Telehack: the website that simulates the 1980s internet
09 Nov 2026 · By the Boring Websites editors
Telehack is a free, browser-based terminal that drops a visitor into a working imitation of the early internet, the ARPANET and Usenet as they looked between roughly 1985 and 1990. There is no install and no sign-up: a green cursor blinks on a black screen at telehack.com, and typing a question mark reveals a menu of period commands that lead to thousands of simulated machines.
In brief
- Telehack is an online simulation of a stylized ARPANET and Usenet, depicting the network circa 1985 to 1990, per its Wikipedia entry.
- It holds roughly 26,600 simulated hosts carrying period files, programs, games, and bulletin board systems, and has drawn more than 50,000 users.
- An anonymous engineer using the handle Forbin launched it on March 7, 2010, writing the whole thing in Perl.
- It answers through a web browser, a telnet client, SSH, and even a simulated dial-up line, so the way in is part of the exhibit.
What Telehack actually is
Telehack presents itself in one plain line. Its official documentation describes it as a simulation of a stylized arpanet and usenet, circa 1985 to 1990, and calls it a full multi-user simulation with more than 26,600 simulated hosts, programs to run, and simulated historical users. The screen is the whole interface. There are no windows, no mouse targets, and no menus in the modern sense.
What makes it more than a screensaver is that the environment is shared. Participants occupy a common world rather than a private sandbox. The machines, files, and message boards persist between sessions, so one person poking at a simulated host leaves a trace that another person can later find. The result reads less like a game level and more like a frozen snapshot of a network that millions of people once used in earnest.
The period it recreates matters. The late 1980s sat before the World Wide Web, when reaching another computer meant knowing its name, dialing in, or hopping across the ARPANET by hand. Telehack rebuilds that texture: the terse prompts, the host-by-host navigation, and the sense that the network was a place to be visited rather than a feed to be scrolled.
Three doors into the same room
Part of the charm is that Telehack can be reached the way such systems actually were reached. The Wikipedia summary notes that the simulation accepts connections through a web browser, a telnet client, and SSH. The browser version runs a terminal emulator over a WebSocket, so nothing needs installing, but the older paths still work for anyone who wants the authentic friction.
The front page spells out the alternatives. It invites a visitor to open a shell and connect by typing telnet telehack.com, or to use secure shell with ssh on port 2222 as the guest user. Coverage of the project has also mentioned a simulated dial-up line answering at 14.4 kbit per second, reproducing the modem handshake that once preceded every session. The method of arrival becomes part of the history lesson.
None of these doors demand an account. Registration is optional, and a guest can wander freely. For people who do register, the system allows a public key to be attached to secure the login, a small nod to how real accounts were protected. The point is that the barrier to looking around is close to zero, which is rare for anything this deep.
A map of 26,600 machines
The scale is the first surprise. TechRadar counted some 24,000 simulated machines when it wrote about the project in 2011, and later accounts put the figure near 26,600, according to the Wikipedia article. Each host carries period material, and the network as a whole holds files, executable programs, games, and bulletin board systems. It is a museum where every exhibit is also a working door.
Navigation uses the vocabulary of the era. Hosts are enumerated with the netstat command, connections are opened with telnet or by dialing a number, and the users logged in to a given machine are listed with finger. Someone who learned these commands on a real system in 1989 would need no retraining here, and someone who never did can learn them in a few minutes of typing a question mark and reading the one-line help.
The command list itself rewards browsing. Alongside the networking tools sit small diversions and utilities: a Star Wars animation rendered in text, traceroute, a typing-speed test, a unit converter, an uptime readout, and a Usenet reader. The help command gives short descriptions, and any command can be stopped with control-C. The whole thing is case-insensitive, so the uppercase spelling in the manuals is a convention rather than a requirement.
The programs are not just for show. Many of the simulated hosts carry executable files that run inside the environment, from utilities to small games, so the network is browsable and playable at once. A visitor who finds an interesting machine can often do more than read about it, and the surprise of a working program on a fake 1980s host is a large part of why people keep clicking into one more connection.
Commands that double as a history lesson
The tool names are not decoration. Each one is a real utility that shaped how people used the network before the web, and typing them teaches the mental model of the era. To find a machine, a visitor lists what is out there, picks a name, and opens a connection by hand. There is no search box standing between the user and the network, because in 1988 there was none.
Finger is the clearest example. On the real early internet it reported who was logged in to a given host and sometimes a short personal note they had left. Telehack preserves that behavior, so curiosity about a machine starts with asking who is home. It is a reminder that identity on the old network was tied to specific computers, not to accounts that followed a person everywhere.
The Usenet reader may be the richest exhibit. Usenet was the sprawling discussion system that predated web forums, organized into topic hierarchies rather than feeds. Reading simulated posts inside Telehack conveys how conversation worked when messages propagated host to host over hours or days. The pacing alone explains a great deal about why early online culture valued patience and careful writing.
Even traceroute earns its place. On the real internet it reports the chain of machines a message crosses to reach its destination, and seeing that chain inside Telehack makes the network feel like geography rather than magic. The old commands share a quiet assumption that the user wants to understand the system, not merely use it. That assumption is the history lesson, delivered without a single paragraph of exposition.
Why one engineer built a dead network
Telehack is the work of a single anonymous developer known by the handle Forbin, who has described himself as an engineer working in Silicon Valley. He began the project on March 7, 2010, and built it part-time over the months that followed, as recorded on Wikipedia. The handle itself is a reference: Forbin is the scientist in the 1970 film Colossus: The Forbin Project, a fitting alias for someone resurrecting a machine world.
