Hitting the Books: What life on the internet was like at 300 bits per second

As distressing a prospect it may sound, our world did exist before social media. Those were some interesting times with nary a poorly lit portion of Cheesecake Factory fare to critique, exactly zero epic fails to laugh at and not one adorable paw bean available for ogling. There weren't even daily main characters! We lived as low-bandwidth savages, huddled around the soft glow of CRT monitors and our cackling, crackling signal modulators, blissfully unaware of the societal upheaval this newfangled internet would bring about.

In his new book, The Modem World: A Prehistory of Social Media, author and Assistant Professor in the Department of Media Studies at the University of Virginia, Kevin Driscoll examines the halcyon days of the early internet — before even AOL Online — when BBS was king, WiFi wasn't even yet a notion, and the speed of electronic thought topped out at 300 baud.

Excerpted from The Modem World: A Prehistory of Social Media by Kevin Driscoll. Published by Yale University Press. Copyright © 2022 by Kevin Driscoll. All rights reserved.

Early on, the heartbeat of the modem world pulsed at a steady 300 bits per second. Streams of binary digits flowed through the telephone network in 7- and 8-bit chunks, or “bytes,” and each byte corresponded to a single character of text. The typical home computer, hooked up to a fuzzy CRT monitor, could display only about a thousand characters at once, organized into forty columns and twenty-four rows. At 300 bits per second, or 300 “baud,” filling the entire screen took approximately thirty seconds. The text appeared faster than if someone were typing in real time, but it was hardly instantaneous.

In the late 1970s, the speed at which data moved through dial-up networks followed a specification published by Ma Bell nearly two decades before. Created in the early 1960s, the AT&T Data-Phone system introduced a reliable technique for two-way, machine-to-machine communication over consumer-grade telephone lines. Although Data-Phone was initially sold to large firms to facilitate communication between various offices and a single data-processing center, it soon became a de facto standard for commercial time-sharing services, online databases, and amateur telecom projects. In 1976, Lee Felsenstein of the People’s Computer Company designed a DIY modem kit offering compatibility with the AT&T system for under $100. And as newer tech firms like Hayes Microcomputer Products in Atlanta and US Robotics in Chicago began to sell modems for the home computer market, they assured consumers of their compatibility with the “Bell 103” standard. Rather than compete on speed, these companies sold hobbyist consumers on “smart” features like auto-answer, auto-dial, and programmable “remote control” modes. A 1980 ad for the US Robotics Phone Link Acoustic Modem emphasized its warranty, diagnostic features, and high-end aesthetics: “Sleek... Quiet... Reliable.”

To survive, early PC modem makers had to sell more than modems.

They had to sell the value of getting online at all. Today, networking is central to the experience of personal computing — can you imagine a laptop without WiFi? — but in the late 1970s, computer owners did not yet see their machines as communication devices. Against this conventional view, upstart modem makers pitched their products as gateways to a fundamentally different form of computing. Like the home computer itself, modems were sold as transformative technologies, consumer electronics with the potential to change your life. Novation, the first mover in this rhetorical game, promised that its iconic black modem, the Cat, would “tie you into the world.” Hayes soon adopted similar language, describing the Micromodem II as a boundary-breaking technology that would “open your Apple II to the outside world.” Never mind that these “worlds” did not yet exist in 1979. Modem marketing conjured a desirable vision of the near future, specially crafted for computer enthusiasts. Instead of driving to an office park or riding the train, modem owners would be the first truly autonomous information workers: telecommuting to meetings, dialing into remote databases, and swapping files with other “computer people” around the globe. According to Novation, the potential uses for a modem like the Cat were “endless.”

In practice, 300 bits per second did not seem slow. In fact, the range of online services available to microcomputer owners in 1980 was rather astonishing, given their tiny numbers. A Bell-compatible modem like the Pennywhistle or Novation Cat offered access to searchable databases such as Dialog and Dow Jones, as well as communication services like CompuServe and The Source. Despite the hype, microcomputers alone could sometimes seem underwhelming to a public primed by visions of all-powerful, superhuman “world brains.” Yet, as one Byte contributor recounted, the experience of using an online “information retrieval” service felt like consulting an electronic oracle. The oracle accepted queries on virtually any topic — “from aardvarks to zymurgy” — and the answers seemed instantaneous. “What’s your time worth?” asked another Byte writer, comparing the breadth and speed of an online database to a “well- stocked public library.” Furthermore, exploring electronic databases was fun. A representative for Dialog likened searching its system to going on an “adventure” and joked that it was “much less frustrating” than the computer game of the same name. Indeed, many early modem owners came to believe that online information retrieval would be the killer app propelling computer ownership into the mainstream.

Yet it was not access to other machines but access to other people that ultimately drove the adoption of telephone modems among micro- computer owners. Just as email sustained a feeling of community among ARPANET researchers and time-sharing brought thousands of Minnesota teachers and students into collaboration, dial-up modems helped to catalyze a growing network of microcomputer enthusiasts. Whereas users of time-sharing networks tended to access a central computer through a “dumb” terminal, users of microcomputer networks were of- ten themselves typing on a microcomputer. In other words, there was a symmetry between the users and hosts of microcomputer networks. The same apparatus — a microcomputer and modem — used to dial into a BBS could be repurposed to host one. Microcomputers were more expensive than simple terminals, but they were much cheaper than the minicomputers deployed in contemporary time-sharing environments.

Like many fans and enthusiasts, computer hobbyists were eager to connect with others who shared their passion for hands-on technology. News and information about telephone networking spread through the preexisting network of regional computer clubs, fairs, newsletters, and magazines. At the outset of 1979, a first wave of modem owners was meeting on bulletin board systems like CBBS in Chicago and ABBS in San Diego to talk about their hobby. In a 1981 article for InfoWorld, Craig Vaughan, creator of ABBS, characterized these early years as an awakening: “Suddenly, everyone was talking about modems, what they had read on such and such a bulletin board, or which of the alternatives to Ma Bell... was most reliable for long-distance data communication.” By 1982, hundreds of BBSs were operating throughout North America, and the topics of discussion were growing beyond the computing hobby itself. Comparing the participatory culture of BBSs to amateur radio, Vaughan argued that modems transformed the computer from a business tool to a medium for personal expression. Sluggish connection speeds did not slow the spread of the modem world.

True to the original metaphor of the “computerized bulletin board,” all early BBSs provided two core functions: read old messages or post a new message. At this protean stage, the distinction between “files” and “messages” could be rather fuzzy. In a 1983 how-to book for BBS software developers, Lary Myers described three types of files accessible to users: messages, bulletins, and downloads. While all three were stored and transmitted as sequences of ASCII characters, Myers distinguished “the message file” as the defining feature of the BBS. Available day and night, the message file provided an “electronic corkboard” to the community of callers: a place to post announcements, queries, or comments “for the good of all.” Myers’s example routine, written in BASIC, identified each message by a unique number and stored all of the messages on the system in a single random-access file. A comment in Myers’s code suggested that eighty messages would be a reasonable maximum for systems running on a TRS-80. A caller to such a system requested messages by typing numbers on their keyboard, and the system retrieved the corresponding sequence of characters from the message file. New messages were appended to the end of the message file, and when the maximum number of messages was reached, the system simply wrote over the old ones. Like flyers on a corkboard, messages on a BBS were not expected to stay up forever.