Shelf / 75 saved pieces
reading
Essays worth keeping close. Open one to read the original website without leaving Ombogo Jr.
Followed RSS feeds refresh when you visit and with every publish.
75 pieces
Meaning search uses Google. Enter a phrase, then press Enter.
Follow the thought.
- unikernels were hard. key word: were. Justin Cormack, who worked on MirageOS back in the day, called me at the pub to talk about why unikernels are coming back. The friction that made them hard is gone, the operating system is design debt, and if you care ab…
- Bun v1.4.3 Fixes 166 issues, addressing 122 👍. bun check, a TypeScript type checker built into Bun that is 3x to 6.4x faster than tsc 7, --check for bun run, bun build and bun test, 67–89% less idle CPU, Bun.FetchSession, experime…
- if your team is too busy doing their 'normal job' to experiment with AI, you're preparing them to be replaced When Ralph was published last year, I sat on it. People who have been reading this blog for a while know this. I first came across the idea of loop engineering in early January last year, sat on it for four months, and t…
- --help on your CLI is all you need for LLM context until you don't. Here's something I've been meaning to write up for a while. It's kind of late, but I'm going to lean into the whole MCP vs. CLI argument because I have a perspective others might find interesting.What's the actual differ…
- Humans Are Just Stochastic Parrots Ryan Dahl turns familiar criticisms of language models back on humans: pattern matching, confident mistakes, and borrowed ideas. Published 10 October 2025.
- Ryan Dahl — Tiny Clouds Ryan Dahl's personal essays on AI, software, JavaScript, and the way people think.
- If your team is happy, are you doing a good job? Brilliant jerks, ZIRP-era managers, and how psychological safety lost the plot. Part 3 of my conversation with Dr. Cat Hicks.
- Collecting favours Notes on reputation; A thing I learnt from dad
- How Optical Testing Constrains CPO Deployment at Scale Technical challenges of scaling high-volume optical test - insertions, single/double-sided probing, OE / CPO tests, and key equipment suppliers
- Shopify went back to native. I think the bigger shift is formal verification Agents have collapsed the cost of building native apps twice. But I think they will also make formal verification the default way software gets written.
- Talk of the Town - October 2026 Playing with Jev & home-cooked transit dev
- Do you actually like hard problems? I spent years becoming the engineer people reached for. Now they still reach for me, but for different reasons.
- What is Codemode More than a year ago I wrote a few posts here that recommended people not to load custom tools into their context (or MCP servers) but to just use more scripts. Most importantly I wrote that Code Is All You Need and I wr…
- Databricks ai_decide Is Not an LLM -- And That's the Point Databricks shipped a SQL function that makes structured decisions on unstructured data in sub-second latency. It returns probabilities, not paragraphs. Here's why that distinction matters more than it sounds.
- 💥 Main Branch: The One Where Your Coverage Check Was Lying Coverage uploads no longer fail CI on new branches, and dynamic workflows land in Copilot CLI and the Copilot app.
- GitButler A Git client for working on multiple branches at once, organizing commits, and keeping changes easy to review.
- Snowflake and Databricks Converged on AI Agents in 2026. Here Is What Changes for Data Engineers Both platforms now ship agents as governed objects, managed MCP servers, AI gateways and SQL extraction functions. The work moving to data engineers is context, permissions and cost control.
- Scaling National Climate Resilience: How Uzcosmos Is Transforming Water Governance with PlanetScope The Uzcosmos Agency in the Republic of Uzbekistan leverages Planet data to drive operational decision-making across government sectors.
- Beyond the Soda Straw: Why Modern Defense Needs Continual GEOINT Learn why a new geospatial intelligence (GEOINT) foundation is required to keep pace with modern warfare.
- A Complete Overview of AI Data Movement: Scale-Up/Scale-Out Networks and Memory Tiers The network is the compute: How networks process tokens, factors that limit network scaling, and memory trends such as disaggregation
- Deser: Rethinking Rust Serialization Serde is an amazing serialization library for Rust and it has been a huge reason why I felt productive with it for years. However already while at Sentry I got quite frustrated with some of the limitations with it but ac…
- The AI margin collapse is gathering pace OpenAI has cut Luna prices 90% in two months, DeepSeek is still king on cached input, and even Anthropic is now cutting Opus prices. What that means for the frontier labs.
