A problem at a physics laboratory
CERN, the European particle physics laboratory near Geneva, was a crowded and constantly changing place in the late 1980s. Thousands of visiting scientists brought different computers, different software and different ways of storing documents. Finding information about a project, or who had worked on a part, could be hard. A British software engineer there, Tim Berners-Lee, had already built a small personal program in 1980 called Enquire to keep track of links among people and projects. In March 1989 he wrote a proposal for a larger system to help the laboratory manage information, and a revised version followed in 1990 with help from his colleague Robert Cailliau.
The proposal was modest in tone. It described linked documents that could be read from any kind of computer, with no central catalog to maintain. Its key insight was to combine two existing things: hypertext, the idea of documents with clickable links, and the internet, which already connected many computers, including those at CERN.
Three building blocks
The web is often described as a layer that sits on top of the internet. Three simple technologies made up that layer, and they are still its foundation.
- A way to name things. Each page or resource gets an address, now called a URL, so that a link can point to it unambiguously.
- A way to ask for things. HTTP is the set of rules a program uses to request a page from a server and receive it. The details are covered in the history of internet protocols.
- A way to write pages. HTML is a simple markup language that describes headings, paragraphs and links in a plain text file. The story of this format is told in how HTML markup shaped web pages.
Berners-Lee built a first server and a browser-editor on a NeXT computer, a workstation of the era, in late 1990. That first program, which he called WorldWideWeb, could display pages and also edit them. Later it was renamed Nexus to avoid confusion with the system as a whole. The first web servers were few, and the early web consisted mostly of pages describing the project itself and CERN’s own information.
Browsers carry the web outward
The web began to spread beyond CERN in 1991, when the software was made available to other institutions and described in public online discussions. A browser is the program on a user’s computer that requests pages and displays them. Several early browsers appeared, some with only text and others with graphics.
A line-mode browser that worked on almost any terminal was written at CERN, and universities in different countries wrote their own programs. A text-based browser from the University of Kansas called Lynx, and a graphical one called ViolaWWW, attracted attention among academics. These early efforts were useful but reached mostly technical users. Setting up a browser in the early 1990s often meant compiling software and understanding network settings.
Mosaic and the leap to ordinary users
A turning point came in 1993 with Mosaic, a graphical browser from the National Center for Supercomputing Applications at the University of Illinois. Its developers included Marc Andreessen and Eric Bina. Mosaic ran on common computer systems, was relatively easy to install, and could show images inline with text, which made pages feel more like magazines than technical documents. Many accounts identify it as the program that introduced the web to a wide audience, though browsers by others came before it and its success also depended on timing.
In 1993 CERN also stated publicly that the web technology could be used by anyone without fees. That decision is often considered important for the web’s rapid spread, because competing systems at the time, such as Gopher from the University of Minnesota, faced concerns about licensing.
The browser business
Andreessen left Illinois and, with others, helped found the company that released the Netscape Navigator browser in 1994. It became very popular within a short time. Microsoft, which had at first given little attention to the internet, responded with Internet Explorer in 1995 and bundled it with its operating systems in the following years. The resulting competition, sometimes called the browser wars, pushed both companies to add features quickly, and it also raised legal questions about bundling software. Another browser, Opera, appeared in the mid-1990s as well.
Features that people now take for granted arrived in this period: scripting within pages, forms, cookies for remembering visitors, and encrypted connections for shopping. Each solved a problem and introduced new ones, including privacy and security trade-offs.
In 1994, Berners-Lee founded the World Wide Web Consortium (W3C) to help guide shared standards, so that pages would work across different browsers. Standards bodies and commercial rivals have pulled in different directions ever since, and agreement has often been uneven.
Finding things on the new web
Once pages multiplied, finding them became the next challenge. Early users relied on hand-compiled lists such as web directories, and then automated web search tools arrived in the mid-1990s. The web also changed what people expected from computers, a theme in the broader story of the personal computer and the internet.
Keeping the story in proportion
The invention of the web is sometimes told as the work of one inventor. Berners-Lee’s role in proposing and building the core design is well documented, but the web depended on earlier research in hypertext, on the internet that already existed, and on many people who wrote servers, browsers and standards. It also depended on a policy choice to make the core technology freely usable. Seen this way, the early web was a combination of good timing, open design and a group of people willing to share their work.