Mentioned in This Article
1. Ethereum upgrade waits on outside audits
The Ethereum Foundation has asked Ethereum's core developers to postpone any decision on Glamsterdam, the network's next Ethereum upgrade, until independent audits of its gas changes are published. As of November 10, Glamsterdam has no new testnet date, after four reports identified live smart contracts that could break under the planned rules.
The request leaves the upgrade in a holding pattern. Developers are neither pressing ahead with the current design nor formally rewriting it. Instead, two outside security firms are working out how many contracts already deployed on Ethereum (ETH) would stop behaving as intended if the changes shipped as written, and how serious the consequences would be.
Ethlabs, which has been tracking the same reports, added its voice on November 10 and recommended that no decision on Glamsterdam be taken until the picture is clear. Both organizations are pointing to the same sequence: the audit findings come first, a public review of those findings follows, and only then do developers choose between revising the specification and setting a fresh date for the next test network.
2. Why is Glamsterdam delayed?
Glamsterdam is delayed because some contracts that are already running on Ethereum could malfunction once the upgrade changes how gas is charged. The Ethereum Foundation has received four reports describing deployed contracts that could break, and it does not want developers to commit to a path until auditors have measured the scale of the problem.
Marius van der Wijden, who wrote the foundation's blog post on the subject, said the reports identify deployed contracts that could break under two specific proposals, EIP-8037 and EIP-8038. Those proposals alter the gas cost of certain operations and change the way the network charges for state storage, meaning the data that contracts and accounts keep permanently on the chain.
The foundation said it is working through the reports and getting in touch with the teams whose contracts are affected. It did not name those teams.
The concern is a familiar one in Ethereum's history. A smart contract, once deployed, generally cannot be edited by the network's developers, and many contracts cannot be changed by their own authors either. Code written years ago may have been built around assumptions about what a given operation costs. When a network upgrade changes those costs, logic that was sound when it was deployed can begin to fail, even though nobody has touched the contract itself.
3. What EIP-8037 and EIP-8038 would change
Gas sits at the center of the dispute. Every action on Ethereum, from a simple transfer to a complex interaction between contracts, consumes gas, the unit the network uses to measure and price computation and storage. Users pay for the gas their transactions consume, and each block can only hold a limited amount of it.
Glamsterdam would raise that per-block gas limit to 200 million. That is more than three times the current limit of 60 million, and it would allow considerably more activity to fit into each block.
EIP-8037 and EIP-8038 are the parts of the upgrade that reprice what happens inside those blocks. As described in the foundation's update, they change the gas cost of operations and the way the network charges for state storage. State is a particular concern for a blockchain because data written to it must be kept by the machines that run the network, so the price attached to creating and using it shapes how quickly that burden grows.
Repricing of this kind is where existing contracts are exposed. A contract may forward a fixed amount of gas to another contract, or assume that a particular step will always fit within a certain budget. If the price of the underlying operation moves, that assumption can fail. The four reports now under review describe deployed contracts that fall into this category under the current Glamsterdam specification.
4. How the testnet schedule slipped
Testing of Glamsterdam had been moving along a planned sequence of test networks before the problem surfaced. The upgrade went live on the Sepolia testnet on October 6, giving developers a first public environment in which to observe the new rules.
The next step was meant to be Hoodi, another test network, where a Glamsterdam fork had been scheduled for October 26. Developers scrapped that fork. The cancellation followed work by the security firm ChainSecurity, which showed that gas spent before a failed call could be returned under the new rules.
That behavior matters because contracts are often written with the expectation that gas already consumed stays consumed, whether or not the call that follows succeeds. A rule change that hands such gas back can alter how a contract's accounting works in practice, which is the kind of unintended side effect a testnet phase exists to catch.
Since the Hoodi fork was called off, no replacement date has been set. The foundation's position is that a new Hoodi date should not be chosen until the audit results are in and have been examined in the open.
5. Ethlabs backs waiting for clarity
Ethlabs has reached the same conclusion as the foundation. The organization said it has been monitoring and triaging the reports for more than two weeks, and on November 10 it recommended that no decision about Glamsterdam be made until there is clarity on their impact.
It also sought to put the risk in context. According to Ethlabs, none of the issues can be exploited today. They would only apply if Glamsterdam were to ship under its current specification, which means contracts on Ethereum are not exposed to them at present. The question is what would happen after the upgrade, not what is happening now.
Ethlabs was explicit about what it would support if the findings are bad. Should the audits show problems at a scale it considers unacceptable, the organization said it would back changes to the specification or a significant review of Glamsterdam's scope. That leaves open the possibility that the upgrade is not only adjusted but reduced, with some of its contents reconsidered.
6. A deliberate pace draws support
Caution has its defenders among people who build on the network. Kain Warwick, the founder of Synthetix, addressed the October testnet cancellation in a podcast appearance and pointed to the sheer number of contracts that already live onchain. He spoke approvingly of what he called Ethereum's "slow, cautious approach."
His point reflects the trade-off developers face. A larger gas limit and updated pricing are meant to improve the network, but Ethereum also carries a large body of deployed code that other applications and users depend on. Each contract that breaks after an upgrade is a problem for whoever relies on it, and in many cases there is no simple way to repair the contract after the fact.
For that reason the decision is being handled through the network's usual public process instead of being settled quickly. All Core Devs calls are the meetings where the teams that build Ethereum's client software agree on upgrade plans. The foundation has asked the participants in those calls to wait until the audit findings are public and reviewed before deciding either on revisions to the specification or on a new Hoodi date.
Work has not stopped in the meantime. Several possible revisions to the specification are already being prototyped, so that developers have concrete options to weigh once the auditors report.
7. What contract developers are being asked to do
Teams with contracts on Ethereum have been given a task of their own. The foundation has asked contract developers to check their code against eight invariants that it has published in a public code repository. An invariant, in this setting, is a property that is supposed to remain true no matter how a contract is used. Testing code against the list is a way for a team to find out whether its own contracts depend on gas behavior that Glamsterdam would change.
Developers who find a problem are asked to report it to the foundation through a dedicated email address set up for Glamsterdam. Those reports feed the same effort the auditors are engaged in: building a fuller count of what would break, so that the eventual decision rests on evidence and not on estimates.
The request also spreads the work. Outside auditors can study the rule changes and survey deployed contracts, but the authors of a contract are usually best placed to know what assumptions were built into it. Asking them to test against a fixed list gives the foundation a second source of information alongside the formal reviews.
8. What happens next
The audits are being carried out by ChainSecurity and Dedaub, and the published timelines for their results do not fully agree. The Ethereum Foundation's schedule says both engagements end on November 13, with reports published then. Ethlabs says publication will come by Monday, November 16. The agenda for a testing call on November 16 anticipates completion between November 16 and November 19.
Two meetings are fixed. On November 16, the auditors and the foundation are due to present to developers. An execution-layer call follows on November 19, and no decision has been committed to in advance of it.
From there, the outcome depends on what the reviews show. Developers will weigh the findings on those calls before either revising the upgrade or setting a new testnet date. If the number of affected contracts turns out to be manageable, the path could be a targeted revision followed by a rescheduled Hoodi fork. If the audits find problems at a scale judged unacceptable, the options Ethlabs has described, changes to the specification or a significant review of scope, come into play.
Until then, Glamsterdam remains live on Sepolia, absent from Hoodi and without a timetable, while the people responsible for it wait to learn how much of Ethereum's existing code the upgrade would disturb.