People who built the network
Researchers, engineers, operators, and community leaders whose choices shaped the internet's open architecture.
BTW Media Editorial Collection
The internet was not born as a finished system. It was built across decades by researchers, engineers, operators, institutions, and communities who chose openness, interoperability, and shared stewardship before the world knew what the network would become.
History of the Internet is BTW Media's editorial archive of the people, protocols, institutions, and turning points that shaped the global network.
Researchers, engineers, operators, and community leaders whose choices shaped the internet's open architecture.
TCP/IP, DNS, routing, email, the Web, number resources, and standards processes made global interoperability possible.
Interviews, profiles, and records preserve how the internet was imagined, governed, expanded, and defended.
This collection brings together interviews, profiles, source-backed reporting, and historical analysis about the builders and witnesses of the internet age.
This page follows how individual decisions, technical principles, public institutions, and regional communities became civilization-scale infrastructure.
For the people recorded here, this is a record of contribution. For readers, it is a map of how the internet became one of humanity's shared systems.
Series
From packet-switching theory to the World Wide Web, this map connects the people, protocol decisions, and governance transitions that shaped the global network.

1961
Leonard Kleinrock (1961): Published the earliest packet-switching and information-flow work that provided mathematical foundations for breaking messages into packets.

1962 / 1963
J.C.R. Licklider (1962 / 1963): Articulated the "Intergalactic Computer Network" vision that inspired ARPA's networking goals and the idea of a globally-connected computing commons.

1964
Paul Baran (1964): RAND memorandum On Distributed Communications laid out distributed network designs intended to be survivable — a key conceptual precursor to later packet networks.

1965
Donald Davies (1965): Coined the term "packet" and advanced practical packet-switching architecture at the UK National Physical Laboratory.

1968
Douglas Engelbart (1968): Demonstrated the oN-Line System (NLS) at the "Mother of All Demos" — showing hypertext, shared editing, the mouse and other interactive concepts that shaped future online tools.

1969
Lawrence (Larry) Roberts (1969): As ARPA program manager he designed and led the ARPANET construction, turning packet-switching proposals into the first operational research network.

1969
Jon Postel (1969): Became central to the RFC series and began his long stewardship of protocol documentation and numbering — the Internet's numbering authority.

1969
Steve Crocker (1969): Instigated the ARPANET "Network Working Group" and created the Request for Comments (RFC) series — the primary open documentation mechanism for Internet protocols.

1972–1973
Vint Cerf & Bob Kahn (1972–1973): Co-designed the Transmission Control Protocol (TCP) and later TCP/IP, enabling communication across multiple networks — the birth of internetworking.

1973
Robert Metcalfe (1973): Invented Ethernet at Xerox PARC, making local networking fast and scalable for office and campus environments.

1973–1975
Louis Pouzin (1973–1975): Developed CYCLADES in France, a datagram-based packet network that directly influenced TCP/IP's design.

1973
Danny Cohen (1973): Pioneered early packet voice transmission and endianness concepts, key to interoperability in TCP/IP systems.

1975
Yngvar G. Lundh (1975): Promoted and implemented early packet networking in Scandinavia, connecting European research networks.

1976
Elizabeth "Jake" Feinler (1976): Managed the ARPANET Network Information Center (NIC) at SRI, maintaining the official host name directory before DNS.

1977
Pål Spilling (1977): Helped establish early ARPANET links in Europe and co-authored the first Internet (TCP/IP) experiments between continents.

1977
Bob Kahn, Vint Cerf & Jon Postel (1977): Conducted the first three-network TCP/IP test (ARPANET, SATNET, PRNET), proving the architecture's viability.

1978
Steve Crocker, Jon Postel & Joyce Reynolds (1978): Led early IETF-like protocol coordination and RFC series continuity — the core of today's open standards process.

1979
Radia Perlman (1979): Developed early routing algorithms that evolved into the Spanning Tree Protocol (finalized 1985).

