Search the archive

Register

About Test Double's Pragmatic AI approach

THE GEMINI APP IS NOW ON MAC

The Gemini app is now a native macOS experience, with integrated AI assistance accessible via a keyboard shortcut, allowing users to share their screen for instant context on local files and charts.

MAGIKA

Magika is an AI-powered file type detection tool that rapidly and accurately identifies over 200 content types using a lightweight deep learning model, used extensively for improving user safety.

CAL.COM GOES CLOSED SOURCE: WHY AI SECURITY IS FORCING OUR DECISION

Cal.com is moving its main product from open source to closed source after five years. This difficult decision was made primarily due to escalating security threats from AI, which can now rapidly find and exploit vulnerabilities in open-source code. While the core product is closing, a separate, open-source version called Cal.diy will be released under an MIT license for hobbyists and developers.

WHY AI HASN'T REPLACED HUMAN EXPERTISE—AND WHAT THAT MEANS FOR YOUR SAAS STACK

Stack Overflow reports that advanced questions on its platform have doubled since 2023. AI handles the easy stuff, but 75% of devs turn to humans when they don't trust AI output. Devs value the comments and discussion around answers more than the answers themselves, which is something AI can't replicate by collapsing discourse into a single confident paragraph.

FULLY REMOTE)

TLDR's Applied AI team is tasked with making every process at TLDR legible to code, runnable by anyone, and composable into larger workflows. Join a small, fast moving team using the latest AI tools with an unlimited token budget. Learn more [21].

CAVEMAN

Caveman is an LLM skill/plugin that enables terse, "caveman-like" communication to cut agent output tokens by up to 75% while maintaining technical accuracy.

CYBERSECURITY LOOKS LIKE PROOF OF WORK NOW

New AI models like Anthropic's Mythos suggest cybersecurity is becoming a "proof of work" system where securing systems requires outspending attackers in token-based vulnerability discovery.

GOOGLE BROKE ITS PROMISE TO ME. NOW ICE HAS MY DATA.

The Electronic Frontier Foundation has filed complaints against Google for breaking its promise to notify users before sharing their data with law enforcement.

HOW I LEARNED UNITY THE WRONG WAY

Years of "vibe coding" Unity by copying tutorials without understanding fundamental programming principles led to numerous interview failures for this dev, forcing him to rebuild his knowledge from the ground up to get genuine expertise and warning others against this common, increasingly prevalent pitfall.

✅ Truly serverless → Run your first query in ~4 minutes

✅ Up to 95% cheaper than traditional vector/text search databases

EVERYTHING SHOULD BE TYPED: SCALAR TYPES ARE NOT ENOUGH

Relying only on scalar types creates subtle semantic bugs that bypass compilers because they cannot distinguish between different domain concepts, a risk not fully mitigated even by grouping parameters into structs. The solution is to wrap every meaningful domain concept in a distinct "newtype," ensuring the compiler enforces semantic correctness and leading to safer, self-documenting code with compile-time validation.

MANAGING CONTEXT IN LONG-RUN AGENTIC APPLICATIONS

Slack Engineering overcomes LLM context window limitations in long-running multi-agent security investigations by using three complementary context channels. This system uses a Director's Journal for structured memory, a Critic's Review to assess and prune inaccurate findings, and a Critic's Timeline to construct a plausible event chronology.

I SHIPPED A TRANSACTION BUG, SO I BUILT A LINTER

THE AI CODING AGENT MANIFESTO

The AI Coding Agent Manifesto establishes core values emphasizing contracts over conventions, verification over generation, and explicit communication to guide AI-driven software development. The manifesto then outlines twelve principles, such as humans defining code skeletons, documenting agent conventions, separating code generation from judgment, and treating AI-generated code as "guilty until proven innocent."

HOW TO WALK THROUGH WALLS

The hacker mindset involves seeing through a system's superficial rules and abstractions to understand its deeper, underlying mechanics, allowing for unconventional and highly efficient outcomes. This approach shows the power of understanding how a system truly operates.

Read the changelog

CLAUDE CODE ROUTINES

Claude Code Routines automate unattended, repeatable tasks, such as code review or alert triage, by running saved configurations autonomously on Anthropic's cloud infrastructure. Users can trigger these routines via a schedule, an API call, or GitHub events like pull requests, and manage them through the web interface, CLI, or desktop app.

TURN YOUR BEST AI PROMPTS INTO ONE-CLICK TOOLS IN CHROME

Google Chrome is launching Skills, a new feature that allows users to save and reuse their most effective AI prompts as one-click tools for recurring tasks across different web pages. These personalized tools streamline workflows like product comparisons or document analysis.

PLAIN

Plain is an opinionated Python web framework, forked from Django and redesigned for the "agentic era," that provides an explicit, typed, and predictable foundation for both human developers and AI agents.

MY PR HAS BEEN WAITING A YEAR, OR THE EXPONENTIAL CURVE BEHIND OPEN SOURCE BACKLOGS

Open-source projects like Jellyfin face massive backlogs of unmerged PRs due to bottlenecks, such as a single maintainer nearing 100% utilization, which results in exponentially growing wait times and a "batch size death spiral" where large PRs clog the system. Combat this problem by optimizing the existing maintainer's time. Some solutions include capping PR size, gating quality, limiting work in progress, setting review cadences, and building a reviewer tier.

THE ORIGINS OF GPU COMPUTING

Government-funded academic research spanning 30 years, focused on parallel computing, graphics, and stream processing, directly led to the development of GPU computing, which is now vital for the modern artificial intelligence revolution and holds a lot of economic value. This foundational academic expertise was transferred to industry, evolving languages like Brook into CUDA at Nvidia.

