Internet Archives
A documentary archive of what the internet lost: the companies, software, browsers and sites that disappeared, and the evidence that they were ever here.
Preview only. This one isn't open for use.
What it is
Sites close, companies get bought and shut down, and file formats stop opening. What is left is scattered across dead links, half-complete captures and memory. Internet Archives is my attempt to build the record properly: every subject dated, every claim sourced, and every gap in the evidence stated instead of smoothed over.
It pairs the archive with a film programme. The first film is in production, and nothing has been published yet.
Walkthrough
- Start with the headlineThe page opens on a single line, The internet forgets, with a strip showing recorded events by year from 1990 to 2026.
- Browse the archiveThere are 80 subjects. Filter them by what happened: shuttered, acquired, abandoned, erased or still running. Or filter by kind: company, software, browser, website or protocol.
- Walk the timelineThe timeline places 269 recorded events on the year they happened, split by decade: 49 in the 1990s, 86 in the 2000s and 119 in the 2010s. Years with nothing on record stay empty instead of getting padded.
- Open a subjectEach record has its dates, a plain description of what happened, and how it ended. Examples include GeoCities, Netscape Navigator, Adobe Flash Player, ICQ, Google Reader and Vine.
- See what is recoveredEvery record states how many artifacts have been recovered. Right now that number is zero, and the site says so openly.
- Follow the filmsThe film section explains what is in production. Once a film is released it lives on the site permanently, with a full transcript and links to every artifact it uses.
- Hear about the next oneAn email box for one message per release, nothing more, and unsubscribing takes a single click.
Design decisions
- Honesty is the main feature. The site shows 0 artifacts recovered instead of hiding it, and sparse years look sparse.
- A dark, quiet monospaced look so the evidence does the talking. A separate study explored the logo, a ring that looks like it is fading out.
- If JavaScript is switched off, the full list of subjects is printed on the page, so nothing is unreachable.
How it's built
- The prototype is a single HTML file with its own small scripts, no framework, and nothing loaded from a server except fonts.
- Video slots in the page are placeholders with poster images, ready for the film clips once they exist.
- The plan for a production version is to give every record its own page, so each one can be linked to, indexed and cited.