June 2026

Still lots going on while I’m looking for a job. I’ve been using Claude Design and Claude Code to build Catchup . It’s gone very well - using OpenSpec I’ve incrementally built it and got it to an MVP state. Because it uses Supabase as the backend I’m considering where to deploy it IF I do deploy it. Currently just implementing E2E tests with Serenity using the Screenplay pattern to get things to a happy place, and I’m intending to do some code review and hopefully post about findings.

I’m trying a similar process to what I used for Catchup with another pet project using the Google suite: Stitch for UI design and Antigravity for implementation. There are some subtle differences between the Google and Anthropic offerings. The implementation plan in Antigravity is nice.

Things I’m seeing/reading/watching/listening to in June 2026…

Videos

• Zig 2026: Andrew Kelley Explains the No-AI Policy, Moving off GitHub, and Why Zig Isn’t 1.0

  • A fascinating interview with the creator of Zig covering why they operate on a strict no-AI pull request policy (treating LLM contributions as “negative value slop” that drains limited review time), their deliberate migration from GitHub to Codeberg due to CI constraints and corporate stability, and why they reject the “worse is better” mindset to chase an uncompromising 1.0 release built to last 50 years.
  • Andrew Kelley seems like such a grounded person he’s become an instant inspiration to me.

• Agentic AI Coding Practices for Speed with Quality — Adam Tornhill

  • A presentation from Jfokus 2026 revealing how unmanaged AI coding creates an immediate velocity spike that is fully cancelled out after two months by a massive surge in code complexity. Tornhill details using custom Code Health MCP servers to keep AI in a self-correcting refactoring loop, enforcing strict 100% test coverage pipelines, and the emerging “CLEAR” principles for agentic software architecture.
  • Slides: https://drive.google.com/file/d/117SIISp-KXTq4MPU7_x1ssfDS0Pj6SvU/view?usp=drivesdk

• The Story of C++: The World’s Most Consequential Programming Language | The Official Story

  • A detailed documentary tracing the origins, design evolution, and massive real-world impact of C++, featuring interviews with Bjarne Stroustrup and other pioneers.

• IntelliJ IDEA: The Documentary | An origin story

  • The official origin story of JetBrains and IntelliJ IDEA, documenting the journey of how a small Prague startup built one of the industry’s most successful IDEs.

• NVIDIA GTC Taipei 2026 Keynote | Full Replay

  • Jensen Huang’s full keynote presentation at GTC Taipei 2026, outlining NVIDIA’s latest advances in next-generation hardware, physical AI, and autonomous agent systems.

• Qwen 3.6 vs Gemma 4: I Built the Same App With Both Locally

  • An engineering stress test comparing local models Qwen 3.6 (27B) and Gemma 4 (31B) side-by-side as they build the exact same cross-platform markdown editor using the OpenCode agent.

• How Fast Can You Parse 1 Billion Rows in Java? – Insane Speed Test • Roy van Rijn • GOTO 2025

  • Roy van Rijn walks through the optimization process for the 1 Billion Row Challenge (1BRC) in Java, explaining how techniques like custom memory mapping, branchless parsing, and SIMD execution can be used to push parsing speeds to the hardware limits.

• OpenSpec & Spec Driven Development with Claude Code

  • An insightful exploration of Spec-Driven Development (SDD) using Claude Code, demonstrating how to guide AI agents using structured markdown specs to build a Rails application MVP from scratch.

Podcasts

• Tech Truth: Agile Evolution & the Future of SW Engineering — Martin Fowler & Kent Beck (GOTO)
  • A legendary unscripted reunion looking back at nearly 25 years of the Agile Manifesto and Extreme Programming. They explore how AI (“the Genie”) acts as an endlessly patient tutor for rapid exploratory work, but draw a firm line between writing syntax for computers and the irreplaceable human skill of communicating with stakeholders to understand the business domain.
• Eric Ries: Why good companies go bad, and how to stop it
  • Andrew Haschka (Field CTO at GitLab APJ) breaks down the “AI Paradox”: why organizations pouring capital strictly into code generation tools hit a wall because writing code accounts for less than 20% of the software delivery lifecycle. He advocates for shifting focus to governance dashboards and managing agentic flows to unblock systemic bottlenecks.
• 5 Quality Gates That Let You Ship 250% Faster with AI Coding Agents — Ryan Cormack (The AWS Developers Podcast)
  • Motorway Principal Engineer Ryan Cormack outlines the practical engineering blueprint used to scale AI coding assistants across 120+ engineers. He discusses building custom tooling on top of the Agent Client Protocol (ACP), orchestrating parallel exploration agents via Git + S3 pipelines, and channeling Werner Vogels’ “Renaissance Developer” mindset where engineers own systems end-to-end.
• Kubernetes and retiring at the top with Kelsey Hightower
  • The amazing story of Kelsey Hightower’s career from his first job to retiring early. Very inspirational. Sounds like a techie whose attitude and proactiveness made his own “luck”.
• Geoffrey Cain’s new book - Steve Jobs in Exile: The Untold Story of NeXT and the Remaking of an American Visionary, the 12 years Jobs spent outside of Apple turned him into the leader the world came to know.
  • The Vergecast - How Steve Jobs became Steve Jobs
  • Founders - #420 Steve Jobs in Exile

