Category: Ancient Web

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

Posted on by

Ancient Web: Medieval Cookery Turned 20 Years of Food Research Into a Working Reference Library

If you want to know what medieval food actually looked like, a useful website needs to do more than tell you that people ate a lot of bread.

Visit Medieval Cookery

Daniel Myers built MedievalCookery.com into the kind of specialist reference that only makes sense after somebody has spent years caring about the difference between a modern themed recipe and a dish that can actually be traced to a period source.

The site has been running for more than twenty years. Its table of contents includes transcribed medieval recipes, reconstructed menus, statistics drawn from cookbooks, online source books, a searchable recipe system, a dictionary of Middle English cooking terms, food-related paintings, spice sources, fruit varieties, recommended books, and research notes.

That is a much deeper structure than a normal recipe site.

A visitor can move from the wording of an old recipe to an interpretation that can be cooked, then back out into the manuscript, vocabulary, ingredient history, or menu context behind it.

Historical cooking is partly a translation problem

Old recipes rarely behave like modern ones.

Measurements may be vague or absent. Ingredient names change. Cooking temperatures are not expressed as oven settings. Techniques that were obvious to the original cook may never be explained. A word can refer to something different from the modern ingredient that looks most similar.

MedievalCookery.com preserves both sides of that problem.

The homepage can serve up a random historical recipe such as an early-16th-century Dutch preparation while other sections provide the tools needed to understand the terms and sources around it. Myers has also published transcriptions from manuscript material and written about the process of turning historical instructions into readable modern research without pretending the original text said more than it did.

In a 2023 post, Myers wrote that he had started the site more than twenty years earlier and was thinking seriously about how to preserve the work as his available time and energy changed. That is one of the recurring problems of the independent Web: a useful archive can become larger than one person’s ability to maintain it.

The fact that the site is still online matters.

CacheRat’s 1,967 Ancient Web Domains research list contains many sites where years of hobby research quietly turned into infrastructure for everyone else working on the same obscure subject.

Medieval Cookery is exactly that kind of site—with more ginger, grains of paradise, and alarming quantities of breadcrumbs.

Dig through MedievalCookery.com’s recipes and research tools

Posted on by

Ancient Web: The Mojave Phone Booth Became an Internet Landmark Because Somebody Kept Calling It

The Mojave Phone Booth was just a working payphone in the middle of nowhere until the Internet noticed that you could call it.

Visit the original Deuce of Clubs Mojave Phone Booth site

The Deuce of Clubs archive preserves how that discovery turned into one of the stranger stories of the late-1990s Web.

The site’s navigation is practically a timeline: how the booth was discovered, an interview with Mojave historian Dennis Casebier, visits in August and November 1997, another trip in 1998, a 1999 “Booth Blizzard,” reader mail, press coverage, a satellite image, and increasingly elaborate attempts to document what happened around one telephone in the desert.

The basic premise was irresistible.

Somebody found the number of an isolated phone booth in the Mojave Desert and started calling it to see whether anyone would ever answer. Eventually someone did. Other people learned the number. Some called. Some drove into the desert. Some waited for the phone to ring.

The Web supplied the missing ingredient: an audience large enough to turn a geographically useless telephone into a social destination.

This was a viral object before social media

The booth’s fame spread through personal websites, email, zines, radio, press coverage, word of mouth, and links from other strange corners of the Web.

Deuce of Clubs did not merely summarize the story after the fact. The site accumulated the event while it was happening.

Visit reports preserve directions, photographs, conversations, mistakes, weather, visitors, and desert logistics. The mail archive records people discovering the site from elsewhere online and then immediately trying the number themselves. Later updates document the booth’s removal and the strange afterlife of the story.

That makes the site valuable in a way a modern encyclopedia entry cannot reproduce.

It captures the texture of an Internet phenomenon before “going viral” became a standard phrase and before a platform could compress the whole thing into a trending topic for forty-eight hours.

The original page even acknowledges its own age. In a 2014 update, its author hesitated to change the site because it felt like a primitive artifact. That instinct was correct. The primitive artifact is now part of the evidence.

CacheRat’s 1,967 Ancient Web Domains research list contains plenty of pages that are interesting because of what they describe. The Mojave Phone Booth site is rarer: the website itself helped make the thing famous.

A payphone rang in the desert, and the Internet decided somebody should answer it.

