Glamsterdam Could Slip to May 2027, and 50 Companies Just Found Out Why
Ethereum's next upgrade is reworking gas pricing, and at least 50 companies were told their live smart contracts might break. The fix could push Glamsterdam to May 2027. Here's who actually pays for that delay, and why immutability was always a sales pitch more than a technical guarantee.
At least 50 companies got a call from Ethereum developers about their smart contracts. The message wasn't great. The code they already shipped might stop working.
The culprit is gas repricing inside Glamsterdam, the combined upgrade of the Gloas consensus layer and the Amsterdam execution layer. Developers are reworking how the network charges for computation. Change the price of an opcode and you change the cost of every contract that touches it. Push the numbers far enough and contracts that assumed a fixed gas budget run out of gas and revert.
That's not a bug in anyone's code. It's the network moving the floor under software that was working fine yesterday.
The Timeline Already Moved
Glamsterdam was tracking toward a late 2026 activation. It could now land in May 2027 if the compatibility work goes smoothly. Six months, give or take, added to a schedule that already ate two years of planning.
If that sounds familiar, it should. Dencun was supposed to arrive in 2023 and shipped in March 2024. Pectra was penciled in for late 2024 and landed in May 2025. Ethereum's upgrade cadence has a consistent shape. Announce, slip, ship, repeat. Developers aren't lazy. The work is genuinely hard. But the pattern matters because every slip has a price tag attached to it, and the people paying aren't the ones setting the schedule.
Here's the part that should get your attention. The 50 companies that got contacted are the lucky ones. Someone rang them. Someone explained what's changing. they've time to audit, patch, and redeploy. Now ask yourself how many protocols out there are running contracts with hardcoded gas assumptions and a team that checks the research forum twice a year. The answer isn't 50. It's a lot more than 50.
Gas repricing is a silent failure mode. There's no error message that says "the network changed its pricing model." There's a transaction that used to cost 80,000 gas and now needs 105,000, and a user staring at a revert wondering why the app is broken.
Immutability Was Never a Property
Let's clear something up. Immutability is a claim about storage. It's not a claim about cost.
Your bytecode can't be edited. Fine. But the price of executing that bytecode is set by a social process involving client teams, researchers, and a governance layer that has no formal vote and enormous power. When the gas schedule changes, your contract's economics change. Nothing in the code was touched. Everything about running it got more expensive.
People in this industry love to talk about code as law. Code is law right up until the gas schedule gets repriced and the law becomes a suggestion. That's not a scandal, it's just the truth nobody puts on the conference slide.
So who wins here?
Auditors and gas optimization shops, first. Every repricing event is a fresh wave of billable work. Tooling teams that build simulation and fork-testing infrastructure win too, because the only real defense against a pricing change is testing your contract against it before mainnet does the testing for you. Layer 2s win in a quiet way. They run their own upgrade schedules and can absorb execution-layer changes without asking anyone's permission.
And the competitors win. Every chain that pitches stability over velocity gets a fresh talking point, and this time they didn't have to invent it.
Who loses? Solo developers with no one to call. Small DeFi protocols with a two-person team and a $40,000 audit budget already spent. Anyone who built a contract in 2022, assumed the gas math was settled, and moved on to the next product. They didn't get a phone call. They'll get a support ticket.
Follow the incentives, not the press releases. The incentive here's to ship the upgrade correctly, and correctly means slow. That's the right call. It's also a reminder that Ethereum's biggest competitive risk isn't a rival chain. It's the cost of changing its own mind.
Permissionless means exactly what it sounds like. Nobody needs a license to deploy. But nobody gets a grandfather clause either. If you want the freedom to ship code without asking, you accept the flip side, which is that the network won't ask before it reprices your execution.
What You Do If You Ship Code
Three things, and none of them are glamorous.
First, stop hardcoding gas assumptions. The 2,300 gas stipend on a plain transfer has survived for years, but surviving isn't the same as guaranteed. Any contract that relies on a fixed gas number to make a decision is a contract with an expiration date nobody printed on the label.
Second, fork-test against proposed changes. Not the final spec. The proposals, the messy drafts, the ones that might get pulled. If you wait for the mainnet fork to test, you're testing production.
Third, price the delay into your roadmap. If you're building something that depends on a specific EIP landing in a specific quarter, you're not building a product, you're building a bet on a schedule that has missed four times in three years.
The core developers are doing this the right way. They could've shipped the repricing, let the breakage happen, and patched complaints later. Instead they picked up the phone and contacted at least 50 companies, and they're willing to eat a six-month slip to do it. That's the system working.
But don't confuse careful with costless. The state isn't protecting you, and neither is a protocol roadmap. The code doesn't ask for a license, and it doesn't ask for permission before it changes the price of yours.
May 2027 is the current target. Treat it as a direction, not a date. And if your contract has a hardcoded gas number in it tonight, go find it. You've got somewhere between six months and eighteen to fix it, and nobody's going to call you.
Explore More
Key Terms Explained
The compiled, machine-readable version of a smart contract that runs on the blockchain's virtual machine.
A blockchain platform that enabled smart contracts and decentralized applications.
The part of a blockchain that processes transactions and runs smart contracts.
A change to a blockchain's protocol that creates a new version.