The motivation was personal. According to the author, the initial aim was to let his son experience the atmosphere of 1980s computer intrusion without interfering with any real systems. A simulated ARPANET is a safe place to practice: nothing breaks, no law is troubled, and the stakes are entirely imaginary. What started as a way to show a child the old network grew into a public artifact used by tens of thousands of strangers.
In a 2011 interview with Andy Baio, published at waxy.org, the creator discussed the project while keeping his identity private. Baio summed up the result with a memorable phrase, calling Telehack playable archaeology. It is a reconstruction assembled from surviving traces of the real early network, not an invented fantasy setting, which is exactly why it feels specific rather than generic.
What is running under the hood
The engineering is as period-appropriate as the content. Telehack is implemented in Perl and runs as a single process organized around an event loop based on epoll, per the technical notes gathered on Wikipedia. One program holds the entire shared world in memory and services every connected visitor, which is a compact way to run a network of tens of thousands of hosts.
Inside that process sit several smaller machines. Two language interpreters are embedded: one for Z-code, adapted from the Rezrov package distributed through CPAN, and a BASIC interpreter written by the author. The Z-code engine means classic interactive-fiction games can run inside the simulation, so a visitor can boot a text adventure on a fake 1980s host and play it as a contemporary would have.
The emulation goes further down. Emulation of the 6502 and VAX architectures is handled within the same process rather than delegated to separate programs. The 6502 powered home computers like the Apple II and the Commodore 64, and the VAX was a workhorse of institutional computing, so their presence lets period software actually execute rather than merely appear as static text. It is a stack of simulations nested inside one Perl script.
What a visitor actually does
There is no scoreboard and no quest marker, which throws some first-time visitors. The activity is exploration. A session might start by listing hosts, checking who is logged in to one of them, reading a decade-old Usenet thread, and stumbling into a BASIC prompt or a playable game. The reward loop is curiosity rather than achievement, closer to walking through an archive than clearing a level.
The social layer is real in a quiet way. Because the world is shared and persistent, messages, files, and traces left by other explorers accumulate over time. People have used the simulation to practice command-line technique in a forgiving setting, and writers covering it have valued it for conveying the working culture of the period, not just its syntax. The feel of the thing, the patience and the terseness, is the real exhibit.
That patience is a feature worth sitting with. Modern interfaces race to anticipate the next tap, while Telehack waits for a typed command and answers only what was asked. Fans of deliberately unhurried corners of the web, such as The Bureau of Very Slow News, will recognize the appeal of a site that refuses to rush. The slowness is not a bug from an older era; it is the texture the project set out to preserve.
Who it is for, and who it is not
Telehack suits a particular kind of curiosity. People who grew up with a command line, or who wish they had, will find a familiar grammar waiting. Teachers covering the history of computing have a live exhibit rather than a slideshow, and writers or designers chasing period detail can check how the old tools behaved instead of guessing. The appeal is strongest for anyone who enjoys reading a system rather than being guided through it.
It is a poor fit for anyone expecting instant payoff. There is no onboarding wizard, no progress bar, and no gentle nudge toward the next thing to do. A visitor who types a few commands, sees only green text, and leaves has not done anything wrong; the site simply rewards patience and a willingness to type the help command often. The friction is deliberate, and it filters for a certain temperament.
Why it still matters fifteen years on
Telehack has outlasted most sites of its vintage. Its Wikipedia infobox lists its status as active, and coverage has continued well beyond its launch, with the French technology writer Korben revisiting the simulation as recently as 2025. For a one-person hobby project begun in 2010, surviving and still drawing visitors in the mid-2020s is a quiet achievement, as noted in the Wikipedia record.
Its staying power also reflects genuine care in the build. When it first reached a wide audience, it drew extended discussion on Hacker News, where programmers picked apart how the shared simulation worked and traded commands to try. That attention from a technical crowd is a useful signal: the thing holds up under inspection by people who would notice shortcuts, because the author did the unglamorous work of making the illusion consistent.
As a piece of preservation, it argues for a particular method. Rather than freezing the old network behind glass as screenshots or prose, Telehack keeps it executable. A curious person in 2026 can type the same commands a student typed in 1988 and get a plausible answer. That is a rare form of digital conservation, and it is why Baio's label, playable archaeology, has stuck so firmly to the project.
It also fits a wider revival. Interest in command lines, bulletin boards, and the pre-web internet has grown as those systems slip out of living memory, and Telehack gives that nostalgia something to do beyond looking at old photographs. The project treats the early network as worth inhabiting rather than merely remembering, and after more than a decade online it remains the most generous way to spend an afternoon pretending it is 1989.
FAQ
Is Telehack free to use?
Yes. Telehack runs in a web browser at telehack.com with no cost and no required account. Registration is optional, and a guest can explore the simulation without signing up at all, per its Wikipedia entry.
Is Telehack a real hacking tool?
No. It is a self-contained simulation of a stylized ARPANET and Usenet, not a gateway to real systems. Its creator built it specifically so that the atmosphere of 1980s computer intrusion could be explored safely, without touching any live network, as he explained in the waxy.org interview.
How do I connect to Telehack?
Three ways. The simplest is to open telehack.com in a browser. The site also invites connections by typing telnet telehack.com in a shell, or by using secure shell on port 2222 as the guest user, matching how such services were reached at the time.
Who created Telehack and when?
An anonymous Silicon Valley engineer using the handle Forbin launched Telehack on March 7, 2010, and developed it part-time afterward. He wrote it in Perl and has kept his real identity private, according to the project's Wikipedia article and the 2011 waxy.org interview.