1981–1983
Paul Mockapetris (1981–1983): Invented the Domain Name System (DNS) to replace the central HOSTS.TXT file. Introduced hierarchical, distributed naming via RFC 882/883.

1981
Danny Cohen (1981): Published "On Holy Wars and a Plea for Peace," formalizing the concept of byte ordering ("endianness").

1982
Daniel Karrenberg (1982): Helped build EUnet, the first pan-European Internet Service Provider (ISP) and in 1989 was one of the founders of RIPE.

1982–1983
Vint Cerf, Bob Kahn, Jon Postel & Steve Crocker (1982–1983): Oversaw the January 1, 1983 "flag day" switchover from NCP to TCP/IP — the operational birth of the modern Internet.

1983–1984
Radia Perlman (1983–1984): Developed the Spanning Tree Protocol (STP), which made Ethernet networks reliable and loop-free.

1984
Mike Muuss (1984): Created ping, one of the simplest and most enduring diagnostic tools in networking.

1984–1986
Eric Allman (1984–1986): Wrote Sendmail, the first general-purpose Internet email transport system, crucial for global email interoperability.

1985
Brian Carpenter (1985): Led development of key Internet standards and chaired IETF working groups, promoting IPv6 and architectural consistency.

1986
John Klensin (1986): Significant early work on SMTP, FTP, and internationalization of Internet protocols.

1987
Joyce K. Reynolds & Bob Braden (1987): Contributed to RFC editing, TCP/IP documentation, and Internet research group coordination.

1989
Alan Emtage (1989): Developed Archie, the first Internet search engine, enabling users to find files across FTP servers.

1990 / 1991
Tim Berners-Lee (1990 / 1991): Invented the World Wide Web at CERN — creating HTTP, HTML, and the first web server and browser, converting network infrastructure into a global publishing and information system.

1991
Linus Torvalds (1991): Created the Linux kernel, which became a cornerstone of Internet infrastructure, web servers, and open-source computing.

1991
Jianping Wu (1991): Led the development of CERNET, China's first academic Internet network, linking universities across the country.

1991
Scott Bradner (1991): Key figure in IETF standardization and Internet governance; helped define early Internet policy and protocol stewardship.

1992
Kilnam Chon (1992): Built KREONET, the Korean Research Network, pioneering Internet connectivity in Asia.

1993
Marc Andreessen (1993): Co-developed Mosaic, the first mainstream graphical web browser, which popularized the Web and inspired Netscape.

1993
Robert Cailliau (1993): Collaborated with Berners-Lee at CERN to promote and refine the World Wide Web and its first browser/server setup.

1994
Mitchell Baker (1994): Led the Mozilla project, defending open web standards and community-driven browser development.

1994
Xing Li (1994): Architect of CERNET, China's first academic network (1994), and early IPv6 advocate for Asia-Pacific.

1995
Paul Vixie (1995): Principal developer of BIND, the most widely used DNS server software; improved DNS reliability and security.

1996
Larry Landweber (1996): Created CSNET and helped connect over 25 countries to the early Internet, bridging academia and policy.

1997
Radia Perlman (1997): Authored Interconnections, formalizing network protocol design principles and influencing Internet routing education.

1998
Jon Postel (1998): Demonstrated control over the DNS root servers and led the IANA transition, symbolizing his lifelong role.

1998
Vint Cerf & Bob Kahn (1998): Helped found ICANN and promoted the institutional frameworks for managing Internet naming and numbering globally.