Explore the surviving Mojave Phone Booth archive

Posted on by

Ancient Web: Robin Hanson’s Great Filter Essay Is Still Sitting on a 1998 Academic Webpage

One of the Internet’s most durable ideas about extraterrestrial life is still sitting on a plain academic webpage dated September 15, 1998.

Read Robin Hanson’s original Great Filter essay

Robin Hanson’s The Great Filter — Are We Almost Past It? starts from the problem usually called the Fermi paradox or the Great Silence: if technological life can arise and spread through space, why do we not see convincing evidence that somebody already did it?

Hanson reframes that silence as information about probability.

Between ordinary dead matter and a civilization capable of lasting expansion through the universe, there must be at least one step—or several steps—that is extraordinarily difficult. That obstacle is the Great Filter.

The uncomfortable question is where it sits.

Behind us is good news; ahead of us is not

If the hardest steps were things Earth has already passed—perhaps the origin of life, complex cells, multicellular organisms, intelligence, or technological civilization—then humanity may have already survived most of the filter.

If those biological steps turn out to be easy and common, the missing civilizations become harder to explain. The filter may then lie later in the sequence, possibly in the future of technological societies.

That is why Hanson makes the counterintuitive observation that discovering abundant independent extraterrestrial life could be bad news for our long-term prospects. Each apparently easy earlier step pushes more of the unexplained difficulty toward whatever comes next.

Whether or not one accepts every assumption in the argument, the framework became a useful way of thinking about astrobiology, existential risk, SETI, and the Fermi paradox.

The original webpage is also a beautiful old-Web artifact in its own right.

It is not a retrospective explaining an idea after it became famous. It is the 1998 essay itself, complete with long sections of argument, citations, a bibliography, and links to related academic material from the same era. There is almost no interface between the reader and the idea.

That directness has SEO value now because people still search for Great Filter, Fermi paradox, Robin Hanson, extraterrestrial civilization, and SETI while the primary document remains publicly readable.

CacheRat’s 1,967 Ancient Web Domains research list contains a surprising number of pages like this: concepts that escaped their original websites and entered general culture while the source pages quietly remained online.

The argument traveled much farther than the design did.

Read the 1998 Great Filter paper on the original page

Posted on by

Ancient Web: New Zealand’s Rare Breeds Site Has Been Tracking Livestock Genetics Since 2002

A livestock breed can disappear for the same basic reason an old website disappears: once too few copies remain, one bad turn can finish the job.

Visit the Rare Breeds Conservation Society of New Zealand

The Society’s website has been online since 2002 and is still actively updated. It documents rare and minority livestock in New Zealand while supporting the much more practical work of keeping those animals genetically and reproductively viable.

The site covers breed information, a breeder directory, a registry, rescue projects, conservation watchlists, auctions, fibre resources, and a gene bank. It also explains why preserving old livestock genetics is not merely sentimental.

Commercial agriculture tends to reward a relatively small number of high-production breeds. That can push older or locally adapted animals toward dangerously low population levels. The Society argues that retaining those breeds preserves genetic diversity that may matter when climates, diseases, production systems, or consumer needs change.

New Zealand created some especially strange preservation problems

The site’s most interesting material concerns animals that developed distinctive populations after long periods of isolation.

It points to Arapawa sheep and goats, Kaimanawa horses, Auckland Island pigs, and other populations with unusual histories in New Zealand. Some descended from domestic animals released or abandoned in remote places and then adapted to local conditions over generations.

That makes conservation complicated. These are not simply museum pieces with pedigrees sitting safely in a binder. They are living genetic populations that need enough animals, responsible breeding, records, and in some cases stored germplasm to remain viable.

The website turns that work into a public reference system.

Its news pages still record new breed pages, registry changes, auctions, and conservation updates in 2026. The National Library of New Zealand also catalogs the site as an electronic resource dating from 2002 onward, which is a nice reminder that long-running websites can become archival objects while they are still alive.

CacheRat’s 1,967 Ancient Web Domains research list regularly surfaces old specialist sites where the Web is doing exactly what it was good at: giving a small organization enough space to maintain an absurdly specific body of knowledge for decades.

In this case, the stakes are higher than broken GIFs. The thing being preserved can actually go extinct.

