You can have your composer.lock and not make others eat it too! | Kevin Ullyott
How a one-line .gitattributes directive stops your PHP package's composer.lock from cluttering consumers' vendor/ directories and triggering false-positive vulnerability scans.
The 45% Hidden ROI of Automation No One Puts in the Business Case: The Vertical Spillover Effect | Digital Leadership
Research shows that nearly half the gain from capability-building flows upward to managers. The same “vertical spillover” applies even more strongly to
OneDisplay
You're using a MacBook with an external monitor. You want one screen. The big one.But macOS doesn't let you disable the built-in display without closing the lid.So you're stuck choosing between:Clamshell mode. Lid closed, charger always needed, worse airflow.B…
I’ve been a Tailscale user for almost five years. Today, I stumbled across NetBird again and largely replaced Tailscale with it.
Before I tell you about my customized Pi coding agent setup, I want to briefly mention how any vanilla coding agent is more than enough. I built most of Tuna...
Own Something and Be Happy
Remember how fun it was to burn and trade mix CDs back in the day? What do you mean "no"?
Rough Patches
Only a few close friends and family know this, but in 2023, I was quite certain I was done with the venture and technology business. Despite our financial success, my first firm Notation did not su…
When factories automated their assembly lines, the repetitive tasks went to machines, but humans stayed to oversee, fix problems, and make decisions. That’s exactly what’s happening to software…
Hardware breakpoints can trigger eBPF programs when specific memory addresses are accessed, leveraging CPU hardware support for low overhead. By utilizing these hardware breakpoints, we can efficiently monitor PostgreSQL’s internal variable updates, such as tr…
TL;DR: Paul Graham thinks it's bad when companies start competing on brand vs. functionality. I think he's wrong. It's actually good for consumers and innova...
If you need just one metric to judge anything in your professional life, it's the success metric.
The Final Form of Software Development
What if the final form of software development was just watching code and proof popping up while you sip a drink? Letting AI agents write assembly directly alongside Lean proofs sidesteps the whole compiler-trust problem. With a peek at real EVM 256-bit additi…
The recursive descent
Traversing the syntax forest, getting lost in the depths of recursion
>“We have no idea how we one day may generate revenue… [but] once we build this generally intelligent system, we will ask it for a way” - Sam Altman Fron...
Writing code using an AI agent is awesome and makes you super duper quick. Recent frontier LLMS write even somewhat decent code. So yey to that!
An interactive exploration of how solar panels work: from photons and the bandgap of silicon to Wright’s Law and the political will problem.
The hurrier I go, the behinder I get.
Back in December 2023, I got curious about Steven Bartlett's then-new book The Diary of a CEO: The 33 laws of business and life and ordered a copy for myself. I dived right in. On page 184 this graph hit me:
The best AI coding tool already supports open models. Two scripts and you're running Claude Code for free — locally or through OpenRouter.
Is it still all about the people? - Holistic Testing with Lisa Crispin
Modern agile software development was all about the people and led to high performing teams, where are the people in the AI discussions
Learn how memory allocation works by implementing your own malloc() and free() functions in C. A deep dive into heap management, fragmentation, and...
LLMs find the right factors but miss the frame
A clustered analysis of LLM-generated thesis statements for Dwarkesh Patel's AI Blog Prize questions.
Spencer Dailey:
Local LLM speed calculator
A back-of-envelope tokens/sec estimator for running LLMs locally. Pick your memory, model size and quantization to see how fast it'll actually run.
When cameras were first introduced, French painter Paul Delaroche reportedly opined “From today, painting is dead.” That was nearly two hundred years ago, and painting is doing just fine. But the nature of painting changed. Photography took over the work of fa…
Ask for an email recommendation on Hacker News or Reddit, and everyone agrees that Google spies on you and Microsoft is clunky, so the default answer is always the same: Fastmail. AI tools will parrot…
cuda-oxide is an experimental Rust-to-CUDA compiler that lets you write (SIMT) GPU kernels in safe(ish), idiomatic Rust. It compiles standard Rust code directly to PTX — no DSLs, no foreign language bindings, just Rust.
Ratty — A GPU-rendered terminal emulator with inline 3D graphics 🐀🧀
Ratty is a GPU-rendered terminal emulator with a spinning rat cursor, multiple 3D presentation modes and support for inline 3D graphics. Inspired by TempleOS and built with Rust and Ratatui.
Mythos 'Discovered' a CVE Already in Its Training Data - and That’s Still Worrying
Anthropic made headlines claiming Claude Mythos achieved the “first remote kernel exploit discovered and exploited by an AI.” We went looking for how - and found a 20-year-old bug hiding in plain sight.
A sophisticated campaign is abusing the Obsidian note-taking app to deliver a new RAT, PHANTOMPULSE, to targets in the finance and crypto sectors using social engineering and malicious plugins.
Mythos finds a curl vulnerability
yes, as in singular one. Back in April 2026 Anthropic caused a lot of media noise when they concluded that their new AI model Mythos is dangerously good at finding security flaws in source code. Apparently Mythos was so good at this that Anthropic would not re…
We Are Not Going to Agree on AI
On one hand, I know a developer producing 30,000 lines of code a month. On the other, I know a developer who says AI is stupid. Each swears by their stance and has evidence to back it up. One has a wo
Aren't AI hyperscalers supposed to pay for these upgrades?
I vibe-coded a GPU-aware Kubernetes TUI for 7 months, archived it, and started over. Here's what AI gets wrong when projects grow complex.
Running local models on an M4 with 24GB memory | jola.dev
Experiments with getting usable outputs out of local models on a standard Macbook