April 2025 Archive
7921.
Hacker News Hug of Deaf (susam.net)
7922.
TSMC Warns of Limits of Ability to Keep Its AI Chips from China (bloomberg.com)
7923.
'Views' Are Lies (theverge.com)
7924.
Tailscale's surprising interaction of DNS settings and 'exit nodes' (utcc.utoronto.ca)
7925.
The Evolution of Graph Learning (research.google)
7926.
7927.
MacLynx beta 6: back to the Power Mac (oldvcr.blogspot.com)
7928.
Firefox Release 137.0 (mozilla.org)
7929.
Discrete Mathematical Modeling [pdf] (sites.math.washington.edu)
7930.
A 'fungi tile' with elephant skin texture to cool buildings (ntu.edu.sg)
7931.
Gemini 2.5 Pro is another game changing moment (old.reddit.com)
7932.
A Problem About Pigeons Powers Complexity Theory (quantamagazine.org)
7933.
Proof or Bluff? Evaluating LLMs on 2025 USA Math Olympiad (arxiv.org)
7934.
Infinite Mac (infinitemac.org)
7935.
TikTok deal put on hold after China objects over tariffs (reuters.com)
7936.
Lightweight threads and channels are better than async/await syntax
7937.
Aircraft supplier Howmet may halt orders if hit by Trump tariffs, letter says (reuters.com)
7938.
Call for expansion of Royal Navy surveillance after Kremlin spy devices found (theguardian.com)
7939.
Reconstructing the Knuth-Morris-Pratt Algorithm (cambium.inria.fr)
7940.
High-level discussions to phase out petrol and diesel vehicles (in India) (economictimes.indiatimes.com)
7941.
Try: Test anti-framework via CL Condition System (github.com)
7942.
For the first time in 25 years, California has a snowpack trifecta (latimes.com)
7943.
China hits back at US tariffs with export controls on key rare earths (reuters.com)
7944.
Ukraine's autonomous killer drones defeat electronic warfare (spectrum.ieee.org)
7945.
Show HN: An open-source code scanner to find issues in prompts and LLM calls (github.com)
7946.
Why an Australian journalist is teaching meditation in America's toughest jails (abc.net.au)
7947.
Probability Theory and Computational Mathematics [pdf] (tropp.caltech.edu)
7948.
Testing Frameworks and Mini-Languages (matttproud.com)
7949.
7950.
Meta's benchmarks for its new AI models are a bit misleading (techcrunch.com)