Illuminating Horror Stories from the Browser API Haunted House
summary
Describes legacy browser API decisions that still cause security and performance pitfalls: document.write blocks the HTML parser so scripts at the top of a page delay image and CSS loading, and HTML5 drag-and-drop uses a shared dataTransfer and dropEffect flow that can let one window infer a successful drop in another (the cross-window remove-and-add case originating in IE-era design). Also surveys modern platform efforts—Web Components and Houdini—and argues that browser feature adoption oscillates between high- and low-level primitives and will likely remain limited in scope despite long lifetimes.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 1:48HTML standard quirks
- 3:34Blocking scripts and legacy
- 9:55Web standards support
- 15:11Drag-and-drop discovery
- 18:00HTML5 drag-and-drop
- 27:10Why drag-and-drop is weird
- 32:26Web standards time capsules
- 36:16Web Components and low-level
- 40:10Future of web standards
related talks
44mBattling Browser Bugs for "Fun" and (Non-)Profit
31mMagic tricks with Houdini
22mThe (Abundant) State of the Web
35mHTML5 APIs - The New Frontier
50mThe Browser is a Marvelous Thing
26mThe Browser is not a document reader!
25mHoudini - how you too can write a browser redndering engine!
26mProgressive HTML Rendering 101