period of time (1-3 months) the blobs are deleted to avoid unnecessary data bloat. However, this doesn’t necessarily mean that the transaction data is lost, just that it’s not stored on-chain. Rollup operators, users or other parties can still store the data off-chain.
If you’ve been following the Web3 sector and Ethereum’s evolution, in particular, you may recall that sharding was once envisaged as a Layer 1 scalability solution for the world’s second largest blockchain protocol. So is EIP-4844 part of that vision? Well, the short answer is no. Instead, it’s more in line with the ongoing shift towards L2 solutions like rollups.
And indeed, danksharding and proto-dankshardng have little in common with traditional sharding, which is a technique for splitting a large database into several smaller ones to reduce load and eliminate potential bottlenecks. Instead, danksharding utilized distributed data sampling across blobs, which is much easier to implement and is complementary to the rollup push.
So what can we expect in the future? Well, EIP-4844 is expected to usher in an exciting new chapter for Ethereum scaling, but that’s only the beginning. The next step will be realizing full danksharding, which would involve dramatically expanding the blobs attached to blocks, from one in proto-danksharding to 64. Other changes would also be needed to ensure that the new large blobs can be handled.
Full danksharding is currently still years away, but work on implementing EIP-4844 is progressing well and can reasonably expect to see proto-danksharding realized in the foreseeable future. And with that comes the exciting prospect of a much more scalable Ethereum, with cheaper transactions made possible by a robust rollup ecosystem.
