What You Need to Know About EIP-1559
On August 5th, 2021, Ethereum activated its London hard fork, one of the most anticipated upgrades to the network in years. Bundled in with other changes was EIP-1559, which addresses Ethereum gas fees, mining rewards, block size, currency supply and more.
An EIP is an Ethereum Improvement Proposal, which consists of suggestions for improving the way Ethereum works by altering the code the blockchain runs on. Anyone can submit an EIP, but they tend to get more attention if they recommend usable fixes for major network problems. It also helps if they’re proposed by individuals or groups with a track record on Ethereum.
EIP-1559 was created by a group of famous Ethereum names, headed up by Vitalik Buterin, and is intended to modify Ethereum’s restrictive and expensive gas fee price-setting system. Profound and far-reaching, EIP-1559 was much discussed before activation and much analyzed after. In this post, we’ll talk about what it is, what it does, and what its effects will be on the Ethereum ecosystem.
EIP-1559 101: what is it?
EIP-1559 proposes to address several problems that make Ethereum harder to use. Ethereum is general-purpose, not a single-use blockchain like Bitcoin, whose sole purpose is to support the BTC cryptocurrency. Thus, Ethereum has multiple dapps and smart contracts running inside the blockchain, which trigger transactions when conditions are met or users instruct them. Ethereum miners charge gas fees to enter transactions into a block and validate it.
Until EIP-1559, gas fees were set by a simple first-price auction. Users offered a bid for including their transaction in a block, and miners selected the transactions offering the highest bids and validated them.
From the miners’ point of view, they were rewarded for including transactions in blocks through gas fees, as well as block minting rewards, creating new ETH with each new block. Naturally, they tended to look for the highest possible fees, so blocks were usually filled first with large, well-funded and non-time sensitive transactions.
Bulk transactions from slow-moving projects fitted the bill nicely; fast-moving, small, heterogeneous transactions did not. But as DeFi, which relies on such transactions, became more popular, it became apparent that Ethereum’s fee structure needed attention. And with block sizes fixed, the only elastic factor in the equation was the gas price, as that rose so did the transaction fees.
EIP-1559 identifies the following four areas of concern:
1: Mismatched transaction fee volatility
Prices aren’t just high, they’re volatile — and they’re mismatched with true network costs. It’s like paying anything between $2.50 and $700 for cab fare: on average it’s far too high, but it also has nothing to do with the cost of operating the system.
2: Slower experience for users
Gas per block is limited. Transaction volumes fluctuate as well as prices. The result is that transactions are often waiting for several blocks before being processed. That’s no big deal if it’s just a person waiting on the other end; Ethereum block time is 10 to 20 seconds, so missing three or four blocks might take a minute. But if it’s one step in a multi-stage, algorithmically-managed DeFi transaction, a minute’s wait may as well be a week. There’s no inherent reason for this to happen; it’s the consequence of a misaligned reward system rather than of some law of nature.
3: Inefficiency of first-price auctions
The process of setting fees by first-price auction is inaccurate, inefficient and clumsy. Complex fee estimation algorithms attempt to bring some fairness to the process, with limited success. Meanwhile, over-payment remains common. The goal of the auction, to find the ‘true price’ of the transaction, is missed. It would be possible to use a second-price auction system in which every transactor included simply paid the same fee as the lowest gas fee included in the block, but as Vitalik Buterin points out, such auctions may be “efficient in a traditional economic analysis, but have the flaw that they are vulnerable to collusion.”
4: Technical issues around block reward
PoW blockchains like Ethereum and Bitcoin reward miners for validating blocks by generating new coins (mining reward) and paying them out of users’ fees (transaction reward). But Ethereum is planning a transition to Proof-of-Stake, which will eliminate block rewards for miners. When that happens, miners won’t be needed but block validators will; they will be rewarded solely out of transaction fees. The pre-1559 fee schedule would tend to reward validation of empty blocks (as many as 2% of Ethereum blocks may be empty), and open attack vectors to the network. It’s a source of instability and there is no good solution without addressing fees.
