fp.
newest
Open in hackernews
The Input Stack on Linux: An End-to-End Architecture Overview
https://venam.net/blog/unix/2025/11/27/input_devices_linux.html
15
•
venamresm__
•
1h ago
All recent bash commits are by one person
https://cgit.git.savannah.gnu.org/cgit/bash.git/log/
1
•
behnamoh
•
34s ago
•
0 comments
Ask HN: What open source projects are you grateful for?
1
•
jayzalowitz
•
1m ago
•
0 comments
Signal's president warns AI agents are an existential threat to messaging apps
https://fortune.com/2025/11/27/ai-agents-are-an-existential-threat-to-secure-messaging-signals-pr...
1
•
giuliomagnifico
•
3m ago
•
0 comments
Show HN: Hatch v1.16.0 – workspaces, dependency groups and SBOMs
https://hatch.pypa.io/1.16/blog/2025/11/24/hatch-v1160/
1
•
ofek
•
5m ago
•
1 comments
Why Have Vaccines Become a Religion?
https://www.midwesterndoctor.com/p/why-have-vaccines-become-a-religion
1
•
bilsbie
•
6m ago
•
0 comments
How Microsoft's developers are using AI
https://www.theverge.com/tech/831379/microsoft-developer-ai-usage-stats-notepad
2
•
manveerc
•
7m ago
•
0 comments
Major journal under fire for omitting Pfizer's failed flu data in seniors
https://blog.maryannedemasi.com/p/major-journal-under-fire-for-omitting
1
•
bilsbie
•
8m ago
•
0 comments
YouTube testing new 'Home' page with more control over suggested content
https://9to5google.com/2025/11/26/youtube-testing-new-home-page-with-more-control-over-suggested-...
2
•
geox
•
9m ago
•
0 comments
LLM unpredictability isn't a model problem – it's a process problem
https://pub.towardsai.net/uncertainty-architecture-a-modern-approach-to-designing-llm-application...
1
•
oddish-tv
•
11m ago
•
1 comments
Disallow code usage with a custom `clippy.toml`
https://www.schneems.com/2025/11/19/find-accidental-code-usage-with-a-custom-clippytoml/
1
•
austinallegro
•
12m ago
•
0 comments
Show HN: PrinceJS v1.7.7 Update. Down to 2.2 kB and top (13yo dev)
1
•
lilprince1218
•
15m ago
•
0 comments
Major AI conference flooded with peer reviews written by AI
https://www.nature.com/articles/d41586-025-03506-6
2
•
EA-3167
•
16m ago
•
0 comments
How to do Localization and Internationalization properly on the Web
https://community.qbix.com/t/how-qbix-handles-text-translation-and-international/783
1
•
EGreg
•
17m ago
•
1 comments
Did We Just Solve the 100-Year War Between Quantum Theory and Gravity?
https://pajuhaan.medium.com/did-we-just-solve-the-100-year-war-between-quantum-theory-and-gravity...
1
•
pajuhaan
•
20m ago
•
1 comments
Show HN: Cool fonts you can use almost anywhere
https://fontgen.cool/
3
•
liquid99
•
20m ago
•
0 comments
Cree syllabics still used today
https://www.thecanadianencyclopedia.ca/en/article/cree-syllabics
3
•
ilamont
•
24m ago
•
0 comments
Show HN: Zenus – A note-taking app with Local, Server and Client mode
https://github.com/skorotkiewicz/zenus
1
•
modinfo
•
24m ago
•
0 comments
Stochastic Bohmian Framework for Retrocausal Quantum Communication
https://github.com/DOSAYGO-STUDIO/quacomms
1
•
keepamovin
•
25m ago
•
0 comments
Pointer Pop Quiz
https://dave.cheney.net/2025/11/27/pointer-pop-quiz
1
•
todsacerdoti
•
26m ago
•
0 comments
Show HN: LLM Inference Performance Analytic Tool for Moe Models (DeepSeek/etc.)
https://github.com/kevinyuan/llm-inference-perf-model
1
•
kevin-2025
•
27m ago
•
0 comments
Jeff Dean on AI Trends at Stanford AI Club [video]
https://www.youtube.com/watch?v=AnTw_t21ayE
2
•
guiambros
•
28m ago
•
0 comments
Electron vs. Tauri
https://www.dolthub.com/blog/2025-11-13-electron-vs-tauri/
2
•
todsacerdoti
•
30m ago
•
0 comments
LaSuite Keynote
https://www.youtube.com/watch?v=X6c-mYOWrvA
1
•
maelito
•
31m ago
•
0 comments
Study: First Visualization of the Internal Structure Behind AI Decision-Making
https://news.kaist.ac.kr/site/newsen/html/news/?mode=V&mng_no=55090&skey=&sval=&list_s_date=&list...
1
•
giuliomagnifico
•
34m ago
•
0 comments
Desperately Seeking Squircles (2018)
https://www.figma.com/blog/desperately-seeking-squircles/
1
•
williamjsdavis
•
35m ago
•
0 comments
iOS games for iPads that cannot be upgraded beyond iOS 9
https://cjstewart88.github.io/vintage/
3
•
walterbell
•
35m ago
•
0 comments
All your LLMs ranked by speed every minute
https://metrik-dashboard.vercel.app/
2
•
mbouassa
•
36m ago
•
1 comments
Firm pioneers 3D printing copper coolers directly onto processors
https://www.tomshardware.com/3d-printing/firm-pioneers-3d-printing-copper-coolers-directly-onto-p...
1
•
Teever
•
38m ago
•
0 comments
Join the Parasite Rebellion on T-day
https://usop.substack.com/
2
•
richardatlarge
•
38m ago
•
0 comments
Ask HN: Why do people say LLMs create bad code "quality"?
2
•
chaidhat
•
40m ago
•
3 comments
Load More