Category: Ancient Web

The most ancient times of the internet: 90s, early 2000s.

Posted on by

Ancient Web: The Garfield and Friends Supapage Did the DVD Bonus Features It Never Got

When Garfield and Friends finally received chronological DVD box sets, one fan looked at the lack of bonus features and effectively said: fine, I’ll make them myself.

Visit The Garfield and Friends Supapage

The result is not merely an episode guide. The Supapage collects production trivia, broadcast differences, old commercials, song lyrics, international versions, mysteries from the show and details about what viewers actually saw when the series was running or syndicated.

One section digs into the Klopman Diamond, a recurring bit of Garfield lore, and records an answer obtained from animation writer and historian John Cawley. Another tries to sort out the “real story” behind the farm setting. There are pages on the Buddy Bears, Screaming with Binky, CBS bumpers and period advertising.

The site’s episode material is particularly useful because television history is often messier than a DVD menu makes it look. The page discusses Mark Evanier’s episode list, syndicated packages and the fact that many later episodes disappeared from ordinary reruns. It also compares broadcast versions with the DVD releases and notes differences in international prints.

Fan research catches what official releases flatten

A modern streaming catalog can tell you an episode’s title and season number. It is much less likely to explain which bumper aired around it on CBS, which episodes were missing from a syndication package, whether a segment changed in an overseas print, or what advertising surrounded the show during its original run.

That material lives at the edges of television history. It survives because viewers recorded tapes, remembered oddities, compared releases and cared enough to write down discrepancies that were never important to a distributor.

The Supapage understands its own role clearly. It was designed to supply the context and extras that the official discs did not. That makes the slightly homemade structure appropriate: this is somebody’s research desk made public, with every interesting loose end allowed to become another link.

Fan sites like this are one reason CacheRat’s 1,967 Ancient Web Domains research list includes entertainment pages alongside manuals and technical archives. Media history is not only the finished show. It is broadcast order, regional edits, bumpers, advertising, syndication and what viewers noticed at the time.

The DVDs preserved the episodes.

The fan page preserved some of the stuff around them.

Explore The Garfield and Friends Supapage

Posted on by

Ancient Web: Ruud Helderman Teaches Text Adventures by Building One in C

Ruud Helderman’s text-adventure tutorial begins from an excellent premise: if you want to understand how interactive fiction works underneath the authoring tools, build the machinery yourself in C.

Read How to Program a Text Adventure in C

The tutorial is not presented as a beginner’s introduction to the C language. It assumes the reader already knows basic programming and instead uses a small adventure game to explain how the pieces of an interactive world fit together.

Its chapters progress through the main loop, locations, objects, inventory, passages, distance, movement in the cardinal directions, generated code, additional attributes, conditions, opening and closing objects, and finally the command parser.

That order is valuable because the finished game is not dropped on the reader as one opaque codebase. The system grows a layer at a time.

The earliest program is barely more than a small “Little Cave Adventure.” From there, rooms become data, objects acquire state, exits connect places and user input has to become something the game can interpret.

Why use C for a text adventure at all?

Helderman is candid that dedicated interactive-fiction tools are usually a better choice if the actual goal is to publish a game. He also notes the historical association of adventure systems with languages such as MDL and ZIL rather than pretending C was the natural center of the genre.

The point of C is educational friction.

When there is no adventure engine hiding the implementation, the programmer has to decide what a location is, how an object knows where it is, how movement rules are represented and how a parser turns words into actions. Those problems are small enough to understand but rich enough to teach data structures and program organization.

The tutorial also touches on standard input/output, Glk and the possibility of translating the approach into languages such as Java or Python. The result is useful even for somebody with no intention of releasing a text adventure in 2026.

This is the kind of narrowly practical programming page that earns a place in CacheRat’s Ancient Web research list. It preserves not only source code but a way of teaching: choose one comprehensible project and expose every layer until the reader can see why the program is shaped the way it is.

The commercial glory days of parser games may be behind us.

A cave still makes a perfectly good place to learn C.

Continue through Ruud Helderman’s text-adventure tutorial

Posted on by

Ancient Web: A Samurai Pizza Cats Fan Page Has Been Frozen Since 1996

This Samurai Pizza Cats fan page says it was last modified on August 18, 1996. That date has now spent far longer being old than the page spent being current.

