People have been leaving public messages since the first artists painted hunting scenes on cave walls. But it was the invention of electricity that forever changed the way we talked to each other. In 1844, the first message was sent via telegraph. Samuel Morse, who created the binary Morse Code decades before electronic computers were even possible, tapped out, “What hath God wrought?” It was a prophetic first post.
World War II accelerated the invention of digital computers, but they were primarily single-use machines, designed to calculate artillery firing tables or solve scientific problems. As computers got more powerful, the idea of time-sharing became attractive. Computers were expensive, and they spent most of their time idle, waiting for a user to enter keystrokes at a terminal. Time-sharing allowed many people to interact with a single computer at the same time.
Part 0: The Precambrian era of digital communication (1969–1979)
Soon after time-sharing was invented, people started sending messages to other users. But since every computer spoke its own unique machine language and had its own way of storing and retrieving data, none of these machines could talk to each other. The solution to this problem came out of the Pentagon’s Advanced Research Projects Agency (ARPA), and was thus dubbed the “ARPANET.” When two different computers connected to each other through an “IMP” (Interface Message Processor, the first router) in 1969, it was a massive breakthrough.
[
Now, instead of sending a message to a friend who was probably sitting next to you in the same computer lab, you could send it to someone in a different city. In 1971, Ray Tomlinson wrote the first inter-computer messaging program, SNDMSG. Because he had to differentiate between the receiver’s username and the name of the computer they were using, he needed a character that wouldn’t be part of either. He hit “SHIFT-P” on the Model 33 Teletype, got an @, and the rest was history. Email was born.
The friendly orange glow
At around the same time, a self-contained computer network called PLATO was also changing the world. PLATO was an educational system that began in 1960 and was nearing its fourth iteration. It was responsible for many computer firsts, such as the first flat-screen plasma display, which launched in 1972 with PLATO IV. These touch-enabled, 512×512 graphical displays looked like they came from the future. And while it couldn’t talk to ARPANET, every PLATO user at every terminal could communicate with each other all over the world.
In 1971, PLATO was the home of the first “phishing” scam, when student Mark Rusted created a fake login screen that stole users’ passwords. (He was politely asked not to do it again.) Because of this, the next revision of PLATO added a special keystroke combination, SHIFT-STOP, that would guarantee that the user saw a real login screen. Years later, Microsoft would use the same idea for Windows NT with CTRL-ALT-DEL.
The PLATO environment hosted one of the first public message boards, called Discuss. In 1973, Discuss was home to the world’s first online free speech controversy, which threatened the future of PLATO itself. Richard Nixon’s firing of the special prosecutor had many Americans talking about impeachment. Stuart Umpleby, a graduate student and PLATO user, posted a message that some saw as a call for action across universities to support the removal of the president.
Knowledge of this post quickly reached the Pentagon, and the director of PLATO, Don Bitzer, was contacted by the National Science Foundation (NSF). He was told that the Nixon White House was threatening to cut the NSF’s funding for both PLATO and the ARPANET. Bitzer stayed calm and said that while he hadn’t read the message thread in question, he would allow it to continue only if it was at the level of educational discourse but not if it was a political call to action. The NSF agreed, and Bitzer added a disclaimer to that effect to the opening screen of Discuss. Umpleby later posted a response:
It is hard to believe that a few comments in one program could cause such a reaction. What is also interesting, however, is that apparently on the basis of only one comment, the Pentagon understood the importance of computer-based communications media. Months and even years of talking and attempting to persuade social scientists had produced at best indifference. Such differences in reaction testify far more eloquently than a scientific article why those who are the establishment are there and why social science has been so ineffective.
Meanwhile, the ARPANET wasn’t standing still. From modest growth in its early years, by 1973, it connected over 42 computers through 34 routers. An experimental program called FORUM, created at the Palo Alto, California-based nonprofit Institute for the Future, offered a way for ARPANET users to leave public messages. Seventy-five percent of the traffic on the ARPANET was still email, but advances in email software allowed the creation of “mailing lists,” which anyone on the network could join and receive updates from anyone else.
