Ethereum’s Glamsterdam Upgrade Slips Into 2027 as Review Debt Catches Up
• November 21, 2026 3:30 pm • CommentsEthereum’s developers are writing and testing protocol changes faster than they used to. Glamsterdam is showing why that does not automatically mean the network can ship its biggest upgrades faster.
CryptoSlate reported that the earliest realistic mainnet date for Glamsterdam has moved to mid-January 2027. The upgrade was once targeted for the first half of 2026, but unresolved review work, late specification changes and security questions disrupted the testing schedule.
The outlet found that developers gave Glamsterdam roughly 320 days from scope to launch even though Ethereum’s three previous upgrades needed between 433 and 475 days. Glamsterdam is also larger than those releases by several measures.
Its first public-testnet period produced 10 specification changes in six weeks. Older upgrades between Altair and Deneb recorded none after their first public testnet, making the current review burden unusually visible.
Ethereum Foundation researcher Toni Wahrstätter put the problem plainly: the schedule underestimated how much material other developers and security reviewers would need to understand.
We said Glamsterdam would ship early 2026. Earliest mainnet date now: mid-January 2027.
Nobody did a bad job, but I think we fooled ourselves, twice.
Here's me trying to figure out what happened and why. 🧵
— Toni Wahrstätter ⟠ (@nero_eth) November 20, 2026
The numbers explain the bottleneck. Glamsterdam carries roughly 6,080 lines of execution-layer specifications, far above the previous record set by Prague.
Two central components—enshrined proposer-builder separation and block-level access lists—add thousands more lines of proposal specifications, while the other included proposals widen the review surface again.
Development output is up sharply. Wahrstätter found that monthly contributions to Ethereum’s technical specifications have risen between twofold and 3.5-fold since 2024.
Glamsterdam also completed 12 development networks in 160 days, compared with seven over 285 days for Pectra.
But faster production can create review debt. Specifications, client code and tests may arrive quickly while the number of people capable of checking their interactions does not scale at the same rate.
Glamsterdam reached public testing before its specifications had settled, then recorded 10 specification changes in six weeks. Earlier upgrades typically saw few or none after their first public testnet.
EIP-8037 is a good example of why the last mile matters. The proposal changes how Ethereum charges for creating new blockchain state.
Its specification is substantial, and CryptoSlate reported that it requires roughly 30,000 lines of tests—more than any other proposal in the package. Vitalik Buterin has separately laid out his view of that proposal as developers debate the tradeoffs.
My views on EIP-8037:https://t.co/WUeR5lxJiC pic.twitter.com/eUnHJiUsgk
— vitalik.eth (@VitalikButerin) November 16, 2026
Security concerns around EIP-8037 and EIP-8038 already helped push back Glamsterdam’s planned October 26 activation on the Hoodi testnet. Core developers are now considering December 16 or 17 for another Hoodi attempt, with the next major decision expected November 26.
The original timetable was aggressive even before the late changes. Ethereum’s three most recent upgrades took between 433 and 475 days from finalized scope to mainnet.
Glamsterdam was given about 320 days despite being larger. Fusaka met its target only after developers removed Ethereum Object Format, while Pectra’s original scope was split.
Glamsterdam did not receive a comparable reduction.
None of this means the upgrade is failing. It means Ethereum has reached a point where writing more code is not the same thing as safely absorbing more change.
A protocol securing hundreds of billions of dollars cannot treat review as a box to check after implementation.
The healthier outcome may be a smaller, more predictable release discipline: set targets from historical performance, freeze specifications earlier and drop unfinished features before public testing.
Glamsterdam’s delay is frustrating. Shipping an upgrade reviewers do not fully understand would be far worse.
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.