Visit the Samurai Pizza Cats fan page

The site identifies itself as the first page on the Pizza Cats and credits Patrick Atoon. It is a compact handmade guide to the English-language cartoon and its Japanese source, Kyatto Ninden Teyandee, built at a time when simply finding background information about an imported or localized anime series could require another fan to type it out for you.

The navigation covers the premise, recurring characters, the Cat Ninden, episode information, opening and closing themes, a fan-club oath, “Pizza Cat-isms,” credits and acknowledgments. There is even a Dutch version.

The episode guide is especially revealing. It describes 48 episodes and openly notes gaps where descriptions were still needed. Its numbering follows Canadian broadcast order rather than pretending there was one universally obvious sequence.

That is the sort of messy detail fan sites preserved well. Television syndication, regional broadcasts and localization often produced conflicting episode orders long before streaming services presented one clean list.

Before fandom became a database

A page like this did not have a wiki engine underneath it. The webmaster decided what newcomers needed to know, typed character descriptions, organized episodes, transcribed songs and invited other fans to fill missing information.

Its explanation of the series also preserves terminology from the period. The English title sits beside Kyatto Ninden Teyandee, Little Tokyo and the show’s original character framework. Those names now provide exactly the kind of long-tail search surface that disappears when old fan pages are reduced to a generic sentence saying “this was an anime from the 1990s.”

The page’s age is visible not because somebody recreated a retro theme, but because nobody replaced the original assumptions. It was made for people who needed an introduction to the show, and then it simply stayed online.

That makes it a good specimen from CacheRat’s larger 1,967 Ancient Web Domains research list. Fan pages often preserve the practical history around media—broadcast order, terminology, early community knowledge and incomplete contemporary understanding—not just the media itself.

There are slicker ways to research Samurai Pizza Cats now.

There are not many cleaner ways to see what researching it on the Web looked like in 1996.

Explore the surviving Samurai Pizza Cats fan page

Posted on by

Ancient Web: WebNeko Still Lets a Tiny Cat Chase Your Mouse

There is a tiny cat on WebNeko whose entire job is to chase your mouse pointer, and this is sufficient justification for the page continuing to exist.

Visit WebNeko

The page calls the current incarnation “It’s Neko! (3.0).” Move the pointer and the sprite follows it around the browser window. You can add another Neko, try a cat-herding variation, or swap the default cat for contributed skins ranging from black, gray and calico cats to robots, rainbows, socks and stranger interpretations.

Nothing about this needs an account. There is no feed, progression system or engagement loop. The reward is that a little pixel animal runs toward the cursor.

More importantly for Web archaeology, the page still preserves the copy-and-paste culture that once made small browser toys spread from homepage to homepage. Its instructions show a simple JavaScript embed and explain how visitors could put Neko onto their own sites.

Then the historical sediment gets even better.

Xanga gets instructions; MySpace gets blamed

WebNeko includes specific instructions for using the cat on Xanga. It also warns that MySpace and some other services block the <script> tag, preventing users from embedding Neko there.

That tiny compatibility note preserves an entire phase of the social Web. Personal publishing was moving from independent pages toward hosted profile services, but people still expected to decorate those profiles with code, counters, music, animated cursors and little creatures that served absolutely no business purpose.

The site says the Neko concept and artwork passed through many forms of software and traces its first incarnation back to X Window System programs. WebNeko is therefore less an invention than another migration of a persistent desktop toy into the browser.

The contributed skins are part of the story too. Once a simple toy was easy to copy and modify, users did what users always do: changed the graphics, made themed versions and passed them around.

This is the lightweight side of what CacheRat’s Ancient Web research list is meant to preserve. Not every old page needs to contain a manual, archive or major historical document. Some pages matter because they show what people thought websites were allowed to do before every interaction had to justify itself to a product manager.

A cat can chase a mouse. That is enough.

Go bother the cat at WebNeko

Posted on by

Ancient Web: IORR Took a 1980 Rolling Stones Fan Club Online in 1996

Before the Web had fandom platforms, IORR already had fifteen years of Rolling Stones fandom behind it.

Visit It’s Only Rock’n Roll

It’s Only Rock’n Roll, usually shortened to IORR, was founded by Bjørnulf Vik in Norway in November 1980. The fan club moved online in 1996, which means the website did not create the community so much as give an existing one a new piece of infrastructure.

