SERIES,
one thread.

A series that splits the CWV measured by web-vitals into OpenTelemetry traces and metrics, and re-measures in a reproduction environment every point where a value disappeared without an error on its way from the browser to a Grafana panel

Examining how SWE-bench builds and grades its tasks, designing an experiment to compare models in Claude Code, and deciding whether to switch based on resolved rate and cost.

Why coldpath was built, how its V8 coverage analysis works, and how it traced the JavaScript dependencies of Toss Securities without source maps.

A record of forking number-flow to support older browsers and improving its animation performance in modern browsers.

The third cache layer that did not fit in the caching chapter of Frontend Performance Optimization Deep Dive. From the general theory of proxies, lifecycles, and strategies to applying it to a Next.js blog and measuring it with GA4

What should you decide first when building a link preview server? Following the principles from the reasoning behind the runtime choice to SSRF, encoding, and cache stampedes, and running every piece of code written along the way

A record of opening up and measuring Kubernetes in a kind cluster, so that frontend developers running SSR do not have to treat it as a black box

Essays on what remains for developers when AI writes the code, examined through the lens of judgment and responsibility

'use client', 'use server', 'use cache'. Following what a single line at the top of a file turns into at build time and at runtime, all the way down to the source

From the retreat of the Edge Runtime and Cloudflare reimplementing Next.js to the launch of the React Foundation. Five posts on where Next.js stands in 2026

A record of analyzing the performance of four services in production, along with the process and feedback from their developers