I RAN GEMMA 4 AS A LOCAL MODEL IN CODEX CLI

An evaluation of Gemma 4 as a local agentic coding model in Codex CLI, compared against cloud models like GPT-5.4, showed that first-pass reliability, not raw token speed, was the most important performance metric, as cloud models were the fastest and most reliable. While cloud models still offer the best overall performance, Gemma 4's tool-calling ability makes local agentic coding viable.

ships in. See how it works

5NF AND DATABASE DESIGN

Traditional presentations of the Fifth Normal Form (5NF) are confusing due to unmotivated examples and illogical splitting, but designing schemas from business requirements and logical models naturally produces fully normalized physical schemas through patterns like the AB-BC-AC triangle or ABC+D star, making explicit 5NF decomposition unnecessary.

INTRODUCING A NEW SPAM POLICY FOR "BACK BUTTON HIJACKING"

Google is making “back button hijacking,” where sites manipulate browser history to trap users, an explicit spam policy violation, with enforcement starting June 15.

FIGMA DESIGN TO CODE, CODE TO DESIGN: CLEARLY EXPLAINED

Figma's MCP server enables AI agents to accurately bridge the gap between design and code, and vice versa, by transforming raw design data into a structured, LLM-friendly format.

MULTI-AGENTIC SOFTWARE DEVELOPMENT IS A DISTRIBUTED SYSTEMS PROBLEM

Multi-agentic software development is fundamentally a distributed systems problem with inherent impossibility results, like FLP and Byzantine Generals, that smarter LLMs alone cannot overcome, thus necessitating explicit coordination protocols and languages rather than waiting for future model improvements.

A BUG ON THE DARK SIDE OF THE MOON

JUXT used its spec tool, Allium, to find a critical four-byte omission in the Apollo Guidance Computer code that failed to release a gyroscope resource lock during a mid-torque "cage" event. This shows the power of behavioral specification in surfacing deep-seated bugs that thousands of people missed for decades.

WHAT WE LEARNED BUILDING A RUST RUNTIME FOR TYPESCRIPT

Encore developed a 67,000-line Rust runtime for TypeScript to overcome Node.js's single-threaded limitations and establish a multi-language core for future extensibility. By offloading infrastructure concerns and integrating tightly with Node.js via N-API, this architecture achieves up to 9x higher throughput and 80% less latency compared to other frameworks.

MOVING RAILWAY'S FRONTEND OFF NEXT.JS

Railway migrated its production frontend from Next.js to Vite and TanStack Router to resolve slow build times and better support its client-side architecture. The transition of over 200 routes was completed with zero downtime, resulting in a faster development loop and improved type safety across the application.

WRITE LESS CODE, BE MORE RESPONSIBLE

The discourse around AI-assisted programming often advocates for a balanced approach: using AI for tedious tasks while developers personally craft the enjoyable parts to guarantee quality. However, there are growing concerns about a potential decline in project quality and maintainability due to vibe-coded applications.

THE ECONOMICS OF SOFTWARE TEAMS: WHY MOST ENGINEERING ORGANIZATIONS ARE FLYING BLIND

Most organizations "fly blind" regarding their software teams' economics because they don't have the financial visibility to understand both their substantial costs and the value they must generate to be truly viable.

GITHUB STACKED PRS

GitHub Stacked PRs let developers break large code changes into small, interdependent pull requests, simplifying code reviews and speeding up the merge process. The system integrates natively with GitHub, providing a UI and a gh stack CLI for local management, and allows merging an entire stack with a single click.

FUSEJS

Fuse.js is a lightweight, zero-dependency JavaScript library that provides powerful fuzzy-searching capabilities for both the browser and server-side environments. Using the Bitap algorithm, it offers typo-tolerant matching and advanced features like weighted keys, logical operators, and tokenized searching within small-to-medium datasets.

HOW WE OPTIMIZED AGENT DEVICE FOR MOBILE APP AUTOMATION

Callstack developed the “agent-device” CLI to enable AI agents to control mobile devices for automated testing, achieving over 50% reduction in LLM token usage by replacing screenshots with trimmed, text-based accessibility-tree snapshots. This method aggressively trims the mobile UI hierarchy to only visible elements, improving the signal-to-noise ratio for the agents and keeping the context window useful longer.

THE UPHILL CLIMB OF MAKING DIFF LINES PERFORMANT

GitHub's “Files changed” tab struggled with large pull requests because each diff line contained 8+ React components, leading a 10k-line diff to consume 1GB of heap and create over 400k DOM nodes. The team solved this by rewriting the feature to use only two components per line, flattening state, and adding virtualization, resulting in a 78% cut in INP and a 50% reduction in memory.

BUILDING A CLI FOR ALL OF CLOUDFLARE

Cloudflare is rebuilding its Wrangler CLI into a unified "cf" tool that provides consistent infrastructure-as-code management across its entire catalog of over 100 products. This new TypeScript-based architecture guarantees synchronization between CLIs and SDKs.

WHAT THE AXIOS SUPPLY CHAIN ATTACK REVEALED ABOUT LOCKFILES AND PNPM 10

The Axios supply chain attack showed the importance of lockfiles for making sure predictable dependency versions and pnpm v10's safeguards for controlling install-time script execution to mitigate similar supply chain risks.

I JUST WANT SIMPLE S3

The author, looking for a simple, fast, and reliable S3 solution for basic needs, reviews several options like Minio and SeaweedFS, ultimately finding success with Versity GW due to its excellent performance when using a local filesystem backend.

HOW TO MAKE FIREFOX BUILDS 17% FASTER

buildcache's Lua plugin system can now cache Firefox's Python-based WebIDL code generation step, speeding up warm builds and showing its potential for optimizing other deterministic build processes.