- The Spec Can Come Later When I say things like "delete the code, keep the spec," many people seem to interpret that as a rule: write the specification first, and only then are you allowed to program. That's not what I mean. It's also not how an…
- Counting 5.2 Billion Trees Across Two Nations See how researchers are empowering nations with scalable, high-accuracy methodologies to count trees and support global restoration goals and climate tracking.
- 🔐 Main Branch: The One Where Your Git Push Might Break GitHub SSH security improvements, Node 20 Actions retirement, and Node 24 for JavaScript actions.
- Fool's Expertise I listened to the recent Ezra Klein interview with Nvidia CEO Jensen Huang, which seems to be something of a Rorschach test for AI doomerism: if you are inclined to see doom, Jensen sounds dismissive, but if you (like me…
- Where do AI norms come from? A dialogue with Dr Cat Hicks on ethics, learning, and what we owe each other. Part one: start by listening.
- 🚧 Main Branch: The One Where Workflows Get Boundaries - Issue #41 Ubuntu 26.04, GitHub Actions workflow execution protections, and typed decisions from Jev
- What Sun got wrong We have the whole Oxide team coming to Emeryville this week for our annual OxCon meetup. For a remote company, an in-person meeting is uniquely energizing, and in preparation for that, we made some t-shirts featuring Oxi…
- the senior engineer death spiral on proving yourself, burning out, and being a good teammate
- Saving Jet Fuel I'll walk through Scikit-decide's optimal flight planning examples.
- Co-Packaged Optics for 224Gbps+ Scale-Up: An All-In-One Overview A system breakdown of challenges for every technology component of CPO: packaging, PIC, laser, and EIC
- Confessions of an Unrepentant Slop Snob A framework for thinking about when AI involvement is additive or a violation
- Interpreting Pangram Yesterday David Sacks wrote a tweet and within a few minutes people did, what they usually do, and they asked Pangram if it was AI. And Pangram said it’s entirely AI generated. To which David replied that these AI detect…
- The contagion of fear Bryan Cantrill, The Observation Deck — a confession he says he has never spoken of, even to those closest to him.
- An Advanced System Architecture Breakdown of OpenAI's Jalapeno Inference Accelerator Chad Wallace on how OpenAI's Jalapeno chip traded peak performance for latency and efficiency — system-level co-design, with AI in the design loop.
- Changelog – September 10, 2026 Val Town Access, Deno 2.9, Business & Pro trials, YouTube series, and more
- Planet Labs' Open Satellite Feed I'll walk through Planet Labs' new open satellite imagery feed.
- Have the frontier labs mixed up AI safety and security? The recent agent sandbox escapes at OpenAI and Anthropic look less like a technical failure and more like a philosophical one - treating security controls as though they only need to work most of the time.
- Bun v1.4.2 This release fixes two regressions impacting Elysia and AsyncLocalStorage, a hang in @discordjs/ws, CMYK JPEG decoding in Bun.Image, a rare JIT crash, and a GC crash on musl.
- ICEYE's Open Satellite Feed I'll walk through ICEYE's new open SAR imagery feed.