Explore New Zealand’s rare-breed conservation archive

Posted on by

Ancient Web: VGMusic Still Hosts 1,900 Nintendo 64 MIDI Files

Before streaming sites made every game soundtrack instantly searchable, fans often rebuilt the music themselves one MIDI sequence at a time.

Browse VGMusic’s Nintendo 64 archive

VGMusic—the Video Game Music Archive—has been online since December 19, 1996. Its larger collection contains more than 31,000 MIDI files across dozens of game systems, but the Nintendo 64 directory is enough to show how enormous the project became.

As of its September 2023 index update, that directory contained 1,900 MIDI files.

The page is not just a pile of downloads. Each entry can include the game, song title, file size, the person who sequenced it, and comments. Banjo-Kazooie alone spills into dozens of arrangements and variations. Farther down the page are games such as Wave Race 64, War Gods, WrestleMania 2000, and Yoshi’s Story.

That metadata matters.

MIDI archives documented fan labor

A MIDI file is not an audio recording of a cartridge. It is a sequence of musical instructions that a synthesizer or sound module interprets.

For fans in the 1990s and early 2000s, recreating game music as MIDI meant listening closely, transcribing parts, approximating instruments, and fitting music designed for specialized game hardware into the General MIDI ecosystem found on PCs and sound cards.

The resulting file might sound different on a Sound Blaster, a Roland module, or a later software synthesizer. That variability was part of the experience.

VGMusic preserved the files, but it also preserved who made them.

That makes the archive useful beyond nostalgia. Researchers looking at game-music fandom, amateur sequencing, MIDI culture, web distribution, or the history of fan-created digital media can see thousands of credited contributions organized by platform and game.

The site’s structure has barely any interest in modern presentation tricks. You choose a console, choose a game, and find the file. The Nintendo 64 page is more than two thousand lines long because sometimes a giant index is exactly what the job requires.

CacheRat’s 1,967 Ancient Web Domains research list includes many archives whose value comes from patient accumulation rather than clever design. VGMusic is one of the clearest examples.

Nearly thirty years after the archive was established, the MIDIs are still sitting there waiting to be played.

Dig through the Nintendo 64 MIDI collection

Posted on by

Ancient Web: Historian.net Put the History of Writing Inside a Shockwave-Era Website

Historian.net has one of those entrances that immediately dates a website: the front page still tells visitors that a Shockwave plug-in is required.

Read Historian.net’s History of Writing

Fortunately, the actual History of Writing page survives as ordinary HTML and images, which makes it much more useful in 2026 than the site’s old multimedia doorway suggests.

The page attempts a broad illustrated path through the development of written communication. It starts with prehistoric images and early counting systems, moves into Neolithic tokens and Mesopotamian record keeping, then follows pictographs toward cuneiform and later turns to Egyptian hieroglyphs and phonetic signs.

Some of its historical framing is stated more confidently than a modern scholarly treatment might choose, so it is better read as an older educational synthesis than as the last word on the origins of writing.

That distinction actually makes the page more interesting.

The pictures show how the Web taught before video took over

The page uses small GIFs and JPEGs to put token shapes, pictographs, cylinder seals, hieroglyphs, and character forms directly beside the explanation. It is not trying to be cinematic. The images exist because the reader needs to see how a picture can become a stylized sign and how signs can begin representing sounds instead of objects.

The section on Mesopotamia describes clay tokens used for accounting and the transition toward impressed and inscribed symbols. Later sections discuss cuneiform, Egyptian consonantal signs, and acrophony—the use of a picture-sign for the initial sound of the word it represented.

Even the page’s imperfections tell us something about early educational publishing online.

A person could assemble a visual lecture, attach source images, write long explanatory text, and make it globally accessible without turning it into a course platform, a video channel, or a subscription product. The result might be uneven, but it could remain readable decades later.

The contrast with Historian.net’s Shockwave-dependent front page is almost too perfect. The proprietary presentation layer aged badly. The plain HTML article underneath survived.

CacheRat’s 1,967 Ancient Web Domains research list is full of that lesson: the supposedly primitive parts of the old Web are often the parts that still open.

Explore the illustrated History of Writing page

Posted on by

Ancient Web: UNSW Explained Why Sailboats Can Go Faster Than the Wind

A good technical webpage can survive for decades because the question it answers never gets old: how exactly does a sailboat move toward the wind instead of simply being blown away from it?

