A solid tradition // Week 23 — 2026(opens in new tab)
A week in which some things happen and some things do not happen, much like other weeks. Current situation: The pain of having children who become driving teenagers is the...
Curated river of news
Reader
Latest posts from blogs I follow
2972
entries
60
pages
Finding the Middle in AI Narratives(opens in new tab)
A pendulum swung from a collapsing clock tower to an immovable monolith, its heavy bob now resting at center/images/finding-the-middle-in-ai-narratives.webp/images/finding-the-middle-in-ai-narratives.webp Another major AI vibe shift is happening. The tech is moving so fast that our collective rea...
Communities of Not(opens in new tab)
There is a strange thing that happens in communities that gather around abstinence from something: identity from opposition. At their best these communities are not just negative: childfree spaces can be about autonomy, choice and acceptance, anti-car spaces about safer ...
Sprinter Van Phone Mount + Better CarPlay-Compatible Cable Situation(opens in new tab)
This is the stock look of my 2021 Sprinter van front console area: If it’s not obvious, there’s no great place for a phone. You’ve got the cup holders. Those are actually sorta workable, except for this boss battle: wired-only ...
From open source hits to OpenAI (Changelog Interviews #682)(opens in new tab)
This week I'm talking with Max Stoiber, currently working on ChatGPT's plugin directory and app platform at OpenAI. We discuss the hundreds of open source projects nobody remembers alongside the big ones like react-boilerplate and styled-components, how Spectrum became part ...
Self-Host Weekly (5 June 2026) (opens in new tab)
42 million Plex users can't be wrong
We are in the golden age of Open Source. Surf the wave until it crashes.(opens in new tab)
While a few people complain that AI is "stealing Open Source", training on Open Source code and giving nothing in return, I'm currently living the complete opposite: we are in
Day 3: The DGX Spark Unpacked. GB10, Unified Memory, sm_121, and the One Reason This Hardware Exists(opens in new tab)
A practical teardown of NVIDIA DGX Spark's GB10 Grace Blackwell Superchip, unified memory, sm_121, NVFP4 tensor cores, memory reporting, and decode limits.
The Story of C++ : The World's Most Consequential Programming Language | The Official Story(opens in new tab)
The New Van(opens in new tab)
I got something I’ve wanted for years and years! A camper van! I’m a camper van guy now! It’s a Mercedes-Benz Sprinter, and even more technically a Winnebago Revel. I scoured Craigslist, Facebook Marketplace, and RV-specific sites for a long ...
Kubernetes Autoscaling(opens in new tab)
Workload and node autoscaling in 2026
Anti-AI nostalgia and the cult of the past(opens in new tab)
How I made a 45-video course without wanting to die(opens in new tab)
Tips and tricks from me to you.
AI Predicts the Text of Answers(opens in new tab)
A stream of text reasoning its way through a locked room, landing on the one glowing stone/images/ai-predicts-answers.webp/images/ai-predicts-answers.webp There's a common argument about AI that says it doesn't understand anything about the world. That no matter how impressive it looks, it's ...
Thoughts on AI Adoption Speed(opens in new tab)
A human figure reaching upward, lifted by rising machinery — magnified, not replaced, by AI/images/ai-adoption-speed.webp/images/ai-adoption-speed.webp One of the most talked about AI topics is the speed of AI adoption in companies, and the reasons for it. I want to give ...
A pond of interesting problems(opens in new tab)
Async Rust: deep dive into cooperative scheduling and Tokio's architecture(opens in new tab)
Threads were designed to parallelize compute-intensive tasks. However, these days, a lot of applications (such as a network scanner) are I/O (Input / Output) intensive. Thus, threads have two significant
A survey of inlining heuristics(opens in new tab)
Compilers, especially method just-in-time compilers, operate on one function at a time. It is a natural code unit size, especially for a dynamic language JIT: at a given point in time, what more information can you gather about other parts ...
Wandler: Local OpenAI-Compatible Inference With Transformers.js and WebGPU(opens in new tab)
A practical Wandler deep dive with a local M1 Max WebGPU demo, real latency numbers, architecture diagrams, and getting-started commands.
The Case for Talking to Your Computer(opens in new tab)
I've always been a quiet person who prefers to write than to talk (though this may not be obvious based on the number of YouTube videos I've published). So it feels a little weird to say that I think ...
herdr(opens in new tab)
A tmux-like and agent-aware terminal multiplexer.
A Tale of Two Quality Cultures: The Importance of Quality Integration(opens in new tab)
Let's paint a picture of two very different environments, whose root differences can be largely traced back to differences embedded deep in their quality culture. Let's begin with what sadly is more likely to be the default state. "Quality isn't ...
5 CLI Tools That Actually Changed How I Work in 2026(opens in new tab)
Adorable tiny skulls // Week 22 — 2026(opens in new tab)
Sunday came, Sunday went, but the notes can be week notes any day they want to be. Current situation: Monday 25 May: Memorial Day, also hospital day. Having to work...