The first of these mailing lists was created for discussion of the ARPANET itself. But other lists on more general (if nerdy) topics gained traction quickly. One of the most popular was SF-LOVERS, which lasted until 2001.
These mailing lists were the birthplace of the first flame wars, where users would harass each other with harsh words and insults. To try to prevent discussions from devolving into flame wars, Kevin MacKenzie proposed a new thing called a smiley in April 1979. He claimed that he got the idea from an article in Readers’ Digest.
At this time, electronic communication was still the domain of a privileged few. Access to the ARPANET was limited to a handful of universities and research institutions, and PLATO only ever supported a few thousand terminals at once. It would take an unprecedented leap in computer technology to bring electronic messaging to the rest of the world.
Part 1: The Jurassic era—Early personal computers grow the scene (1978–1996)
When Intel invented the microprocessor in 1971, it believed the device would primarily be useful for calculators and other embedded uses, like traffic lights. But when Ed Roberts released the Altair 8800 in 1975, it unleashed a torrent of pent-up demand for personal computers that has never abated.
In 1978, Randy Seuss and Ward Christensen took a jury-rigged clone of an Altair computer, connected it to a modem, and wrote the software that would change the world. CBBS, which stood for Chicago Bulletin Board System, was a server that anyone in the world could call up on their own computer, using their own modem, through regular phone lines.
CBBS was instantly popular and spawned dozens of imitators. Since long-distance charges applied for calls outside one’s hometown, local BBS sites bloomed in cities all over North America, Europe, and Japan. BBS systems at first delivered only text, which was fine since that’s all personal computers could offer. In later years, support for the ANSI standard added color and special characters like those found on the IBM PC and clones. But when you called a BBS, it didn’t matter what computer you had or what computer the BBS was running on. An IBM PC user could call up an Amiga-based BBS with no problems.
You called a BBS using a modem program like Procomm or my favorite DOS app, TELIX. You’d find the numbers in local computer magazines or by downloading a BBS list from the first BBS you managed to connect to. Once you added the phone number, the program would dial it using the modem, and you’d be presented with a login screen. If you were a new user, you had to enter your name, choose a password, and maybe add some optional personal information. Being young and naive, I used my real name, but most people used “handles,” fake names that sounded cool at the time. I did use a handle sometimes, though. I was “TheJaguar!” Yes, including the exclamation point.
Most BBS systems were run as a hobby out of people’s homes, so they were free to access. The disadvantage was that BBS operators, called sysops, had to pay for additional phone lines for each person who wanted to connect to the system at the same time. Popular activities that kept users dialed in included file downloads (both legal and otherwise), turn-based online games like Trade Wars 2002, and my personal favorite, the message boards.
Some sysops set up “Call-Back Verification” (CBV), which would automatically hang up and call the user back, just to make sure people weren’t spamming new accounts. The sysop of one board I liked, RAVE, did manual CBV by personally phoning everyone who signed up. Because of this, I became good friends with him, and I later became the co-sysop for RAVE BBS.
Messaging on BBS systems was fun. Newer software, like QuickBBS or Remote Access, supported “quoting,” which let you automatically copy the message you were replying to. This was helpful since messages were presented in date order, without threading. While flame wars still popped up occasionally, the supreme power of the sysop, combined with the fact that you could actually meet up with users in person since you were all in one city, kept the temperature down to a medium simmer. Things got more complicated in 1984 when Tom Jennings invented FidoNet, a clever way for BBS operators to share messages cheaply across geographic boundaries.
FidoNet gathered up messages into compressed bundles, then sent them to other FidoNet BBS “nodes” late at night when long-distance charges were lower. Suddenly, BBS users could communicate with people all over the world. This had advantages and disadvantages. What was once a fun hobby turned into unending friction between node operators since every sysop had different ideas of how things should be run. The high price of running FidoNet nodes led to a series of proposals for recouping costs, none of which anyone could agree on. This led to folks jokingly referring to the network as “Fight-o-Net.”