Transaction fees, block size, and tipping
EIP-1559 proposes to address transaction fees by splitting the fees in half. Now, Ethereum calculates gas fees using two separate methods: base fee, and tip. Base fees are calculated algorithmically based on the degree of congestion on the network, and crucially, they use the blocks prior to the one under consideration, meaning they’re based on the congestion of the network, not the fullness or otherwise of the block.
How base fees are calculated
Base fees don’t consist of a fixed number of gwei. Instead, base fees are calculated based on the fees to be included in the previous blocks and then adjusted up and down based on how congested those blocks were.
Totally empty blocks result in a 12.5% reduction in base fee, totally full ones a 12.5% increase. The result looks like this:
Base fees are susceptible to quite rapid increases and decreases:
There is a strong incentive not to default to 100% full blocks every time, because there is a rapid rise in base fees following a 100% full block.
In addition, miners now have no incentive to raise base fees by validating junk transactions or empty blocks, because they don’t receive base fees. Since EIP-1559 was implemented, base fees are burned by being sent to wallets to which no-one has the private keys.
The implications of this for Ethereum as an economic entity and especially for the core token are profound, since high volumes of small transactions will tend to make Ethereum deflationary, and indeed over $100 million in ETH has been burned since EIP-1559 went live with the London hard fork.
Users can also offer a tip, a direct incentive to miners to process their transaction, and miners do keep tips. So competitive bidding is alive and well on Ethereum block space.
Block size is now adjustable
However, gas fees are no longer the only elastic element in the equation. Ethereum blocks are now flexible too. Ethereum always has had more block size flexibility than some blockchains: where Bitcoin, for instance, has a hard 1MB block size limit, Ethereum’s block size limit was measured in gas rather than data and held steady at 15 million gas, with average blocks accounting for about 10 million gas and between 20kb and 30kb in size.
That block gas limit has now risen and has become variable. Base fee calculations aim to encourage a block size of about 15 million gas. When the block is smaller than this, base fees fall automatically; when it’s larger, they rise up to a prearranged limit of 2x average block size or 30 million gas, giving a maximum block size around 90kb.
Why not simplify it?
If the problem is pressure on block space, and there are plenty of people willing to pay, why limit blocks to 30 million gas? Why not just double block size, or increase it by an order of magnitude? This has been a popular proposal, because it would be much simpler than EIP-1559, however, according to Vitalik Buterin increasing the gas limit to 100M tomorrow will cause the average-case block processing time to increase from the current ~400 ms to ~3.2s. This would lead to centralization and trouble syncing, make data harder to retrieve and thus slower and more costly, and reduce the network’s resilience in the face of attacks.
A more flexible block size has specific advantages too. In the past even the most well-functioning block producers would leave a small amount of gas unused because there was too little space remaining to fit a another transaction. Now, blocks should more or less fit with their contents.
Summary
One of the major issues for Ethereum users face is high and volatile transactions fees. EIP-1559 increases transparency and visibility into how fees are determined. In theory this should help users to avoid overpaying for transactions, and reduce wait times in cases where users underestimate fees required by the network to process a transaction. It remains to be seen if the claimed benefits will materialize in practice, but we’ll have to wait a little while for the most popular wallets to update their UX for EIP-1559 transactions.
Loom Network’s mainnet Basechain doesn’t implement EIP-1559, and there are no plans to do so at this time. However, if you’re transferring assets between Basechain and Ethereum on the Staking Dashboard you’ll probably notice post-EIP-1559 UX changes in your wallet of choice when sending transactions on Ethereum. MetaMask recently rolled out an update to support EIP-1559, so if you’ve updated your MetaMask recently you might want to watch this short video from The Defiant which demonstrates how the updated MetaMask UX works.
Loom Network is building an ecosystem of blockchains for the next generation of DeFi protocols, NFTs, and high-performance multi-chain dapps. At the core of this ecosystem is our Basechain network — already live in production, audited, and battle-tested.
New to Loom? Start here.
Want to stake your LOOM tokens so you can earn rewards while helping secure Basechain? Find out how.
Got experience running Linux servers, and interested in running a Basechain validator node? Reach out to us in our Telegram channel and we’ll tell you all about the current requirements.