1999
Al Gore (1999): Sponsored the 1991 High Performance Computing and Communications Act, which funded Internet expansion — earning the nickname "the Information Superhighway."
Internet Milestones
A decade-by-decade guide to the experiments, protocols, standards, networks, and institutions that turned internetworking into shared global infrastructure.
1962-1965
Packet switching becomes a practical networking idea
Paul Baran and Donald Davies separately formalize packet-switched communication, giving later research networks a way to move data without relying on circuit-switched telephony.
Paul Baran, Donald Davies
RAND, National Physical Laboratory
1969
ARPANET sends its first message
The first host-to-host ARPANET message travels between UCLA and SRI, turning packet networking from research architecture into an operating experiment.
Leonard Kleinrock, Charley Kline, Bill Duvall
UCLA, SRI, ARPA
1971
Email becomes the first durable social application of the network
Ray Tomlinson adapts network mail between machines, making person-to-person communication one of the internet's earliest reasons to exist.
Ray Tomlinson
BBN, ARPANET
1974
TCP paper defines internetworking as an open architecture
Vint Cerf and Bob Kahn publish the TCP paper and give the network a protocol model for connecting heterogeneous networks into an internet.
Vint Cerf, Bob Kahn
DARPA
1977
Three-network TCP/IP test proves internetworking across media
Packet radio, satellite, and ARPANET paths carry traffic through a common protocol design, showing that an internet can survive different physical networks.
Vint Cerf, Jon Postel, Bob Kahn
DARPA research community
1983
TCP/IP flag day makes the modern internet operational
ARPANET hosts move from NCP to TCP/IP on January 1, 1983, making the protocol suite an operational foundation rather than a research proposal.
Vint Cerf, Bob Kahn, Jon Postel, Steve Crocker
DARPA, ARPANET community
1983-1984
DNS turns names into distributed internet infrastructure
Paul Mockapetris and the early naming community replace host tables with the Domain Name System, giving the expanding network a scalable naming layer.
Paul Mockapetris, Jon Postel
USC/ISI, IANA
1986
NSFNET expands academic networking into a backbone
The NSFNET programme links supercomputing centres and research institutions, creating a backbone that helps move the internet beyond its original defence research context.
National Science Foundation, regional academic networks
Late 1980s
Routing growth forces inter-domain coordination
As networks multiply, operators move toward scalable inter-domain routing practices that later make BGP and routing security central to internet operations.
Network operators and routing researchers
IETF, operator community
1989 / 1990
Archie makes distributed network resources discoverable
Alan Emtage creates Archie, showing that the internet also needs discovery systems, not only transport and addressing.
Alan Emtage
McGill University
1989 / 1991
The World Wide Web opens the internet to publishing and navigation
Tim Berners-Lee proposes and implements HTTP, HTML, URLs, and early browser/server software, turning the network into a global information space.
Tim Berners-Lee, Robert Cailliau
CERN
1992
Regional number coordination takes institutional form
Regional internet growth creates the need for durable address allocation institutions, with APNIC emerging in the Asia-Pacific coordination context.
Regional operators and registry builders
APNIC, RIR community
1993
El Salvador joins the global internet story
Local builders connect national academic and technical communities into the global network, showing how internet history is made outside the original centres of research funding.
Lito Ibarra
El Salvador internet community
1998
ICANN is formed to coordinate global identifiers
ICANN becomes the institutional home for DNS and identifier coordination debates, moving internet governance into a more formal multi-stakeholder structure.
Jon Postel, Vint Cerf, global governance community
ICANN, IANA community
1998
IPv6 is specified in RFC 2460
The IETF publishes RFC 2460, preserving the internet's ability to scale addressing beyond IPv4 exhaustion and making protocol stewardship a generational task.
Brian Carpenter, IETF IPv6 community
IETF
1986 / 1992
Latin American academic networks become regional infrastructure
Builders in Chile, Venezuela, Argentina, and neighbouring communities turn research connectivity into lasting regional internet capacity.
Florencio Utreras, Ermanno Pietrosemoli, Olga Cavalli
Latin American academic and governance communities
2000s
African technical communities expand capacity and participation
Training, software communities, and advocacy broaden who can build, operate, and govern internet infrastructure.
Dorcas Muthoni
African technology and internet communities
2000s
Operational analysis becomes part of internet stewardship
Routing, addressing, measurement, and security analysis become essential to explaining where the global network is resilient and where it is fragile.
Geoff Huston
APNIC Labs, operator community
2016
IANA stewardship transition completes
The transition marks a major governance shift from US government oversight toward the global multi-stakeholder community.
ICANN, IANA community, global internet governance community
2010s / 2020s
Routing security becomes a core public-interest issue
RPKI, MANRS, and operator coordination make routing security part of the internet's institutional memory and daily operational discipline.
Routing security researchers and operators
IETF, RIRs, operator community
2024
Internet history becomes a source-backed archive
Measurement archives, oral histories, and editorial collections make internet history itself a field of stewardship for the next generation.
Jim Cowie and internet history researchers
Internet History Initiative
Published Records
Published interviews, profiles, and historical reporting from BTW Media's History of the Internet collection.

