ETH upgrades are not launching today: Glamsterdam is still on the Devnet—I’ll continue waiting around 2750
First, my stance: I’m leaning more toward ETH’s mid-term upgrade path being bullish; for the short term, I’m still on the sidelines. I won’t treat the roadmap as a ready-made buy order.
The latest roadmap page on ethereum.org is very clear: Glamsterdam is currently in the devnet testing stage. The next step is the Sepolia testnet fork on October 6; the mainnet only gives an “expected Q4 2026” with the exact date not yet confirmed. The page also specifically reminds that while the scope has been frozen, mainnet-related changes are still possible before launch, and the Meta EIP is still a draft.
The most common market mistake is turning “scheduled for an upgrade” into “already live,” then directly converting technical expectations into a coin-price target.
The value of this upgrade isn’t low. One of the core parts is ePBS, which incorporates the handoff between proposers and builders into the protocol. According to official explanations, it can expand the execution-payload propagation window from about 2 seconds to about 9 seconds, reducing reliance on third-party relays. Another feature, Block-Level Access Lists, will pre-list the state dependencies involved in transactions, paving the way for parallel reads and execution and enabling faster synchronization. At the same time, state creation and access will be repriced—aiming to control database bloat rather than simply pushing the Gas limit higher and higher.
In simple terms, this is a combined engineering effort for scaling, censorship-resistance, and node sustainability. But for it to matter in practice, three things must all happen: the devnet works, clients achieve compatibility, and the mainnet date gets confirmed.
On the screen, ETH is currently around $2,749, with a 24-hour range of 2,607–2,768.57. OKX perpetual funding rate is about +0.0018%, with open positions around 627,700 ETH and a notional value of about $1.725 billion. After the price stands above 2,700, the heat is still high, but the funding rate hasn’t spiraled—suggesting sentiment is bullish, yet not broadly overcrowded.
In my previous plan, I asked for 2,750–2,755 to absorb and then for a 15-minute reclaim of 2,762. Even though price spiked to 2,768.57, it didn’t complete a stable confirmation; it then pulled back to around 2,733. So the old plan can’t be written as already executed, and definitely can’t be claimed as profitable.
If I were trading on my own, I’m currently keeping a 0 position. I’ll only use 2% of my principal for a spot test long if 2,740–2,746 show shrinking volume and stop falling, and then the 15-minute candle reclaims above 2,756. First target: 2,768–2,772; second: 2,790–2,810. If it touches 2,768, I’ll cut one-third. If it drops back to 2,745, I’ll cut half. If the 15-minute closes below 2,732, I’ll exit everything.
If instead it directly stands above 2,769 with volume and then retests 2,758 without breaking it, I’ll add at most 1%. If it falls back to 2,748, I’ll close the follow position immediately. Conversely, only when 2,732 breaks down on high volume and the retest back to 2,745 fails, will I use up to 0.5% of principal to try a low-leverage short; targets are 2,715 and 2,695. A renewed reclaim of 2,755 is the stop-loss.
A successful fork on Sepolia can only increase the probability of mainnet launch, but it can’t replace price confirmation. If the test is delayed or the scope changes again, I’ll lower my position expectations and keep waiting.
$ETH
The above is only my personal market observations and does not constitute investment advice.
First, my stance: I’m leaning more toward ETH’s mid-term upgrade path being bullish; for the short term, I’m still on the sidelines. I won’t treat the roadmap as a ready-made buy order.
The latest roadmap page on ethereum.org is very clear: Glamsterdam is currently in the devnet testing stage. The next step is the Sepolia testnet fork on October 6; the mainnet only gives an “expected Q4 2026” with the exact date not yet confirmed. The page also specifically reminds that while the scope has been frozen, mainnet-related changes are still possible before launch, and the Meta EIP is still a draft.
The most common market mistake is turning “scheduled for an upgrade” into “already live,” then directly converting technical expectations into a coin-price target.
The value of this upgrade isn’t low. One of the core parts is ePBS, which incorporates the handoff between proposers and builders into the protocol. According to official explanations, it can expand the execution-payload propagation window from about 2 seconds to about 9 seconds, reducing reliance on third-party relays. Another feature, Block-Level Access Lists, will pre-list the state dependencies involved in transactions, paving the way for parallel reads and execution and enabling faster synchronization. At the same time, state creation and access will be repriced—aiming to control database bloat rather than simply pushing the Gas limit higher and higher.
In simple terms, this is a combined engineering effort for scaling, censorship-resistance, and node sustainability. But for it to matter in practice, three things must all happen: the devnet works, clients achieve compatibility, and the mainnet date gets confirmed.
On the screen, ETH is currently around $2,749, with a 24-hour range of 2,607–2,768.57. OKX perpetual funding rate is about +0.0018%, with open positions around 627,700 ETH and a notional value of about $1.725 billion. After the price stands above 2,700, the heat is still high, but the funding rate hasn’t spiraled—suggesting sentiment is bullish, yet not broadly overcrowded.
In my previous plan, I asked for 2,750–2,755 to absorb and then for a 15-minute reclaim of 2,762. Even though price spiked to 2,768.57, it didn’t complete a stable confirmation; it then pulled back to around 2,733. So the old plan can’t be written as already executed, and definitely can’t be claimed as profitable.
If I were trading on my own, I’m currently keeping a 0 position. I’ll only use 2% of my principal for a spot test long if 2,740–2,746 show shrinking volume and stop falling, and then the 15-minute candle reclaims above 2,756. First target: 2,768–2,772; second: 2,790–2,810. If it touches 2,768, I’ll cut one-third. If it drops back to 2,745, I’ll cut half. If the 15-minute closes below 2,732, I’ll exit everything.
If instead it directly stands above 2,769 with volume and then retests 2,758 without breaking it, I’ll add at most 1%. If it falls back to 2,748, I’ll close the follow position immediately. Conversely, only when 2,732 breaks down on high volume and the retest back to 2,745 fails, will I use up to 0.5% of principal to try a low-leverage short; targets are 2,715 and 2,695. A renewed reclaim of 2,755 is the stop-loss.
A successful fork on Sepolia can only increase the probability of mainnet launch, but it can’t replace price confirmation. If the test is delayed or the scope changes again, I’ll lower my position expectations and keep waiting.
$ETH
The above is only my personal market observations and does not constitute investment advice.
