Hermez is a decentralised layer 2 zk-rollup. This means that it’s a scalability solution that allows users to transact at a very cheap cost and with better throughput than Ethereum layer 1.
But why is it so hard to scale blockchains in the first place?
The so-called “scalability trilemma” states that having a system that is decentralised, scalable and secure is an intrinsically hard problem (it is much easier to have any two out of these three properties).
Hermez Network is built upon the security of the Ethereum network, so we can focus on having better scalability and maintain decentralisation: Hermez can process thousands of transactions per second while, at the same time, allowing anyone to participate in the validation process as a coordinator.
The coordinator is a key piece of the Hermez Network. It is the actor that will collect transactions from the network’s users and compute a validity proof using zero-knowledge cryptography. This is called a "batch." Then the coordinator will post this proof in the Ethereum mainnet with the compressed data of all the transactions in the batch.
Because the validity proof and the compressed transaction data are much smaller than the total aggregate of the individual transactions, the amount of gas needed is quite small, and the cost per transaction is reduced significantly. When you divide this cost among all the transactions that have been “rolled up”, zk-rollups can be extremely gas-efficient.
But scalability solutions sometimes rely on having a single party to do the work of the coordinator or equivalent.
In Hermez, we ascribe to permissionless decentralisation - encapsulated in the fact that anyone can become a coordinator - to avoid the many pitfalls of giving power to a single actor.
Among all the potential coordinators: how will the protocol decide who should roll up the next batch?
The Auction Mechanism: Proof-of-Donation
A coordinator collects transactions, computes the validity proof, and posts it with the transaction data onto the Ethereum chain.
To accomplish that they pay for the computation of the proof (electricity, hardware…) and for the gas of posting proof and data on mainnet, and in exchange receives the fees from the transactions. The more transactions they are able to process in a batch, the more fees they will receive.
Hence, we want the coordinator candidates to bid in an auction for the right to be the coordinator during a slot. Slots are time periods 40 Ethereum blocks long, which is around 10 minutes.
The candidate who wins the auction for a particular slot, will have the opportunity to roll up the batches and collect fees for its work during this time period.
In order to ensure that everyone can be a coordinator, the Hermez team will open source all the necessary software and tooling to become one.
What will happen with the proceeds of the bid?
- 30% will be burned
- 40% will be donated to Gitcoin Grants
- 30% will be used to reward network participants
(These percentages and the donation address are subject to change and will be modifiable by the network governance - community council)
Why this particular distribution?
The reason to burn and donate most of the tokens is to avoid potential game-theoretical scenarios.
For example, an actor with a big participation in the network can afford to bid disproportionately high amounts for the slots because they will receive this money back through the cunning capture of incentives.
This would create unfairness in the bidding process, discourage other participants and eventually result in centralisation.
Nevertheless, burning most tokens did not seem like the most efficient way to utilize the value generated by the network, and we wanted to direct some of it to giving back to Layer 1. After all, there is no Layer 2 without Layer 1!
After careful consideration, we believe that Gitcoin Grants and its quadratic funding mechanism of allocating funds is hard enough to game for a coordinator so it doesn’t introduce additional risk, compared to burning the tokens - so the match is made in heaven: giving back to the ecosystem while avoiding potential attack vectors for the network!
If you are interested in becoming a coordinator, stay tuned for an article that we will publish in the following weeks. In it, we will detail the steps and show the software necessary to start bidding, validating and collecting fees!