The internet today connects over 4.5 billion users and has irreversibly permeated modern work, play and culture. But rewind a few decades, and networked computing was non-existent outside specialized technical realms. Just how did we get from closed private networks to today‘s open participatory web? This article will analyze the landmark developments behind the internet‘s birth, evolution and explosive adoption.
The Precarious Beginnings: Building a Robust Defense Network
The backstory behind the internet begins in the high-stakes geopolitical context of the 1960s Cold War. The Soviet Union‘s launch of Sputnik sparked fears that the US was falling behind technologically and militarily. This spurred major investments into R&D for strengthening US technical capabilities through agencies like the Defense Advanced Research Projects Agency (DARPA).
DARPA was specifically tasked by President Eisenhower in 1958 to prevent another "Sputnik surprise" by developing cutting-edge military technologies. This included advancing surveillance, weapons systems, flight control and – crucially for the internet – robust, survivable communications for coordinating far-flung military operations.
The initial ARPANET network was motivated by concerns over surviving nuclear attack. Image source
Packet Switching: Paul Baran‘s Breakthrough Concept
In 1964, an engineer at RAND Corp named Paul Baran developed a novel technical solution to a challenge posed to him: "How might a communications network continue functioning even if 20% of it gets destroyed in an attack?"
At the time long-distance phone calls relied on dedicated analog circuits routing continuously between callers. Severing any link breaks communication. Baran proposed a radical upgrade – digital packet switched networks that chop messages into discrete chunks with address headers. Each packet independently navigates the network by getting passed from node to node without a predetermined rigid route.
Packet switching dynamically routes packets around network damage unlike circuit systems. Image source
This allows graceful degradation in nuclear attack scenarios rather than total failure. Surviving nodes quickly route around damage automatically. Baran published his now seminal volumes outlining this concept of "distributed networks" in 1964.
And Donald Davies independently devised a similar packet switched approach in 1965 while working on Britain‘s early NPL network. The groundwork was laid for constructing a network with no central hub.
Kicking Off ARPANET: UCLA Gets Connected!
The team at DARPA began implementing Paul Baran‘s radical vision for packet switching over the coming years under the leadership of Lawrence Roberts. The proposed network aimed to connect researchers across the country to timeshare expensive mainframe computing resources as well as foster collaboration. By 1969, Roberts spearheaded development of the first prototype nodes installed at UCLA, SRI, UC Santa Barbara and the University of Utah.
On October 29, 1969 at 10:30pm, history was made when the first host-to-host connection was made from SRI to UCLA via the Network Control Protocol (NCP) over the initial ~50 Kbps lines. The first message intended to login from SRI gets stuck after the first two letters "Lo" due to a software crash! But after an hour, the login completes and networked computing is born with applications like email and file transfer!
The original Interface Message Processor at UCLA housed network pioneering history! Image source
ARPANET grew steadily from there through the 1970s incorporating more university campuses and defense research sites across all time zones in the US. Long distance radio and satellite links stitched networks across broader geographic regions. With packet switching, the network easily scaled beyond the initial four nodes without reliance on a single central hub and could sustain partial outages.
Growth of ARPANET Sites:
| Year | ARPANET Hosts | Description |
|---|---|---|
| 1969 | 4 | UCLA, SRI, UCSB, Utah |
| 1971 | 15 | Added Harvard, MIT, RAND Corp etc |
| 1972 | 37 | Added international sites like London‘s NPL |
| 1977 | 111 | Most major institutions connected |
| 1983 | ~1000 | IPv4 protocol rollout, transition to TCP/IP |
Table: Expansion timeline of early ARPANET sites through 1983 before the TCP/IP transition
Building Internetworking Gateways: Enter TCP/IP!
Through the 1970s, numerous other networks like SATNET and PRNET were built using custom protocols tied to specific applications. Each network was its own siloed entity much like intranets today. The original host-to-host NCP protocol was limited to the growing ARPANET itself.
Fragmented protocol landscapes before internetworking unified connectivity. Image source
To address this, visionary Bob Kahn at DARPA aimed to develop universal translation gateways enabling heterogeneous networks to exchange data. Teaming up with UCLA Professor Leonard Kleinrock and others, he developed the Transmission Control Protocol (TCP) in 1973. This broke messages into ordered data packets reliably delivered between two systems. When combined with the Internet Protocol (IP) to universally address and route between networks, TCP/IP became the answer!