That distinction helps explain the density of the surviving site. IORR is not a retrospective page assembled from Wikipedia-era summaries. It contains tour material, show reports, news, member discussion and band information accumulated while events were actually happening.

Tour pages reach back through the 1990s. Fans submitted reports from concerts. Forum threads became their own archive of rumors, set lists, travel plans, ticket problems, arguments and the thousands of small observations that official histories rarely bother preserving.

The site describes itself as entirely nonprofit, without advertising or content marketing. It also claims millions of yearly visitors, tens of thousands of active posting users, thousands of contributors and more than a million published Web pages. Those are IORR’s own figures, but even without the numbers the scale is obvious from the surviving indexes.

A 1990s forum became something worth preserving

One of the most revealing current pages is IORR’s discussion of its next-generation forum. The original “Tell Me” forum was built around a separate SQL-based system dating from the 1990s. By 2026, the site says it contains more than a million posts and is approaching thirty years of continuous history.

Rather than simply deleting it during a modernization, IORR introduced a newer parallel forum while explicitly planning to keep the old one available because of the historic material inside.

That is the right instinct. Forums are often treated as disposable software when the software is only the container. The actual archive is the accumulated memory of thousands of people talking about tours, venues, recordings and shows close to the time they happened.

This is one reason CacheRat’s 1,967 Ancient Web Domains research list includes fan sites alongside technical archives and museums. A long-running fan community can preserve a cultural record no official archive would ever reproduce in the same form.

IORR started before the Web existed, survived the Web’s first decades, and now has to preserve its own first-generation online infrastructure.

That is a pretty good definition of Internet history happening slowly enough that nobody noticed.

Explore IORR’s surviving Rolling Stones archive

Posted on by

Ancient Web: Rich Whitehouse Has Been Reverse-Engineering Weird Stuff in Public Since 2009

Rich Whitehouse’s website is what happens when somebody keeps a technical notebook online long enough for its odd little experiments to become useful history.

Visit Rich Whitehouse’s site

The site has been accumulating posts since 2009 and is still active in 2026. Its subjects repeatedly circle around reverse engineering, game technology, obscure data formats and the sort of technical questions that begin with “I wonder how this old thing actually worked” and end several hours later inside a debugger.

A 2011 post on FX Fighter, for example, describes digging through a 32-bit DOS game’s formats with IDA Pro and the DOSBox debugger. A 2015 piece on the Sega Saturn version of Quake examines its level data and the unusual fact that the game was rebuilt around Lobotomy Software’s SlaveDriver engine rather than simply porting the original PC engine.

In 2017 Whitehouse wrote about old production data from the Genesis version of Disney’s Aladdin. In 2018 he produced DOOMBA, a Noesis script that could turn tracking data from a Roomba’s simultaneous localization and mapping system into randomized DOOM maps.

Yes: vacuum cleaner telemetry became level geometry. The old Web occasionally rewards persistence.

Noesis ties the whole archive together

Whitehouse is also the author of Noesis, a tool for previewing and converting an enormous range of game and media data formats. Its documentation covers plugins, archives, batch export, command-line operation and a data viewer, while continued version updates show that this is not merely abandoned freeware attached to an old homepage.

Noesis makes the site more than a blog about reverse engineering. Posts about formats and experiments connect directly to a working tool that lets other people inspect the same kinds of assets.

Later research includes an extensive Sega VR emulation project examining the unreleased headset. That combination—old hardware, undocumented formats, emulator work, source digging and public notes—is exactly the kind of material that becomes increasingly valuable as original development environments disappear.

The site reports tens of millions of page hits since February 2009. That number is the site’s own counter, but the more convincing measure of usefulness is simpler: technical posts written years ago still contain names, formats, engines and implementation details somebody can search for when a strange file lands on their disk.

CacheRat’s Ancient Web research corpus is full of specialist pages whose value comes from that accumulation. Whitehouse’s site is an unusually lively example because the archive never really stopped growing.

Some people maintain blogs. Some people feed Roomba navigation data into DOOM.

Browse Rich Whitehouse’s technical archive

Posted on by

Ancient Web: The Napoleonic Guide Turned One Obsession Into 4,000 Pages

Somebody decided Napoleon needed a website and then apparently refused to stop until the thing had more than 4,000 pages.

