Have you ever written a smart contract that worked perfectly for years, only to realize a quiet network update could make parts of it stop working overnight? That is the situation many Ethereum developers face right now. A planned upgrade called Glamsterdam is bringing gas repricing changes that could disrupt a small but real group of Layer 1 contracts. The Ethereum Foundation issued a clear heads-up on this issue, and the timing feels important because the network keeps pushing for higher capacity.
I have followed these kinds of upgrades for a while, and this one stands out. It is not a dramatic overhaul of the consensus rules. Instead it focuses on how much gas certain operations cost. That might sound dry, yet the practical effects can be very concrete for anyone maintaining contracts that rely on fixed gas assumptions.
Why Ethereum Is Changing Gas Costs Now
Ethereum has grown a lot since the last major adjustment to state operation prices. That previous shift happened during the Berlin upgrade back in 2021. Since then the state has expanded significantly while validators have supported higher block gas limits. The network wants more throughput. Raising capacity without adjusting the cost of resource-heavy operations would create problems over time.
The Glamsterdam upgrade includes two specific proposals: EIP-8037 and EIP-8038. The first one changes how the network charges for creating new state. Think of new accounts, storage slots, and deployed contract bytecode. It introduces separate state-gas accounting. The goal is to slow down rapid growth of the blockchain state as capacity increases.
EIP-8038 takes a different approach. It raises the cost of accessing existing state. Operations such as SLOAD, SSTORE, cold account access, EXTCODESIZE, and EXTCODECOPY will become more expensive under the new schedule. Developers designed these numbers around a performance target that could support roughly three times the current base throughput. In my view that kind of forward planning makes sense, even if it creates short-term friction for some contracts.
The Foundation has been clear about the motivation. Repricing these operations is necessary before the network can safely increase capacity further. Without that step, state growth could outpace what the network can handle comfortably.
How Developers Identified The Risk
Researchers did not rely on theory alone. They replayed historical mainnet transactions under the proposed pricing schedule. The results fell into several groups. Some transactions stayed completely unchanged. Others succeeded but used different amounts of gas. A portion failed in ways that could be fixed simply by raising the gas limit. Then there was a smaller set that continued to fail even after researchers supplied substantially more gas.
That last group is the one that triggered the public warning. The Foundation described the potentially broken contracts as small in number. Direct outreach to the most affected builders has already started. Still, anyone maintaining Layer 1 contracts should treat this as a prompt to check their own code rather than wait for a personal message.
Most contracts remained unaffected during transaction replays, while many flagged cases could be corrected by raising their gas limits.
I find this approach practical. It gives the community concrete data instead of vague concerns. At the same time it leaves room for individual projects to investigate their specific situations.
Which Contracts Face The Highest Risk
Certain patterns appear again and again in the problematic cases. Contracts that rely on fixed gas stipends stand out. Solidity’s historical 2,300-gas stipend used by transfer or send is one clear example. Operations that used to finish comfortably within that allowance may now consume more gas because of the updated state-access costs.
Hardcoded call limits create similar issues. Logic that checks gasleft() and makes decisions based on remaining gas can behave differently once the cost of individual operations changes. Presigned transactions that carry fixed gas limits also sit in the risk category. If the actual cost rises above the pre-approved amount, the transaction simply fails.
These patterns are not exotic. Plenty of older contracts used them for good reasons at the time. The problem is that the underlying cost assumptions no longer hold. Raising the supplied gas limit solves many of the failures. In other cases developers will need to review the call sites more carefully and adjust the logic itself.
- Contracts using the 2,300-gas stipend through transfer or send
- Code with hardcoded gas limits on external calls
- Logic that depends on the exact value returned by gasleft()
- Presigned transactions that cannot adjust their gas parameter
Perhaps the most interesting aspect is how quiet these failures can be. A contract might continue to deploy and even accept some transactions while specific paths suddenly start reverting. That kind of partial breakage can be harder to notice than a complete outage.
Impact Beyond Smart Contracts
The warning does not stop at contract code. Wallets, RPC providers, indexers, and node tools all need updates. Any software that uses cached constants for gas estimation risks underestimating the required amount. When that happens the transaction fails on the user side even though the contract itself might be fine.
Tools that call eth_estimateGas and related functions must account for the revised state costs. Ordinary transfers offer a simple illustration. Transfers to existing accounts keep the familiar 21,000 gas figure. Transfers that create new accounts will incur an additional state charge. Software that assumes every transfer costs exactly 21,000 gas will produce incorrect estimates in those cases.
Regular users should not need to change anything manually. Updated wallets and infrastructure providers are expected to apply the correct estimates automatically. Still, the transition period always carries some friction. In my experience these moments are when clear communication from core developers helps the broader ecosystem stay calm.
Practical Steps For Contract Maintainers
Ethereum developers launched a long-running test environment called Platåberget, also known as glam-devnet-8. The network already runs the new repricing schedule. That gives maintainers a concrete place to test their contracts under realistic conditions.
There is also a checker tool that accepts a contract address and identifies historical transactions that diverge under the proposed rules. When the checker flags a transaction, the first response is usually to raise the gas limit and see if the failure disappears. If it does, the fix is straightforward. If the failure persists, the developer needs to examine the specific call sites and the surrounding logic.
I recommend treating this as a short audit rather than a full rewrite. Most projects will only need small adjustments. The ones that relied heavily on tight gas stipends or gasleft() checks will require a bit more attention. Starting the process now, while the public testnets are still ahead, removes a lot of last-minute pressure.
- Enter the contract address into the available checker tool
- Review any flagged historical transactions
- Test the same interactions on Platåberget
- Raise gas limits where that resolves the issue
- Rewrite call sites that still fail after the limit increase
The sequence feels manageable. It turns an abstract network change into a series of concrete actions that fit into a normal development cycle.
