Computer science recap for week 18/2025

From 28/4/2025 to 4/5/2025

Articles #

• Fifty Years of BASIC, the Programming Language That Made Computers Personal 213 comments in 11 discussions

• Jepsen: Amazon RDS for PostgreSQL 17.4 160 comments in 3 discussions

• [D] ICML 2025 Results Will Be Out Today! 142 comments in 1 discussion

• Bloom filters 92 comments in 3 discussions

• Rye principles 87 comments in 5 discussions

• What the heck is AEAD again? 71 comments in 2 discussions

• BBC 1 1985 - A Computer Originated World (1999) 66 comments in 1 discussion

• Reversing the Fossilization of Computer Science Conferences 61 comments in 1 discussion

• [R] The Leaderboard Illusion 51 comments in 1 discussion

• Ups in Talks with Startup Figure AI to Deploy Humanoid Robots 48 comments in 2 discussions

• Show HN: Create your own finetuned AI model using Google Sheets 41 comments in 1 discussion

• Stop treating `AGI' as the north-star goal of AI research 37 comments in 2 discussions

• Nouveau: The Rule Based Language Family 35 comments in 2 discussions

• Learnable matrices in sequence without nonlinearity - reasons? [R] 35 comments in 2 discussions

• Designing the Language by Cutting Corners 28 comments in 3 discussions

Projects #

• An i386 Hobby operating system with graphics, multitasking, networking and an i386 C compiler 237 comments in 3 discussions

• Xiaomi MiMo Reasoning Model 189 comments in 1 discussion

• Distributed training of LLMs on consumer GPUs 27 comments in 1 discussion

• (Game Boy Emulator) Blargg Test #4 (04-op r,imm) Failing. 19 comments in 2 discussions

• I made my own Bison 10 comments in 2 discussions

• SEFA: A Self-Calibrating Framework for Detecting Structure in Complex Data [Code Included] [R] 10 comments in 1 discussion

• [P] Predicting the 2025 Miami GP 5 comments in 1 discussion

• minidisc: Zero-config service discovery for Tailscale networks 4 comments in 1 discussion

• Prolog Notes

• Felix86: Play x86-64 games on RISC-V

Videos #

• AI Learns to Play Crash Bandicoot [R] (Deep Reinforcement Learning) 9 comments in 1 discussion

• Breaking down math expressions to IR instructions without using trees 9 comments in 1 discussion

• Sydney's law, exponential growth in AI model capability

• Bring­ing ISA se­man­tics to Lean and Lean-MLIR — Léo Stefanesco

• PGP by Leslie Fish (WorldCon '96)