TCP/IP saw rapid adoption beyond ARPANET in the US defense network. By 1983, ARPANET itself switched over completing its transition into just another part of what was now the early internet. The modern global network starts taking shape!
The internet topology in 1983 after TCP/IP adoption starts unifying previously disparate networks Image source
The Dawn of Global Interconnectivity
By the 1980s, TCP/IP saw widespread adoption including on a rapidly expanding ARPANET. But accessing and sharing information was still cumbersome. This was about to dramatically change with the 1989 arrival of Tim Berners-Lee‘s World Wide Web protocol!
Tim Berners-Lee Invents the World Wide Web
At CERN in 1980, Tim Berners-Lee wrote his first computer program on an RC666 and then spearheaded transitions from mainframe terminals to desktop systems in the years after. By 1984 he built a system called Enquire to organize personal notes using associations much like hypertext links. In 1989, he circulated a proposal for improving remote document access at CERN.
Berners-Lee‘s breakthrough concept extended hypertext links used in stand-alone help systems and earlier experimental networks like Xanadu. He proposed using HyperText Markup Language (HTML) as a portable document format for encoding links. HTTP handled transferring HTML over networks. By Christmas 1990, he had built the first web browser and HTML editor so colleagues could access his new system!
The NeXT workstation used as first web server at CERN! Image source
The ease of linking related documents revolutionized information access over the internet. Berners-Lee helped standardize HTML & HTTP through the IETF and W3C while CERN committed the technologies to royalty-free open access. This kickstarted rapid adoption first across academic and technology circles which soon started expanding exponentially as personal computers became affordable!
Mosaic – The First Graphical Browser
After Berners-Lee‘s text-only browser, user experience improvements were still needed for more intuitive access. A team led by Marc Andreessen at National Center for Supercomputing Applications (NCSA) took the next leap by developing Mosaic in 1993. It introduced embedded image support along with point-and-click navigation. Their compiler also enabled easy porting across Microsoft Windows, Macintosh and UNIX workstations.
Mosaic provided the first accessible graphical web interface! Image source
This made the web far more comfortable to use over text-heavy interfaces. What started as a simple research network now gained mainstream traction beyond scientists or computer experts! Mosaic kickstarted the first browser wars with commercialized successors like Netscape Navigator dominating mid-1990s market share. The stage was now set for ordinary households to soon enter the internet age!
The Early Days of the Public Internet
The remarkable growth of the 1990s public internet pivoted on a key development – the 1993 decision allowing commercial ISP access ultimately leading to today‘s competitive landscape spanning Google Fiber, Comcast, Vodafone, Reliance Jio and countless local providers worldwide.
Privatization and the Dot Com Boom
From 1987 through early 1993, the NSFNET served as the US internet backbone connecting academic sites under a government grant. But seeking to open access beyond academia, the Clinton administration controversially opened it to private ISPs. This fueled rapid growth of companies like AOL, Prodigy, Netcom and UUNet selling dial-up subscriptions en masse. The Mosaic browser and Netscape Navigator brought friendly graphical access shortly after.
Global Internet Access Growth:
| Year | Hosts | Domains | Users |
|---|---|---|---|
| 1993 | 1.3 M | 2,000 | ~5M worldwide |
| 1996 | 16 M | 290,000 | ~40M / ~1% US penetration |
| 2000 | 72 M | 30 M | ~400M / ~15% US penetration |
Table: Explosive 1990s early web growth especially post-privatization
With pent-up mainstream interest around this novel communications platform combined with the ubiquity of personal computing and cheap modems, a consumer market was born!
Dotcom mania saw an explosion of speculative investment flooding ambitious internet entrepreneurs. While unfettered speculation led to a traumatic 2001 crash, the era produced category-defining businesses showcasing the commercial potential:
- eBay revolutionized person-to-person exchange of goods
- Amazon successfully recreated mail-order business at internet scale
- PayPal built online payments processing from the ground up
Topics awaiting expansion:
- Web 2.0 concepts fueling interactive platforms and social media
- Cloud computing reshaping internet infrastructure
- The smartphone and apps revolution
- Current trajectory predictions on decentralization & Web 3.0
