Active Preservation Initiative

The Mosaic Archive Project

A comprehensive preservation initiative dedicated to recovering, cataloging, and rendering web pages as they appeared in the Mosaic browser — the first widely-used web browser from 1993.

Status: Active
Started: 1996
Pages: 1.2M+
Version: 3.2.1

Project Overview

The Mosaic Archive Project represents one of the most ambitious digital preservation efforts ever undertaken. Launched in 1996 by a consortium of NCSA alumni, web historians, and preservationists, this initiative aims to capture and preserve the entire visual experience of the early web as rendered through Mosaic — the browser that brought the World Wide Web to the masses.

What makes this project unique is not just the collection of HTML source code, but the faithful recreation of how pages looked and felt in the Mosaic browser environment. Every color, every font rendering, every broken image placeholder — all preserved with meticulous attention to period accuracy.

Mosaic - [Welcome to the Information Superhighway] X
File Edit Options Go Help

☆ Welcome to the Information Superhighway ☆

This page was created using NCSA Mosaic, the first graphical web browser.

Best viewed at 640x480 resolution in 16 colors.


Links:

NASA Homepage

World Wide Web Consortium

Cool Links Page


This page is part of the Mosaic Archive Project.

Done 1995-03-15

What We Preserve

  • Full HTML source code as served in the 1990s era
  • All inline and external images (GIF, BMP, early JPEG)
  • Embedded MIDI files and audio elements
  • Browser rendering fingerprints — how Mosaic interpreted each page
  • HTTP headers and server responses from original crawls
  • Broken links and dead URLs — documented and catalogued
  • JavaScript fragments (even if Mosaic couldn't execute them)

Why Mosaic?

Marc Andreessen and Eric Bina's Mosaic browser, released in 1993, was the gateway through which millions of people first experienced the World Wide Web. Before Mosaic, the web was an academic curiosity accessed through command-line tools. Mosaic made it visual, intuitive, and accessible.

By preserving pages as they appeared in Mosaic, we're not just saving HTML code — we're saving the experience of the early web. The way tables were used for layout, the limited color palettes, the chunky serif fonts, the "under construction" GIFs — these aren't just aesthetic choices, they're cultural artifacts of a specific moment in human history.


Technical Specifications

Our preservation pipeline uses a combination of period-accurate emulation and modern archival technology to ensure every page is captured with maximum fidelity.

Browser Engine
Mosaic 2.0/2.1/3.0
Emulated rendering using Mosaic-compatible layout engine
Color Depth
16 / 256 / 16-bit
Preserves original color depth limitations per era
Font Rendering
Times New Roman
Period-accurate font stack with MS Serif fallback
Image Formats
GIF89a, BMP, JPEG
Native support for all Mosaic-compatible formats
Archive Format
WARC + Mosaic-X
Custom extension of WARC for Mosaic-specific metadata
Storage
847 TB Total
Distributed across 12 geographically diverse locations

Preservation Pipeline

mosaic-archive ~ pipeline
archive$ mosaic-crawl --url "http://www.example.com" --date "1995-06-15"
[1/6] Resolving DNS... 192.0.34.163
[2/6] Fetching HTTP response... 200 OK (14.2KB)
[3/6] Extracting images... 3 GIF, 1 BMP found
[4/6] Emulating Mosaic rendering... ✓ Complete
[5/6] Generating snapshot hash... sha256:a3f8c2...
[6/6] Writing to WARC archive... ✓ Indexed
archive$ mosaic-verify --hash "sha256:a3f8c2..."

Archive Gallery

A sampling of pages preserved by the Mosaic Archive Project. Each entry represents a faithful recreation of how the page appeared in Mosaic at the time of capture.


Project Team

A dedicated team of historians, engineers, and preservationists working to keep the early web alive.

DA
Dr. Dana Andrews
Project Director
SK
Sven Karlsson
Lead Engineer
RL
Dr. Rei Lin
Digital Historian
JP
Javier Prieto
Crawler Architect

API & Developer Access

The Mosaic Archive Project provides programmatic access to our collection through a RESTful API. Researchers, developers, and educators can query the archive by date, URL, content type, and more.

mosaic-api ~ example
api$ GET /api/v2/mosaic/pages?year=1995&format=gzip
{
"total": 284931,
"pages": [
{ "url": "http://www.ncsa.uiuc.edu/",
"captured": "1995-04-12T08:23:01Z",
"size": 4217,
"mosaic_version": "2.0b4"
}
]
}
api$

Access the full API documentation for authentication, rate limits, and data export options.

"}