Search the archive

JAVASCRIPT™ TRADEMARK UPDATE

Deno is pursuing a case to cancel Oracle's trademark of "JavaScript," arguing it is a generic term for the programming language and that Oracle's use of it is misleading and restrictive.

ON ERROR HANDLING IN RUST

Error handling in Rust should not be done with the common practice of using large, all-encompassing error enums in crates, but rather by using structs to represent individual errors.

HOW WE ACCIDENTALLY SOLVED ROBOTICS BY WATCHING 1 MILLION HOURS OF YOUTUBE

These devs trained a neural network on 1 million hours of YouTube videos to predict the next moment in reality, allowing robots to understand physics, perform tasks in new environments, and potentially surpass LLMs for real-world grounding.

UNDERSTANDING REACT RE-RENDERS: WHAT TRIGGERS THEM AND WHY THEY MATTER

React re-renders components when their state changes, triggering a re-render of the component itself and, by default, all its child components. Wrapping components in `React.memo` can prevent re-renders if props haven't changed, but this comes with complexities and isn't always the best solution. A more effective approach is to isolate state and the components that depend on it into smaller, dedicated components.

REFLECTING JSON INTO C++ OBJECTS

C++26's reflection features can be used to convert a JSON file into a C++ object at compile time. This post provides a simplified example of parsing key-value pairs and then expands it to handle multiple pairs and different data types (integers, strings, and nested objects) within a JSON object.

LEARNINGS FROM BUILDING AI AGENTS

An AI code review agent was initially too noisy, flooding pull requests with low-value comments. To address this, the Cubic team updated the architecture by requiring explicit reasoning logs, streamlining the toolset, and specializing tasks with micro-agents. These changes led to a 51% reduction in false positives and halved the median comments per pull request.

WHY SENIOR DEVELOPERS ARE MORE VALUABLE THAN EVER

THE IMPORTANCE OF KINDNESS IN ENGINEERING

The most successful engineers are kind, as shown in their helpfulness, empathy, clear communication, and consideration for their team's and users' needs.

MAGNITUDE

Magnitude is an AI-powered browser automation framework that uses vision AI to control browsers with natural language. It lets users navigate, interact with, extract data from, and verify actions within web interfaces. The framework supports both high-level tasks and low-level actions and can intelligently extract data based on DOM content using a Zod schema.

VERCEL SHIP 2025 RECAP

Vercel Ship 2025 introduced the AI Cloud, expanding beyond the Frontend Cloud to support AI-native application development. Vercel also announced the AI Gateway for simplified model access, Fluid and Active CPU pricing for efficient AI compute, and Vercel Sandbox for securely running untrusted code generated by AI agents.

HOW MUCH SLOWER IS RANDOM ACCESS, REALLY?

There is a performance difference between summing array elements in sequential order versus random order. Experiments were conducted on both a MacBook Pro and a Linux desktop, showing that random access becomes much slower as array sizes exceed cache and RAM capacity, with a 4-16x slowdown observed within RAM limits, and a more drastic slowdown when exceeding RAM on Linux. The standard shuffle algorithm is inefficient for large datasets.

THE TIME IS RIGHT FOR A DOM TEMPLATING API

A declarative templating API would be useful to the web platform to address the lack of ergonomic and efficient DOM manipulation tools. The current absence forces devs to rely on external libraries. A standardized templating API would improve developer experience, security, and rendering times.

STATE OF DEVS 2025

The State of Devs survey focuses on workplace issues, health, and hobbies of developers.

AI DATASET GENERATOR

A Next.js application that uses OpenAI to generate realistic datasets for demos and learning.

WHY IS THE RUST COMPILER SO SLOW?

Slow Rust compilation times in Docker can be investigated with cargo-chef for dependency caching and profiling with rustc and LLVM.

NEXT.JS 15.1+ IS UNUSABLE OUTSIDE OF VERCEL

Next.js 15.1+ introduces metadata streaming that negatively impacts SEO and creates vendor lock-in with Vercel, making it unusable on other platforms.

