Lin Oshitani

I have written a blog post to share my experience of writing a Game Boy emulator in OCaml! linoscope.github.io/writing-a-game…

昨日、南砺市長の田中幹夫氏 @tanakatoga とNyx Foundationのvita @keccak255 がNanto構想に関する共同記者会見を行いました。

最近のAIのヤバさと、そのブロックチェーンとの関係について書きました。一年ぶりのブログ投稿です。 AIとブロックチェーン - Satoshi Nakamoto Fan Club bitcoinspace.hatenablog.com/entry/2026/02/…

Introducing defi-cli Agents love CLIs. This gives them a single tool to query lending rates, compare yield, get bridge and swap quotes — across protocols and chains.

🚀 Introducing PrivateX402 — a privacy-preserving payment channel protocol for multi-agent AI systems. Joint work with @ElusAegis. 1/👇for design, privacy, and scaling properties & article link.

Aztec's privacy-first Stage 2 ZK rollup creates some pretty unique challenges for app devs One important challenge to address is how to secure bridged tokens on Aztec from soundness bugs This post explores one way to do that, at the cost of some privacy forum.aztec.network/t/request-for-…

1/ “Who is the top JIT liquidity searcher in Ethereum?” → a question answer by our paper w/ @0xAlphaist (accepted to a FC workshop: fc26.ifca.ai/defi/) We mapped JIT activity in Uniswap v3 and linked searchers <-> builders from Jan 2024 to Sept 2025. One name dominates: jared accounts for >50% of all observed public JIT execution.

@ConorMcMenamin9 @donnoh_eth @LayerZero_Core @Polkadot Great minds think alike🙂

@donnoh_eth @LayerZero_Core @Polkadot @linoscope hahah, who said it first

> Zero introduces the first multi-core world computer looks like @LayerZero_Core's marketing team hasn't read @Polkadot's marketing team content

Amazing work by Lin here, and happy to have been able to collaborate on this. If we are serious about running the world on zk, there's still a lot of work to do. - Current gas prices do not reflect proving complexity today - Rollups(and later L1) need to build zk gas as part of derivation/consensus to make sure blocks are provable in a reasonable amount of time and resources - Unfortunately zk cycles are not a great proxy for proving time in most zkVMs today. I hope this continues to improve @taikoxyz and @Nethermind are working to implement zk gas into Taiko's derivation. This is needed for a decentralized network to achieve stage 1 securely.

@jdetychey I wouldn't say so - If we add more opcodes/precompiles, we can just rerun benchmarks to get the up-to-date proving costs.

@linoscope Is it fair to see this as an argument to ossify the list of opcodes and precompiles?

What does one more opcode actually cost to prove in zkVMs? We measure the proving time per gas for each individual EVM opcode and precompile using real multi-GPU benchmark runs. Summary and link to the post in 🧵👇

Full post 👇! This work was supported by @taikoxyz as part of the Taiko <> Nethermind strategic partnership. Big thanks to the Taiko team for the support!

Fusaka Livestream is coming on December 3. Learn what node operators, L2 teams, and developers/users should expect. Speakers from Nethermind, @ethereumfndn, @go_ethereum, @ErigonEth, @prylabs, @taikoxyz, @Offchain and more. Sign up 🔗 nethermind.io/nethermind-fus…

EIP-7917 improves validator scheduling predictability and enables faster preconfirmations. Co-authored by @linoscope, this proposal ships with Fusaka and strengthens coordination between validators and applications. Learn more ⬇️ x.com/ben_a_adams/st…

@mteamisloading @potuz_eth @barnabemonnot @casparschwa @notnotstorm @_charlienoyes @terencechain @nero_eth @ralexstokes @soispoke If it is practical, then would be a natural addition to the solutions mentioned here: x.com/linoscope/stat…

@mteamisloading @potuz_eth @barnabemonnot @casparschwa @notnotstorm @_charlienoyes @terencechain @nero_eth @ralexstokes @soispoke My impression is that it's still far from being practical in the L1 protocol, so not betting the future on it. But I am no expert and could be wrong, and will be happily corrected about this.

I'm very torn on ePBS tbh ePBS+FOCIL to me feels like enshrining broken incentives and a patch to them perhaps it is from a deeper lack of conviction in Vitalik's few big builders w/ small verifiers + many CR contributors roadmap