Not all of us have completely given up hope that Ethereum can make its way back into the light.
1 Like
Well the hope never dies :))) I think we are at a moment of maximum chaos :)))
It can still censor users based on other factors like region, usage of a VPN, country of credit card used, etc. which is something that centarlized services already censor on.
Well. Some privacy-friendly RPC providers, e.g. Publicnodes, DRPC, Llama Nodes, etc., offer RPC services for free / without registration as a public good. Some of them are okay (I don’t want to speak for them but it’s discussed) to host PIR nodes. So far, each node is much heavier infrastructure-wise than an archive node, and practical schemes require high-end GPUs for subsecond scans of the full state for queries or hints generation for pre-processing. VPN usage is easy to hide in the era of mass OpenClaw/ Tailscale, region-based restrictions are meaningless on the internet post-dial-up era (you are, likely, too young to remember that if your landline phone number starts from +7 it’s not easy to call to a BBS’s number starting with +1.. with tcp/ip geographic restrictions is a meme)
Also, we have learned repeatedly that just describing a good idea without an implementation in the wild is a good way to never see a good idea implemented
Agree. I have no interest in theorycrafting (PIR is not easy but simpler than launching the first evm2evm bridge in prod or launching xDai.. no doubts tbh). JFYI PIR is solved as a research primitive end-to-end for multiple scheme (with different trade-offs); it just needs incentives to productize and launch. Since most RPC providers are in a red ocean and PMF for private RPC is almost nonexistent, there should be external incentives to make it a norm. We don’t need to change protocol to make it happen. A subproduct of PIR is ultra-light client which is only proving data needed for atomic operation (that’s what I started my comment)
This post makes a compelling case that state is the hardest scaling bottleneck — no magic bullet like ZK-EVMs for execution or PeerDAS for data. Tiered state, temporary storage, and UTXOs are promising long-term directions.
But right now, the base fee sits between 0.09 and 0.6 gwei. At these levels, anyone can grow permanent state for essentially nothing. Address poisoners added 270M+ attempts in January 2026 alone, each creating permanent state entries that every full node must store forever — for fractions of a penny.
EIP-8037 reprices state creation to 131,488 gas per new account at 100M gas limit. EIP-2780 adds a 25,000 gas surcharge for account creation via value transfers. Both are great — but at 0.1 gwei base fee, EIP-8037’s repriced account creation still costs $0.027. That’s not a deterrent at scale.
These long-term state architectures won’t matter much if the current fee market continues to let anyone grow permanent state for free. A minimum base fee floor is the simplest first step — it prices the existing state externality today while these solutions are developed. Blob gas already has a protocol-enforced floor via EIP-7918. Execution gas is the only fee market left without one.
I wrote up a full analysis here: Solving Spam, State Bloat, and Fee Volatility with a Single Parameter