LIQUID GLASS, BUT IN CSS

This article explains how to recreate Apple's new “Liquid Glass” design language from WWDC25 using CSS and SVG filters. The effect can be broken down into layers with box-shadows, backdrop-filter blur, color adjustments, and complex SVG displacement maps for refraction simulation. However, the implementation shown only works in Chrome, requires a lot of GPU resources, and is only practical for hero elements rather than full UI due to performance constraints.

A GUIDE TO SCROLL-DRIVEN ANIMATIONS WITH JUST CSS

Scroll-driven animations can be implemented in CSS, allowing animations to be controlled by user scrolling instead of time. There are three key components: the target element, the keyframes defining the animation, and the timeline (either `scroll()` or `view()`) that links the animation to scrolling. The `scroll()` timeline triggers animations based on scrolling, while `view()` activates them when an element enters the viewport.

OPENTELEMETRY FOR GO: MEASURING THE OVERHEAD

Everything comes at a cost, even logging. This article benchmarks logging overhead in a Go HTTP server that handles over 10k requests/second. It finds that enabling OpenTelemetry increased CPU usage by 35% (from 2 to 2.7 cores), memory by 50% (10MB to 15-18MB), and p99 latency from 10ms to 15ms while adding 4MB/second of network traffic for telemetry export.

START ALONE, THEN TOGETHER: WHY SOFTWARE MODELLING NEEDS SOLITARY BRAINSTORMING

Software modeling sessions often suffer from the "CTO effect," where dominant voices override other useful perspectives. To counter this, it's better to start modeling sessions with 10-15 minutes of silent, solitary brainstorming to encourage everyone to generate ideas independently first.

PREPARING FOR GOOGLE ZERO

Google is declining as a reliable traffic source due to AI-powered search. Its new approach uses creators' work to compete against them without compensation, breaking the traditional SEO model that rewards valuable content with traffic. To adapt, businesses needing distribution need to start looking into niche communities, referral programs, conferences, audio content, and word-of-mouth marketing.

GEMINI CLI: YOUR OPEN-SOURCE AI AGENT

Gemini CLI is an open-source AI agent that brings Gemini directly to your terminal. It has generous free usage limits, with access to Gemini 2.5 Pro and up to 60 free model requests per minute. It also has Google Search integration and Model Context Protocol support.

BUILD AND HOST AI-POWERED APPS WITH CLAUDE - NO DEPLOYMENT NEEDED

The Claude app now allows users to build, host, and share interactive AI-powered apps directly within the platform. These apps are built using artifacts that interact with Claude through an API. When users access these shared apps, their usage is charged to their own Claude subscription, removing scaling costs and API key management for the developer.

PLJS

PLJS is a lightweight and fast JavaScript language extension for PostgreSQL 14+ powered by QuickJS. It has integrations, type conversions, and configurable functions with documentation available.

OPENAI CHARGES BY THE MINUTE, SO MAKE THE MINUTES SHORTER

To reduce the cost and time of OpenAI transcriptions, speed up your audio using ffmpeg before uploading it. Transcribing 2x or 3x sped-up audio reduces the number of audio tokens charged by OpenAI with barely any impact on transcription quality. In the example shown in this article, there was a 33% cost reduction with 3x speed compared to the original audio duration using the gpt-4o-transcribe model.

THE PROBABILITY OF A HASH COLLISION

This post explains the math behind hash collisions, which occur when a hash function assigns the same value to different inputs. It not only shows the exact formula to predict a collision, but also has a few simpler approximations that can be used for large datasets.

A SHORT HISTORY OF WEB BOTS AND BOT DETECTION TECHNIQUES

Simple bots are easily identified. Nowadays, there are more sophisticated bot detection methods like IP reputation checks, TCP/TLS fingerprinting, and JavaScript analysis. As headless browsers have evolved and the use of orchestration frameworks has grown, so have advanced detection techniques like behavioral analysis and latency checks to detect bots attempting to mimic human behavior.

WHAT PROBLEMS TO SOLVE - BY RICHARD FEYNMAN

Richard Feynman advises a former student to find satisfaction and value in tackling solvable, even humble, problems where they can make a meaningful contribution, rather than pursuing grand, unattainable goals.

CLAUDE CODE ROUTER

Claude Code Router is a tool that allows users to route Claude Code requests to different models, customize requests with plugins, and integrate with GitHub Actions.

