===== The Request Tax: Re-evaluating 20 Years of Web Performance Dogma =====
  * **Speaker**: Alex Moon  
  * **Room**: CC 236  
  * **Time**: Sun 3:00 pm – 3:30 pm  

  * **Format**: Lecture (30 Min + Q&A)  
  * **Difficulty**: Introductory  
  * **Track**: Development & Dev Tools  
  * **Additional Tags**: Networks  
  * **Presenter Location**: In-person  
  * **Experience**: first time speaking / several-th time speaking  
  * **At**: anywhere / at lfnw  

==== Description: ====  
For nearly two decades, web developers have lived by a simple rule: **reduce your HTTP requests**. In the era of HTTP/1.x, that advice made perfect sense — TCP handshakes, head‑of‑line blocking, and connection limits meant that bundling everything together was the only way to achieve acceptable performance.

But the web has changed. With HTTP/3, QUIC, and modern CDN architectures, Alex has observed real‑world cases — especially in Headless WordPress and WPGraphQL — where **smaller, granular requests outperform monolithic ones**, thanks to better caching behavior and more efficient delivery at the network layer.

This talk investigates what Alex calls the **“Request Tax”** and asks whether our long‑held performance dogma still applies. Attendees will explore:

  * **The Wire** — How HTTP/3’s UDP‑based streams and multiplexing reshape the cost model of a request  
  * **The Cache** — Why giant all‑in‑one fetches can sabotage CDN efficiency, and how granular requests improve cache reuse  
  * **The Data** — Comparisons between legacy patterns and modern architectures, backed by research and Alex’s own testing  

This session doesn’t promise a silver bullet. Instead, it offers a clear‑eyed look at where old rules still matter, where they fail, and why it may finally be time to stop fearing the humble HTTP request.

**Target Audience:**  
  * Web developers  
  * Sysadmins  
  * Web performance practitioners