Illuminated fiber-optic strands representing Ethereum network capacity and compatibility paths

Ethereum’s Glamsterdam Upgrade Hits a 15,000-Contract Compatibility Question

• November 17, 2026 7:22 pm • Comments

Ethereum’s next major upgrade has run into the kind of problem that only appears after millions of contracts have learned to depend on the network’s old rules.

The proposed Glamsterdam changes are meant to make Ethereum scale more safely. Developers now say the same gas-pricing changes could disrupt 10,000 to 15,000 existing smart contracts at the high end of current estimates.

That figure is an upper-bound compatibility estimate, not a count of apps that have already failed. Glamsterdam is still in testing, and the dispute is about what could happen if the proposals reach production unchanged.

Decrypt reports that the debate centers on EIP-8037 and EIP-8038, two proposals that change how Ethereum charges for creating and accessing permanent network state. Developers want those costs to reflect the growing hardware burden carried by nodes as Ethereum expands.

The proposals are tied to a much larger scaling goal: moving from roughly 60 million gas per block toward as much as 200 million. More capacity could lower costs and let Ethereum process more activity, but it would also accelerate database growth unless the network charges more accurately for operations that create or read permanent state.

Independent audits complicated that plan. ChainSecurity, Dedaub and Ethereum Foundation researchers found roughly 200 classes of potential compatibility problems, including contracts that could run out of gas or behave differently because they were written around today’s pricing assumptions.

The Ethereum Foundation says core developers delayed Glamsterdam’s Hoodi testnet fork while researchers investigate eight protocol invariants that the two EIPs would change. The Foundation asked contract teams and auditors to test deployed code and report contracts that depend on those assumptions.

Its November update said four reports had already identified deployed contracts that could break, while audit teams continued reviewing a much larger set of possible issues. The Foundation also said specifications could be revised if the impact proves too broad, and it scheduled the findings for discussion on the November 19 execution-layer call without promising a decision in advance.

No mainnet outage is underway, and the review has not established current user losses. The warning is more specific: a base-layer rule change can alter immutable code even when nobody changes the contract itself.

The Ethereum roadmap frames gas repricing as a hardware-sustainability measure. State creation and state access have become more expensive for nodes as Ethereum’s database has grown, while some operation prices still reflect a much smaller network.

EIP-8037 would separately meter the cost of creating new state, including accounts, storage slots and contract code. EIP-8038 would raise prices for reading and updating state so a block cannot cheaply force nodes through an unreasonable amount of database work.

That engineering logic is why developers are reluctant to drop the proposals. Pulling them could delay Glamsterdam and parts of Ethereum’s post-quantum roadmap by four to six months, while shipping them unchanged could undermine confidence among builders who expected deployed contracts to keep behaving the same way.

The cleanest result is the least dramatic one: publish the audit findings, identify contracts that genuinely depend on the old assumptions and revise the specifications where the risk is unacceptable. Glamsterdam’s test process found the conflict before mainnet, which is exactly when Ethereum needs to solve it.

Join the conversation!

We have no tolerance for comments containing violence, racism, profanity, vulgarity, doxing, or discourteous behavior. If a comment is spam, instead of replying to it please click the icon below and to the right of that comment. Thank you for partnering with us to maintain fruitful conversation.