The Website Moved Its Filing Cabinet. It Sent the Moving Bill to Every Link.
A publisher renovates its site. The document is still there, the staff still believe in it, and the new menu looks splendid. The old address now leads nowhere. In a 1998 essay, Tim Berners-Lee treated this familiar renovation as a choice about who does the work. A site can move files, replace software and change departments behind a stable public URI. When it changes the URI instead, every reader, bookmark and distant publisher must rediscover the document. The office got its fresh floor plan; the rest of the Web inherited its missing address.
Read Record
The public WebCrawler in 1994 answered questions from a shared index. Brian Pinkerton had built another mode: start with likely results, send a robot down their unexplored links and add whatever it found before answering. That sounds like the perfect cure for a search index with holes. Pinkerton kept it out of public hands. A shared crawl could serve thousands of queries; a fresh crawl for every curious user would ask the young Web to pay the network bill again and again.
Read Record
In April 2009, OpenStreetMap contributors could not save new edits through the project API while the service moved to version 0.6. The online editor was disabled. When editing returned, every change had to belong to a changeset, and clients had to check which version of a map element they were replacing. A community map paid for a better account of its own making with a conspicuous pause in the act of making it.
Read Record
A blog author could write a reply on her own site and still want it seen beside the article she answered. The old choices made her pick a house: put the comment on someone else's page, or keep the reply at home and hope readers followed the link. Webmention gave the link a messenger. In its 2017 W3C Recommendation, the sending site tells the receiving site where the reply lives. The receiver checks the link and decides what appears on its page. The messenger crossed the street; the front door stayed with the owner.
Read Record
The Site Was Down. Google Had a Copy. Then the Link Vanished.
For years, a Google result could offer two doors to one page: the publisher's site and a snapshot stored by Google. If the site failed, the snapshot might still open. The publisher could ask Google to remove that copy. In 2024 Google removed the reader-facing cached link itself. A little secondary doorway had exposed a long argument about who controlled a page after it left its own server.
Read Record
In 2006, some embedded content in Internet Explorer appeared on the page but waited for a manual activation before a reader could use it. Microsoft connected that change to its Eolas patent case. A legal and engineering decision had acquired a user interface: click, or use the keyboard, and only then interact. Two years later another Microsoft update removed the extra step.
Read Record
On 16 December 2002, Creative Commons launched licenses that could speak to three audiences: a human reader, a lawyer and a machine. The machine was supposed to find works offered for reuse without making every person negotiate permission from scratch. The first recommended web markup carried RDF/XML inside an HTML comment. Six years later, Creative Commons' own ccREL paper called that choice cumbersome and replaced the recommendation. The legal permission was only one part of the work. A machine also had to find the label attached to the right object.
Read Record
The Web SQL Database draft offered web apps a local database and a familiar query language. By 2010, the W3C had two implementations to point at. Both depended on SQLite. That made the standardization test almost comic: two browsers could agree because they were asking the same engine, while the rule they were supposed to implement independently remained unwritten.
Read Record
In 2003, the case for CSS had a sales problem. A stylesheet could separate a page's appearance from its markup, but clients and designers still saw late-1990s table layouts as the practical route to an interesting site. Dave Shea offered a sharper argument: everyone could design the same page, provided nobody touched its HTML. CSS Zen Garden turned a standard's promise into a public contest of visible results. The restriction gave designers room to show their craft, and gave the keeper of the common page control over the field they had to play on.
Read Record
