Speech Models That Fit in 500kb

Moonshine Micro runs STT and TTS in under 500kb — no GPU, no API, no network calls. Here's where that changes what you can ship.

Your curl Has a Hidden Latency Profiler

One flag turns curl into a diagnostic tool that breaks down DNS, TCP, TLS, and transfer time.

Route LLM Calls by Complexity, Not by Habit

Kimi K3 just proved open-weight models can hit frontier quality. Here's the routing pattern that lets you mix self-hosted and premium APIs.

Dependabot Now Waits Before Upgrading Your Deps

GitHub just gave Dependabot a default cooldown period. If you auto-merge dependency PRs, this quietly changes your threat model.

Stop Volume-Mounting Code Into Containers

Compose watch mode syncs files, rebuilds on dependency changes, and restarts on config edits — all without bind mounts.

Math.tanh Now Fingers Your Operating System

Since Chromium 148, Math.tanh returns values that differ by OS. Any website can detect it in milliseconds.

SQLite Strict Tables Prevent Silent Type Bugs

SQLite's loose typing has been hiding bugs in your data for years. STRICT tables fix that with one keyword.

Why Your Lock-Free Queue Starves Threads

Lock-free MPMC queues guarantee progress but not fairness. Bounded waiting fixes the starvation you didn't know you had.

TypeScript 7 Ships With a Native Compiler

TypeScript 7 rewrites the compiler in Go. Type checks that took minutes now finish in seconds. Here's what changes for your project.

The Constraint That Kills Booking Race Conditions

Double-bookings happen when SELECT-then-INSERT races itself. PostgreSQL's EXCLUDE constraint prevents overlapping ranges at the DB level — no locks needed.

A Single PR Can Leak Your Private GitHub Repos

Security researchers just published GitLost — a proof that GitHub's AI coding agent can be tricked into leaking private repository contents through a single malicious pull request. The attack needs no stolen credentials or zero-days. It uses prompt injection: hidden instructions embedded in PR descriptions, code comments, or

Scale to Zero, Wake on Queue — KEDA in K8s

Stop paying for idle pods If you've got worker services sitting idle 90% of the time, you're burning cluster capacity for nothing. KEDA lets those deployments scale all the way down to zero — then wake up the instant work shows up at the door. What KEDA