I tried Gleam for Advent of Code
blog.tymscar.com
SSE sucks for transporting LLM tokens
zknill.io
Analysis finds anytime electricity from solar available as battery costs plummet
pv-magazine-usa.com
Useful patterns for building HTML tools
simonwillison.net
macOS 26.2 enables fast AI clusters with RDMA over Thunderbolt
developer.apple.com
Java FFM zero-copy transport using io_uring
mvp.express
Photographer built a medium-format rangefinder
petapixel.com
Ask HN: How can I get better at using AI for programming?
EasyPost (YC S13) Is Hiring
easypost.com
What is the nicest thing a stranger has ever done for you?
louplummer.lol
How exchanges turn order books into distributed logs
quant.engineering
Show HN: I audited 500 K8s pods. Java wastes ~48% RAM, Go ~18%
github.com
Researchers seeking better measures of cognitive fatigue
nature.com
A Lisp Interpreter Implemented in Conway's Game of Life (2021)
woodrush.github.io
Using Python for Scripting
hypirion.com
A 'toaster with a lens': The story behind the first handheld digital camera
bbc.com
Z8086: Rebuilding the 8086 from Original Microcode
nand2mario.github.io
Show HN: LinkedQL – Live Queries over Postgres, MySQL, MariaDB
github.com
Will West Coast Jazz Get Some Respect?
honest-broker.com
The SWAR (SIMD-within-a-register) numbers are strictly better than the SIMD versions as well as the standard library baseline. Why is that? SIMD should be strictly faster if the machine supports it, since the SWAR max bitwidth is 64, while SIMD starts at 128 bits.
The Java SIMD API used here must not result in using actual SIMD machine code.