Mozilla Persona promised websites a login without collecting another password. It deliberately gave them the user's email address, a useful account key. When Mozilla retired its hosted service in 2016, that key remained on the website. The machinery that proved the visitor owned it had an end date.
Read Record
In late 1996, Daniel Stenberg wanted a bot in an IRC channel to answer currency questions. The rates were already on the web. The missing piece was a small program that could fetch them without a person opening a browser. He found Rafael Sagula's HttpGet and started changing it. Four years later, that narrow errand had become curl and then libcurl: a transfer engine other programs could call. The names changed as the job outgrew each one.
Read Record
In September 2009, Google prepared more than 100,000 invitations to Wave, its live conversation and document editor. The same announcement admitted a host could not yet remove a participant or configure editing permissions. The invitation was a finished product. The exit was still on the work list.
Read Record
The `scp` Command Kept Its Name. OpenSSH Replaced the Machinery.
For years, `scp` looked like the simplest way to move a file across an SSH connection. The command was familiar; the old SCP/RCP transfer protocol beneath it asked a remote shell to expand wildcards and made some filenames a quoting puzzle. In April 2022, OpenSSH 9.0 kept the command and changed its default engine to SFTP. The same parcel counter opened onto a different conveyor. Old servers and old scripts made the switch less tidy than a release-note headline.
Read Record
Gears gave a web app its own local database and kept stories available when the network went away. Google Reader was the launch example. Three years later, the Reader team removed offline access while Gears was still alive. A year after that, Chrome removed Gears itself. The pieces of the technology moved into browsers; the reader's feature did not travel with them.
Read Record
Git's object names look like labels pasted onto a filing cabinet. They are closer to the bolts holding the cabinet together. A tree names its contents; a commit names a tree and its parents. Change the hash used for those names and the references inside the objects change too. Git chose SHA-256 in a 2018 transition design, but its 2026 manual still says repositories using the two object formats cannot interoperate. The cryptographic decision was the short meeting. Moving a connected history was the long job.
Read Record
The FCC Gave AOL Three Exits. The One Without Interoperability Opened in 2003.
A merger condition put a gate between AOL Time Warner and a particular kind of video messaging. The company could build an interoperable presence system, sign agreements to connect rival systems, or persuade regulators that the market had changed. It chose the third route. The bridge between messaging networks stayed on the drawing board while the gate came down.
Read Record
In 1996, Andrew Tridgell and Paul Mackerras described a file update problem with two machines and a slow link. One held the new file; the other held an older, similar copy. Sending the new file whole was the obvious move and often the expensive one. Their algorithm made the receiving machine send a compact map of the blocks it already owned. The sender then shipped references to those blocks and only the missing bytes. The bandwidth bill shrank; the computers took on the work.
Read Record
A buyer looking for one concert did not need to tour a ticket seller’s home page first. Tickets.com could show the event, then send the buyer straight to Ticketmaster’s event page when Ticketmaster held the sale. Ticketmaster took that shortcut to court. Its filings trace four years of attempts to make the route itself part of its control over ticketing.
Read Record
In May 2019, the leaders of OpenTracing and OpenCensus agreed to join their instrumentation projects as OpenTelemetry. They called the existence of two similar projects the biggest problem. They also wanted the old code in read-only mode before year-end. By May 2023, OpenTelemetry was still publishing a specific OpenCensus exit plan, complete with bridge libraries, a July archive date and an exception for Python. A merger can settle where new work belongs. It does not pull old libraries out of running services.
Read Record
On 25 January 2002, English Wikipedia switched from files to a PHP and MySQL system built for the project. The encyclopedia kept its current pages. The converter carried each page’s latest version, leaving much of its earlier editing trail for a different job. Eight years later, an old backup would reveal that some of the trail had been waiting in logs all along.
Read Record