The Broader Context Of Layer 1 Scaling
Gas repricing sits inside a larger set of Glamsterdam work. Block-level access lists and changes to block construction are also part of the conversation. Together these pieces aim to bring Layer 1 scaling back into clearer focus. For a long time the narrative centered on Layer 2 solutions. That approach remains important, yet improving the base layer still matters for security, simplicity, and the overall health of the ecosystem.
The roadmap currently targets Glamsterdam for the fourth quarter of 2026. No fixed mainnet activation date has been announced. The final schedule will depend on stable client implementations and successful public-testnet deployments. Additional devnet testing will come first, followed by forks on Sepolia and Hoodi.
This measured pace is typical of Ethereum upgrades. It gives the community time to adapt while still moving the network forward. I have seen similar patterns in earlier hard forks. The projects that prepare early usually experience the smoothest transitions.
Why Fixed Assumptions Become Fragile Over Time
Smart contract development often involves making assumptions about costs and resource limits. Those assumptions feel solid when they match the current network rules. Years later the same numbers can become liabilities. The 2,300-gas stipend is a classic example. It solved a real problem at the time. Under a new cost schedule it can turn into a source of unexpected reverts.
The same pattern appears with gasleft() checks. Developers once used them to implement graceful degradation or to decide whether a call was safe. When the cost of the underlying operations shifts, the remaining gas no longer means what the original author expected. The logic still runs. It simply produces different outcomes.
In my experience this is one of the subtle challenges of long-lived blockchain code. The execution environment is not static. Protocols evolve. Cost models change. Contracts that treat gas figures as immutable constants eventually collide with those changes.
A useful habit is to treat every hardcoded gas value as a potential future liability. Prefer patterns that adapt to the current network parameters when possible. When fixed values are unavoidable, document them clearly and plan for periodic reviews around major upgrades.
What Regular Users Should Know
Most everyday users will not notice any difference once the upgrade lands, provided their wallets and infrastructure providers have updated their estimation logic. Transfers to existing accounts keep the same gas cost. The extra charge appears only when a transaction creates new state. Updated software should handle that automatically.
The risk surface is concentrated on the developer side and on tooling. That is why the Foundation’s warning focuses on contract maintainers and infrastructure teams. Users benefit from the eventual capacity increase without having to manage the transition details themselves.
Still, it is worth staying aware of the timeline. When public testnets start running the new rules, early adopters and power users may encounter occasional estimation quirks until every wallet has shipped its updates. Those moments tend to be short-lived once the major providers move.
Looking Ahead To Higher Capacity
The performance target behind the new gas schedule is ambitious. Supporting roughly three times the current base throughput would represent a meaningful step for Layer 1. Achieving that goal safely requires the cost of state creation and access to reflect the real resource consumption more accurately. Otherwise the network risks accelerating state growth in ways that become hard to reverse.
EIP-8037 and EIP-8038 are the tools chosen for that adjustment. They are not the most glamorous part of the roadmap, yet they form necessary groundwork. Without them, further capacity increases would rest on shakier foundations.
I find the combination of technical care and open communication encouraging. The Foundation published the warning early enough for the community to respond. The test environment is already available. The checker tool lowers the barrier to identifying affected contracts. These pieces together create a reasonable path for the ecosystem to adapt.
A Few Personal Observations
Having watched several Ethereum upgrades, I notice a recurring theme. The technical changes that look small on paper often matter most to the people who maintain production contracts day after day. Gas repricing falls into that category. It does not rewrite the consensus rules or change the fee market in a flashy way. It simply updates the cost of doing certain things. For contracts that made tight assumptions about those costs, the update is significant.
The good news is that most of the flagged failures appear fixable with modest effort. Raising gas limits solves a large share of them. The remaining cases require more careful review, yet the number of such contracts is described as small. That combination should keep the overall disruption limited if maintainers act in time.
Another observation concerns the role of tooling. Wallets and gas estimators sit between users and the network. When they lag behind protocol changes, users experience the friction first. Keeping those layers updated is therefore just as important as fixing the contracts themselves.
Finally, the existence of a dedicated long-running testnet for this work is a positive signal. It gives the community a stable place to experiment rather than forcing everyone onto short-lived temporary networks. That kind of infrastructure investment pays off during moments like this.
Preparing Without Panic
The practical recommendation is straightforward. Check your contracts. Use the available tools. Test on Platåberget. Adjust gas limits or call sites as needed. Coordinate with any wallet or infrastructure partners that interact with your contracts. Do these steps early and the upgrade becomes a non-event for your project.
There is no need for dramatic announcements or emergency patches in most cases. The changes are known. The test environment is live. The timeline still has room. Treating the warning as a normal part of the development cycle keeps the response proportional to the actual risk.
Ethereum continues to evolve. Capacity increases remain a priority. Gas costs are one of the levers available to make those increases sustainable. The current proposals represent a careful use of that lever. Contracts that adapt will continue to function as expected. The ones that ignore the shift may encounter avoidable friction later.
For anyone who maintains Layer 1 code, the message is simple. The gas landscape is about to change in measured but meaningful ways. A modest amount of preparation now will prevent surprises when Glamsterdam eventually reaches mainnet. That preparation is already possible today. The tools exist. The warning has been given. The rest is up to the builders who care about keeping their contracts reliable across network upgrades.
In the end these moments remind us that blockchain protocols are living systems. They grow. They adjust. The code that lives on them must be ready to grow and adjust as well. The current gas changes are one more example of that ongoing process. Handled thoughtfully, they clear the path for higher capacity without leaving too many contracts behind.