At its peak in 1996, FidoNet had over 39,000 nodes, with approximately eight million users worldwide. It was used in the first Gulf War, allowing soldiers to send email home before the Internet was available. FidoNet is still around today, albeit drastically smaller, but it can still be useful on occasion. For example, Vietnamese doctors have used FidoNet to share medical knowledge for HIV patients that would be censored by the government if it was sent over the Internet.
The rise of Usenet
Meanwhile, ARPANET had merged with PRNET and SATNET in 1977 to form what was increasingly being called the “Internet.” Other networks joined in the fun, like the Unix to Unix Copy Protocol (UUCP) network, which was eventually renamed the Users’ Network, or simply Usenet.
Usenet, which came alive in 1979, was a public message board divided into different “newsgroups” on various topics. The first was net.general, which was used to discuss general topics. These were later organized into hierarchies, like comp.* for computer topics, sci.* for science topics, rec.* for recreational activities, and misc.* for everything else. You could create new posts or reply to existing ones, and later, Usenet software sorted replies into a tree-like threaded hierarchy.
Usenet was structured without any central authority. Admins at each site, which were generally Unix servers at universities or corporate laboratories, would decide which newsgroups to carry, and the software would automatically dial up other servers to send every message posted to these newsgroups. The original UUCP actually showed you the complete path each message had taken, with each server separated by exclamation points (this was called a “bang path”). Later, UUCP was replaced with the Network News Transfer Protocol (NNTP), which ran over the TCP/IP protocol that the Internet standardized on.
The volume of posts was low enough that it made sense for every site to carry complete copies of every message in every group. But a loose organization of admins (the infamous “Usenet Cabal”) could still influence what new groups could get traction. Because of this, a new hierarchy—“alt.*”—was created to serve as a place for “sensitive” discussions, like alt.drugs. There were no real rules for creating new alt.* groups, so busy server admins could decide whether to carry the whole alt.* hierarchy or discard it. Most of them carried it, because it was fun, with groups like alt.conspiracy and alt.fan.warlord—the latter group devoted to showcasing the longest and silliest Usenet “signatures.”
Usenet was also the home to many communication firsts. The system was relatively open—there were a few “moderated” groups that required posts to be approved, but they were easy to bypass. In fact, the group alt.hackers forced users to do so in order to post. It wasn’t long until this lax security was abused.
On April 12, 1994, lawyers Canter and Siegel sent copies of their “Green Card Lottery” post to every single newsgroup at once. This wasn’t the first time this had happened—the practice had already been dubbed “spam” by Usenet posters, after the Monty Python skit—but it was the first commercially motivated spam. It was an ominous foreshadowing of things to come.
Usenet veterans came to dread every September, when undergraduate students would discover this amazing new communications medium and flood the newsgroups with their naive questions. But in the early 1990s, the outside world was knocking at the door. Larger multi-line BBS sites, which charged a monthly or even hourly fee to access, started adding Usenet feeds to make their service more attractive.
I signed up for one of these services, called Mind Link! While I had dabbled in FidoNet, the fact that it took at least two days to get a reply made it less enticing than local BBS messaging. But Usenet had people like me, who spent most of their time with their computer dialed into a modem. I racked up multiple $200 to $300 monthly online bills just from posting silly things on Usenet. I still used my real name, but I added my handle into my email address: [email protected]. Most of my posts make me cringe when I read them today, but I do have fond memories of posting to alt.fan.jeremy-reimer.
The “Eternal September” arrived in 1993, when American On-Line (AOL), the most popular online service in the world at the time, with 1.25 million subscribers, added Usenet access. Along with an estimated 60,000 BBSes in the US alone, with an estimated 17 million users worldwide, a lot more people were getting online. But it was a mere foreshock of what was about to come.
