Chris Zacharias
How visual media on the web has always worked Subject → Camera → Image → Storage ←Network ←Application ←Interface ←Display ←Viewer
2007: Mobile devices introduced
- One image can be displayed on different types of screens
- Can no longer predict where/why/how visual media will be displayed
- Heavier penalties if you approximate incorrectly
- Time consuming to optimize one image for multiple viewports
- To keep up, visual media has to become dynamic
- Adapt to network, device etc.
- Output needs to be driven by signals from the environment
Subject → Camera → Image → Storage → [NEED NEW LAYER HERE] ←Network ←Application ←Interface ←Display ←Viewer
Why is one needed now?
- Distance from visual media to brain is decreasing (becoming more personal)
- Movie screen → Television → Computer → VR
- The Web is primary conduit of visual media
- Over 80% of the average web page is visual media (images, video, fonts, styles)
- The Web wasn’t designed to do this
tag was almost scrapped from the first HTML spec - concern that images would cause the Internet to be used for porn
- Other media types unaccounted for until approximately 2010
- First YouTube HTML5
- Requires a “layer cake” of technology and shims
- E.g., How 3D currently works in the browser — not natively handled: HTML Document → JS Loader → Canvas Element → JS Logic → WebGL content → glShader
- Web technologies still take cues from print media (margin, padding, stylesheets)
- The Web needs to advance as a graphics platform
How do we get there?
- Start with the infrastructure
- Then adapt browsers, software, and services
- Explore new metaphors
- Window = scene
- DOM = scene graph
- Document = stage
- Elements = actors
- Styles = properties
- Scripts = scripts and behaviours
- Keep best parts of the web
- View Source - means anyone can learn
- Browser is the development environment
- Preserve indexing and searchability
- HTTP resource loading