Fil's Unbelievable Garbage Collector
fil-c.org
Why RDF Is the Natural Knowledge Layer for AI Systems
bryon.io
Escaping the odds and a formula for life (2024)
farhadg.com
What Is the Fourier Transform?
quantamagazine.org
Rasterizer: A GPU-accelerated 2D vector graphics engine in ~4k LOC
github.com
Stripe Launches L1 Blockchain: Tempo
tempo.xyz
io_uring is faster than mmap
bitflux.ai
Evolving the OCaml Programming Language (2025) [pdf]
kcsrk.info
Building Supabase-Like OAuth Authentication for MCP Servers
hyprmcp.com
WiFi signals can measure heart rate
news.ucsc.edu
Wikipedia survives while the rest of the internet breaks
theverge.com
Classic 8×8-pixel B&W Mac patterns
pauladamsmith.com
Forking Chrome to render in a terminal (2023)
fathy.fr
Atlassian is acquiring The Browser Company
cnbc.com
The Scam of Age Verification
pornbiz.com
What If OpenDocument Used SQLite?
sqlite.org
Experimental Hook-and-Loop Attachment System for Walls and Floors
core77.com
How to Build a High-Performance UR5 Inverse Kinematics Solver with IK-Geo
alexanderelias.com
Type-safe and user-friendly error handling in Swift 6
theswiftdev.com
A PM's Guide to AI Agent Architecture
productcurious.com
Almost anything you give sustained attention to will begin to loop on itself
henrikkarlsson.xyz
It's interesting to see how much Haskell has changed since this was written. It feels weird not having every Monad be an Applicative Functor. (Sidenote, I really like that you can define a Monad instance in Lean 4 and automatically get those. I guess Haskellers derive this stuff anyway, so it's not the biggest deal)