Part 2: The Eocene era—the rise of the World Wide Web (1993–2010)
In 1991, English computer scientist Tim Berners-Lee, working at CERN, unveiled his latest creation. It was a “universal linked information system” loosely based on Ted Nelson’s ideas about hypertext. Berners-Lee called it the World Wide Web. At first, it was a curiosity—the web server and browser both ran only on the relatively obscure and expensive NeXT computer—but in early 1993, CERN released the Web protocol, the HTML specification, and the server and browser source code for free. Later that year, NCSA launched its Mosaic web browser for Unix. Ports for Windows, Macintosh, and the Amiga came out a few months afterward. Now, anybody with a personal computer and an Internet connection could browse this new thing called the “web.”
But where could people get an Internet connection? What companies were around in 1993 that had multiple dedicated phone lines and server hardware and the appropriate modem software? Why, commercial BBSes, of course! In a very short period of time, the overwhelming demand for access caused most multi-line BBS systems to pivot to being the first Internet service providers. My local Mind Link! BBS sold for $3.5 million to a new startup, iSTAR Internet, and I suddenly had an Internet account and a free copy of this new software called Netscape Navigator.
It was an exciting time. The web was a completely green-field space where anyone could have their own URL and their own website that they hand-coded in Notepad using raw HTML. But it was new dynamic programming languages, like Perl and PHP, that made things really interesting.
Sure, Usenet still existed, and you could now browse newsgroups without hourly fees over the Internet, using advanced news reader software that ran on your computer. But the web was so much better in every way—it was so much more visually appealing and easy to use—that people immediately found ways to create message board sites that ran in a web browser.
The first of these was WIT, invented in 1994 following the first World Wide Web conference. WWWBoard was released a year later. Most early forum software used the Perl language and stored posts in individual flat files directly on the server’s hard drive (MySQL wasn’t even out yet!). This was clunky and prone to breakage, but when it worked, it was like magic.
Many other forum software programs followed, most of them open source. UBB came out in 1996, followed by Ikonboard and VBulletin in 1999. A big one was phpBB, released in 2000. This was a free, full-featured message board system that was easy to set up, used the MySQL database for storage, and supported tons of fun (if easily abused) features like user avatars and signatures.
Most online forums were run as a hobby and would focus on specific topics or fandoms. The largest phpBB board, the anime-themed Gaia Online, was founded in 2003 and peaked at over 26 million registered users. Other sites, like Ars Technica itself (founded in 1998), had their own forums in addition to the main site. The front page readers didn’t always mesh with the much smaller group of forum users, but the Ars Forums enjoyed a lively and active culture. Memorable moments and memes included The Cheeto, Leafycaust, Moonshark, Wharrgarbl, “The Lounge Says No!”, and many others.
At the time, I was going by the handle “Lord Baldrick” on the Ars forums, primarily posting nonsense about operating systems (which I still don’t know anything about, really) in the Battlefront. The intense heat from those discussions motivated me to start my own forum. I launched the site, known as “OSY,” in 2001. It was hosted at first on my desktop computer hooked up to my brand-new (and still somewhat exotic) cable modem.
My initial, naive thoughts were that complete freedom and openness were key. After all, wasn’t “sunlight the best disinfectant”? What followed was a speedrun through all the reasons moderation was necessary. My users would post new threads as fast as possible, trying to “take over the front page,” which was fun for about a minute. Spammers would post links to ads or phishing sites. New, anonymous users would post disgusting and possibly illegal things that I had to take down. But if I was too heavy-handed with my moderation, my users had no reason to stop by instead of going to one of the “serious” forums. It was a constant struggle, and for a brief time, I just let the board run without posting to it or even looking at it.
I eventually migrated OSY to a Linux server, and then a tiny AWS instance (at the time, AWS was cheaper than my electric bill for the server) so I didn’t have to babysit hardware anymore. Its popularity peaked in 2007 at over 1,400 registered users, with about 50 who were very active with more than 1,000 posts. There were at least 50 million web forum users in the United States by this time, and maybe double or triple that internationally. It was intoxicating to think that I could personally run a website that looked and felt similar (for the forum part, at least) to a giant site like Ars Technica.
But something happened to my forum, which also ended up happening to the Ars forums—and to every other web forum. It all started with a tiny device, small enough to fit inside the pocket of a pair of blue jeans, worn by a man who really liked black mock turtlenecks.
