Canonical · Foundational Tech · Paper

Information Management: A Proposal

Tim Berners-Lee · 1989

"A linked hypertext system over a network would let information at CERN (and everywhere) be shared without central control."

The idea

CERN had a chronic problem: it was a research institution with constant staff turnover, and critical technical knowledge kept disappearing when people left because there was no consistent way to link related documents together. Berners-Lee proposed a system of 'hypertext' — documents containing links that point to other documents, regardless of what computer or format they lived on — accessible over CERN's internal network. The specific institutional problem was narrow, but the solution he designed had no reason to stay narrow: a linked, decentralized information system.

Why it works

The mechanism's most important property is what it deliberately lacks: a central index or master database of all the links. Earlier hypertext concepts assumed some central authority had to track and validate every link. Berners-Lee's system let anyone create a link to anything without asking permission or registering it anywhere — the cost of that decision is that links can break (the 404 error is a direct consequence), but the benefit is that the system has no bottleneck and no single point of failure, so it can scale to a planet's worth of documents without anyone in charge. That tradeoff (accept some broken links, gain unlimited decentralized scale) is the single decision underneath the entire World Wide Web.

The takeaway — recall it first

Check your understanding

Why did Berners-Lee's hypertext system deliberately avoid having a central index of all links?

The Anatomy of a Large-Scale Hypertextual Web Search Engine

"Ranking web pages by the link structure of the web itself (PageRank) produces dramatically better search than keyword matching."

Sergey Brin, Lawrence Page · PaperContinue→