BRINGING REACT'S `VIEWTRANSITION` TO VANILLA JS

2000 LINES OF CODE

In 1982, a developer protested his team's management's decision to measure software progress by lines of code by reporting that his optimization of an application resulted in -2000 lines of code.

⚒️ Build for free with Nylas APIs

BUILDING THE TWITTER 1.0 WEB EXPERIENCE

Twitter Lite was a Progressive Web App designed to provide a fast, responsive, and data-efficient Twitter experience across all devices, especially in emerging markets with low-end devices and unreliable internet. Starting in 2015, Twitter Lite aimed to create a universal web app from a single codebase, leading to a 65% increase in pages per session and a 75% increase in tweets sent. This article goes over the technical aspects, like code splitting, service workers for caching, image optimization, lazy loading, and responsive design principles.

FAST(ER) REGULAR EXPRESSION ENGINES IN RUBY

This article compares the performance of different regular expression engines in Ruby, including the default Onigmo, re2 (Google), rust/regex, and pcre2. Benchmarks show that rust/regex is the fastest overall, while re2 is generally faster than Ruby's default, except when dealing with Unicode text.

WORKER THREADS IN NODE.JS: A COMPLETE GUIDE FOR MULTITHREADING IN JAVASCRIPT

Worker Threads allow you to spawn multiple threads of execution within a single process, with each worker running in its own isolated V8 environment to execute heavy JavaScript code without blocking the main event loop. They are ideal for CPU-intensive tasks like image processing, data parsing, complex mathematical computations, and machine learning inference.

WRITING TOY SOFTWARE IS A JOY

Writing "toy programs" is a way to rekindle joy in software development, especially as AI and commodification threaten the craft. These programs focus on 20% effort for 80% functionality.

PROGRAMMER YELLING AT THE CLOUDS ABOUT VIBE CODING

While AI can be helpful for simple tasks like auto-completion, it struggles with complex problems and often provides solutions no better than existing resources like StackOverflow. Reviewing AI-generated code requires a deep understanding, potentially more than reviewing code from human colleagues.

WHY ENGINEERS HATE THEIR MANAGERS (AND WHAT TO DO ABOUT IT)

Engineers often have frustration towards their managers due to common management anti-patterns. These include disrupting focus with frequent meetings, making uninformed technical decisions, taking credit for engineers' work, and scheduling excessive unnecessary meetings.

live to code along

VSCAN

VSCan is a tool that analyzes Visual Studio Code extensions for security vulnerabilities. You can input an extension's name or ID, and VSCan performs a deep analysis of its code, permissions, and metadata. The tool then generates a report detailing potential risks and recommendations.

AUTUMN

Autumn is an open-source platform that makes pricing and billing easier by acting as an intermediary between Stripe and your application. It supports various pricing models like subscriptions, credit systems, and usage-based models, without requiring extensive webhook management.

INTL-T

Intl-T is a fully-typed, object-based i18n translation library for TypeScript with features like autocomplete, node-based translations, and support for JSON files and dynamic remote imports. It integrates easily with React and Next.js and provides components for React injection and server-side rendering.

THE BITTER LESSON IS COMING FOR TOKENIZATION

Tokenization-free LLMs are possible. Tokenization is efficient but can lead to information loss and various issues that have to be addressed later on with different methods. The post explains recent advancements in byte-level models like ByT5, MambaByte, MEGABYTE, SpaceByte, and ends with a detailed analysis of the BLT architecture, which uses an entropy-based dynamic patcher to compress sequences and allocate compute more efficiently.

BASIC FACTS ABOUT GPUS

This article goes over the intricacies of GPU performance, focusing on maximizing efficiency through understanding the compute and memory hierarchy. It explains how to analyze kernel performance, categorized as memory-bound, compute-bound, or overhead-bound, and introduces the Roofline model for visualization.

MCP IS EATING THE WORLD—AND IT'S HERE TO STAY

MCP is gaining traction despite previous attempts at connecting LLMs to external tools falling short. This is because MCP benefits from better LLMs, a vendor-neutral protocol design, and straightforward tooling, making it easier for developers to adopt it for building and sharing tools.