Markets
BTC $58,986 +2.00% ETH $1,662 +2.23% BNB $540 +3.03% XRP $0.9150 +2.60% SOL $69.10 +2.39% LINK $8.56 +2.96% BTC $58,986 +2.00% ETH $1,662 +2.23% BNB $540 +3.03% XRP $0.9150 +2.60% SOL $69.10 +2.39% LINK $8.56 +2.96% TOTAL $2.05T BTC.D 57.6%

Ethereum

Ethereum’s Glamsterdam Upgrade Slips to 2027 as Review Capacity Lags Code Output

A foundation researcher says the network's largest upgrade yet cannot reach mainnet before mid-January 2027, and he blames scheduling rather than effort.

Ethereum’s next major upgrade will not reach mainnet in the first half of 2026 as planned. On 20 November, Ethereum Foundation researcher Toni Wahrstätter said Glamsterdam cannot go live before mid-January 2027 after delays disrupted its testing schedule.

His assessment is less about effort than about scheduling. “Nobody did a bad job, but I think we fooled ourselves, twice,” he wrote, acknowledging that developers have repeatedly underestimated how long large upgrades take even as their output has risen.

The numbers support the second half of that sentence. Monthly contributions to Ethereum’s technical specifications have grown between twofold and 3.5-fold since 2024. Glamsterdam completed 12 development networks in 160 days, against seven devnets over 285 days for Pectra. By the measure of raw activity, the team is moving faster than ever.

The bottleneck has moved downstream. Developers can now produce specifications, code and tests faster than colleagues can review them and absorb the implications. Glamsterdam’s execution-layer specifications run to about 6,080 lines, compared with 2,318 for Prague, the previous record. Two headline changes — enshrined proposer-builder separation and block-level access lists — account for roughly 11,900 and 9,300 lines respectively, while the other 16 proposals together cover about 3,600. EIP-8037, which alters how the network charges for creating state, carries about 1,700 lines of specification but requires roughly 30,000 lines of tests, more than any other proposal in the upgrade.

Glamsterdam’s scope is what makes the review problem acute. Enshrined proposer-builder separation would move a market structure that currently runs through relays and middleware into the protocol itself; block-level access lists would let transactions declare the storage they touch, opening the door to parallel execution. Both are structural changes to how blocks are assembled rather than parameter tweaks, and both touch code paths that validators and builders depend on.

History is unkind to the timetable. Every Ethereum upgrade since Dencun has taken more than 400 days from scope freeze to mainnet, and the three most recent took between 433 and 475 days despite smaller scope. Glamsterdam was given 320 days. The pattern behind missed deadlines is consistent. Fusaka met its target only after developers removed the Ethereum Object Format, a major component. Pectra required splitting its original scope across separate upgrades. Glamsterdam took a different route: proposals kept arriving in successive testing environments while work on its two main components continued, and no comparable scope reduction protected the schedule. The result was public testing that began before the specifications had stabilized, which generated more work at the stage where time is scarcest.

For users, the practical effect is a longer wait for changes that touch block building and state management. For the foundation, the finding is an argument about planning discipline rather than productivity: the constraint is no longer how quickly code can be written, but how quickly a large and growing body of it can be understood.

Wahrstätter’s line counts put the review burden in concrete terms. The two largest components account for more than 21,000 specification lines between them, against roughly 3,600 for the remaining 16 proposals combined, and the testing requirement for a single state-charging change dwarfs its specification.

Share this article

Viber uses viber://forward?text= when the app is installed.