← antimony labs

History / The open Web

Who let everyone build the Web?

Pick what catches your eye.

1983
Connected computersExplore ↗ 1989
Scattered knowledgeExplore ↗ 1990
The hyperlinkExplore ↗ 30 April 1993
Code to shareExplore ↗ 1993
A window to the WebExplore ↗ The consequence
Anyone can add a siteExplore ↗
Play with a link ↓

1983

Connected computers

AI illustration: Connected computers
AI illustrationView full image ↗

Different networks can talk. Documents will travel along those connections.

Read more & sources

Could a document reach a different kind of computer?

ARPANET switched to TCP/IP in January 1983. The protocols provided a way for different networks to communicate. Years later, the Web could use that infrastructure instead of building a new global network for its documents.

What this doesn't explain. The Internet already carried services such as email. Connecting computers did not, by itself, make their information easy to find or read.

SourcesInternet pioneers explain the TCP/IP transition ↗

A common network carried requests between independently operated computers.

Internet pioneers explain the TCP/IP transition ↗Berners-Lee on the Web's design, 1996 ↗
Full article →
How was this image made?

The creative brief sent to the image model. These are instructions, not a record of what actually happened.

Original prompt & settings (JSON) ↗
↑ Choose another stop

1989

Scattered knowledge

AI illustration: Scattered knowledge
AI illustrationView full image ↗

CERN's projects keep changing. Their knowledge needs a way to stay connected.

Read more & sources

Why invent this at a physics laboratory?

At CERN, people, projects and software changed faster than their documentation could keep up. Berners-Lee proposed links between information that lived in different systems. His requirement was practical: let separate collections grow, then connect them without reorganizing everything.

How the proposal describes following links through his earlier Enquire system.
“rather like in the old computer game "adventure"”
Read the original →

What this doesn't explain. This began with a particular institution's needs. An open public publishing medium was not an inevitable outcome of a better filing system.

SourcesRead Berners-Lee's 1989 proposal ↗

Scattered documentation made connections between existing systems a design requirement.

Read Berners-Lee's 1989 proposal ↗
Full article →
How was this image made?

The creative brief sent to the image model. These are instructions, not a record of what actually happened.

Original prompt & settings (JSON) ↗
↑ Choose another stop

30 April 1993

Code to share

How it connects · schematic

Others are already building the Web. CERN releases its code for anyone to reuse.

Read more & sources

A working invention. But who was allowed to use it?

By April 1993, people outside CERN were already building Web software. CERN's declaration put its basic line-mode browser, server and common-code library into the public domain, explicitly allowing anyone to reuse, modify and distribute that code.

CERN's transcription of its 30 April 1993 release statement.
“permission is given to anyone to use, duplicate, modify and distribute it.”
Read CERN's transcription →

What this doesn't explain. This release concerned CERN's software. It did not make every website's content free, pay for Internet access, or remove every barrier to publishing.

SourcesCERN's account of the software release ↗See the signed release statement ↗The Mosaic team's recollections, University of Illinois ↗

Permission to reuse CERN's code removed a software-licensing obstacle for outside builders.

CERN's account of the software release ↗
↑ Choose another stop

1993

A window to the Web

How it connects · schematic

Mosaic makes the Web easier to explore, bringing more people into it.

Read more & sources

What makes a network worth exploring?

NCSA's Mosaic helped people navigate the Web through a graphical interface. Versions for common computer platforms made it accessible to a wider audience. The software release and the browser addressed different obstacles: permission to build, and a reason to participate.

What this doesn't explain. Mosaic was not the first Web browser. Early versions were already circulating before CERN's April declaration; Windows and Macintosh versions followed later in 1993.

SourcesThe Mosaic team's recollections, University of Illinois ↗Berners-Lee's timeline of the early Web ↗

Mosaic turned the Web's existing links and documents into an interface more people could use.

The Mosaic team's recollections, University of Illinois ↗

A growing audience gave publishing on the Web greater practical value.

The Mosaic team's recollections, University of Illinois ↗
↑ Choose another stop

The consequence

Anyone can add a site

How it connects · schematic

Independent sites join through shared rules. No central editor has to add every page.

Read more & sources

Why can a small site belong to the same Web as a university?

A publisher can run a compatible server and connect to other resources through links. The architecture does not require one central editor to enter every new page. Shared rules let independently built sites participate in a larger whole.

Tim Berners-Lee, reflecting on the design in August 1996.
“anyone could start a server, and it could run happily on the Internet without regard to registration with any central authority.”
Read the original →

What this doesn't explain. That freedom does not ensure equal reach, reliable information or permanent links. Access providers, hosting services and other institutions still affect who can publish and who gets seen.

SourcesBerners-Lee on the Web's design, 1996 ↗Berners-Lee on decentralization ↗

An address in one document can point to another publisher's resource without combining their databases.

Berners-Lee on the Web's design, 1996 ↗

Permission to reuse CERN's code removed a software-licensing obstacle for outside builders.

CERN's account of the software release ↗

A growing audience gave publishing on the Web greater practical value.

The Mosaic team's recollections, University of Illinois ↗
↑ Choose another stop

A miniature Web · yours to try

Your server

Field notes

Not published

Museum server

Collection catalog

Online

A link stores an address. It does not copy the destination or keep that server running.

Why this matters

A link stores an address, not a copy. Independent sites can connect without combining their collections, but a destination can disappear. This example runs in your browser. Real publishing needs a server and network access.

Read the design rationale →