Visit The Napoleonic Guide

Edited by Richard Moore, The Napoleonic Guide describes itself as an independent reference to Napoleon Bonaparte and the Napoleonic era. Its copyright line runs from 1999 to 2017, and the surviving navigation shows what happens when a Web project grows by accumulation rather than by redesigning itself every eighteen months.

There are biographies, battle summaries, campaign material, timelines, maps, quotations, family history, letters, Napoleon’s will, assassination plots, marshals, generals, armies, weapons, uniforms, medicine, naval warfare, battlefields and a glossary. Then it keeps going into the French Revolution, Goya, historical fiction, Sharpe, Hornblower, miniatures, reenactors, quizzes and book reviews.

The chronology alone is broken into month-by-month material. This is not a landing page about Napoleon. It is a hand-built reference shelf.

The bibliography is part of the value

One useful feature is the source material behind the site. Its references point readers toward firsthand accounts and established histories rather than presenting thousands of pages as anonymous Web fact. Works associated with Rifleman Harris, Sergeant Bourgogne, Friedrich Schaumann, William Surtees and George Simmons sit alongside historians such as David Chandler, Philip Haythornthwaite and Charles Oman.

That matters because military history on the Web can become a swamp of repeated claims whose original source disappeared six copies ago. A site that exposes its reading list gives a researcher somewhere else to go.

The page structure is also a surviving example of how large specialist sites used to scale. Instead of turning everything into a database-backed feed, the Guide divides a huge subject into narrow human-readable indexes. You can enter through a battle, a person, a regiment, a weapon or a date and keep moving sideways.

Modern search engines often reward pages that answer one query and then attempt to sell something. The Napoleonic Guide feels like the opposite impulse: answer one question, notice seventeen adjacent questions, and make pages for those too.

That is exactly why specialist survivors are useful inside CacheRat’s 1,967 Ancient Web Domains research list. A domain may be old, but the real value is the dense vocabulary and structure accumulated by somebody who knew the subject well enough to keep subdividing it.

Four thousand pages later, Richard Moore had built less of a website than a small Napoleonic reference room.

Explore The Napoleonic Guide

Posted on by

Ancient Web: Matthew Furman Kept Building Websites Like Websites

Matthew Furman’s homepage is a useful reminder that the old Web was not only a visual style. It was a way of organizing the Internet around one person’s interests, with guestbooks, counters, side projects and links leading into increasingly specific corners.

Visit Matthew Furman On-Line

The current site still behaves that way. It offers a separate mobile version instead of pretending every screen is the same, maintains a guestbook, and even explains that its visitor counter had to be reset after the old one broke. The footer identifies Matthew Furman On-Line as a project running from 2009 through the present rather than a museum piece frozen twenty-five years ago.

That makes it more interesting, not less. Furman kept building in a personal-Web tradition after the rest of the Internet increasingly moved toward standardized profiles and platform pages.

The guestbook preserves some of that transition history in miniature. Old entries stretch back more than a decade, while the site notes that Microsoft FrontPage Server Extensions eventually disappeared after a server upgrade. Recent visitors have arrived through small-Web discovery tools and commented specifically on the site’s old-fashioned presentation.

Then there is the payphone museum

Furman’s larger network includes The Pay Phone Museum, a photo archive built around a piece of public infrastructure that has been quietly vanishing from streets, stores, stations and parking lots.

The museum describes payphones as an endangered species and asks contributors to send photographs and stories. Updates continued into 2025, including new uploads and older pictures Furman discovered from 2016 through 2018 that had never made it onto the site. The archive has grown into dozens of pages of surviving and vanished phones.

This is precisely the kind of subject that benefits from an independent obsessive archive. A corporation can publish a history of telecommunications. It is less likely to preserve a photograph of one battered payphone outside a particular laundromat before somebody removes it.

Furman has also maintained other archive-style projects, including a WAV-file museum. Taken together, the sites show why personal publishing still matters: one webmaster can decide that something mundane is disappearing and start collecting evidence before anybody with a preservation grant notices.

CacheRat’s larger Ancient Web research list contains plenty of pages like this—sites that are valuable less because they were famous than because a person kept noticing things.

There is still room on the Web for that.

Browse Matthew Furman’s surviving personal site

Posted on by

Ancient Web: Pinball Rebel Built an Encyclopedia Around the Real Machines

The Internet has no shortage of virtual pinball. Ken’s Pinball Machine Pages are interested in the heavier version: steel balls, painted backglasses, switches, coils, instruction cards, burned lamps and cabinets somebody eventually has to repair.

Visit Ken’s Pinball Machine Pages at Pinball Rebel

The site is part personal collection, part reference library and part field guide to finding machines that can still be played. Its A-to-Z machine pages cover individual tables with photographs, game information, playing tips, videos and, where available, repair or restoration notes.

That specificity matters. A pinball machine is not merely a title and release year. It is a physical system with manufacturer-specific mechanisms, artwork, rules, sounds and failure modes. Pages for individual machines preserve model names and production details that are useful to collectors precisely because they are too specific for a generic arcade-history article.

Ken also built custom pinball cards and collected material around jukeboxes, drive-in theaters, fortune-teller machines, soda advertising and other coin-operated or roadside machinery. The result feels less like a modern content site than a gameroom whose owner kept adding shelves.

The locator turns the archive back into a map

One of the best surviving pieces is the World Wide Pinball Locator. The project says it began years ago at the now-defunct pinball.nu and was kept alive with help from other contributors. The current page reports thousands of machines at thousands of locations across six continents.

That is an unusually practical form of preservation. A photo can document a machine after it disappears, but a locator can still send somebody to one while it is alive, powered on and accepting money.

The site is also careful about what it means by pinball. These pages are about the physical machines, not computer recreations. That distinction sounds obvious until the Web starts returning emulators, mobile games and virtual tables for searches that used to mean several hundred pounds of wood, glass and electromechanical trouble.

Pinball Rebel is the kind of specialist site that makes CacheRat’s 1,967 Ancient Web Domains research list useful as more than a pile of old URLs. The interesting part is not merely that a site survived. It is that somebody spent years identifying machines, photographing them, explaining them and recording where the surviving hardware can still be found.

The ball drains eventually. The documentation does not have to.

Explore the surviving Pinball Rebel machine archive

Posted on by

Ancient Web: Underground Homes Built a Rabbit Hole Below Ground

Some websites answer one question and stop. Underground Homes starts with “what if the house were under the dirt?” and keeps digging until somebody is living behind a 27-ton former missile-silo door.

Visit Underground Homes

The site is an independent collection of pages about earth-sheltered living, underground construction, pit houses, subterranean cities, costs, design ideas and unusual existing homes. It moves freely between archaeology, modern residential construction and the wonderfully specific category of people who bought decommissioned nuclear facilities and decided the launch complex needed a kitchen.

One page profiles old missile sites converted into houses. Edward Peden’s Kansas home, for example, occupies part of a former Atlas E missile base. The site describes a roughly 6,000-square-foot converted area and a 27-ton door leading into the former missile room.

That is a difficult fact to make boring.

The Web used to reward extremely specific curiosity

Another section covers pit houses, tracing the basic idea of partially buried dwellings through ancient and Indigenous building traditions. A separate underground-cities page jumps from tunnels and shelters to places such as Coober Pedy, where mining and extreme heat encouraged people to build living spaces directly into the ground.

There are also practical pages about what counts as an underground home, construction costs and the tradeoffs involved in putting walls or roofs below grade.

The interesting part is the accumulation. None of these subjects necessarily needs its own media company. Together they form the sort of deep topical rabbit hole that independent websites used to excel at: somebody becomes interested in one odd class of building and keeps adding pages until the site becomes a small reference library.

For researchers, the terminology is useful on its own: earth-sheltered homes, pit houses, underground cities, Atlas E missile silos, subterranean housing and underground construction. These are better search hooks than flattening the site into generic “alternative housing” language.

For webmasters, there is another lesson. A narrowly focused independent site does not need to compete with every architecture publication on Earth. It only needs to become unusually useful for the people who type the weirdly specific query.

That is one of the recurring patterns behind CacheRat’s 1,967 Ancient Web Domains research list. The old Web is full of entire domains built around questions that would now be buried inside three social posts and a video algorithm.

This one kept digging.

Explore the Underground Homes reference site