A live stream walkthrough and commentary on a benchmarking article examining whether React Server Components (RSC) actually improve performance. The host analyzes various data fetching approaches—client-side rendering, server-side rendering without streaming, SSR with data fetching, and RSC with suspense/streaming—comparing metrics like LCP, interactivity time, and hydration gaps. Key finding: the real performance gains come from streaming and suspense, not RSC architecture itself. The host argues that frameworks like TanStack Start and SolidStart achieve similar results without RSC. A guest (Tanner) joins to discuss the 'isomorphic-first' vs 'server-first' framework philosophy, contrasting Next.js App Router's server-first model with isomorphic frameworks.

5h 59m watch time

Sort: