We’re watching two completely different philosophies play out in real-time, and honestly, I find it fascinating.
The Ethereum Approach: Measure Twice, Cut Once
Glamsterdam is scheduled for May/June 2026, and after months of research and community discussion, the core devs have confirmed exactly two EIPs: EIP-7732 (enshrined Proposer-Builder Separation) and EIP-7928 (Block-Level Access Lists). That’s it. Everything else is still under consideration.
Here’s what gets interesting: FOCIL (Fork-Choice Forced Inclusion Lists) was originally part of the plan—it’s the piece that lets validators publish a list of transactions that must be included, giving them a defense against builder censorship. Perfect complement to ePBS, right? But the Base engineering team pushed back, warning that launching ePBS + FOCIL together could delay the whole upgrade beyond 2026. So what did Ethereum do? They split it. ePBS ships in Glamsterdam, FOCIL waits for Hegotá.
This is classic Ethereum: extensive research, formal specification, multi-client testing, rough consensus. It’s thorough, it minimizes surprises, and yes—it’s slow.
The Solana Approach: Ship Fast, Iterate Faster
Meanwhile, Solana’s Alpenglow upgrade is already in testnet. They’re replacing the entire consensus mechanism—Tower BFT and Proof of History are out, Votor and Rotor are in. The result? Finality drops from 12.8 seconds to 100-150 milliseconds. Block propagation hits 18ms under typical conditions.
Think about that for a second. Ethereum is carefully finalizing scope while Solana is shipping a complete consensus re-architecture that’s 100x faster. Testnet deployment happened Q3 2025, governance vote coming Q4 2025, mainnet targeted for Q1 2026.
This is classic Solana: rapid iteration, test under load, fix issues in production. It’s exciting, it delivers capabilities quickly, and yes—it’s risky.
So Who’s Right?
I’ve been building on both chains for years, and here’s my take: both strategies are valid, but they’re optimizing for completely different things.
Ethereum is optimizing for institutional confidence. If you’re Goldman Sachs tokenizing bonds or Fidelity launching RWA products, you want a settlement layer that never surprises you. You want predictability, formal verification, multiple client implementations. You don’t want to wake up to breaking changes or consensus bugs. The slow, deliberate approach makes sense when you’re positioning as infrastructure for the financial system.
Solana is optimizing for developer excitement and user-facing improvements. If you’re building consumer apps, payments, gaming—you need capabilities now, not in 2-3 years. Alpenglow’s 100ms finality unlocks real-time applications that weren’t possible before. The rapid shipping approach makes sense when you’re competing for mindshare and need to deliver.
The Trade-offs Are Real
Ethereum’s approach means fewer consensus bugs and L1 outages (when was the last major Ethereum mainnet halt?). But it also means developers waiting years for features they need, potentially choosing other chains.
Solana’s approach means cutting-edge capabilities shipping fast. But complete consensus re-architecture = massive attack surface. Votor and Rotor are both brand new. Yes, they’re improving from past network halts, but production testing will be the real validation.
The Uncomfortable Question
Here’s what I keep thinking about: at what point does “roadmap-driven development” become a liability rather than a strength?
Bitcoin took 4 years to activate SegWit because of consensus requirements. Ethereum ships major upgrades every 18-24 months. Solana ships consensus changes every 6-9 months. If Ethereum takes 2 years to ship ePBS while Solana delivers 100x faster finality in that same timeframe, how does that affect where developers build?
On the other hand: Chrome beat Firefox in the browser wars partly by shipping features fast while Firefox was careful. But Bitcoin’s ultra-conservative approach created the trust that enabled institutional adoption. So which model matters more for blockchain?
Looking at the Technical Details
Glamsterdam’s ePBS moves proposer-builder coordination from relays into the protocol—that’s fundamental infrastructure improvement. BALs (Block-Level Access Lists) declare state access upfront, enabling better parallelization. These aren’t flashy features, but they’re important for long-term scalability and MEV mitigation.
Alpenglow’s 100ms finality is immediately user-facing. Real-time payments, sub-second DeFi trades, gaming applications that feel native—this unlocks new use cases.
My Prediction
I think both chains will coexist because they’re serving different markets. Ethereum becomes the settlement layer for high-value transactions, institutional RWAs, and applications where security/stability are paramount. Solana becomes the consumer layer for payments, gaming, social apps where speed and low cost are paramount.
The question isn’t “which approach is better?” It’s “which approach is better for what?”
What do you all think? Is Ethereum’s careful approach building institutional trust, or is it losing developer mindshare? Is Solana’s rapid shipping creating competitive advantage, or dangerous technical debt? And most importantly—which chain’s philosophy do you prefer to build on?
P.S. I’m running validators on both chains, so I have skin in this game from both sides. Would love to hear from other builders about which development philosophy you prefer and why.