Visit The Physics of Sailing

Joe Wolfe’s University of New South Wales page attacks that problem with the tools a physics student would expect: Newton’s laws, vector addition, relative velocity, lift, drag, torque, buoyancy, and the force produced by a keel moving through water.

The page dates to 2002, with an early-2003 modification noted at the bottom, and it still works as a compact teaching document.

It begins with deceptively simple questions. How can a boat sail upwind? How can a boat go faster than the wind? Why do extremely fast boats seem to spend so much time sailing with the apparent wind coming from ahead?

Then it draws the arrows.

The interesting part is the relative wind

The page explains that a sailboat going directly downwind is limited because its speed reduces the wind it feels. As the boat approaches the true wind speed, the relative wind falls toward zero and the sail loses the force pushing it forward.

At an angle, the geometry changes.

The boat’s velocity and the true wind combine to produce an apparent wind. A sail can deflect that airflow and generate a force with both sideways and forward components. The keel resists most of the sideways motion by pushing against the water, leaving a useful forward component to accelerate the boat.

That is where the page becomes more than a sailing FAQ. It is a practical lesson in vectors.

Wolfe uses eighteen-foot skiffs on Sydney Harbour as an example of boats fast enough that their crews can experience apparent wind from well forward even while making progress downwind. He also uses a hand held into the airflow while bicycling to make the idea of lift less abstract.

Nothing about the page needs animation-heavy software or a modern framework. A few diagrams, equations, photographs, and careful paragraphs do the work.

That durability is exactly why these educational pages matter. CacheRat’s 1,967 Ancient Web Domains research list contains many academic pages that remain useful long after the departments around them have redesigned everything else.

The physics did not need a redesign.

Read Joe Wolfe’s Physics of Sailing

Posted on by

Ancient Web: Greg Egan Has Kept a Science-Fiction Home Page Alive Since 1997

Greg Egan’s website is what happens when an author treats the Web as part of the work instead of as a disposable marketing layer around it.

Visit Greg Egan’s Home Page

The site says it was created on October 25, 1997. Nearly three decades later, it is still active and still recognizably a personal website.

Egan describes himself simply as a science-fiction writer and computer programmer. The site contains a bibliography, links to works available online, illustrations and applets that supplement his fiction, introductory science articles under the Foundations heading, more technical notes, and an applet gallery.

That mix is unusually revealing.

A conventional author site might give you a biography, book covers, tour dates, and a purchase button. Egan’s site gives you mathematical and scientific scaffolding for the ideas behind the fiction.

The author controls the context

This matters especially for a writer whose work regularly leans into mathematics, physics, computation, consciousness, cosmology, and speculative biology.

A novel such as Diaspora, Permutation City, Schild’s Ladder, or Dichronauts can send a reader looking for explanations that ordinary promotional copy is not equipped to provide. Egan’s site can answer with diagrams, notes, applets, and references built by the same person who wrote the story.

The site also preserves a very old-Web idea: the home page as sovereign territory.

Egan links to his own work, keeps his own bibliography, publishes science material directly, provides an RSS feed, and states his contact boundaries in his own words. He even explicitly warns readers that photographs circulating online under his name are not actually him.

That kind of direct authorship is increasingly rare. Modern writers are often represented online through publisher pages, retailer listings, social accounts, search summaries, and scraped biographies. Here, the author is still the primary source.

The site was being revised in September 2026, proving that “Ancient Web” does not have to mean abandoned. Sometimes the more interesting artifact is a site that survived because its owner never saw a reason to replace a useful structure with a trendier one.

CacheRat’s 1,967 Ancient Web Domains research list includes plenty of dead pages and fossils, but sites like Egan’s are arguably better evidence of what the independent Web got right.

It is old, current, specific, and still owned by the person it is about.

Explore Greg Egan’s science, fiction, and applet archive

Posted on by

Ancient Web: Russell Hoffman’s Nuclear Glossary Packed 300 Sources Into a 2002 Web Primer

The old Web had a particular talent for turning one person’s concern into an entire reference library.

Visit the Internet Glossary of Nuclear Terminology

Russell Hoffman’s Internet Glossary of Nuclear Terminology is a strong example. First published in July 2002, it tries to explain nuclear power and nuclear risk through a web of short pages, pictures, terminology, and source material rather than through one linear essay.

The index alone is a time capsule of what the project wanted to cover: atomic energy, background radiation, contamination, decontamination, downwinders, fuel cycle, nuclear navy, nuclear reactor, spent fuel, waste repositories, meltdown, health effects, nuclear war, and Nevada.

Hoffman says he researched more than 300 documents and at least as many websites for the presentation, with images drawn from more than 50 books, magazines, and websites representing different sides of nuclear issues.

That does not make the site neutral. Hoffman identifies himself plainly as a “Concerned Citizen,” and the structure reflects a strong interest in nuclear hazards and policy disputes.

That honesty is part of what makes the site useful in 2026.

A reference work can also preserve an argument

Modern explainers are often scrubbed into a corporate voice that hides who assembled them and why. This glossary does the opposite. One person gathered the material, chose the categories, selected the images, and left the point of view visible.

That makes it useful on two levels.

A reader can use it to encounter older terminology and technical subjects surrounding reactors, radiation, fuel handling, contamination, nuclear weapons, and waste. A researcher can also use it as evidence of how nuclear technology was being discussed on the independent Web in the early 2000s.

The presentation is unmistakably from that period. It announces a preferred 1024-by-768 screen resolution. It expects visitors to browse topic by topic. It does not bury the project inside an infinite feed or an SEO article factory.

It is simply a large hand-built information object.

CacheRat’s 1,967 Ancient Web Domains research list keeps turning up sites like this: projects where a webmaster decided a subject deserved hundreds of links and pages whether or not a publishing company agreed.

Twenty-four years later, the nuclear glossary is still there.

Browse Russell Hoffman’s nuclear glossary

Posted on by

Ancient Web: Neocomputer Repaired Atari E.T. Three Decades After Release

Somebody eventually looked at one of the most mocked video games ever made and asked the more useful question: what, specifically, is wrong with the code?

Visit Neocomputer’s E.T. repair project

The Neocomputer project on Atari’s E.T. the Extra-Terrestrial did not stop at repeating the familiar story about the 1982 cartridge, the rushed development schedule, terrible reviews, and the New Mexico landfill. It treated the game as software that could be analyzed.

That led to something far more interesting than another “worst game ever” list.

The project was created by David Richardson, also known as Recompile, and published in 2013. Richardson examined the Atari 2600 ROM, documented several of the behaviors players hated, and made targeted changes rather than rewriting the game into something else.

The most famous problem was falling into wells.

One pixel could ruin your afternoon

In the original game, E.T. could fall into a pit when part of his sprite overlapped it even though the character appeared to be standing safely beside the opening. The Neocomputer analysis traced the behavior to collision logic and changed the relevant code so that the game judged E.T.’s position more sensibly.

The project also adjusted energy consumption. The original game drains energy constantly as E.T. moves, making exploration punishing. The patched version changes that behavior so ordinary walking is less expensive while actions such as running, falling, and hovering still consume energy.

Another tiny modification corrected E.T.’s color closer to the tan appearance of the movie character by changing hexadecimal values in the ROM.

These are small changes, which is exactly why the project matters.

E.T. was written by Howard Scott Warshaw under an exceptionally short development schedule in 1982. Decades of retelling eventually compressed the entire game into a punch line. Richardson instead demonstrated that specific frustrations could be identified in the program and altered without discarding the underlying design.

PCWorld documented the project in 2013 and later identified Richardson as the sole creator of the repair. Warshaw himself praised the work and said the changes addressed problems he would have liked more development time to fix. The patched ROM also escaped the original webpage and continued circulating through emulation and preservation communities.

That afterlife is important because the Neocomputer page itself has become unreliable to reach. The work did not disappear with one server. It has been discussed in Atari communities, technical articles, academic writing about game archaeology, and later fixed-ROM releases.

There is a nice old-Web lesson buried in there. A hobbyist can publish a careful piece of reverse engineering on a small personal site and, if the work is specific enough, it becomes part of the historical record of the thing it studied.

CacheRat’s 1,967 Ancient Web Domains research list turns up plenty of pages whose influence is easier to see in the links and citations they left behind than in their current traffic or appearance. Neocomputer’s E.T. project is one of those.

The game that supposedly belonged in a landfill ended up getting a code review thirty years later.

Return to the original Neocomputer E.T. project