Ethereum's Glamsterdam upgrade slips to 2027 after testing setbacks
An Ethereum Foundation researcher says the upgrade cannot reach mainnet before mid-January 2027, blaming an aggressive schedule, a huge scope and late specification changes.
· 2 min read
- Ethereum Foundation researcher Toni Wahrstätter said Glamsterdam cannot reach mainnet before mid-January 2027, after a first-half 2026 target.
- The upgrade is the largest yet, with about 6,080 lines of execution-layer specifications versus 2,318 for Prague.
- Core developers are expected to decide on Nov. 26 whether a December Hoodi testnet activation can go ahead.
Ethereum's Glamsterdam upgrade has been pushed into 2027, according to CryptoSlate. On Nov. 20, Ethereum Foundation researcher Toni Wahrstätter said it cannot reach mainnet before mid-January 2027. The foundation had targeted the first half of 2026.
In his assessment, Wahrstätter said developers have repeatedly underestimated how long major upgrades take, even though their output has risen. His summary was blunt: "Nobody did a bad job, but I think we fooled ourselves, twice."
The data shows a clear rise in activity. Monthly contributions to the technical specifications are up between twofold and 3.5-fold since 2024. Glamsterdam went through 12 devnets in 160 days, while Pectra needed seven devnets over 285 days. Wahrstätter argued that this speed has created its own problem, because developers can now write specifications, code and tests faster than colleagues can review them.
Glamsterdam is also the biggest upgrade so far. It holds about 6,080 lines of execution-layer specifications, against 2,318 for Prague, the previous record. Enshrined proposer-builder separation (ePBS) and block-level access lists (BALs) account for roughly 11,900 and 9,300 lines in a separate count of proposal specifications, while the other 16 proposals add about 3,600. EIP-8037, which changes how Ethereum charges for creating new state, has about 1,700 lines of specifications but needs roughly 30,000 lines of tests.
History suggests the schedule was tight. Every upgrade since Dencun has needed more than 400 days from locking its scope to mainnet, and the three latest took between 433 and 475 days. Glamsterdam was given a target of 320 days. Almost every upgrade with an announced target has missed it. Fusaka was the exception, but only after Ethereum Object Format was removed, and Pectra had its scope split across separate upgrades. Glamsterdam did not see a comparable cut.
As a result, it reached public testing before its specifications were stable. Upgrades from Altair to Deneb saw no specification changes after the first public testnet, Pectra had two and Fusaka one, but Glamsterdam logged 10 in six weeks. Activity on the upgrade rose 34% after its Sepolia activation, with client commits up 25%.
Security researchers then found problems involving EIP-8037 and EIP-8038, which change how certain operations are charged and could upset assumptions in existing smart contracts. Developers postponed the Oct. 26 activation on the Hoodi testnet. The gap between the first public testnet and mainnet is now expected to exceed 100 days, versus the usual 4 to 10 weeks.
At their Nov. 19 meeting, developers tentatively aimed for Dec. 16 or 17 for Hoodi testing, with mid-January 2027 the earliest mainnet date if the specification holds. Wahrstätter urged developers to set targets based on past performance and to fix deadlines for dropping unfinished features. He added: "I'm sure we can ship faster, but for that, we must ship smaller." Core developers are expected to decide on Nov. 26 whether the December Hoodi activation can proceed.