Posts

• I Tried Both Claude Design and Google Stitch. Here’s the Honest Comparison
  • A side-by-side practical engineering comparison evaluating the design capabilities, context handling, and code generation limits of both LLMs.
• The Gold Standard of Optimization: A Look Under the Hood of RollerCoaster Tycoon
  • An incredible deep dive analyzing Chris Sawyer’s legendary feat of writing 99% of RollerCoaster Tycoon in raw x86 assembly language to compress a massive simulation into 90s hardware constraints.
• Screenplay Pattern in Playwright
  • Moving past traditional Page Object Models (POM) to implement the Screenplay Pattern in Playwright, focusing on actor-centric behaviors for cleaner, more maintainable end-to-end automation test suites.
• Bringing Gemma 4 12B to your laptop: Unlocking local, agentic workflows with Google AI Edge
  • Google’s official announcement detailing the release of Gemma 4 12B, optimized to run fully on-device on standard laptops to enable offline multimodal and agentic AI workflows.
• Bringing emulation into the 21st century
  • David Tyler builds a Kubernetes-orchestrated, microservices-based Intel 8080 CPU emulator to run Space Invaders—where each individual CPU instruction runs as its own microservice in a different programming language. A hilarious and educational critique of cloud-native over-engineering.
• A malware-laced “coding assessment” is hitting job searchers take home tasks
  • A security warning highlighting a rising phishing campaign targeting job applicants with malware hidden in technical take-home assessments.
• Hundreds of productivity books distilled to 6 lessons from Daniel Pink
  • Daniel Pink synthesizes key takeaways from reading hundreds of productivity-focused books, boiling them down to six actionable principles.
• How Claude Code works in large codebases: Best practices and where to start
  • Anthropic’s guide on integrating Claude Code into large, complex projects, offering strategies, best practices, and practical starting points.
• 7 coding patterns I stole from senior engineers
  • A collection of clean coding patterns and idioms commonly used by experienced developers to write robust, readable, and maintainable software.
• Stop Naming Your Variables “flag” – The Art of Boolean Prefixes
  • A practical post arguing against generic variable names like “flag” and advocating for semantic, self-documenting boolean prefixes like “is”, “has”, “can”, and “should”.
• Agentic Engineering
  • An article exploring “Agentic Engineering”—the practice of design, developer collaboration, and workflows built around orchestrating autonomous AI agents.
• CTOs agree: Cognitive debt is the new technical debt
  • Discussing why cognitive load (the mental effort required to understand code) is becoming a more critical bottleneck in modern development than traditional technical debt.
• Superpowers vs Agent Skills: Faster shipping, safer reasoning
  • Om Mishra breaks down the distinction between general LLM capabilities (“superpowers”) and target-specific, structured developer tools (“agent skills”) to achieve reliable AI reasoning.
• Agentic Engineering (System Design One)
  • A system design perspective on Agentic Engineering, outlining patterns, architectures, and standard protocols for agentic software delivery systems.
• Project Valhalla, Explained: How a Decade of Work Arrives in JDK 28 (JVM Weekly vol. 180)
  • Artur Skowronski reviews the milestone integration of JEP 401 (Value Classes and Objects) into OpenJDK, detailing the history and performance implications of this massive change.

Other

• Mailpit
  • A small, fast, low-memory, zero-dependency, multi-platform email testing tool and API for developers that acts as an SMTP server and provides a modern web interface to view and test captured emails.
• Google AI Edge Gallery
  • A curated gallery of on-device machine learning examples, demo apps, and model artifacts to help developers run and experiment with local AI across macOS, iOS, and Android.
• AI DDD Mainframe Modernization Patterns GitHub Repository
  • A pattern catalog from Project Rosetta featuring twenty-eight software design patterns combining Domain-Driven Design (DDD) with blackfield mainframe modernization strategies.
• TinyBase
  • A small, fast, reactive data store and sync engine designed for local-first JavaScript applications.
• Supertone Play
  • An AI-powered text-to-speech (TTS) and voice cloning platform designed to generate emotionally expressive, hyper-realistic voice narratives.
• Asimov’s Science Fiction Digital Magazine Store
  • The digital storefront for Asimov’s Science Fiction, one of the world’s leading Hugo and Nebula award-winning science fiction publications.
• Codility for Programmers
  • A platform hosting programming challenges and sample tests to practice coding skills, algorithmic problem-solving, and interview preparation.
• Trilium Notes
  • A free, open-source personal knowledge base application designed for organizing massive hierarchical note structures.
• Ben Eater’s Breadboard Computer Project (bbcp)
  • A step-by-step educational guide, parts list, and video series for building an 8-bit programmable computer from scratch using simple logic gates on breadboards.
• BBC Global Experience Language (GEL)
  • The BBC’s official design system, detailing guidelines, design components, and accessibility practices to build consistent, user-friendly digital services.