frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

Is MCP becoming the common foundation of smart, decentralised search?

https://contextetech.com/en/articles/mcp-protocol-smart-decentralised-search
9•contextetech•1h ago

Comments

troupo•28m ago
As always, the answer to an article whose title is a yes/no question: no
hobofan•23m ago
> Three things still have to be built, server by server

> citation

Nope, as long as the MCP server exposes enough metadata to be cited (mostly trivial), citation is a job of the client/LLM.

> trust

Also just a function of the client. The server can't aid here at all as it's inherently untrustworthy in the system.

> measurement

Maybe. Though it's debatable how much measurement _each_ server needs, and whether that actually gives meaningful insight to optimize. This might as well be measured outside of the servers, in the client/a centralized logging system that also sees surrounding context.

randomblock1•13m ago
No, because HTTP already is. You can't have every MCP added. Every harness has a search tool, so just fetch data as needed. MCP isn't for search, it's for interacting. Sometimes that's search but mostly not.

DeepSeek: DeepSeek v4.1 Flash 1017

https://openrouter.ai/deepseek/deepseek-v4.1-flash-1017
118•theanonymousone•1h ago•9 comments

"Tame" boar and deer are instinctively adapting to the wolf in the Netherlands

https://nos.nl/video/2635315-wildcamera-s-tonen-hoe-zwijnen-en-herten-zich-aanpassen-aan-de-wolf
20•felipevb•1h ago•20 comments

Ultramarine 1 Flash

https://www.paperinstruments.com/ultramarine
247•i_rush_carriers•1d ago•71 comments

Is MCP becoming the common foundation of smart, decentralised search?

https://contextetech.com/en/articles/mcp-protocol-smart-decentralised-search
9•contextetech•1h ago•3 comments

Tesla Roadster

https://www.tesla.com/roadster
1435•shekhar101•1d ago•2512 comments

DeepSeek-v4.1-Flash-1017

https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017
145•theykk•1h ago•61 comments

Show HN: Llmspeed – feel how fast a local LLM runs before you buy the hardware

https://llmspeed.dev/
74•albinstman•5h ago•28 comments

Intuitive concurrent programming in any language

https://www.lf-lang.org/
44•mpweiher•1d ago•7 comments

Tell HN: Remember to migrate away from .name domain

38•nikanj•2d ago•19 comments

Recipes Are Moving from Maven Central to the Code Genome Project

https://docs.openrewrite.org/reference/latest-versions-of-every-openrewrite-module
6•javahippie•1d ago•1 comments

Llama.cpp on a jailbroken PS5: run any GGUF model natively on the console

https://github.com/cobanov/PS5LM
290•Anon84•15h ago•72 comments

You could have invented transformers from memory

https://alexwang.ai/posts/associative-memory/
308•speckx•15h ago•72 comments

Claude reveals ECDSA curve origins

https://www.anthropic.com/research/ecdsa-curve-origins
875•ahlCVA•15h ago•325 comments

FLUX 3 Image Klein

https://bfl.ai/blog/flux-3-image-klein
166•gmays•14h ago•46 comments

0/1 Knapsack Problem in PyTorch

https://jedrzej.maczan.pl/2025_11_21_dp_knapsack_sliding_hirschberg
23•yu3zhou4•2d ago•1 comments

Apple ML Research: Muse Spark Open with MLX on Mac Studio

https://machinelearning.apple.com/research/mlx-distributed-inference
320•antipaul•20h ago•161 comments

eBay Restricts Listings of Radio Equipment, but Exempts Ham-Band-Only Gear

https://www.arrl.org/arrlletterissue?issue=2026-10-15#eBay-Restricts-Listings-of-Radio-Equipment_...
222•7402•16h ago•198 comments

Butlerian Jihad

https://dune.fandom.com/wiki/Butlerian_Jihad
230•wetwater•7h ago•124 comments

Designing more efficient quantum circuits with Claude

https://www.anthropic.com/research/designing-efficient-quantum-circuits
142•pretext•16h ago•32 comments

JPEG XL vs. AVIF for web use-cases

https://jakearchibald.com/2026/jpeg-xl-vs-avif/
262•farlight•18h ago•182 comments

I built the RealWorld app in Lean 4

https://leangineer.com/blog/realworld-in-lean/
58•saviorand•2d ago•5 comments

Flexport Is Hiring a Biz Ops Leader in Office of CEO

https://www.flexport.com/company/careers/
1•thedogeye•9h ago

Monte-Carlo Simulations

https://eli.thegreenplace.net/2026/monte-carlo-simulations/
121•ibobev•3d ago•22 comments

90% of US working hours are spent on tasks not covered well by AI benchmarks

https://www.mercor.com/research/labor-index/benchmark-map/
503•speckx•17h ago•426 comments

A story of WiFi secrets, UART lockdowns and a damaged flash memory chip

https://adepts.of0x.cc/eufy-homestation2-firmware/
44•michidk•18h ago•3 comments

Nine-Inch Snails

https://www.theatlantic.com/science/2026/10/giant-african-snails-florida-invasion/688850/
66•zdw•11h ago•33 comments

How the Claude Code team does code review

https://claude.dev/blog/solving-code-review-at-anthropic/
349•pretext•15h ago•329 comments

C++26: The New Debugging Header

https://www.sandordargo.com/blog/2026/09/30/cpp26-debugging
97•jandeboevrie•15h ago•23 comments

Wireless magnetic millirobots that deliver medicine inside the beating heart

https://hub.jhu.edu/2026/10/12/magnetic-millirobots-cardiac-treatment/
95•giuliomagnifico•3d ago•16 comments

Show HN: Gemini 4 Argon plays StarCraft

https://starskirmish.com/bench/v0-2/
222•__cayenne__•19h ago•101 comments