3 months ago
- GlitchyZorua aimed to archive the Ibiblio Icon Browser's thousands of GIF icons from the 1990s, but faced challenges due to server-side imagemaps.
- The site used server-side imagemaps instead of client-side ones, hiding direct URLs for icons, making bulk downloading difficult without manual clicking.
- A method was devised to infer hit areas based on a grid layout (72×89 pixels per icon) and generate URLs to fetch icons via HTTP HEAD requests and redirects.
- All icons were successfully downloaded and republished on a new static website with pagination and search, improving accessibility and archiving.
- The approach avoided expensive HTTP round-trips common in older web protocols, though licensing remains unclear under fair-use doctrines of the time.