How I'm Thinking About the Anthropic and OpenAI IPOs(opens in new tab)
A charcoal architectural sketch: a lone figure steps out of a broken, tangled sienna structure on the left—current state—into a clean purple architectural corridor that races toward bright open light on the right—ideal state/images/anthropic-openai-ipos.webp/images/anthropic-openai-ipos.webp Some ...
From Kubernetes Dashboard to Headlamp: Understanding the Transition(opens in new tab)
For many people, Kubernetes Dashboard was their first window into Kubernetes. It offered a simple visual way to see what was running in a cluster, inspect resources, and build confidence without relying on the command line. For years, it helped ...
Social RSS (?)(opens in new tab)
What’s up with this? The spirit, of course, is “I want my Google Reader back.” I should be careful though: I imagine if we really did have it back exactly as it was, it would probably not be as cool ...
Terminal Trove May 2026 Wrap Up (opens in new tab)
A wrap up of all the things that happened in Terminal Trove in May 2026.
Front-End’s Missing Metric: The TBT Window(opens in new tab)
The TBT Window is the FCP-to-TTI interval used to calculate Total Blocking Time. If FCP or TTI moves, TBT can change even when long tasks do not.
Let the agents democratize open source(opens in new tab)
#482: Mr. Beast's episode(opens in new tab)
Topics covered in this episode: CVE-2026-48710: A Maintainer's Perspective daily-stars-explorer Markdown to pdf with pandoc and typst postman2pytest Extras Joke Watch on YouTube About the show Brian #1: CVE-2026-48710: A Maintainer's Perspective Marcelo Trylesinski suggested by Lee Luocks Short version: ...
Checking assembly with Z3(opens in new tab)
Short post today. New ZJIT contributor dak2 submitted a PR to fix an overflow bug in fixnum division in ZJIT. We did the division fine, but lied about the type of the result in the case of dividing FIXNUM_MIN by ...
Weird projects I shipped with AI(opens in new tab)
AI Contributions and Maintainer Load in Open Source(opens in new tab)
You wake up, brew the coffee, open GitHub, and there it is. Another pull request on your open source project. Thirteen thousand lines added. No issue filed first. No discussion. Just "here, please review this for me." Over the past ...
Butter bells, fresh from the kiln(opens in new tab)
Revisiting the closed canon(opens in new tab)
A post I wrote in 2023, the closing of the canon, predicted that LLM answers would replace search results, dramatically lowering traffic to individual sites, thereby removing the incentives to even write along with building and maintaining websites. Unfortunately, this ...
Self-Host Weekly (29 May 2026) (opens in new tab)
We've officially reached the papal stage of the AI debate
mlxcel: A Rust-Native Inference Engine for Apple Silicon, Tested on My M1 Max(opens in new tab)
Day-one deep dive into mlxcel v0.1.0, a Rust-native MLX inference engine. Real M1 Max benchmarks vs mlx-lm and Ollama on Llama 3.2 3B and Qwen 2.5 7B, with architecture diagrams and an honest take on TurboQuant.
New DuckDB-Iceberg Features in v1.5.3(opens in new tab)
DuckDB-Iceberg now has a number of new features supporting Iceberg Tables and Iceberg REST Catalogs: `MERGE INTO`, `ALTER TABLE`, partition transforms, V3 support, and others!
Reading Observability Tools? That’s a Robot’s Job(opens in new tab)
Last Thursday Honeycomb was kind enough to hand me the closing keynote slot at O11yCon, a conference whose entire premise is that observability matters and the people in the room are the ones who would know. I used that slot ...
Fleet: A Control Tower for Claude Code(opens in new tab)
A tmux dashboard for running many Claude Code agents at once — it shows which one needs you, sorted by urgency, so you stop polling and start responding.
We should be more tired than the model(opens in new tab)
Lately, I’ve been feeling like I’m losing control over the code I write when I work with agentic code generation. When I finish an agentic session, I get all the outward signs of having written code, but none of the ...
Did your editor font go default serif on WordPress 7.0?(opens in new tab)
Mine did. Like this: I guess WordPress 7.0 assumes you are using a theme.json file these days. I’m not doing that yet on any of the sites I work on. If you want to embrace that future, you could add ...
Keep calm and embrace (Rust) monorepos(opens in new tab)
Use monorepos My very unscientific analysis of the Rust ecosytem has come to the conclusion that Rust developers are wasting way too much time managing packages and dependencies. Whoever has
Day 2: Anatomy of an LLM Inference Request. From Prompt to Answer, Step by Step(opens in new tab)
A beginner-friendly walkthrough of tokenization, prefill, KV cache, decode, batching, TTFT, and why memory bandwidth shapes local LLM performance on NVIDIA DGX Spark.
How I use Claude Code as my personal assistant from my phone(opens in new tab)
For the past few weeks I've been using Claude Code from my phone and while it's not a 100% replacement for something like OpenClaw I've found it covers 80% of what I want without too many glaring secu...