Astro has a pretty cool new feature that blends the best of static pre-rendering and dynamic server rendering. Matthew Phillips returns to show off Server Islands and explain how they work.

[0:00] Preamble
[8:15] Welcome Matthew Phillips & Component Architecture
[18:30] Compilers & The Perception of Complexity
[29:00] Suspense, Reactivity, and Framework Ecosystems
[36:30] Server Islands: Introduction & Demo
[41:30] SI: App vs. Site or Static vs. Dynamic
[53:30] SI: Clarifying How It Works & Limitations
[58:30] SI: Excalidraw! - Server Islands in Astro 
[1:04:30] SI: Excalidraw! - PPR in Next
[1:10:00] SI: Excalidraw! - Out of Order Streaming
[1:16:00] SI: Inspiration & Background
[1:23:15] SI: HTTP Considerations
[1:28:15] SI: The Spectrum of Solutions
[1:32:15] SI: Live Performance Demos are Impossible
[1:38:00] SI: Inspecting Network Timelines
[1:49:45] SI: PageSpeed Insights
[1:57:45] SI: The New JAMStack (Does PPR Matter?)
[2:05:15] SI: Q&A - Signals in Astro? Speculation Rules? 
[2:11:30] SI: Q&A -  Streaming JSON? PPR in Astro?
[2:14:30] SI: Conclusion & Matthew's Parting Words
[2:19:30] Intermission: A Unified Story
[2:31:45] Intermission: How The Giants Fall
[2:41:30] Intermission: There's Something Here
[2:47:30] Solid HackerNews - /hydrate Script Experiment
[3:09:00] Run Once Suspense (Version 2)
[3:20:45] Derivations as First Principles: Introduction
[3:24:45] DaFP: createMemo & createProjection 
[3:28:00] DaFP: Writable Variants
[3:32:00] DaFP: Diffing Revisited 
[3:37:00] DaFP: Trello Example
[3:47:15] DaFP: Why Derive? & Colorless Async
[3:55:30] DaFP: Deriving Signals from props
[4:01:30] DaFP: Conclusion
[4:09:15] This Week in JavaScript: Astro & Remix
[4:11:45] TWiJ: Stack Overflow Survey
[4:17:15] TWiJ: Atila Fassina's Content 
[4:19:45] TWiJ: Astro & Netlify

Thanks @zombiefacesupreme for the timestamps!!
[4:22:00] Conclusion & Small Gift

Ryan Carniato

A live stream discussion with Astro core contributor Matthew Phillips about the newly introduced Server Islands feature in Astro. The conversation covers how Server Islands work by allowing different parts of a page to be cached and loaded at different times, distinguishing static shell content from dynamic personalized content. The hosts compare Server Islands to Next.js Partial Pre-Rendering (PPR), explore the tradeoffs of each approach, and discuss broader topics like component architecture, reactivity systems, compiler-based frameworks, and the distinction between 'dynamic' vs 'interactive' content as a new way to think about websites vs apps.

Server Islands w/ Matthew Phillips