Part 3: The Holocene era—social networks take over the world (2010–2024)
Social media sites had a slow start at the beginning. The “Six Degrees” profile uploading site went almost unnoticed when it debuted in 1997, as did Friendster in 2001. MySpace, launched in 2003, had a flurry of activity and gained a large number of users, but it was kind of a flash in the pan.
Even the mighty Facebook, which started out as The Facebook in 2004, took a while to really get going. By 2007, it had grown to 20 million registered users—a respectable number, to be sure, but not even as many as Gaia Online. But when Steve Jobs announced the iPhone that year, everything changed.
Sure, it had a small screen, and it was expensive, and it didn’t have a keyboard like a Blackberry, or even cut and paste, but most of us who weren’t Steve Ballmer knew that this thing was going to take off. A coworker of mine bought the original model, and everyone made fun of him for having it “glued to his hand.” The next year, we all bought 3G models and did the same thing. Android came along at a more affordable price, and within a few years, nearly everyone in the world had a tiny Internet computer that they checked constantly.
Facebook released a mobile app in July 2008, and Twitter followed closely behind. Steve Jobs may have launched the iPhone thinking that web-based applications would be perfectly adequate, but it was the App Store that people really wanted. Apps were fast and convenient, and they could even remind you instantly when you had new messages.
Web forums, particularly phpBB, were slow to adapt to the brave new world of mobile. Browsing a phpBB forum on an iPhone wasn’t a fun experience, requiring a lot of pinching and zooming. A mobile-friendly version was in the works, but it was delayed, and a native app seemed like it was far from reality. In the meantime, there was Facebook, Twitter, and, in 2010, Instagram, all with shiny apps that integrated closely with smartphones. You could take a picture, add a cool filter, and post it to the entire world with a few taps of your fingers. In comparison, forums seemed antiquated. As the social networks’ numbers went into the stratosphere, traffic to web forums started to plummet.
Facebook became the dominant social media site, gathering a startling percentage of the world within its service. As of the end of 2023, there were just over 3 billion Facebook users out of 8 billion humans on the entire planet.
But the road to world domination has not been without its problems. Moderation, which was already difficult with millions of Usenet and web forum users, became almost impossible at this scale. Facebook relied on automation to filter most of the junk, but also used human moderators, who had the unenviable job of identifying gross and horrifying posts so that the rest of us didn’t have to see them. This worked only so well, and it occasionally failed completely in languages and countries where Facebook didn’t have many moderators. It’s hard to fathom that an Internet message company could be partly responsible for genocide, but it happened.
As our world gets more divided and angry, many people blame social media for this state of affairs. Mark Zuckerberg has been forced to appear multiple times in front of the United States Congress to defend his empire. What started as a fun hobby for nerds has led to a world where everyone feels anxious and uncertain as they “doomscroll” infinite feeds on their mobile phones.
Messaging users over time, regular scale. Everything else becomes a blip compared to social media.
Jeremy Reimer
Messaging users over time, logarithmic scale. Now patterns start to emerge. Some data has been interpolated or extrapolated. Data sources available upon request.
Jeremy Reimer
Epilogue: The future of online messaging
Today, many folks look back with fondness on the early days of computer-based messaging. Depending on their age, they may wax nostalgic for BBSes, Usenet, or web forums. Surprisingly, all these technologies still exist, although they are either barely used or are full of spam. It’s hard not to think that something may have been lost.
For my part, I decided to take a hard break from all social media and online forums in 2023. I’ve found the experience to be liberating, decreasing my anxiety significantly while boosting my mood, energy, and available free time. But I still crave interacting with other humans online, and I miss the old days of running my own forum.
I’ve found that applications like Discord, where you can join invite-only groups of friends and colleagues, have helped fill this gap. In the smaller Discords, you feel like you can actually get to know the regular posters. Perhaps the future isn’t one of endless growth for all-powerful corporations but a return to smaller, more personal “third spaces” where we can feel comfortable. In the meantime, you can always send me an email.