- Bun v1.4.1 Fixes 202 issues, addressing 236 👍. HTTP/2 in Bun.serve, Bun.write() streams Response bodies to disk, self-contained workspace node_modules, bun install --offline and --prefer-offline, WebSocket pause() and resume(), cr…
- Val Town Newsletter 27 A summer of writing from the Val Town team
- SaaS Isn’t Dead. Sameness Is. Everyone suddenly wants to tell you SaaS is dead. The Klarna-replaced-Salesforce story spread because it fit the mood, whatever the exact boundary between “replaced the CRM” and “consolidated the data underneath it” turn…
- Most of Your Architecture Was Just Expensive Code For decades, architectural knowledge lived inside the implementation, because the implementation was expensive to reproduce. That expense is disappearing. And as it goes, we're going to find out how much of what we calle…
- Q&A from the slop trenches This post is published on geospatialml.com, the blog I write with Isaac Corley. Read the full post →
- The real AI risk is inside the labs Amodei in his latest blog post wrote a mix of agreeable things and things that I believe misrepresent where the real risk of AI is located. I want to focus my attention on why, among all the risks, open weight models con…
- Being Linux Torvalds (This blog post was adapted from the transcription obtained from my YouTube video at https://www.youtube.com/watch?v=l6lxgYeVZqs) When Linus Torvalds developed the first Linux kernel, he had studied the Minix sources, he…
- Not just development, distribution of software may change as well Even if you are as averse to semver as I used to be in the course of my programming activity, you can still think of open source software distribution as something that used to follow a fixed number of steps. There is a…
- Sentinel-2 Paint: Recreate Any Image from Real Satellite Imagery This post is published on geospatialml.com, the blog I write with Isaac Corley. Read the full post →
- Claude Code Anthropic's running collection of Claude Code ideas, workflows, announcements, and field notes.
- Control the ideas, not the code Antirez argues that AI-era engineering shifts attention from line-by-line code review to design, testing, QA, and control of the underlying ideas.
- Prompting A practical guide to writing useful prompts for ChatGPT, ChatGPT Work, and Codex.
- Tech Blog — Mark Litwintschik Benchmarks and field notes on big data, cloud infrastructure, PostgreSQL, Spark, Python, and more.
- Ponytail — the lazy senior dev for your AI agent A deliberately minimal ruleset that teaches coding agents to stop at the smallest solution that works.
- Rewriting Bun in Rust Why and how the Bun team rewrote its runtime from Zig to Rust with agents, adversarial review, and a lot of testing.
- Similar but Different: A Benchmark for Measuring Whether Models Actually Use Multispectral Bands This post is published on geospatialml.com, the blog I write with Isaac Corley. Read the full post →
- GLM 5.2 and the coming AI margin collapse (part 1) Why open-weight models could crush inference margins and reorder frontier AI economics.
- Built to a Class, Not a Number Why design classes beat exact figures when the real world refuses to hold still.
- Seismic Double Stress What happens when a structure has to survive two shocks stacked on top of each other.
- A Letter from the Orchestra Pit The work of the musician is not the work of the machine — a letter about craft in the age of agents.
- The Coming Loop Armin Ronacher on harness-level loops, human judgment, and software that increasingly behaves like an organism.
- Why 80-metre wind blades fail Kenyan road geometry The blade fits the physics but not the roads — logistics as the real constraint.
- Turbine or Windmill? The Vertical-Axis Design That Breaks Every Rule A vertical-axis turbine that ignores the textbook and somehow still spins.
- AI demands more engineering discipline. Not less Why nondeterministic software raises the bar for production discipline instead of lowering it.
- Realtime prompting guide Shows how to prompt realtime voice agents, including Realtime 2 reasoning, preambles, tools, and exact entity capture.
- OpenAI Developers plugin Install and use the OpenAI Developers plugin in Codex, Claude Code, or Cursor to set up API access, use current OpenAI docs, and troubleshoot OpenAI API issues.
- API deployment checklist Review Responses API settings for production quality, latency, cost, reliability, and data handling.
- The Harness Is Everything What Cursor, Claude Code, and Perplexity actually built -- and why the model was never the real product.
- The Generative Stack — The Phoenix Architecture Why choosing one generative platform too early could become a decades-long architectural trap.
- The Claude Certified Architect Is Here — And It's Unlike Any AI Certification Before It A close look at Anthropic's architecture-focused certification, what it tests, and who it is for.
- Cognitive Security Geoffrey Huntley on the security boundary moving from systems and data into human cognition.
- The “Night Shift” Agentic Workflow Jamon Holmgren's day-and-night agent workflow for turning a scoped plan into reviewed code while keeping human judgment in the loop.
- Don't Waste Your Back Pressure Why useful constraints and feedback should shape an agent loop instead of merely slowing it down.
- Everything Is a Ralph Loop A compact explanation of the persistent agent loop pattern behind autonomous coding work.
No pieces match that search. Try a title or a source, or clear the search.