Before the Web: What People Actually Did on the Early Internet

A network unlike a webpage

Speaking of the “old Internet” often brings modems, early browsers, and text-heavy screens to mind. That image contains some truth, but it arrives too late if it is applied to the whole history of the network. Before the World Wide Web became a recognisable way to find and move through information, the Internet was above all a working infrastructure: a means of connecting computers and enabling people in different places to use remote resources.

Internet Society’s account places the rapid addition of computers to ARPANET, and work on functional host-to-host protocols and network software, in the years after 1970. That wording matters. It does not yet describe a large public display of content; it describes machines that had to communicate with one another. The associated everyday experience was not “opening the web,” but logging in remotely, moving files, and sending messages.

ARPANET and the task of linking machines

ARPANET is central to accounts of the early Internet because it served as research infrastructure and because technical cooperation formed around it. Internet Society says that participants made extensive use of available means to coordinate their efforts: email, file sharing, remote access, and, later, World Wide Web functions. The sequence helps prevent a backward reading of history. The web appears as a later layer within a wider trajectory, not as an automatic synonym for every computer network.

The Network Working Group, noted by Internet Society in December 1970 under the leadership of S. Crocker, points to another crucial feature: the network was also a collective practice of technical definition. Protocols and software were not invisible scenery; they made it possible for different systems to exchange data. Remembering the early Internet therefore requires attention to operations that may seem less visible today than a page, app, or platform.

Logging in, copying, writing

Remote login and file transfer are listed by Internet Society among the important applications of that period. From today’s perspective, they may sound administrative. Yet they describe a concrete shift: one computer could be used as a resource from somewhere else. A network’s value did not lie only in owning a machine, but in reaching a connected machine, using its capabilities, or retrieving material stored on it.

File transfer made the network a channel through which documents and software could circulate. Remote access allowed work on systems that were not physically in front of the user. Both practices call for an environment that was more technical and less visual than today’s: host names, commands, accounts, and specific procedures belonged to the experience. It was not necessarily simple or uniform, but it opened an operational form of distance that would shape network culture.

There were technical limits as well. Internet Society explains that the model worked well for remote login and file transfer, but that early work on advanced applications—especially packet voice in the 1970s—showed that packet loss could not always be corrected by TCP and had to be handled by the application. This is more than a minor detail: network capabilities depend on compromises among protocols, programs, and uses.

Email as an everyday innovation

Among these tools, Internet Society describes email as probably the innovation with the greatest impact. The caution in “probably” is useful: historical impact is difficult to measure, and a timeline alone is not an explanation of culture. Still, email brings together several changes: communication that need not be simultaneous, direct exchange among network participants, and the ability to coordinate work without being in the same place or acting at the same time.

There is no need to assign every later digital habit to email in order to recognise its importance. In Internet Society’s description, it appears beside file sharing and remote access as a heavily used means of coordinating effort. That concrete role is enough to place it historically: before social networks and mass publishing platforms, an essential part of connected life consisted of writing, receiving, and answering messages within technical and research communities.

When information began to overflow the network

Growth did not merely multiply connections; it also made resources harder to locate. The summary cited by Wikipedia says that as early ARPANET expanded, host names were distributed through a HOSTS.TXT file from SRI International, a method that became less practical. The same source identifies the Domain Name System as a technical answer associated with Paul Mockapetris. It should be used as orientation rather than as a replacement for primary technical documentation, but the problem is clear: a larger network needs scalable reference systems.

During the 1980s and early 1990s, the need to find and organise files and information became more evident. Wikipedia names Gopher, WAIS, and the FTP Archive list as projects intended to organise distributed data. Their historical interest is not that they form an inevitable march toward the web, but that they show several responses to one difficulty. Finding information was a problem before contemporary search engines, and the solutions coexisted, changed, and left uneven traces.

An open history, not an album of milestones

CCCB Lab warns that an Internet chronology is built from familiar milestones but also exposes large gaps in research and reflection. The warning is especially valuable when discussing the “old Internet.” Linear stories can turn complex infrastructures into a neat sequence of inevitable inventions. They can also diminish the people who maintained systems, wrote standards, administered machines, or adapted networks to local settings.

The Digital Archive of the Polytechnic University of Madrid presents work on the beginnings of the Internet in Spain, completed with contributions from pioneers involved in network deployments, bibliography, and personal reflection. Its own description shows that network history is not exhausted by a US account of ARPANET. Infrastructures are deployed, interpreted, and remembered in specific places; studying them requires distinguishing testimony, bibliography, and an author’s conclusions.

The legacy of a function-oriented Internet

The gap between that environment and current use is obvious, yet some functions remain. We still send messages, access distant resources, transfer data, and need ways to locate information. What changed was scale, interface, device diversity, and the cultural weight of these uses. The web made many network operations visible to far wider audiences, but it did not create all of them from nothing.

The strongest way to remember the early Internet is neither to turn it into a golden age nor to reduce it to an awkward prehistory of the present. It was a set of developing infrastructures and practices, with real technical limits and with solutions that did not always survive. Before the web, connecting mostly meant doing something with another computer: logging in, copying, writing, coordinating, and searching. That functional history explains why the Internet was never only a collection of pages.